Navigation – Plan du site
Actualités de la recherche

Les archives sonores de la poésie: production – conservation – utilisation

24-25 November 2016, Maison de la Recherche de Paris-Sorbonne
Decrypting the archive: Scholars and archivists gather to preserve and grant access to poetry recordings
Chris Mustazza

Texte intégral

  • 1 “Lacquer discs are prone to unpredictable and sudden catastrophic failure due to surface delaminati (...)

1As recent reports by the United States Library of Congress have shown, many historical sound recordings are in danger of degradation or destruction, due to the decomposition of the media upon which they are inscribed (Library of Congress; CLIR and Library of Congress). Reel-to-reel tapes, which have a stated shelf-life of 25 years, threaten to disintegrate into chips and powder as they await digitization in sometimes less-than-ideal archival storage. The same is true of acetate transcription discs, aluminum or glass core records that are covered in lacquer, used for many recordings in the 1930s; as the records age, the lacquer substrate peels away from the medium and destroys the recording.1 Given the many challenges to large-scale digitization, preservation, and access—economic, institutional, and intellectual proprietary—how can archivists and scholars respond to this situation with the proper degree of urgency? What are the theoretical implications of remediating physical media to a digital form? And how can these efforts go beyond preservation, providing access to wide audiences and raising new kinds of scholarly questions for researchers? Among the many kinds of recordings at risk are recordings of poets performing their own work, recordings crucial to study of poetry and poetics, as well as literary history at large.

2On November 24th and 25th, 2016, an international group of poetry audio scholars convened at the Maison de la Recherche de Paris-Sorbonne for a colloquium titled Les archives sonores de la poésie. Organized by Professors Abigail Lang (Paris-Diderot), Michel Murat (Paris-Sorbonne), and Céline Pardo (Paris-Sorbonne), the conference examined crucial questions around the theory and practice of archiving audio and video recordings of poetry performances. Preservation, access, use: these were the guiding principles around which the conversations took shape. It was clear from the presentations on a number of archival and phonotextual initiatives, all in different stages of development, that any successful project would need to address all three of these factors. For it’s obvious that one cannot use something that has not been preserved, but the conversation highlighted that the inverse is less accounted for: what’s the point of having an archive—especially a digital archive—if it can’t be accessed, widely and fairly by the world?

  • 2 In addition to my work on PennSound, I am also a Ph.D. student at the University of Pennsylvania, a (...)

3The conference began with a keynote by University of Pennsylvania Professor and co-founder of the PennSound archive Charles Bernstein. Bernstein’s talk, titled “Radio Free Poetry: PennSound@13” presented the history of the archive, emphasizing its commitment to access, and discussed how this founding principle has resulted in millions of downloads per year by listeners from the vast majority of the world’s countries. As associate director of Pennsound I had the privilege of giving the second half of the keynote,2 and talked about PennSound’s plans for the future, including an emphasis on the reconstruction of historical poetry audio archives, as well as work to distribute PennSound as a dataset, for use in what Tanya Clement has termed “distant listening” (cf Franco Moretti’s distant reading). Questions from the audience included how PennSound obtains permission to distribute recordings, the sonic fidelity of the human voice in historical recordings (how can we know, given technological mediations, whether the recordings are true to the sonic textures of the poet’s voice?), and what kind of academic publications PennSound has enabled.

4Professor Clement was in attendance at the conference and presented her ongoing and transformative work titled “High Performance Sound Technologies for Search and Analysis” (HiPSTAS). The “High Performance” in the title is a reference to high-performance computing systems (sometimes known as “cluster computing”) used to “listen” to thousands of hours of audio in a condensed period of time and perceive scales imperceptible to humans. She detailed how the system she had developed can identify applause in poetry recordings with a high degree of accuracy and suggested scholarly questions that could follow: do male poets receive more applause than female poets? Is applause more prevalent in certain historical periods? Does the venue of the reading affect a poet’s reception? These questions around the algorithmic approaches to machine learning and large-scale corpus analysis were paired with a presentation by Jean Carrive, from the Institut National d’Audiovisuel, on automatic transcription of speech in television broadcasts. The presentations raised a number of questions from the audience, including how long it would take for machine learning to produce an interpretation as credible as one by, as the person who posed the question playfully suggested, Raymond Williams. Clement responded that the point is not to make the computer a literary scholar, but rather to leverage it for tasks for which it is uniquely suited. She channeled the Russian Formalists’ concept of ostranenie in her conclusion that machines can help to “make our existing arguments and assumptions strange.”

5Moving from the experimental digital approaches to questions around the reconstruction of the history of poetic performance, the following panel featured a number of presentations of historical attempts to archive the ephemerality of performative immediacy. In addition to presentations about the fledgling days of sound recording (Professor Reinhart Meyer-Kalkus and Patrick Beurard-Valdoye), Professor Vincent Broqua explored Anne Waldman’s use of amateur audio (cassette tapes) in her work to develop the poetry audio archive for the Jack Kerouac School of Disembodied Poetics at the Naropa Institute. Broqua highlighted Waldman’s having developed her approach by working with the poet Paul Blackburn, who is well known for his reel-to-reel recordings of live poetry performances (currently being digitized at the University of California, San Diego). From a media studies perspective, it was interesting to learn that Waldman had been inspired to record poems by Blackburn, yet selected the more “consumer-audio”-grade medium of cassette tapes, as opposed to the higher-quality reels that Blackburn used. Finally, Professor Gaëlle Théval gave a fascinating talk on the performance poetry of Bernard Heidsieck, particularly the value of video recordings of his readings of poems like “Vaduz.” I found it interesting that two of the talks on the panel contained the word “trace” in the title, which could very well be an engagement with Derrida’s famous notion of “la trace.” Perhaps in this case, the written poem always contains the trace of its performance, aural and visual.

  • 3 See: (...)

6The second day of the conference brought a number of theoretical considerations about the affordances of the poetry audio archive. Professor Jean-François Puff tracked the literary and performative afterlives of Paul Eluard’s “Liberté,” including the lyric’s composition and reapplication as a poem of protest, its refrain of “J’écris ton nom” sonically regenerating in performances from wistful to oppositional. A comment from American poet and professor Jennifer Moxley discussed poems that invite these adaptations and reinterpretations, such as W.H. Auden’s “September 1, 1939” and its reemergence after September 11.3 While the two cases differ, they both highlight the ability for performative context to shift the meaning of a poem and add to what Peter Middleton has termed “the long biography of the poem.” The conversation progressed to think about how the poetry audio archive can further the afterlives of recorded poetry. Professor Will Montgomery’s talk then examined the meta-hearing that can occur in the archive, where we are able to hear poets hearing other poets. Professor Daniel Kane’s talk suggested that the poetry audio archive can serve as an ethnographic tool for studying poetry communities, and focused his talk on the New York School poets and intertextuality as community ethnography.

7Access by the general public was the theme for a panel that included Professors Olivier Brossard and Anne-Christine Royère. Brossard examined Richard O. Moore’s series of television poetry documentaries, which included features on Allen Ginsberg, Robert Duncan, and Louis Zukofksy. He claimed that the films become extensions of the texts and performances of the poems, and pointed to their didactic function for other poets, including Anne Sexton, who learned from watching the films, including the episode featuring Charles Olson. Professor Steve Evans presented his “Phonotextual Braid” theorization of the phonotext, calling for attention to the entwinement of “timbre, text, and technology.” The conference concluded with a very lively roundtable discussion on the opportunities for, and obstacles to, the creation of accessible poetry archives. The respondents, William Chamay (Centre Georges Pompidou), Pascal Cordereix (BNF), Eric Giraud (Centre International de Poésie Marseille), Emmanuelle Leroyer (Printemps des poètes), and Claude Mussou (INA), showed off current archival projects, such as BNF’s Gallica project (Gallica), and discussed challenges to open access, such as navigating copyright clearances. All in all, a tremendous breadth of topics was covered, and there was a sense that the participants were galvanized to further this important work.

  • 4 Translated from: “Nous voici embarqués dans un beau projet, qui demande à être partagé, et qui donn (...)

8In a follow-up note, after the conference, Professor Murat wrote, poetically, “We are embedded in a beautiful project, which needs to be shared, and that gives life to poetry.”4 He’s right. Les archives sonores de la poésie was just the first step in an important series that must take place to preserve and make available our poetic cultural histories. The group is now considering the next steps for the preservation of poetry audio archives, as well as future colloquia. And a collection of essays based on the papers given in the conference will be published, in French, by Les presses du réel in 2017. The hope is for more scholars and archivists to become interested in these crucial considerations. In the meantime, let this conference report, and all of its grooves and contours, hereby be inscribed into the record.

Council on Library and Information Resources, The Library of Congress. The State of Recorded Sound Preservation in the United States: A National Legacy at Risk in the Digital Age. CLIR and LoC, August 2010,​pubs/​reports/​reports/​pub148/​pub148.pdf. Accessed 31 January 2017.

“Library Announces National Recording Preservation Plan.” Library of Congress, 13 February 2013.​item/​prn-13-014/​. Accessed 31 January 2017.

Gallica. Bibliothèque nationale de France.​. Accessed 31 January 2017.

Les archives sonores de la poésie: production – conservation – utilisation. 24-25 November 2016, Maison de la Recherche de Paris-Sorbonne.​382947.

Haut de page


1 “Lacquer discs are prone to unpredictable and sudden catastrophic failure due to surface delamination, which is often caused by storage environments with high humidity and/or temperature.”

2 In addition to my work on PennSound, I am also a Ph.D. student at the University of Pennsylvania, as well as an IT director.

3 See:

4 Translated from: “Nous voici embarqués dans un beau projet, qui demande à être partagé, et qui donne vie à la poésie.”

Haut de page

Pour citer cet article

Référence électronique

Chris Mustazza, « Les archives sonores de la poésie: production – conservation – utilisation », Transatlantica [En ligne], 1 | 2016, mis en ligne le 07 février 2017, consulté le 21 février 2017. URL :

Haut de page


Chris Mustazza

University of Pennsylvania

Haut de page

Droits d’auteur

Licence Creative Commons
Transatlantica – Revue d'études américaines est mis à disposition selon les termes de la licence Creative Commons Attribution - Pas d'Utilisation Commerciale - Pas de Modification 4.0 International.

Haut de page
  • Logo DOAJ - Directory of Open Access Journals
  • Logo AEFA - Association Française d'Etudes Américaines