Digital Humanities
@ Pratt

Inquiries into culture, meaning, and human value meet emerging technologies and cutting-edge skills at Pratt Institute's School of Information

Preservation, Distribution, Production in the Digital Humanities Age

What is the most effective way to preserve an object?  The answer to this question will certainly vary, depending both on whom you ask and the object itself.  Among other things, the answer will most likely take into account the historical context of the object, as well as the history—the life—of the respondent, and how that history informs their hopes for the object’s survival and future use.  As an architectural engineer and archaeologist would probably provide differing views on how to preserve a building site, so would the librarian and the poet in regards to the preservation of an audio recording of a poem.

On February 11, 2014, the Digital Humanities Forum at the University of Pennsylvania hosted a Tools and Techniques in the Digital Humanities Luncheon featuring a roundtable discussion on PennSound, “an ongoing project committed…to producing new audio recordings and preserving existing audio archives” (PennSound) of historical and contemporary poetry readings, as well as lectures and discussions on historical and contemporary poetics.  The site also features video recordings, podcasts, and other media.  Born in 2003 into Penn’s English Department, the Center for Programs in Contemporary Writing, and the Kelly Writers House, PennSound is the largest digital audio archive of poetry in the world.  Its vast trove of MP3s are downloaded by the millions every year.  Present at the luncheon were PennSound’s founders, Charles Bernstein and Al Filreis, as well as one of its technical advisors, Chris Mustazza.

A relatively small crowd—about fifteen participants or so, of many librarians, I gathered—sat around a large table in the Meyerson Conference Center in the Van Pelt Library.  The discussion, which lasted about an hour and fifteen minutes, began with the three invited guests—Charles Bernstein, Al Filreis, and Chris Mustazza—taking turns speaking for two to three minutes (and sometimes a little longer) on a PennSound-topic of their choice.  It became apparent quickly that many people in the room had very little exposure to PennSound before discussion began.  Speaking of PennSound’s historical roots, Mr. Bernstein immediately introduced what became the crux of the hour’s discussion:  PennSound’s first goal, he said, was distribution and access to materials, and not preservation by library standards.  In these first minutes, he went on to say that although PennSound is the paramount source for poetry recordings, they have been repeatedly denied grants due to their apparent lack of focus on preservation.  But distribution is preservation, he went on to say.  (Librarians began to fidget in their chairs.)  And rather than planning to do it, we just did it.

In the first twenty-five minutes, Bernstein, Filreis, and Mustazza discussed PennSound’s history, current projects, collaborations, and achievements.  This discussion highlighted the fact that PennSound has not only provided complete readings, but has also indexed and segmented those complete readings so that individual poems can be searched and listened to.  They also pointed out the fact that all rights remain with the creators or foundations of the creators, and not to the University of Pennsylvania. Further, Mustazza introduced an extremely exciting, collaborative DH project between the iSchool of the University of Texas at Austin, which will allow PennSound’s recordings to be visualized in detail and searchable by sound using a software called ARLO.  Throughout this discussion, however, there was a notable tone of defensiveness in the way Bernstein, and more particularly Filreis, expressed their accomplishments.

For example, it was noted that they decided to digitize and produce files in the form of MP3s (versus WAV files, which are standard in archival preservation) because at PennSound’s inception, MP3s were the most accessible—most distributable—audio file format, and despite the “artifactual” quality of MP3s today, they often provide better audio quality than the original recording (even if they’re not always rival to WAV files).  They also noted that the iSchool of UTexas Austin has downloaded and converted the entire PennSound archive into the WAV format.  Neither Bernstein or Filreis expressed worry that upon their retirements the archive would “go away” (for the simple fact of PennSound’s value not only to the scholarly community, but to the world in general), and that if the School of Arts and Sciences (SAS) at UPenn did not shepherd its existence into the future (PennSound files are stored on the SAS servers), then perhaps an organization like The Poetry Foundation would.  And even if it becomes static, it will still be an amazing archive.  At this point, Ian Bogus, the MacDonald Curator for Preservation at the Van Pelt Library spoke up to say, I am worried for the future of PennSound.

Filreis strongly spoke out by saying that they (he and Bernstein) tried to get the UPenn Library on board in 2003, and there was a lot of talk of a plan but nothing ever came of it.  He also described how, despite how nice librarians are, he doesn’t understand why libraries seem to make things more difficult for themselves (and others) than they need to be.  (This comment was also in reference to how the Columbia Library has a trove of recordings that they are not willing to send to UPenn, and so instead Filreis and Mustazza will have to travel to New York City with their digitization equipment in order to digitize them there.)  To this, the librarians present admitted to failing PennSound, and that despite it being a little late in the game, they were here to help now.  (One librarian also said she would contact Columbia and see if they couldn’t send the recordings to the Van Pelt Library.)

This speaks to Bernstein’s early comment that, Instead of planning to do it, we just did it.  Does library preservation—or even more generally, do standard library practices—hinder or minimize the production of scholars?  To scholars and producers like Filreis and Bernstein, this is evidently the case. However, I think more generally, it is also not the case.  Library standards—authority and preservation standards, metadata schemas, etc.—are carefully—and, yes, thus very slowly—created so that scholars and the general population alike can ideally have at least sort-of equal access to the many of the world’s information sources.  Libraries remain the hub of scholarly research in higher education, and continued to be used more than ever by the public.  But in the words of Bernstein himself, “Authority in the defense liberty is not linear” (2003).  Authority control, though carefully (and again I emphasize slowly) created, maintained, and updated, does not necessarily provide democratic access to materials (and of course name authority name files like LCSH are fraught with their own awful racist, sexist, the-list-continues issues).  In PennSound’s case, its producers ran off without the library in tow, and the result—despite having its own preservation issues and lack of rich, bibliographic metadata—has been spectacular.  The library now is only trying to catch up.  As more and more digital humanities projects like this take shape, librarians must be aware of their potential and try to keep pace from the start.


Bernstein, C. (2003, August 01). Electronic pies in the poetry skies Electronic Book Review. Retrieved from

PennSound. (n.d.). About PennSound. PennSound. Retrieved from

Leave a Reply

Your email address will not be published. Required fields are marked *