IN A 1979 “electronic lecture,” later published as the frequently cited essay “History of the Voice,” Barbadian poet-critic Kamau Brathwaite laid out his theory of what he called “nation language.” In outlining the term, he tells his audience/readers:

Now I’d like to describe for you some of the characteristics of our nation language. First of all, it is from, as I’ve said, an oral tradition. […] It is based as much on sound as it is on song. That is to say, the noise that it makes is part of the meaning, and if you ignore the noise (or what you would think of as noise, shall I say), then you lose part of the meaning. When it is written, you lose the sound or the noise, and therefore you lose part of the meaning. Which is, again, why I have to have a tape recorder for this presentation. I want you to get the sound of it, rather than the sight of it.

Brathwaite here relies on the audio archive of a sound performance to capture the noise for his present-day audience. An example he plays is that of Jamaican American poet Claude McKay reciting a sonnet written in “standard English,” wherein the “noise” of McKay’s inescapably Jamaican pronunciation of the very European poem works against the words themselves. Brathwaite argues that “these elisions, the sound of them, subtly erode, somewhat, the classical pentametric of his sonnet.”

Elisions. Subtle erosions. What you would think of as noise. These are the elusive fundamentals of Matthew G. Kirschenbaum’s new book, Bitstreams: The Future of Digital Literary Heritage, which plumbs “the intersection of digital media, textual studies, and literary archives.” A recognized scholar of the materiality of literature, Kirschenbaum has published two previous books that explore different facets of digital technologies: his first, Mechanisms: New Media and the Forensic Imagination (2008), demonstrated the value of computer forensics in reading electronic literature, while Track Changes: A Literary History of Word Processing (2016) urged us to consider the tools of writing more carefully, examining how the shift to word processors influenced literary works. Though he returns to the more engaging “narrative style” of Track Changes, Kirschenbaum invites us to read Bitstreams as “a kind of sequel” to Mechanisms.

This narrative style covers more than the organization of information: Kirschenbaum is both a wordsmith and storyteller at heart. It’s as evident in his necessarily speculative approach to the bits and bytes of his chosen archives as in his style of writing. He engages the reader with witty asides, tangential flights of fancy, and extended fictional examples that help concretize the abstractness of digital archaeology. There are moments that push away nonspecialists — this is still an academic monograph — but these are often followed by less esoteric passages. In his prefatory definition of the concept that gives the book its title, Kirschenbaum dances between highly technical jargon and language more accessible to those outside computing, first pulling us in with the relatively simple definition of a bitstream as “any contiguous sequence of bits for storage or transmission. A file is thus a bitstream,” but swiftly shifting to more technical parlance with “a bitstream is a complete copy and surrogate for all data contained on some unique piece of storage media, sometimes also known as an ‘image.’” He then makes this all relevant, and vital, to nonspecialists by admitting, “I also mean the bitstream to encompass the vast sea of digital data we encounter every day. The bitstream in this more capacious sense is defined by the extent of our bandwidth, the strength of our signal, and the nodes of our networks.” This is the seesaw of Bitstreams, making readers feel we know just enough to seek more. His ideal readers include those “critics, historians, biographers, bibliographers, editors, designers, publishers, librarians, archivists, curators, dealers, collectors, prize committees, educators, and the general readers who together steward literary heritage” and who “need to understand […] the role of computers in the composition, production, distribution, and circulation of today’s texts.”

Bitstreams may not always be aiming to speak to each group in this broad list, but it is always seeking to explore that key issue. If ever there were a text that sparked intense interest in most if not all of those groups, it would be J. J. Abrams and Doug Dorst’s 2013 novel, S. The focus of Bitstreams’s third chapter, S. is a book that “doubles down on the idea of its being a book,” with inserts, marginal notes, library stamps, signs of wear, and more, all packaged in a slipcase and reproduced identically via digital files across 145,000 physical copies. One might call it a very noisy book. Beyond that, I cannot properly describe here the fascinating layers involved in the production of S. (or the very product it becomes), but the text provides Kirschenbaum with an ideal case study of the ways the bitstream has become “an integral part of the workflow of contemporary bookmaking.”

Et tu, liber? If the printed book, the very thing we might fetishize as our rock amid the onslaught of digitality, depends on the bitstream from writing through production, then what is a book? And what stable option(s) remain for storing our stories? Even before the COVID-19 pandemic forced a worldwide uber-dependence on digital communication for all aspects of life, more and more of our memories were already being created and stored in the ones and zeros of our devices. These shifting forms of our creative and commonplace communications seem to be shaping their content as well, contouring our very memories to fit the medium ­— our lives remembered via digital photo albums, social media posts, voicemail, text messages, or even GPS tracking. How can we trace such shifts when every few years our devices age out of their usefulness and most of us have no clue how to save our data for future access (and often no pressing interest in doing so)? And, if we had the desire and know-how, how would we determine the what? What we think of as wheat today might prove chaff tomorrow; what we think of as noise now might be the very sound we seek in future.

While Kirschenbaum doesn’t declare all data golden, his scholarly foundation in bibliography studies makes him mindful of the details that attend the production of literature. As he writes in the introduction to Bitstreams,

[B]ibliography is a way of knowing, a habit of mind whose remit is nothing less than accounting for all the people and things that make meaning possible, each in their own irreducible individuality. This is a book about computers and networks and files and formats, yes, but it is first and last a book about that habit of mind.

The chapters that follow — on Toni Morrison’s digitized archives, the “Poetics of Macintosh” in Kamau Brathwaite’s and William Dickey’s writings, and the palimpsestic production of Abrams and Dorst’s S. — demonstrate the various ways meanings might be deduced (and induced) by critically minding the digital details available in official, alternate, and adjacent archives. Kirschenbaum closes the circle with a coda on the ways our “currently pressing and present” bitstreams might “inflect our archives and our memory” of both the exceptional and the commonplace in our current moment.

With his careful threading together of wide-ranging examples, Kirschenbaum demonstrates how we can incorporate an attentiveness to the digital details of textual production (what he elsewhere calls digital forensics) into the humanities, and convincingly argues why we must do so now. This argument in Bitstreams is twofold: one, that we need to “make a place in the world for this work” of digital forensics and, two, that this place actually already exists, via the long-historied practice of bibliography. Across his chapters, Kirschenbaum shows us how this practice — the “habit of mind” that bibliographers develop in sorting through editions and ephemera — can extend to the software versions and hardware variations that more and more enable literary works. The benefits of meshing what we may recognize as more “traditional” textual scholarship with digital forensics is most evident in Bitstreams’s first chapter, on the Toni Morrison archive, accessible at the Princeton Libraries only via digital technology — hence the chapter title, “Archives Without Dust.” Kirschenbaum juxtaposes the image of a hauntingly “fire-blackened” draft of Beloved with a similarly haunting “glitched” Microsoft Word file of another draft version of the novel. The latter is the primary focus in the chapter, but the scanned burnt pages remain eerily insistent throughout. I imagine many would see this digitization as replacing the fallibility of physical material. Paper burns, while digital copies live forever. But digital copies are also reducible to the material that makes them accessible. The fire that burned the manuscript pages may remind us of the vulnerability of paper, but the increasingly outdated technology of the scans and Microsoft Word documents point to the deficiencies of the digital.

There is a brief moment where Kirschenbaum offers a close reading of the ending of Beloved, but it’s almost an aside, a reminder that literary interpretation is not his main objective. He makes this clear in his chapter on Dickey’s and Brathwaite’s experimental poetry created via Macintosh computers. His analysis proposes to “offer a meditation of sorts on their literary legacies as they descend to us through the interleaved mediations of software and paper, published and unpublished work, and the interventions of certain key individuals.” This is the longest and arguably most technical chapter in Bitstreams, and Kirschenbaum closes by directly addressing the tension between two significant strands of literary heritage: bibliography and literary criticism. Though he reveals his anxieties about how much the “facts of the matter” — fonts choices, application capabilities, hardware specificities — might actually matter to any future reader, Kirschenbaum argues against the relegating of “such details to the realm of inconsequential minutiae.”

It’s an argument he often makes by doing the very work of interpretation that he declares is not his purpose. True, it’s rarely direct interpretation; as he admits in an aside, his scholarly career has been “artfully devoted to glancing off the edges of meanings.” But there is no way to carefully detail the layers involved in producing S. without making each named step important. There is no way to give a meticulous (if mini) history of a single font without imparting meaning to a poet’s choice thereof. There is no way to draw attention to what you might think of as noise without making the noise meaningful, at least in that moment. And there is no way of telling which moment will be significant for our future selves.


Kelly Baker Josephs is professor of English at York College, City University of New York, and professor of English and digital humanities at the CUNY Graduate Center. She is the author of Disturbers of the Peace: Representations of Madness in Anglophone Caribbean Literature (University of Virginia Press, 2013) and co-editor of The Digital Black Atlantic (University of Minnesota Press, 2021).