In this review, Manuel Portela considers Matthew G. Kirschenbaum's Track Changes in light of a "general computerization of the modes of production of writing."
Matthew G. Kirschenbaum, Track Changes: A Literary History of Word Processing, Cambridge, MA: The Belknap Press of Harvard University Press, 2016. 368 pages. Hardcover. ISBN 978-0674417076
"It says, I can do anything you want now—so think, what do you want me to do?" - Anne Rice about writing with a word processor, CBS radio interview, 1985, quoted on p. 50
Track Changes tells the early history of word processing, roughly situated between 1964, when the IBM Magnetic Tape/Selectric Typewriter (MT/ST) was advertised as a word processing system for offices, and 1984, when the Apple Macintosh generalized the graphical user interface in personal computers. The history of word processing both as technological process and mode of textual production is deeply entangled with the changes in the technologies of writing as they reflect and contribute to efficiency and control in increasingly bureaucratic processes of social administration and organization. The literary history of word processing can be situated within this general computerization of the modes of production of writing. Technological innovation and the adoption of word processing across the literary field are examined here through individual practices of mainstream and popular writers, and without any special attention for avant-garde practices, in an attempt to identify significant moments in a heterogeneous and uneven process.
Anne Rice's quote above expresses what many writers in the 1980s celebrated as the complete freedom of the newly-adopted word processing machines. That heightened sense of the potentiality of the writing act as an imaginative field of open-ended interactions with a new kind of writing tool is a material consequence of what Daniel Chandler has described as "suspended inscription." In Kirschenbaum's own words: "Suspended inscription means that the stored record of a text is separate from whatever the medium or surface on which it is ultimately printed or inscribed in more palpable form" (46). Displayed as dots of light on the screen that could be endlessly manipulated in acts of composition and revision, regardless of their forensic materiality in the electronic circuit, writing signs seemed to acquire the immaterial transcendence of the purely symbolic. Once the signs of all writing systems had been abstracted as binary representations traveling recursively between screen surface and magnetic storage, the inscription became suspendable in its inherent processability. Suspension between storage and final inscription was based on the universality of machine code that gave computers their power as symbolic machines. The potentiality of writing could now be abstracted in the bitmapped patterns of light on the screen, a pure signifier replete with the potentiality of language always waiting for the moment of inscription and re-inscription. "Suspended inscription" and "writing with light" (p. 119): these two phrases try to capture the experiential encounter with word processing in this new writing space.
A particular modeling of a writing environment through hardware and software enabled that photo-electronic suspension of writing. But what is word processing exactly? Who were the technical innovators that came up with new solutions for storing, retrieving and processing written characters and for modeling the ergonomics of writing and its editorial processes in particular machines and graphical user interface displays? How did this assemblage of hardware and software change the processing and production of written language? How and when was this technology adopted by literary writers? Which particular machines and programs did they use? How did word processing change writers' awareness of the act of writing? How did it affect their production workflow? How do the various technologies, techniques and habits of writing relate to each other in this changed scene of literary production? What stock of representations and images (in the popular imagination in magazines and advertisements but also in stories and poems) has been produced around the digital processing of words? Where do we find all of this information? To whom should we talk to? These are some of the major research questions addressed by Track Changes, a tour de force in the way it weaves together an extensive body of documentary research, a theory of textual production as social material act, and, last but not least, a skillfully balanced story-telling narration.
One of the guiding principles in Kirschenbaum's approach is that there are several answers to each of those questions and all of them are worth pursuing, because there are multiple narratives, many individual protagonists, and various contexts for the processing of words. Thus we could say that the many meanings of “word processing” are what make word processing a particular technical and social practice of writing:
Does word processing belong to the history of writing or to the history of computing? The answer is not obvious. The conjoining of the word "word" with the word "processing" or "processor" has at various times been used to denote principles of office management, an actual person (typically female), a physical hardware device, or a piece of intangible software. It has also encompassed not only the written production of text with a keyboard but also verbal dictation, shorthand note taking, and the duplication, mailing, and filing of documents. (p. x)
Kirschenbaum is careful (perhaps too careful) not to offer sweeping generalizations and overall theories about the effects of word processing in literary processes (or in human written communication for that matter). In a certain sense he is continuing Kittler's project of analyzing the discourse networks sustained by our systems of communication, but unlike Kittler (who dramatically described the printing of the 4004-microprocessor by the Intel engineers in the early seventies as "the last historical act of writing"), Kirschenbaum sees human agency everywhere. He prefers to show how new writing technologies are part of multiple social and economic processes, how they always interact with earlier technologies, how they become part of a multilayered ecology, and how writers have varied practices when we look closely at their actual writing places and writing spaces. Most writers' testimonies, some of which express anxiety about sharing the act of writing with the machine and becoming a mere "word processor," generally agree on the changes in the actual labor of writing, particularly in how it impacted the division of labor (including its gendered division) as it had been established by typewriting for most of the twentieth century. Changes significantly affected the process of revision both at a micro and macro scale, including the possibility of rearrangement of large pieces of text as well as searching for or changing character strings automatically, but probably also the process of composition.
Kirshenbaum situates the word-processing moment – with the Kittlerian phrase - at "Around 1981" (chapter 3), when the topic entered popular discourse in magazines and newspapers, the market for personal computers had been established, and a significant percentage of writers were already using word processors. Changes in the writing practices are mostly tracked through the scenes of writing, that is, the places where the actual production of writing (word processed or not) takes place. Word processing — seen in its full and, in its early forms, very bulky presence — is always presented at a specific concrete moment of literary production or technological innovation for each of the authors selected. The full density of those word processing scenes, that extend from technical principles to manufacture to business to advertising to actual acts of literary production, is vividly recreated by means of a series of episodes that are juxtaposed or sequenced according to a fractured sense of time that moves back and forward, by means of interruptions and continuations, anticipations and recapitulations. Kirschenbaum's almost ethnographic description and evocation of the lived relation with the tools and devices of writing make some of the most memorable moments in the book.
His deliberate narrative construction — a sui generis combination of archival work with story-telling and film montage techniques — recovers the investments and expectations of particular moments of human interaction. One of those telecommunications vignettes, involving word processing and satellite communication, describes the collaboration between Arthur C. Clarke (in Sri Lanka) and the director Peter Hyams (in Los Angeles) writing the movie adaptation of Clarke's 2010: Odyssey Two, on which they worked in 1983 and 1984 (pp. 67-70). A sense of narrative beginning, progression and closure can be seen also, for instance, in the account of John Hersey's use of the DEC PDP-10 machine and the LINTRN software to write My Petition for More Space, first published in 1974 (pp. 131-138). Its present tense first sentence reads: "The walk from Pierson College to the Beeton Center on the campus of Yale University is a short one" (p. 131). Another promising beginning can be read in the account of the Evelyn Berezin's development and commercialization of the Redactron as the Data Secretary (pp. 149-155): "In the late 1969 a new corporation consisting of nine employees set up shop in a 20,000-square foot manufacturing space in Hauppauge, Long Island." (p. 149). The sense of discovery and experimentation with the medium is lively captured through these and other embedded short-stories.
Thus, historical depth owes as much to the wealth of documentary evidence collected as to a careful weaving of description, technical and historical information about a specific device, narration, literary analysis, and theorization. "North of Boston" (chapter 4), for instance, is entirely devoted to Stephen King and John Updike, both of whom used the Wangwriter. King's short-story "The Word Processor" (1983) and Updike's poem "INVALID.KEYSTROKE" (1983) are analyzed as reflexive engagements with the glowing screen-pixel electronic materiality of characters and of word processing as executable language. Self-awareness of the nature and implications of word processing provide the fictional and poetical structure for each of those works. Kirschenbaum considers them representative of the experience of writers using the first computer or the first word processor, and they "constitute almost their own genre" (89). Analyses of these authors’ and many others' writing practices also shows that there is not a progressive genealogy of writing machines, but rather "a very complex writing environment, with texts originating in various media and migrating back and forth between them in the course of their revision –longhand and typescript, hard copy and disk" (90).
Those multiple and divergent meanings imply a complex entanglement between computational writing tools and contexts for the production of writing, including their role in the rationalization of office practices according to management principles of efficiency, productivity, and automation. They also suggest that computational or writing technologies have to be understood within the specific social and economic relations that they help to sustain and transform. Chapter 7 ("Unseen Hands") focuses on this managerial perspective on word processing as a form of redesigning and controlling the routines of office labor: "The random, the haphazard, the unpredictable, the largely unmeasurable can thus — when understood in this new conceptual light — be replaced by the controlled, the organized, the automatable" (American Management Association, early 1970s, quoted on p. 147). A new stage of writing as a management technology is emerging with the ongoing development of cloud computing and real-time data-mining, a stage in which the use of word processing tools is a way of providing feedback to the algorithms that script our daily interactions with the network. The network itself could be equated with word processing on an entirely different scale, as it turns into a word processing system capable of crunching human writing as part of its operational protocols.
Kirschenbaum's ingenious title is not just a suggestive pun. It encapsulates a theoretical and methodological program for a materialist approach to literary history, whose writing also contains a literary model for writing about writing as technical and social act: the "literary" applies to the object of narration and to the narration itself. Aware that writing is in and of itself a powerful research tool, the text word processes itself with an exhilarating sense that its object has to be not only carefully excavated but also carefully constructed, which is to say processed in words and processed as words. History and story, fact and interpretation, technical procedure and social process, individual case and media theory, voice and perspective have to be woven into a string of readable meaningful words. Externally documented according to the rules of historical and archival research and internally patterned according to narrative rules, the book makes sure its reader feels the text’s pleasure in framing and telling its story.
Track Changes delivers the happy promise of its title not just in what it tells us about word processing and its early adoption by many writers but also in what it shows us through its own writing surface, in a way that is both subliminal and textually marked, and using a voice that is critically situated and yet always empathetic. By the time you come to the narrator's reflexive concluding sentences - "For myself, I have only one more key to press, and then, just like that, the teeming processes that make up my words within this i5 4200U CPU will come to a punctuated stop. And now, just like that … they do." (247) - a myriad changes have been tracked. While we have not literally tracked the particular changes of his manuscript, i.e., the cutting and pasting and reordering and revising and annotating and rewriting that define the actual digital processing of words, the writing process has signaled itself to us through its chapter titles and section structure, and through its recurrent leitmotifs, and its sense of time, space, and character. Inside and outside the machine, characters processing characters, writing that comes alive as word processors become machines, programs, engineers, writers, places.
How different the new work is from Mechanisms: New Media and the Forensic Imagination (MIT Press), his multiple award-wining 2008 book which brought the inscriptional mechanics of the hard-drive to the attention of pixel-fixated screen-dazzled essentialists? With his earlier analysis of hard-drive inscriptions, Kirschenbaum offered a convincing theory about the specifics of the digital computer as a writing technology materially dependent on the interfacing of physics and mathematics (matter and code) that enabled a cascade of symbolic processes from the writing symbols through programming languages through machine language through differential voltages, and back. He conceptualized this double nature of digital code as an expression of the tension between forensic materiality and formal materiality. Screen presentation, data models and programming are formal material instantiations of processes that also have a forensic material instantiation at the nanoscale of magnetic processes. It is the allographic nature of code that enables it to represent multiple symbolic systems and multiple media materialities. Word processing is just another textual practice where that particular ontology of the computer can be described and observed.
The new book is different in style and structure, but not in its more fundamental conceptualization of the technological and social processes involved. It is a continuation of the theory and history of the computer as a writing technology that we find in his earlier work. It is also a continuation of the social text rationale applied to electronic textuality that we first encountered in his earlier analyses of William Gibson's Agrippa and Michael Joyce's afternoon. Like his former book, what mostly stands out in this approach to word processing is Kirschenbaum's ability to theoretically articulate the technological and the social, and to attach that articulation to the detailed richness of writing practices as lived experience. It is a literary history of word processing but one that is able to demonstrate the co-presence of the social and the technical in the literary. The presentation of various historical forms of word processing within their social processes of material production and use, and several close-ups of individual writing scenes — where the entry of a word-processing machine reconfigured the ecology of a writing space — give Track Changes the ability to evoke the historical density of each particular moment. It is in this back and forth movement between narrative strategies and theoretical interpretations that its particular style and structure set it apart from his earlier work.
We can only imagine the amount of detailed research (and the amount of word processing…) that went into this long-awaited and much-needed book. What is most melancholically striking, as we go from one device to the next in increasingly shorter cycles of updating, is the disturbing sense of how much of an archaeological enterprise the literary history of word processing already is as we look back from the ever-instantaneous present of networked mobile cloud computing. And yet, one of the strong candidates to being the first published word-processed novel — Len Deighton's Bomber (1970) word processed by Ellenor Handley, a novel whose heroic and gendered textual processing history is told in chapter 8 — was published only 46 years ago. It is as if the magical coupling of hardware and software, which gave us the possibility of suspending inscriptions written with light, also accelerated the never-ending and palimpsestic accumulation of digital ruins of writing. WordStar, WordPerfect, LetterPerfect, Perfect Writer, Microsoft Word, Scrivener, Google Docs, IBM Magnetic Tape Selectric Typewriter (MT/ST), Redactor III, Olivetti TES 501, IBM Memory Typewriter, North Star Horizon II, Lexitron VT 1303, Wangwriter, Vydec, TV Typewriter, Kaypro II, Osborne, Hemingwrite. Constant innovation and programmed obsolescence generate a paradoxical archaeology of the present moment, whose material and social history, like an ancient script, can only be painstakingly excavated and deciphered.
Kirschenbaum's methods combine archival work in special collections and writers’ archives, oral interviews with writers and engineers, and hands-on descriptions of historical word processing machines. The book conveys the deeply social and gendered nature of writing practices, capturing the heterogeneity of individual authorial processes and the hybridity of multiple scenes of writing from the 1960s to the present. The result is both a scholarly and literary achievement. Informative and enjoyable, technical and literary, Kirschenbaum's media excavation of word processing extends our archaeological knowledge of writing media and literary practice, and it is a timely addition to the knowledge of notational and inscriptional systems that we find in the works of Friedrich Kittler, Darren Wershler-Henry, Lisa Gitelman, Johanna Drucker, Jussi Parikka, and others. The grammatologist of the hard drive has cleared the way for the historian of word processing.