Hyperrhiz 15: Essays

Articulate Filters, Fluent Bodies: Reading Alan Sondheim’s Writing Under (2012)

Andrew Klobucar

New Jersey Institute of Technology


Writing Under (2012) is West Virginia University Press’s recent print collection of Alan Sondheim’s critical and creative writings. Despite the diverse range of topics and extended time period it covers, it presents a remarkably cohesive summary of the artist’s key aesthetic and philosophical interests. This extended review of the volume looks more closely at two of these primary interests, noting first the consistent, yet highly original use of algorithmic filtering in his writing, and second, the important application of this method to critically explore the physical body and concepts of selfhood within electronic media. The essay begins by situating Sondheim’s theory of “articulated filtering” and the use of digital text generation/combinatoric functions within a larger history of modal writing and serialism in the electronic arts. As Sondheim himself shows, the very process of “filtering” can be traced back to many early 20th century poetry experiments, revealing a core lineage of literary invention based on generative functions and mathematical procedures applied to language. This analysis then goes on to present a number of key philosophical queries triggered by Sondheim’s writing into how one’s individual awareness of the self and the material body continue to be challenged by electronic media network technologies. In this latter topic, Writing Under can be usefully compared to Michel Foucault’s late critical work on “technologies of the self” and Theodor Adorno’s postwar reflections as collected in Minima Moralia (1951).

Articulate Filters, Fluent Bodies

If one is even remotely familiar with the electronic literary work of Alan Sondheim, then one has likely been receiving various samplings of his ongoing text, image and sound-based project in his or her inbox for many years — perhaps decades. A quick glance at Writing Under (2012), a critical collection of his works, published for print by University of West Virginia Press, shows that they have been in production (and circulating) almost since the very inception of the Internet nearly three decades ago—certainly longer than the web, itself. In fact, Sondheim’s long, unique career covers an unusually robust and varied span of literary history in the ongoing development of electronic literature, ranging from some of the earliest experiments in text generated poetry in the 1980s to the multimodal, interactive literary apps now available as tablet interfaces.

Despite working in a variety of media forms and formats, it is the earlier, text-based explorations in e-lit, as this volume makes clear, that continue to dominate his own aesthetic and critical interests. Writing Under features a remarkably diverse, yet representative share of what Sondheim calls “articulated filtering” in his approach to writing: a “form of mathematicization of a text or part of text, through which the chosen domain is modified in its entirety by one or another algorithm” (25-26). The aesthetics of “filtering” have been suitably central to the cultural development of electronic media across formats, consistent with the expanded role coding and computation continue to play in the composition process. To read or work with electronic documents, Sondheim notes, is to consider a fundamentally different relationship to how meaning is produced via media, especially when we begin to apply automation and the mechanics of pattern generation to the practice. In many ways, the idea of the filter (articulate or not) effectively overturns traditional understandings of media as a necessary conduit for building meaning in and out of the world on their head. Instead of media serving as a kind of tool or lens, enabling our capacity to understand the world around us, filters invoke media more as data-filled sets of confusion, against which various sifting mechanisms or digital sieves must also be applied.

Text generators continue to constitute one of the more important and popular models of these kinds of filters in electronic media, allowing, as Sondheim shows us, computation itself to operate as something akin to its own literary genre. This notable addition to modern techniques in the literary arts echoes the emergence of pandiatonicism, modal writing and serialism in 20th century music composition, where formal, often mathematically based patterning in tonality began to inspire a broad range of alternative harmonies counter to the more traditional use of the chromatic scale. As early as the 1950s, immediately following the development of the first high speed digital computers, composers like Lejaren Hiller and Iannis Xenakis began to use computation and algorithmic programming to produce entire musical scores. Working with fellow composer and chemist Leonard Isaacson at the University of Illinois, Hiller programmed the university’s “Illiac,” or Illinois Automatic Computer, for specific procedural and stylistic parameters, resulting in the “Illiac Suite” (1957). Histories of modern music still consider this the first score to appear first as a computer program and second as musical notation for performance by a string quartet. Even more recently it appears that Grammy-winning producer Alex Da Kid found a successful song writing partner in IBM’s Watson, to help him generate popular music and beat patterns along with successful lyric combinations to build one of the more successful Billboard chart hits of 2016: the new country-ish ballad “Not Easy.”

Although experiments in text generation were occurring at roughly the same time, none seems as fully developed as the efforts we see in music until we encounter the combinatoric writing programs introduced by Hugh Kenner and Joseph O’Rourke with their “Travesty” text generating program produced in the early 1980s. Travesty, written for PCs, used Markov chains to appropriate different letter combinations out of various source texts to build or generate entirely new ones. The best summary of this project with examples of the types of strings it could produce still remains Charles O. Hartman’s excellent critical survey of early electronic literature Virtual Muse: Experiments in Computer Poetry (Wesleyan UP, 1996). The results, as Hartman shows, tend to be humorous, often nonsensical in content, depending on the number of characters in the text string used to generate the new words. The longer the string (up to nine characters), the more closely the generated words will resemble those of the source text. Shorter text strings produce a more varied combination of characters grouped together, risking less lexical coherence when unusual letter patterns are randomly produced. Regardless of how familiar (or even readable) the resulting words might be, consistent with both the name of the ensuing text and the program itself, each generation proudly announces itself as a “travesty,” a parody of a narrative never actually written. Viewers tend not to expect a work of high literary value to appear; in fact, were anything resembling a polished work of literature by chance generated by the program, most viewers would probably panic outright. Success with “Travesty” shows us ironically how little grammatical or syntactic order is required to evoke some level of coherence. Our filters, regardless of how simple each formal algorithm appears, seem highly articulate.

Travesty’s remarkable capacity to forge new literary and aesthetic relationships with language patterns helps us understand and situate Sondheim’s own ongoing writing project. As Sondheim aptly shows, the very process of literary “filtering” can be traced even earlier in 20th century poetry experiments, revealing a core lineage of literary invention based primarily on generative functions and mathematical procedures applied to language. “Travesty’s” use of Markov patterns certainly looks back to Jackson Mac Low’s incorporation of chance operations in his writings as a poet and playwright as well as the radical fusion of mathematics and literary techniques introduced by OULIPO at mid-century. With each of these examples, we move in 20th century poetry from a procedural use of pattern and form as an aesthetic method to actual computer programming. In its dual form as a computer program and a literary project, however, Travesty brings to this lineage a particularly significant theoretical examination of some extremely complex shifts in the role media plays in the construction of meaning within modern culture. Once again, the authors’ own critical interests in disturbing traditional concepts of narrative and the composition process are evident in the project’s very title; in announcing the work as a “travesty,” Kenner and O’Rourke are also emphasizing its inherent absurdity, reminding readers that whatever combinations may result from their programming, the new texts continue to misrepresent and destabilize all source texts in the form of mockery—works in search of new patterns and possibly missed or overlooked structures.

The conflict between the two elements in play here — the lexical elements as patterned strings of text, arranged seemingly randomly, and the piece’s projected semantic coherence as a single literary work remains highly provocative. At the same time, this engagement underscores an even more fundamental tension in these works in terms of how they relate to or even appear relevant to human social experience. To attempt to read these pieces, to peel back (or filter through) the layers of abstraction they present, encourages a profound, yet necessary agitation between the work on display and its viewers as both individual social beings and cognitive agents. As much as the display itself is able to invoke some degree of sense, whether viscerally or via language, an aesthetic gap nevertheless emerges with the viewer, who in response must negotiate a constant volley of fresh interpretations. Indeed, as our viewer turns to confront these new forms — the lexical patterns refuse to cohere the texts that fight familiarity — she finds herself indefinitely suspended in doubt, uncertainty and perhaps even some anxiety. Key to each work’s underlying structure, of course, is its consistent negation of any and all prior referential meaning. This linguistic events mean nothing beyond the fact that they were executed, generated by an algorithmic “filter” per se. What momentary glimpse into referential meaning they may provide from time to time remains purely grotesque — a mockery of allusion, fleeting at best. What our viewer feels, what she understands, what she senses in her reading of the piece derives fully and immediately from the text at hand. To make it new, these writers as programmers have long maintained, is to forgo reference in favor of a more immanent and materially anchored appreciation of the text as its own object — a thing transmitted for immediate observation.

Aesthetic experiments like “Travesty” tend to prioritize the programmable routines running the events or filters we see in action over the actual texts being generated. At the same time, the poetic or cultural value of these works derives largely from how they conceive and reconceive the reading process as an individual, intellectual cognitive practice. To interpret such works, to engage them as aesthetic patterns and forms is to partake, not so much in an act of communication, as simply more computation, where coherent narratives or relevant social references are readily abandoned in favor of immediate procedures. Again, this latter attribute remains especially evident throughout the history of computation in literature. Where the letter combinations fail to build convincing semantic constructs, the resulting fragmented syntax, along with the consistent lack of referential meaning underscores a state of near constant readerly dissonance. At the same time, considered aesthetically, this conflict constitutes a uniquely dynamic space of critical personal reflection and ontological crisis.

This tension between the “reading” self (or self as reader) and computational subjectivity is particularly key to understanding Alan Sondheim’s ongoing project in electronic literature. Central to his critical use of coding in the literary arts is a dual understanding of the electronic text as a computational system as well as a readerly experience in and of itself in terms of a distinctly embodied, physical mode of subjectivity. Writing Under offers an impressive mix of Sondheim’s writing theory, statements, manifestoes, etc. in relation to the human body as a kind of physical conduit for “experiencing” computation. Particularly valuable is how it brings together several unique insights into how online writing has evolved as a physical practice through the very development of the web itself from the mid-1990s to the present day. Sondheim refers to and understands these various texts as a single work-in-progress he calls the “The Internet Text.” In his introduction to the collection, electronic literature critic and theorist Sandy Baldwin historically links this project to a variety of important, early email listservs, like Postmodern Culture, dating back to 1994. Baldwin himself summarises Sondheim’s ongoing endeavor as “a meditation on the philosophy, psychology, political economy, and psychoanalytics of Internet (computer) communication.” The term meditation is also used by Sondheim in his own description of the collection when parts of it appeared earlier online in the first volume of the Electronic Literature Collection published in 2006: “The Internet Text is a continuous meditation on ‘cyberspace,’ emphasizing language, body, avatar issues, philosophy, poetics, and code-work. It is written daily and presented on several email lists including Cybermind and Wryting. Many of the pieces within it were created through CMC, interactions with computers and online protocols, and programs” (ELC Vol.1, Oct. 2006). As inclusive as this list of discourses seems, the term “meditation” is particularly useful in emphasizing the work’s relationship to concepts of human selfhood: “The Internet Text,” according to Sondheim, “describes the phenomenology of the ‘electronic subject,’ the user who is plugged into the computer as a correspondent or researcher.” Part of the enduring cultural value of Sondheim’s unique efforts here derives from very specific questions regarding, not just how users “used” the Internet as a communication tool (that is with an objective in mind), but again how they confront it in terms of their physical, sensual perception — which is to say, as actual beings: how we sit facing lit screens, a mouse, hard, yet mobile, fluid, in one hand, moving the cursor, directing our line of vision, fingers poised over keys, the thumb resting on the space bar, etc. Sondheim approaches the internet, in this way, as a material practice — reimagining computation itself as a corporeal activity and thus mode of self-development. The preferred formats that Sondheim has continuously used to give shape to and distribute his electronic media work, along with his personal reflections, support this enduring focus on the self in relation to these electronic networks.

Sondheim has continuously depended upon online delivery or distribution-based protocols like email lists and UseNet groups to develop his aesthetic vision, recalling the internet’s earliest incarnation as an electronic document circulation service rather than the modes of mobile, “platformed sociality” and multimedia streaming it offers today. In his own description of his preferred methods and practices, Sondheim specifically emphasizes his continued use of the same early usenet listservs for publication. He has a subscription at Panix.com, using a Pico editor and a live connection. In operation since 1989, Panix network remains one of the oldest, continuous commercial Internet providers. For Panix subscribers, the internet takes the form of a personal UNIX account, accessible via a shell login. In this way, Sondheim’s practice not only harkens back to a vision of the Internet as a primarily text-based communication tool, but also a public telephony resource accessible by and within community centers. The specific parameters forming his writing method includes the server load at the Panix community, the floppy disks he would store and back-up his files on and the terminal itself. There are limitations endemic to Pico editors as well: Pico allows you to work on only one screen at a time, and accordingly does not support editing or even simple “find and replace” searches across multiple files. It also cannot copy text from one file to another. Such restrictions remain central to his project not only in terms of the technology being used, but as core components of his actual authorial position; in his words, “[t]here’s a sense urgency for me in this accountancy which is maintained only by a live connection” (2006); indeed, writing live online, constantly anticipating technological drops in service, yet actively alert to the specific community awaiting your transmissions constitutes its own dynamic. Just as significant is the subsequent archiving method Sondheim has come to depend upon as the foundation of his project. Each file sent through his Pine email is also added to another list of text files he keeps current at http://www.alansondheim.org/. Typically it will be named “zz” to keep it at the end of its sorting and easy to find for further editing.

And what of that community imagined by Sondheim as he sends out piece after piece? How does this increasingly unusual mode of presentation affect our interpretation of the work at hand? First and foremost, to review a Sondheim work in its proper context as an email delivery begins by subscribing to one of the listservs Sondheim co-moderates. His most reliable email lists remain Wryting-L and Cybermind, both currently based out of the University of West Virginia; Cybermind was hosted originally by the University of Sydney. Memberships in either or both of these listservs results in one or more original works by Alan Sondheim being delivered to your email server at least once a day, often on multiple occasions. An inclusive archive of these deliveries organised by month and day from 1999 to the present can be accessed at http://sondheim.rupamsunyata.org/. Alternatively, most current deliveries are available via his Google+ page.

The works, as he himself notes, vary profoundly in form and style — so much so that one instantly seeks a certain equivalence in method in order to overcome the chaos of appearance. Many of them are text based; others offer image or video captures of his experiments with Second Life. Asked to differentiate his output and Sondheim offers terms like “codework, hypertext, online writing, blog or MOO writing” (20). Aside from computation providing the most significant paradigm of both writing and subjectivity throughout this project, one might also note how key the concept and basic structure of networks remain. For Sondheim, networks well underscore the enduring liminality or fluidity of all digital transmissions; in his words, they “are first of all mediated by technological apparatus (including the power grid) and are second of all in-between process and stasis” (25).

Considered simultaneously as individual, discrete documents delivered one at a time and as ongoing nodal streams of personal correspondence, The Internet Text reads easily as an “online” practice. In Sondheim’s words we have “continuous investigation, movement, within diffused sites, applications, networks, inter- and intra- nets, PDAs, cellphones, wireless and Bluetooth, satellite and other radios, cable and other televisions…”(24); in this way, Sondheim means to set up specific conditions, whereby his addressees are able literally to communicate with each other. The minute we open these writings on whatever screens we may have in front us, via our personal Google+ or email accounts, or even the imitated print pages used for this publication, we, too, effectively generate a new Usenet group in its original “format,” encouraging us to imagine the internet again in one of its first incarnations: as the world’s first public, electronic mail and newsletter system.

For Sondheim this context, however transitory, lays bare a very specific set of historical relationships, beginning with the World Wide Web’s original function as a public utility. The Usenet emerged formally in 1979 as the pet project of two graduate students in computer science seeking to build a live, public extension of ARPANET, through which text files could be sent from computer to computer via an open bulletin board system (BBS) available 24/7 to whoever might want to log in. The initial structure of this network supported multiple, topic-based newsgroups open to subscription and able to distribute an unlimited number of announcements, postings and cross-postings thanks to a relatively simple shell script. Before this model emerged, electronic correspondence and news distribution were both possible between any computer terminals formally on the ARPANET exchange, yet uploading and downloading normally took all day, while the network itself had always been conceived as a self-contained, closed system. Usenet offered a new approach to electronic networking in general by opening up these early newsgroups of ARPANET to a wider public. The vision of networking that subsequently took form, as is immediately evident in the earliest dialogues being then distributed, emphasized extending ARPANET's experiments in human networking into a worldwide system of instant, open communication. Suddenly a “WorldNet” became imaginable where, “with a PC in almost every home in a few years,” people could “freely voice their ideas doubts and opinions to gain insight into the very important issue of mass communication” (Skinner, 19 Oct. 1982). Similar values are prevalent throughout Sondheim’s own extensive practice, informing his views of online activity in general. In fact, Sondheim never shies away in both his art and criticism from presenting online writing as an inherently revisionary information system that also signifies a fundamental transition in language use itself.

Sondheim, too, sees in his email lists a way to engage the self, in part, as a physical discipline. Read through Foucault’s theoretical hermeneutics of technology and the body, Sondheim’s relationship to the Usenet might be usefully considered as a very current “technology of the self” — what Foucault says begins as a “certain way of manifesting oneself to oneself and to others.” With Sondheim specifically, we see electronic correspondence as an opportunity to identify and then personally engage literal, physical communities of other subjects: upon opening up his UNIX shell, Sondheim regularly types in the command “who” to obtain: “a list of everyone on now, as well as what software they’re using, for example, are they sending email, working in the Emacs or vi editors, and so forth. This community — one might say, communality — is always in the background, even though I rarely hear from these people. The computer is always already shared” (36). Far from presenting a tool for individual expression, the computer terminal functions here quite literally as a public space, inherently revealing the writer to a greater social body. Comparing traditional print modes of correspondence to email, Sondheim notes the superior capacity electronic formats have to build a sense of presence in the self: “What is the relationship between email and ordinary mail? I am a poor letter-writer in real life; what disappears, seems not only gone forever but an artificial construct, complete with its own rituals. Email on the other hand is instantaneous, of the measure of speech; I tend to create perhaps 40 email posts a day minimum, many back-channel and many to the listservs themselves” (“Blood,” 2006).

Not surprisingly, one of the key metaphors for this particular relationship to writing is that of illness, where the desire to communicate, whatever medium is chosen, is likened to a “compulsion” of sorts — an emotional, as well as physical need to re-connect, and thus “heal” specific vulnerabilities Sondheim has specifically diagnosed. The “book” itself, he tells us, is “an inescapable addiction, raging, regulated, in the absence of drugs,” with its words facilitating a kind of “cauterization of a wound refusing to heal” (22). In a piece entitled “Disorders of the Real,” Sondheim writes,

Containing the blood, detaining it. I'm a bloody mess. Clean up that bloody mess. @create container named Blood "One harbors the flesh against it; the skin chafes, reddens, blisters, scabs, in order that liquidity find its way towards the surface and give vent to the externalizations of form. Form, thrust out. Form, thrust away." (1988)

Sondheim re-uses this text for his opening to a later meditation on “Blood,” presenting a quick paced, “cutting” set of more current descriptions of various injuries, situations where the skin has been pierced to the point of bleeding:

Opening a can, I cut myself deeply at the base of the index finger tonight; the blood flowed forever, through an opening about 3 cm long. I bandaged it tight and the bleeding stopped; after a couple of hours, I removed it and the blood started again, this time black. It's bandaged again; it should coagulate by morning, or tomorrow evening at the latest. (2006)

As sordid as these details first appear in all their carnality, we should note here the overly willful tone carried through to the end in the sense that the injury is primarily an opportunity to explore concepts of agency, presence and self-verification. Sondheim cuts himself and proceeds to control the bleeding, first by bandaging and stopping the flow, then causing the flow to continue by removing the dressing. One can’t help but re-live the injury through the writing, which is precisely Sondheim’s objective here: writing of his injury, he is able to extend it into the text itself, thereby integrating both situations as equally physical, equally communicative: bleeding, writing. The practice of writing appears even more physically, bodily intertwined in a later summary of his “Writing Habits”:

This is the process that’s carried out, day in and day out. (I think I once went for 6 days without producing anything — in 15 years.) I suffer from depression, poor eyesight getting poorer, minor ‘twitches’ in my left frontal lobe, and severe insomnia; it’s the last that governs when I write and perhaps how I go about it. I write at all hours of the day or night; I need to have a complete and final thought that could always be conclusion if I should happen to die at the time; death in fact haunts the text at one end, and sex on the other, with struggling of the inscribed body in their midst. (71)

Looking again to Foucault’s “technologies of the self,” we find a similar relationship to writing as a means to help one keep one’s “self” coherent — to provide a basis of self-integration. Such functions seem especially explicit in Sondheim’s work, where the author often appears quite literally on the verge of disintegration. Without writing to the Internet, without his daily routine of correspondence, he is, he informs us, at risk of falling apart. Further, all efforts to write the self, Foucault reminds us, begin unequivocally from a place of doubt, not certainty. One picks up the stylus, as with any tool, in a gesture of compensation, to offset, possibly repair, specific limitations in one’s physical, non-augmented life. There is something enabling in what Foucault calls “the possible gaze,” where one’s vulnerabilities are made accessible. This concept of the possible gaze seems to parallel Sondheim’s daily need to have what he terms “a complete and final thought that can always be a conclusion” (71). The text put in circulation continues to compensate for deeper questions and apprehensions that refuse to be settled over course of his day-to-day life. Such practices again emphasize a fairly fundamental gap between the actual self and its virtual counterpart in his writing, where the former invokes the latter more at a point of meditation or contemplation, rather than assimilation.

It may be tempting here to compare both the structure and intellectual tone of Sondheim to the collection of aphorisms and discontinuous insights Theodor Adorno gives us in a work like Minima Moralia (1951) — to see, in other words, these “under writings” as a new set of “reflections” from another, equally “damaged life.” Yet where Adorno sought literally to convey an antithesis to Aristotle’s original “Magna Moralia,” a set of ethical teachings by which readers were invited to experience and partake in “the good life,” Sondheim in effect returns to imagine, write (and to live) a work more akin to the original prescription. The damages — or more consistent with the metaphor “symptoms” — we see scattered throughout the fragments, correspondences and reflections before us, unlike Adorno’s critiques, remain confined to the texts themselves. Sondheim does not live to write, as we have typically come to regard the literary ethos of the modern era; Sondheim must write in order to live, his words flowing forth materially, much like any bodily fluid or physical condition, transporting immanent sensations and perceptions symptomatically. We have in the 53 short essays gathered here a collection that is acutely aware of its own incompleteness – a fact Sondheim makes clear from the very beginning, when he describes the work as a mere sample of a much larger, ongoing project, he calls “The Internet Text.” Consistent with its title, while part of the actual Internet, this work literally eschews boundaries as part of its very structure, especially when defined in its broadest sense as a both a digital information network and a corresponding set of communication technology protocols. The protocols, as much as the concept, itself, we can see, are essential to Sondheim’s project. These texts, in short, are conceived, produced and distributed as a means of actively, individually engaging with the Internet itself. Sondheim demonstrates this practice on a daily basis, writing, uploading, streaming, sending and programming his work as electronic, Internet based correspondence, reminding us of his life being lived more than any single set of ideas of concepts. We have thus in Sondheim’s “Internet Text” evidence first and foremost of a living practice — the different texts appearing in our email inboxes as part of our own daily correspondence. Whether we understand such works as personal correspondence or a set of ongoing public bulletins, we rarely lose sight of the fact that we ourselves are being called to respond with our own personal lives. In this shape and manner, we have the Internet Text: text that is simultaneously both writing and Internet, both individual human messages and a communication infrastructure.

Works Cited

Adorno, Theodor. Minima Moralia: Reflections on a Damaged Life. London: Verso, 1951. Print.

Foucault, Michel. Technologies of the Self. Ed. Luther H. Martin, Huck Gutman and Patrick H. Hutton. Boston: Univ. of Massachusetts Press, 1988. Print.

Hartman. Charles O. Virtual Muse: Experiments in Computer Poetry. Middletown, CT: Wesleyan UP, 1996. Print.

Skinner, Greg. Usenet Post. subject: Worldnet Responses. (19 October 1982) <uc.bds. at MIT-EICS at MIT-MC>. Web.

Sondheim, Alan. “Blood.” from The Internet Text. Electronic Literature Collection. Ed. By N. Katherine Hayles, Nick Montfort, Scott Rettberg and Stephanie Strickland. Vol.1 (2006). Web.

— . Disorders of the Real. Barrytown, NY: Station Hill Press, 1988. Print.

— . Writing Under: Selections from the Internet Text. Morgantown, WV: UP West Virginia, 2012. Print.


  1. See for example, Theo Lutz’s famous 1959 recombination of chapter titles and subject listings taken from Franz Kafka’s The Castle or Jean Baudot’s early collection of generated verse entitled La machine à écrire (Montreal: Editions du Jour, 1964). Chris T. Funkhouser’s historical survey of early programming experiments in writing, Prehistoric Digital Poetry: an Archaeology of Forms, 1959-1995 (UP Alabama, 2007) still provides the most inclusive and accurate history of the first computer programs coded and used for text generation.

DOI Permalink