Plain Text

The Poetics of Computation

Dennis Tenen

• Publisher: Stanford University Press
• Year: 2017
• ISBN:
  9781503602342
• Edition: 1
• Topic: Social Sciences
• Subtopic: Media Studies
• Index: Social Sciences

This book challenges the ways we read, write, store, and retrieve information in the digital age. Computers—from electronic books to smart phones—play an active role in our social lives. Our technological choices thus entail theoretical and political commitments. Dennis Tenen takes up today's strange enmeshing of humans, texts, and machines to argue that our most ingrained intuitions about texts are profoundly alienated from the physical contexts of their intellectual production. Drawing on a range of primary sources from both literary theory and software engineering, he makes a case for a more transparent practice of human–computer interaction. Plain Text is thus a rallying call, a frame of mind as much as a file format. It reminds us, ultimately, that our devices also encode specific modes of governance and control that must remain available to interpretation.

CHAPTER 1

> METAPHOR MACHINES

Jean Baudrillard’s sometimes enigmatic comments on the genesis of simulacra in the late 1980s are beginning to come into focus for me only now, when code has already conquered culture. Today, computers interject ubiquitously. We drink our morning coffee from brewing machines that use fuzzy logic. At night we go to bed surrounded by smart alarm clocks and sleep activity monitors. Baudrillard wrote, “At the limit of an always increasing elimination of references and finalities . . . we find ‘the programmatic sign,’ whose ‘value’ is purely tactical.”¹ Like good theater, we experience the machine in immersion, suspending disbelief. This leads to what Baudrillard called the “ever-increasing loss of resemblances and designations.”² The programmatic sign circumvents critical reflection in favor of affect. Its value is tactical in the sense that it appeals to practical reason alone, by invoking purely instrumental terms, such as efficiency, performance, disruption, and innovation.

Code, as we encounter it here, elicits pleasure and catharsis. In this, the programmer and the user occupy distinct and unequal positions in the epistemological hierarchy. Code is not usually meant to be decoded by those it acts upon. Recipients of codified control are spared the friction of signification, remaining instead in the state of asemiosis and therefore nescience. In return, code compels exacting obedience. Baudrillard understands the essence of the programmatic sign to lie in the “micromolecular structure of command and control.”³ The appeal to structure at the molecular level draws readers’ attention to the perils of selective illiteracy. As inscription, code settles within remote materialities, at strata not immediately perceptible to human senses. Users are thereby confronted with the choice to persist in the simulation pleasurably or face the difficulty of microscopic reading, which requires special tools and training. Value is further derived from the resulting imbalance of critical acumen.

The programmatic sign so conceived splits its energies between screen and hard drive. On the screen, at the site of projection, the programmatic sign simulates familiar materialities: buttons, bins, files, folders, drawers, desktops, windows, tiles, wood grain, drop shadows, chrome. At the site of storage, from which the sign emanates, the language changes to the vocabulary of control: central processing units, compilation, commitment, extraction, command, condition, initiation, handling, function, persistence, and execution. In the rift between the sites of storage (what is) and projection (what appears to be), the programmatic sign undergoes a series of structural transformations. What originates from (1) the keyboard as the mechanical action of a switch becomes (2) an electric signal that (3) leaves electromagnetic marks in computer memory, which (4) morph into phases of liquid crystal on-screen, leaving behind (5) letters that emanate outward as light. Programming languages bridge the passage between bodies, archives, and screens, breaking thought up into differentiated units for transfer. Simulation emerges in the reassembly of fractured media into a seemingly continuous integrated whole, whereby texts dissolve into letters and pixels, which then congeal back into holistic literary works.

Roman Jakobson called such construction and deconstruction of meaning the “profuse exchange of ritualized formulas” or the phatic function of language.⁴ Code, like the phatic utterance, facilitates the exchange of information through convention. The words “Hello world” on paper or in plain text format take up 11 bytes. In the Portable Document Format (PDF) they make up more than 24,000 bytes. The added information is directed not toward the receiver of the message but toward the channel itself. Code describes the rules of engagement between author and reader. It makes concrete and constrains the physical capabilities of the medium. The codified phatic function is more profuse than it is in speech, where it is limited to occasional metacommentary: Can you hear me now? It is also less apparent in that it contains other (machine) languages and registers that are physically inaccessible to the recipient.

Programming at its essence is a phatic activity. Code shapes and commands. At the same time, it conjures fantastical metaphors to occlude the structure of shaping and commanding. The simulacrum created by code obscures the incongruence between visible signs and a medium’s underlying material affordances. What you see is not always what you get. We are instead confronted with a composite image, which under examination reveals a complex process of transfiguration between the visible sign and the sign at the site of its inscription.

The simulation is without a referent. It bears little resemblance to material substrata of electronic reading. We believe we are handling books; our ideas about reading and interpretation subsequently rely on that initial physical point of contact with paper. But when reading electronically, we are handling something other than print material. The semblance to paper guides our intuitions about the medium and its associated affordances: to scroll, bookmark, or turn pages. We have far fewer intuitions about the affordances of inscription at the micromolecular level. As we “turn” simulated pages, electric charges embedded into a solid-state medium cross the impenetrable oxide barrier, reaching their destination, the floating gate, through quantum tunneling. Electromagnetic inscription exploits the wave-particle duality of matter and its corollary, the Heisenberg uncertainty principle.⁵ A digital text thus comprises numerous improbable events at quantum scale.

What can be said about practices of reading and interpretation grounded in such remote physicalities? For now, only that they continue to unfold figuratively, removed from the material conditions of knowledge production. Metaphor sustains our lives in digital worlds artificially by analogy to habituated media. We already know what to do with paper; electronic books therefore replicate paper. Replicators dull the discomfort of contact between human and machine. They hinder efforts to master poetics at quantum scale: the ability to inject electrons, draw lattices, manipulate arrays, affect solid states of being. It is tempting to view media ecosystems that host our digital lives as a kind of a natural element like water or air. But we should not forget that computational ecologies are always constructed environments. They are not governed by laws of physics in the same sense that clouds or oceans are. They form part of a massive tactical effort to bring private spaces of inhabitance—think the Home folder on your computer, your family photo album, your digital bookmarks—under the purview of computational control. Simulations encode political structures that should not be naturalized, lest we succumb to the complacency of technological determinism.

My goal in this chapter is therefore to interrupt the frictionless advance of the computational metaphor, to separate resemblances from their designations, the apparent shape from a command in the imperative. What does it mean to turn a page in a medium that sustains neither turning nor pages? I rely on the language of cognitive metaphor theory to tell a story of metaphor’s influence on computation at a formative moment in the history of human–computer interaction. I approach this tradition from within, taking on its language and assumptions in the first few sections of the chapter. In the later sections I narrate the historical transition from literal computing, by which users gave explicit commands to their machines, to the so-called direct manipulation model, in which they began to occupy virtual, figurative environments. The chapter draws on archives from graphic design, literary theory, and computer science, fields that came into direct contact in the 1970s, 1980s, and 1990s. The trajectory from the conversational to the direct model of human–computer interaction points to a speculative possibility by which human and machine couple in seemingly unmediated, affective ways. I acknowledge and finally reject the idea outright in my conclusion.

VERISIMILITUDE

Metaphors structure human habitation within simulated environments. “People do not think like computers,” a group of engineers from Interval Research Group wrote in a patent application describing “methods and systems for providing human/computer interfaces.” According to their description, metaphors “permit more efficient and medium independent communications between people and computers.”⁶ Metaphor mediates in the symbolic transference between human and machine, in the liminal space where two disparate systems of representation meet to exchange information. “Certain encoded information is translated into certain decoded information,” the engineers concluded.⁷

Interfaces translate machine states (the configuration of open and closed circuits) into pictures, numbers, and letters, which are all species of human cognitive states. In essence, a computer is a system for such metaphoric transference of properties. “A user interface is represented on the display screen in the form of metaphoric objects, called icons,” wrote another group of engineers employed by the Xerox Corporation. An icon, they continued, “may be a representation of a virtual object, such as a virtual floppy disk.” Crucially, a virtual object can be accessible either in what the authors referred to as a “host system world” or an “emulating processor world,” even when “virtual floppy disk[s] may have a filing system alien to the host system world.”⁸ An arrangement of information on-screen—the poetics of emulated space—are incompatible with arrangements of electromagnetic charge on disk—the poetics of inscription.

To reflect on metaphoric function in everyday computation, consider the following passage from a patent filed by Xerox in 1991. It describes an “electronic library metaphor,” which includes

a shared books with data base metaphor, a reference books metaphor, and a card catalog metaphor in one system that allows large object oriented data bases to be organized and accessed in an exclusive environment and in addition allows access to screen icons, creates a visual hierarchy of related and shared objects, and allows mutually exclusive access to the metaphors within the library.⁹

The invention ultimately enables the “organizing, accessing, and querying of information unique to physical libraries in an electronic workstation environment.”¹⁰

A related invention describes “methods, systems, and computer program products for the display and operation of virtual three-dimensional books.” Its purpose, according to the authors, “is to mitigate the limitations of the small screen space,” that is, to transform the constrained physical dimensions of screen space into an unconstrained virtual space. To achieve this effect, the system contains “two basic types of information: content information and display/manipulation information.” The authors explain that content “refers to the text and image data for the underlying document,” where display and manipulation protocols refer to the “data defining how the text and image data is to be presented to the user.” The separation of physical and virtual spaces allows users to “touch,” virtual “books and bookcases,” which “fly” from one “space” into another (see Figures 1.1 and 1.2).¹¹

Set aside for the moment the peculiarity of being able to patent a metaphor. Note instead the ambiguity of technical language introduced by the split between “two types of information,” one describing the virtual space and the other the physical space. How does one “access” a metaphor, for example? If someone were to “check out” a book from the electronic library, would she behold a book or its metaphor alone? What is lost or gained in the translation of library space into electronic workstations? What mechanisms would ensure the integrity between the metaphor and its underlying data structure? Metaphor theory strains to answer such questions.

FIGURE 1.1. Visual metaphors extend the affordances of the represented object into virtual space, as illustrated by this drawing from the Card et al. patent showing a sample interface for viewing a three-dimensional book. Source: Stuart Kent Card et al., “Methods, Systems, and Computer Program Products for the Display and Operation of Virtual Three-Dimensional Books,” Patent US7015910 B2, filed December 21, 2000, issued March 21, 2006, sheet 3.

FIGURE 1.2. Card et al. imagine “alternate display techniques for information about a big book.” The virtual object offers new possibilities for interaction. Source: Stuart Kent Card et al., “Methods, Systems, and Computer Program Products for the Display and Operation of Virtual Three-Dimensional Books,” Patent US7015910 B2, filed December 21, 2000, issued March 21, 2006, sheet 4.

Traditional metaphor theorists see metaphor as a linguistic phenomenon. To say “The day stands ti...