Site Unscene: Medial Ideology and the Literary Interface


Who speaks in an interface? How do we understand the work done by the graphical, tactile, audio and spatial features of on-screen and ambient works of literary production? The many aspects of language – subject positions and performative dimensions – do not map neatly onto the structuring features of literary work in on-screen and networked media. This paper explores some of the ways the conventions in a literary interface produce a range of subject positions that pass as neutral user experience – and how these participate in the larger phenomenon of what Matthew Kirschenbaum calls ‘medial ideology’. In addition, it asks how the contemporary category of ‘the literary’ is identified and produced within its own ideological conceptions and their longer traditions.

If a literary interface exists, then what identifies it as distinct from other sites? To what extent does it serve a function in the workings of literature understood as the staging and presentation of literary activity within a networked environment? The answer to these questions requires several theoretical frameworks that describe the performative staging of literary work within a scene of production and reception that is continuous from analogue to networked environments. All of these participate in a ‘medial ideology’ that tends to efface the circumstances of enunciation in order to produce an experience of immediacy and apparency. I will elaborate on these issues with a particular emphasis on the contrast between works that are self-identified as literary and those that are either emphatically outside that tradition and domain or blur its boundaries.

This paper sprang initially from a fascination with the interface of early ‘e-literature’ created in a period before conventions of interactive platforms and applications had become established.1 In particular, the innovative work of the 1970s and 1980s seemed characterised by the emergence of varied and imaginative design solutions for navigation and interaction. These often engaged directly with the themes of the individual literary works or played with meta-concepts of literature and text.2 In addition, older viewing modes from film, book design and exhibit display were appropriated along with some of the devices of interactive games. The conventions of frames, pages, bookmarks, and even simulacral ‘page drape’, used book motifs, as in the popular game Myst, in a combination of nostalgic kitsch and familiar navigational metaphors (Figure 1).3 The concept of an interface can be narrowly defined in such analyses – limited to on-screen features for engaging a reader/viewer with the active unfolding of the experience of the work. But it can also be more broadly defined, to include the user’s embodied relation to the various machines and devices involved in production and reception, and the larger cultural conditions within which these are created and sustained as institutional practices and conventions. Just as the codex book offered a different set of possibilities for navigation and use than the scrolls or tablets which preceded it, and was embedded in a whole meshwork of circumstances linking production and reception, so the graphical, tactile and emerging natural user interfaces add particular characteristics, conceptual and technical affordances as part of their design, as well as indications of their connection to larger social practices.

Figure 1: Myst 25th Anniversary edition with nostalgic design, (
Figure 1: Myst 25th Anniversary edition with nostalgic design,

Interface has been defined variously as a space of interaction, a border zone or a process of exchange within a dynamic, networked system (Galloway; Emerson; Drucker ‘Reading Interface’ and ‘Interface and Interpretation’). When viewed as a space (an entity), it offers its design features for analysis; when considered as a border zone, it can be analysed for its differentiating features; when addressed as a process, it provokes discussion of the social circumstances and systems within which it is situated; when conceived as a component of a dynamic system, it can be read as an index to historical and intellectual conditions. All are relevant, and each framework provides the occasion for a distinct level of analysis.

As this paper developed, the argument took on the dimensions reflected in the punning title now attached, though the original inspiration will form a core of evidence for the argument. But the reframing emphasises the realisation that any digital and/or networked work takes place in a ‘site’ of production/reception of work that goes ‘unseen’, thus becoming obscured from ‘sight’. Though starkly visible on screen or device, the interface of the work hides as much as it conceals. But the pun contained in the term ‘un-scene’ is meant to carry a more active emphasis, identifying not only that which is invisible but also that which is not presented, not staged, not given acknowledged presence. Thus the ‘un-scene’ is not merely that which is not visible, but that which is not made available as part of the bringing forth, or making apparent, as part of the staging of a work.

This concealment is an aspect of the ‘medial ideology’ that forms an intellectual framework for my argument. The term ‘medial ideology’ is borrowed from Matthew Kirschenbaum, who uses it to describe the way the screen (on a laptop, phone, tablet, other device) produces an experience that completely hides the workings of the system (35–38). It points to the aspects of cultural objects that allows them to pass themselves off ‘as if’ they were natural, as if their conventions were simply ‘what is’ in the way we do our business. In this case, the business being done is literary, and the workings of the production of the ‘literary’ as an identifiable category of activity within electronic and digital environments are intimately bound to the way its interface works. For the user, the screen image is the computer. The elements of the interface – icons, menus, applications – appear to be the literal representation of the computer’s protocols and processes. This apparent view of the computer is a fiction, hence the use of the term ‘ideology’ to indicate the ways the screen conceals what it appears to show. The user never experiences the workings of the computational systems, and the user continues to look at the screen rather than see into it in any way. We can extend the concept of medial ideology to point to the ways the workings of a literary interface are embedded in mediating systems and cultural protocols at a larger and more complex level than just the screen and individual device.

Approaching literary production through these critical frames allows us to consider the very identification of the category of ‘literature’ not from an ontological approach – that is, what is or is not literature, as if a set of criteria could be used to distinguish certain kinds of human expressions from others and sort literary from non-literary works – but instead, from a dynamic and performative approach by asking how the interface plays a role in the production of what comes to be identified as literature in a contemporary environment, specifically, that of the Web. This approach has the added advantage that it ducks the questions of judgment on which the literary and non-literary could be assessed and attends instead to the ways interface designs by literary artists play a role in producing that identity for their works. Put more simply, these questions arise: How does an interface contribute to literary production and to the production of the category of ‘the literary’? What ideological activity does the interface engage in in the process and how does it participate in the larger ideology of the self-proclaimed category of literature? How does the interface contribute to the workings of literature, to its self-definitions and identification within the domain of the World Wide Web?

One final critical dimension needs to be made explicit at the outset: the way theories of enunciation play a role in this critical approach (see Drucker ‘Information Visualization’). By invoking enunciation as it comes out of structural linguistics, we can talk about media production in relation to the production of subject positions, as a system of speaking and spoken (enunciating and enunciated) subjects. Everyday instance of human expression – from the built environment to the nano-scale of micro-lithography – is created as a communicative act that is made by someone for someone to serve some interest and purpose (though interest is often hidden within purpose so that justification rests on utility rather than assessment of whose interests are really served). In language and visual arts, the structure of enunciation is readily apparent in the use of pronouns and point of view systems. But every enunciative act belongs to someone and is produced in expectation of being received by someone else, the implied subject of discourse. By asking who is speaking and to whom and with what assumptions in making a critical assessment, much of the ideological workings of an artifact can be revealed. Even a nano-chip is embedded in these systems; in some sense, it ‘hails’ a user within a system through the role it plays. The human-designed environment is constructed as an elaborate system of communicative exchanges, modelling behaviors and positionality through structures that inscribe relations of power through modes of address. The nano-chip may not explicitly call out, ‘Hey you!?’, to a human user, but its place within the system of working parts explicitly enables the constituting relations to take place. The whole idea that a system ‘speaks’ and thus participates in enunciation is foreign precisely to the extent that this unfamiliarity conceals the workings of the apparatus. To varying degrees, the description of this situation will apply to the workings of features of literary interfaces and systems at every level brought to attention here.

In this paper, the focus is on the digital and online interface, mainly within networked environments. However, recognising the link with traditional forms and scenes of literary production is important since these, also, can be described as ‘scenes’ that present, frame and help identify literary activity. So dramatic presentation, oral performance, the differentiating imprints that demarcate literature within the realm of print publishing are extended, rather than forgotten, in digital work. Invoking the concept of the ‘scene’ of a literary work as a foundational condition of its performative appearance for use argues for continuity between analog and digital environments. The theatricality of book and page, the unfolding and revealing of openings and turnings within the structure of a codex, for instance, are rarely considered as features of a literary work – unless it slips into the realm of artists’ books or self-consciously physical and designed artifacts. Awareness of these aspects of literary production and their relation to reception becomes intensified in an on-screen environment, and in particular when conventions are emerging, or being formalised, or, in turn, being played with and referenced as components of the staging platform of the work.

As mentioned, the period of early innovation in the 1970s and 1980s provides a particularly useful point of focus (see Funkhouser) (Figure 2). In those early days artists and writers were playing with and exploring ways of structuring the relation between production and reception, as well as the direct connection between presentation and user, within the space of the interface. This was before the conventions of the Web and Net had taken over, with their identifiable features of navigation. From the beginning, however – and the very beginning will be identified here as the collaboration undertaken by Christopher Strachey and Alan Turing on the production of the Love Letter Generator on the Mark I at the University of Manchester in the early 1950s – we have to recognise that media are not specific to genre (Wardrip-Fruin). In fact, platforms and systems are largely agnostic as to content – though of course some systems are designed to do certain kinds of processing more efficiently than others, and particular programs and platforms are matched to different computational tasks. The enunciative apparatus of the Mark I is the same as in any act of computation: it was not custom-built for a literary game. Just as radio, television or print can be used to communicate content across disciplinary domains, so can computational media. The complex set of operations required for making the Mark I work – punched tapes, switches, operators – meant a long process stretched between any impulse towards a project and the moment of its resulting output. In general (exceptions can always be found), the conventions that distinguish genres by content operate within the enunciative apparatus, as features of formal design, not the structure of the situation of creating or viewing.

Figure 2: Electronic Literature Archive; source for studying much early literary interface design,
Figure 2: Electronic Literature Archive; source for studying much early literary interface design,

Most media, including print, tend to emphasise experience as immediacy (Fuller). That is, they are designed to create the impression that the experience they present is not mediated. The effect is not transparency – not the effect of looking through the interface of page, screen, television set – but rather it is the effect of making ‘sight’ ‘un-seen’ – as if the experience were not mediated. This is medial ideology at work; the function of making mediation disappear by making it appear so natural it hardly registers.

In a community as sophisticated and aware of their historical and critical routes as that focused on electronic literature, the lessons of the avant-garde provide a foundation for their thinking. Such concepts as calling attention to the means, mechanisms or modes of production were part of a generational education, even if the specific texts of a major figure like Viktor Shklovsky were only available in translation relatively late in the twentieth century. Transferring these concepts to digital environments carried a combined charge of aesthetic and political import. Notions of resistance, exposing the workings of ideology, ‘laying bare . . . the device’ (Shklovsky 149) and ‘making it strange’ (Vatulescu 41) became, almost perversely, the regular terms of aesthetic practice. Whether these techniques (if that is what they are) and the assumptions about political efficacy that accompany them are still effective in performing this ideological work is an open question. But that they form a basic intellectual and critical toolkit which later literary work makes use of constantly is evident across critical and creative practices in both analogue and digital formats. The theoretical tenets of Theodor Adorno, for instance, have become almost prescriptive, and whether ‘difficult art’ escapes the very commodification it is purported to resist is questionable (Adorno). The thorny issues of how to create political awareness and agency plagues the early twenty-first century, not only in aesthetic realms, even as our attachment to the possibilities for artistic practice to play a unique critical role continues. Writers and artists without this belief could be readily absorbed into the work force of advertising and commercial production across a full spectrum of industry venues. But experimental and innovative aesthetics retain a commitment to the self-consciousness of the avant-garde’s engagement with exposing workings and mechanisms. The question that remains is whether this disruption is adequate to expose the ideology of mediation and, if so, how the terms of enunciation – the question of who speaks and to whom – might be crucial to reinforcing the category of literature as a realm apart from – or within – that of other cultural productions.

As the first example of the interface as a complex scene of staged production, the Love Letter Generator is not a clear case of literary work. A photograph taken in the early 1950s of Alan Turing shows him with his collaborator Christopher Strachey sitting in front of the console of the Mark I in a sterile laboratory of polished surfaces and functional furnishings (‘Computer “Love Letters”’) (Figure 3). Any concept of the interface that might casually attach to the console in this case has to be complicated by the images of the full mainframe, with its multiple banks of controls and connections, imposing room-sized hardware, and look of belonging to realms of incredibly rarified technical expertise (Figure 4). The sheer opacity of the highly distributed interface of the Mark 1, its monumental physical and electronic complexity, meant that the communicative exchange between the ‘speaking’ system and its spoken subject was mediated through protocols that were at once highly evident (programming was very basic, and every instruction had to be made explicit) and completely obscured (where, how, in what combination of clicking gears and address codes was ‘data’ retrieved to be repurposed). The computer was manipulated by switches programmed from a paper tape, and no storage for programming existed in the machine. The degrees of mediation made the process tedious and cumbersome.

Figure 3: Alan Turing and Christopher Strachey at Mark 1 console, (
Figure 3: Alan Turing and Christopher Strachey at Mark 1 console,

Figure 4: Mark I; where is the ‘interface’ in this complex and distributed process?
Figure 4: Mark I; where is the ‘interface’ in this complex and distributed process?

The Love Letter Generator was designed as a combinatorics game in language generation. The project created absurd-sounding love letters that were signed ‘MUC’ – the Manchester University Computer. They addressed their beloved as ‘Duck Duck’ or ‘Darling sweetheart’ and made proclamations like ‘You are my avid fellow feeling’ (eelawrence). Thus, the entity taking ownership of the output, the apparent speaker, was the ‘machine’. The expressions of affection – absurd, banal, extraordinary and riven with the complications of tragedy and circumstances – were directed to a beloved but from MUC. It is hard to imagine any individual reading the typed output and taking it personally. But the I/you structure of the text is a component of its workings. The game is very clearly framed within the mode of direct address. Presumably, a computer cannot love, cannot address a beloved, and yet it does so with ridiculous abandon. Moreover, the ‘interface’ can hardly be grasped because it is distributed across a bunch of operations, rather than located in any single point of interaction. Created long before graphical user interface, the Mark I was hardly an immediate gratification machine. Punch cards, paper tapes, long time delays for processing and other features were part of its state-of-the-art computation, but these were slow and labor intensive. Each step of the process was carried out separately, so the ‘interface’ involved all of those operations, rather than taking place on a single screen.

Neither Turing nor Strachey made any claims for the literary character of this exercise. But as access to computers became more widespread, the purposes to which they were put included aesthetic production and experiment. User engagement became more and more direct, and in inverse proportion to the amount of mediation required, the systems have come to seem more and more unmediated and immediate as we approach the present and the emerging technology of the natural user interface (Figure 5). At a TED conference in 2006, Jefferson Han demonstrated a natural user interface with its multi-touch, highly haptic properties. He stood pinching and shaping a glutinous mass with his fingers as if the movements of his hands were being transferred directly through the screen, as if there were no screen, just a space of engagement. The idea of a seamless interface that draws on the formats of analogue objects and familiar habits of physical engagement has been part of experiments in the design of reading environments almost since the beginning of digital design and is the stuff of science fiction films and their art directors’ fantasies. Current ‘bump’ table simulations promise just such seamless replication of analog experience in connection to the processing power of networked computing.

We are a long way from the initial ‘blind’ computers, those giant mainframes without graphic user interfaces (GUIs), and the smaller punch card-driven units that lacked screens. When command line terminals started to be used by the designers of Zork/Dungeon, an interactive literary game, they addicted a coterie of eager young adapters (see Classic Reload; Woyke). Screen ideology was fully in place in these environments that often merely showed the glow of green, amber or bluish-white text against the murky dark ground of the cathode ray tube. The screen environment of Zork practises direct address. It speaks to the ‘you’ of the user without any indication of the enunciative frame, and its screen would only provide access to the operating system and file structures through line commands, not through any icons or file structures or other features (Figure 6). In contrast to the Love Letter Generator, Zork, first written in 1977–1979, works in real time. This allows the ‘you’ spoken in the texts to imagine themselves in an active dialogue with some ‘other’ creating the text. The interface is minimal, identified mainly by a simple banding frame that says to the user, you may proceed through this game if you conform to the use of these instructions. The disciplining effects of discourse, of what can and cannot be said within the choices offered for user experience, could not be more clear.

Figure 5: Natural user interface, (

Figure 6: Zork interface,
Figure 6: Zork interface,

Direct text production has one set of enunciative features, but the interaction of a user with a set of options on a screen, menu-driven production of a work, has others. Judy Malloy’s Bad Information database was conceived in the context of the WELL, an early online community making use of internet connectivity in the mid 1980s. It offers the user a set of options in a menu in a declarative mode – here I am, you chose – as if the options carried no obligation, commitment or consequences. While the menu obviously addresses the user, the I/you relationship is never identified, merely assumed (Figures 7 and 8).

Malloy’s work was conspicuous in its historical context. But many of the assumptions that it pointed out have carried over into standard interface design, particularly in its two major defining metaphors – the desktop and windows. Earlier, cockpit simulations had structured the user position into the very sightline and seat assumed to be occupied by a user, thus explicitly mapping the subject’s relation to the bank of controls. Early windows and desktops simply confronted the user with either subdivided screen real estate (windows, one for each function, rendering them distinct and evident) or icons (objects meant to call forth associations by analogy with tasks in the world). GUI conventions have refined over time, but the basic flat facing screen continues to address the user through a set of carefully enunciated options for engagement. As clearly as any conduct book or set of game rules, the rule bars, navigation features and other components of user interface have encoded either task-based options or a view into the structure of information or functions in the system. Jesse James Garrett’s original and much reproduced diagram of these two competing agendas for interface remains structurally sound, even as more interaction is governed by icons with object properties and as the functionalities and knowledge systems on which they draw have complexified exponentially (Figure 9). In a networked environment, the ‘information’ to which the screen provides access cannot, by definition, ever be exposed for its structural organisation, because no unifying environment exists.

Figure 7: Judy Malloy, Bad Information, landing page,
Figure 7: Judy Malloy, Bad Information, landing page,

Figure 8: Judy Malloy, Bad Information, user interface,
Figure 8: Judy Malloy, Bad Information, user interface,

Figure 9: Jesse James Garrett, (

Again, this is familiar ground, though the concept of enunciation may be novel in its analysis of the structuring relation of speaking and spoken subjects. But where does the literary come in? Is a literary interface one that provides, as the Library of Congress or any other online catalogue does, a portal into documents of the cultural record or, as in the case of the archive of Karl Kraus’s Die Fackel, explicitly literary materials? These projects announce their interface and optimise its custom-designed functionality. They anticipate the needs of their users and structure the interactive experience accordingly. This is ballroom dancing, not free-form motion, but in both cases the speaking subject of the interface is an entire team of researchers coordinating their efforts so that any and every ‘you’ can find their way into the structured experience as effortlessly as possible. No attachment to difficulty is necessarily the rule, though some deliberate frustration of user expectations is sometimes built into the design. Jerome McGann did not want users of the Rossetti Archive to be able to simply skip to a gallery of the artists’ images but wanted them to have to work through the structure of the interface with its argument about the interconnection of poetry and visual art (McGann). The didacticism of the speaking subject (that is, McGann’s avatar in his role as director/editor of the project), instructing the user through a particular mode of encounter, may or may not result in making the argument of the work apparent, but it does pull the speaking subject of the project into view. By contrast, most archive or repository projects push the contents of their archives to the fore as if they were simply able to be accessed directly. However sophisticated or well-designed, their purpose is to make the user’s access to the materials as easy as possible. The Die Fackel archive, whose interface was expertly designed by Anne Burdick, provides a highly refined approach to the task of making the materials available, guiding the user experience through a well-articulated structure (Figure 10). The thirty-seven-year run of the journal edited by Karl Kraus (with much material in its issues provided by him) is presented with a landing page that shows the faded spines and discolored cover papers of the entire run of the edition. Though not interactive, the image serves as a striking mnemonic for entry into the archive. The elaborately structured data, decades of work by linguists analysing the language of the journal in a detailed mark-up scheme, can be searched and filtered, while the transcribed text and the facsimile pages are matched in the display. The guiding framework for navigation and use is subdued in color, and the scheme of the interface is artfully designed so that it is inconspicuously present as a frame that recedes when content is displayed. The mediating functions of the interface are effectively effaced. This is a classic illustration of the principles of Jesse James Garrett’s renowned diagram of interface design, and its emphasis on the balance between the need to expose the contents of a site and support the user’s engagement with it through a set of well-identified tasks.

But the general discussion of interface leaves aside the question that is central here: is there a such a thing as a literary interface? If so, what distinguishes it from that of any other site or project? Format features and design elements are not sufficient to distinguish the commercial product of the virtual book being promoted on Alibaba’s highly commercial site from the work of, for instance, Amaranth Borsuk and Brad Bouse in Between Page and Screen (2012). In each case, a trigger creates a projection. In the case of the commercial project, the viewer is situated to receive content. In the case of Borsuk’s project, the viewer has to situate themselves very carefully in relation to a page and a screen to produce the text. The condition of viewing is foregrounded dramatically in Borsuk’s interface, and the requirements for reading include multiple levels of commitment to the project and its mediation, which are also the contents of the text. The work is structurally distinct from that of the display-oriented product offered as a novelty device on Alibaba. Borsuk and Bouse carefuly construct their reader’s subject position, integrating it linguistically and physically into the site of reading through the epistolary exchange between the characters, Page and Screen, and the reader and the source code produce the dimensional effect that hangs in the illusory space between the open printed book and the laptop’s screen (Figure 11).

Figure 10: Die Fackel, Note the spines of the journal in the top edge of the image,
Figure 10: Die Fackel, Note the spines of the journal in the top edge of the image,

Figure 11: Amaranth Borsuk and Brad Bouse, Between Page and Screen,
Figure 11: Amaranth Borsuk and Brad Bouse, Between Page and Screen,

This raises the question of how aesthetic codes enter into the distinction of literary work from commercial products. Let me return to another relatively early work, Hegirascope (1997), by Stuart Moulthrop. Moulthrop’s interface is minimal, though the bright, saturated background colors of the ground under the text clearly distinguishes it from a print artefact (Figure 12). The garishness of contrast in the type and ground has some of the same stridency as a site identified as one of the worst sites on the Internet in design terms – the tourist site of Mercia in the southern part of Britain. The design style of Hegirascope emphasises a minimal approach, with words in short lines shifting on the screen, announcing their constant change in both their statements and their actions. Beyond the word ‘begin’ on the landing page, the site invites the reader in mostly by ‘saying’ almost nothing. The convention of the blue highlight for links invites the reader to click and follow in a path whose larger contours and structures are not available. We do not so much navigate as simply jump, without any orientation as to where in the whole we might be. The Mercia site is flashy, flamboyant, unapologetically garish, and filled with competing graphic agendas and a wide array of modes of address (Figure 13). The tourist board that orchestrates the site addresses its users directly, suggesting that ‘you’ add ‘your’ information to the site. The presence of pop culture imagery, buxom women scantily clad in flag motifs, maps and advertisements for the pleasures of Mercia and Wessex is at the opposite of one extreme from Moulthrop’s site. And yet, both fulfil the demand to meet expectations for their brand identity. The literary site has no advertisements, is conspicuously minimal, presents only the language of its text, and supplies a bare minimum of instructions to its users. These graphical codes of identity are crucial, and literary work boasts its under-capitalised and modest production values as part of its distinguishing features.

Minimal aesthetics, downplayed and elegantly controlled, are the sign of high-end products, designed for discriminating consumers with elegant taste. By contrast to the Mercia tourist site, for instance, consider a site promoting rarified commodities – Tinker watches or other high-end brands. These have much in common with the work of, for instance, Kenny Goldsmith, whose 2002 work Soliloquy has even less visual apparatus around it than Moulthrop’s. In the world of high fine art, the least amount of design, most minimal aesthetic, is equated with the highest level of sophistication. Goldsmith’s minimalism, like that of Rupi Kaur, the most popular Instagram (and print) poet in Canada in the last few years, announces its elegant aesthetic with a white screen, thin dark type and, in the case of Kaur, deftly drawn images (Figure 14). Kaur’s work is personal, affirmative, thematic, and aimed at young women with messages of empowerment. Goldsmith’s is equally personal, intimate and self-involved. The two are clearly in the realm of high art, and though they have radically different audiences, and the conceptual work of Goldsmith is distant in tone and motivation from that of Kaur, they are both self-evidently identified as aesthetic works through the design of their interface as well as the presentation of their contents. Kaur’s navigation devices conform to the Instagram platform, whereas Goldsmith’s were designed within the structure of the work, as a simple forward momentum through the linear unfolding of the relentlessly literal transcription of his recorded text of everything said by him in one week in 1997. Kaur’s editorial control and skill result in more conventional work, but the presentation of each unit of her writing as a single screen has its own in-built constraints. We know how to work the conventions in Kaur’s instagrams. They are all familiar now. But Goldsmith’s custom interface retains certain conspicuous hallmarks of its design moment – and its rarified exclusivity.

Figure 12: Stuart Moulthrop, Hegirascope, (
Figure 12: Stuart Moulthrop, Hegirascope,

Figure 13: Mercia Site, (
Figure 13: Mercia Site,

Figure 14: Rupi Kaur, (

The collection of works of electronic literature provides a large inventory of techniques for access and use. A look at either the ELO (Electronic Literature Organization) or ELMCIP (Electronic Literature as a Model of Creativity and Innovation in Practice) archives, gives a good range of innovative works and their approaches. Rather than organise these according to a historical timeline which, in any case, they do not follow in a developmental way, we can instead list some of the ways that access to contents is organised. One trope of engagement allows the viewer to pick from a front-page menu and work through its contents. So, a grid of thumbnails, a list of titles and a set of icons appears. These say to the user, here, these are your options. The user is presented with a segment of work, and then returned to the home page, or offered a chance to move further through the list. The user is completely at the mercy of the interface, which does not show any of the workings of the site, just its surface elements. Another common approach is to present the user with a text that is animated, either by being rolled over, clicked or merely through a timed release that scrambles the text on screen and rewrites it. More interesting, the work of Shelley Jackson, Patchwork Girl, first published in 1995, combines the metaphors of pastiched body and texts (Figure 15). Evidencing a serious self-conscious critical engagement, the work reflects on and enacts its combination of references. A text is always part of a pre-existing literary corpus, and the cobbled together character of this work forces that recognition while also insisting on the thematics of a feminist aesthetic with regard to authorship and identity.

Figure 15: Shelly Jackson, Patchwork Girl,

Figure 16: Samantha Gorman and Danny Cannizzaro, Pry,
Figure 16: Samantha Gorman and Danny Cannizzaro, Pry,

The conventions of film viewing also play their part in literary interfaces. The recent release of Pry, a work by Samantha Gorman and Danny Cannizzaro, is assembled and experienced much like a film in segments (Figure 16). The viewer watches the scenes unfold on screen, and only intervenes to choose a path or make a decision. Our voyeuristic relation to the scene of action replicates the scenic strategies of film narration. We are the viewing subject. The film is narrated for and through a point of view with which we are meant to identify, and we are implicitly addressed through the navigation apparatus. More explicitly, Andy Campbell’s ‘The Flat’ (2012) insinuates its viewer into a creepy predatory mode of voyeurism. The screen addresses the viewer through a combination of glimpsed scenes, suggestive texts and questions about navigation, as if the viewer/user were able to intervene through response. As all options are programmed into the work, the ‘inter-passivity’, to use Simon Penny’s apt phrase, merely offers an illusion of agency to the user, not a genuine capacity to intervene in any substantive way (47–73).

Much more could be said about the ways the interfaces to literature offer user experience within emerging and established interface conventions, and thus, provide a range of ways in which the subject of the interface is spoken by the structuring principles of its options. In addition, the capacity of the user to ‘speak’ and occupy a subject position is equally constrained, either through menu options and/or navigational features. Nothing in any of these options suggests that literary work has an exclusive set of conventions through which it works, or that its identity as literature (that is, as a category of textual works that exist in distinction from those of commercial, news, informational or other sites) is bound to its interface design. Aesthetic properties that distinguish art from commerce are equally difficult to pin down, and thus the category of the literary interface devolves into that which, by default, becomes part of the workings of a literary textual experience in a digital, networked or electronic format.

A final note, however, is in order, that also invokes a third pun in the first word of my title, which is ‘cite’: the mesh of citational, referential connections within which any and all cultural work is situated. Michael Joyce’s afternoon, first published in 1987, is a fully canonical work within the realm of hypertext literature. Its appearance as an early publication of Eastgate Systems, and use of its hypertext platform, Storyspace, remain milestones. The critical response to the work resulted in an outpouring of hyperbolic rhetorical claims for the distinctiveness of hypertext, for its complete break from print formats, and for its innovative fluidity, combinatorics and structurally distinct organisation. At the time, the work seemed to signal a radical break with all previous literary modes, simply by its design. The modesty of the interface is interesting to revisit. The menu bar at the top of the screen merely announces the identity/title of the work, while a navigation bar within the work offers the options of yes/no, a forward or back arrow and very little else. The anxiety produced in the reader – have I seen it all, where should I go, how to do I know what to do – was disproportionately large in relation to the risks, but the unsettling feeling produced a curious position of the subject as hapless user of a hidden system. At a distance of several decades, all of this seems droll, and even trivial. Interestingly, ten years after the initial publication, attempts to access the work resulted in the display of a screen full of technical instructions. To view, the user had to load various files, be aware of system requirements, conditions of processing and operation systems, graphics cards and other features of one’s computer. The user, once merely the subject of the work, becomes the subject of the system and its functional specifications. Here site becomes cite in the strongest sense: the work reveals its situatedness within the much larger framework of industrial and cultural technological systems. The user is caught in this mesh of citational and referential practices, necessarily at the mercy of its specifications.

Figure 17: Rui Torres, ‘PoemAds’,

This aspect of the literary interface is a potent reminder of the situatedness of literary work within cultural systems of all kinds. Many of the works on ELO or ELMCIP reward the eager viewer with an error message. A plug-in is required, an update, a download, a specific emulation or environment must be created. The result of all of this is that the work does not run, but instead, shows its historical caught-ness in the rapid cycle of upgrade and obsolescence. This positions the literary work squarely within the commercial zones of digital infrastructure. Rui Torres’s hack work ‘PoemAds’ (2011), which relies on techniques of mash-up, taking texts from advertising sites and reworking them into poetic statements, is a striking example of this obsolescence (Figure 17). Perfectly situated within the very terms of its own production, the work ceased to operate, and only displays the call for updates to various commercial products within the grid of its display. The ‘PoemAds’ are now reduced to an advertisement for the plug-ins and products required for them to run. The situatedness of literary work within the larger context of cultural production could not be better demonstrated. The history of avant-garde protest, of the desire to promote the otherness of poetics as a stance of distinction, radical difference, from the larger world of commerce, entertainment, culture industries or other realms of cultural work, is here exposed as a myth of the highest order. The world of poetics and literature exists within that of commerce, industry and its constraints: the literary interfaces with the technoculture of media systems. Claims to a distinct and discrete identify of and for ‘the literary’ exist exactly and only in so far as its workings conform to the larger circumstances of that culture. For all of its centuries’ long attempt to define itself as other than the mainstream of cultural activity, the place of art and fine art are clearly and squarely located within the very same realms as the systems it often pretends to protest with critical distance. The ultimate device laid bare in the workings of literary interface is that of its complicity with the larger systems of cultural production.

In conclusion, we can see that literary work engages with medial ideology in complex ways, both by deliberately engaging in the same kinds of disappearance of its workings as other modes of cultural production while also, apparently and deliberately, playing with attention to its deconstruction within the design and creation of individual works. This latter move allows literary work to assert specific claims about its existence as a category apart from other content produced in a digital environment even though much of this activity is thematic and part of design features within functioning platforms that depend, at the basic level, on the full functionality of larger cultural systems. The main distinction between literary and non-literary work is often the non-commercial character of aesthetic production, a gesture of disregard for commercial success or its terms as part of the conception of the work. This is a naïve approach, however, since it does not account for the ideology of literary production itself – the aspiration to recognition, fame, long-term affirmation of symbolic cultural value and all that attends to it. Talented as Burroughs and Joyce, famous as Springsteen – that is the (unacknowledged) aspirational formulation.

We can put aside the cynicism of this attitude, and attend to the formal features of system disruption, to the enunciation of the subject position that attempts to acknowledge and deconstruct the complicities and complexities of ideological functioning through a combination of formal strategies, thematic concerns and procedural techniques. The question that comes into focus is what the goal of such work might be, and how it is embodied in the performative dimensions of an interface. The drive of literary work is to do something: to embody a cultural position in a fully self-conscious way, inhabit a point of view about something – experience, insight, affirmation, epiphany, tragedy, violence, current events. But the politics of the interface commit us to acknowledging the ideological workings of the system of mediation in which any work operates and exists. Interface is the site of this mediating activity, of what is brought forward into the scene of literary production. If interface is at once a process of mediation and a set of protocols, features, formats for structuring those processes, then the critical task is not to ‘decode’ interface as if some essence or essential meaning could be attached to its components through enumeration of the values on which it is premised, but instead, to create an active engagement with reading the ‘encoding’ of its production. In that production the work of medial ideology occurs. And the literary? As a category of work, it is no more or less immune to or apart from reliance on conditions of production. Literary works in electronic form exist in a condition of full co-dependence on working systems. They depend on the full ‘stack’ of the networked and digital environment for storage, processing, display, transmission and every other function. The protocols of literary work, in the fundamental technical sense, are indistinguishable as media from those of commerce, communication, entertainment and other networked aspects of contemporary life. Media are, as stated earlier, agnostic as to content or genre. The literary realm may be attentive to that agnosticism, and thus to the precarity and viability of the distinction on which literature retains its identity as a category apart. Its own ideology tends to hide that fundamental condition of dependence by focusing on the scenes and staging of its work. In the digital environment, the scaffolding of institutional sites is reworked, and is harder to categorise than in the world of bricks and mortar and real estate, but what is left or replaces this are brands (ELMCIP, ELO, online repositories of high pedigree or, by the same token, without these identifiers). The ongoing claims for a work as literary are still somewhat linked to the legacy of the avant-garde, to difficulty of consumption, disruption, self-conscious attention to the making and working of poetics as a politics and aesthetics of resistance. But other literary forms – popular, mainstream, fan-based – also exist and use similar networks and protocols in spite of differences in demeanour, style and content.

We are perhaps not very far from the debates and frameworks that have been part of our conversations for more than a century, and that still draw on the dialogue of aesthetics and politics formulated within the avant-garde. But that too is an ideology with mythic dimensions that may not hold up in the face of larger dependencies and blindspots about the place of contemporary literature in cultural systems of production and reception. It is not only the reading of stuff and content, but the active negotiation with its production and claims, which remains the task of reading interface and understanding the illusions it produces within its enunciative terms – as well as its promises and potential for substantive revelation. If we are looking to literary work to produce a subject position of radical revolutionary liberation, we may be subscribing to outmoded paradigms. But if we look away from this work, the myth itself disappears from view, becomes a sight/sight fully unseen and this, also, has its liabilities. The peril is in confusing terms of distinction (literature as a distinct set of activities) with those of autonomy (seeing experimental literature as radical alterity). Aesthetic works exist within the same systems of mediation, ideology and enunciation as those of other industries, and in the same contingent conditions of co-dependence. Medial ideology is designed to make us forget this, by hiding it from view.


  1. The Electronic Literature Organization site and the ELMCIP site both contain substantial instances of early interface design (many remediated so they can display effectively in current browsers).

  2. A glance at a list of early e-literature titles includes many like Judy Malloy’s its name was Penelope (1989–1992); Richard Smyth, Genetis: A Rhizography (1996); David Kolb, Socrates in the Labyrinth (1994) and so on, with their references to webs, weaving, networks and forking paths à la Borges. See Dene Grigar and Stuart Moulthrop’s Pathfinders;

  3. See for a discussion of twenty-five years of the project.

Published 19 December 2019 in Volume 34 No. 2. Subjects: Digital humanities, e-literature, Digital literature/s, Literary interface/s.

Cite as: Drucker, Johanna. ‘Site Unscene: Medial Ideology and the Literary Interface.’ Australian Literary Studies, vol. 34, no. 2, 2019, doi: 10.20314/als.1d07532095.