Pervaded by Epistemology

Pervaded by Epistemology

Writing Machines
N. Katherine Hayles
MIT Press, 2002. 224pp. $45. $17.95 (paper)

A review of Writing Machines, building on a number of the book’s earlier reviewers in ebr and elsewhere.

This review is also a reply to earlier ebr reviews by Zervos and Koskimaa. Introductory descriptions about Writing Machines are not included. A less linear, more smooth-flowing Flash-based hypertext can be accessed here (by those possessing current Flash software).


In Writing Machines, Hayles wants to create a literary approach which is ‘alert to the importance of materiality in cultural productions’ (page 19). This leads her to develop media-specific analysis (MSA). Now that there are alternatives to print publication, we can no longer ‘ignore the material specificities of the codex book’ (among other types of publication) (page 32). Litcrit should expand to include images, for the ‘print-centric view fails to account for all the other signifying components of electronic texts, including sound, animation, motion, video, kinesthetic involvement, and software functionality, among others’ (page 20). One questions that arises is in what sense MSA can be considered a theory of literature, and whether the traditional category of literature is itself implicitly problematised by MSA.

texts discussed

Hayles seeks ‘a mode of critical interrogation alert to the ways in which the medium constructs the work and the work constructs the medium’ (page 6). A book is ‘an artifact whose physical properties and historical usages structure our interactions with it in ways obvious and subtle’ (page 22). For example, the material structure of paper (foregrounded in Tom Phillips’ A Humument), determines how the book is read but also ‘profoundly transforms the metaphoric network structuring the relation of word to world’ (page 23). This print effect can be compared to an electronic work like Talan Memmott’s Lexia to Perplexia. The third work Hayles discusses is House of Leaves by Mark Z Danielewski.

Although she argues that all literature is subject to the materiality of its production, Hayles analyzes texts that consciously foreground their materiality; thus, as Zervos observes, the texts of interest to MSA are quite narrow, unlike the range of texts that can be accommodated within cybertextual analysis. However cybertext is a structuralist approach, not conceived of for close readings of specific texts. In my opinion, the potential for MSA to broaden critical approaches to textuality beyond ‘media essential’ (Eskelinen) disciplines can only be developed if it is released from the shackles of prior disciplines and specific contents.

Lexia to Perplexia

According to Hayles, ‘when a literary work interrogates the inscription technology that produces it, it mobilizes reflexive loops between its imaginative world and the material apparatus embodying that creation as a physical presence’ (page 25). Lexia to Perplexia by Talan Memmott is a web-based programmed technotext that self-reflexively discusses its relationship with technology and distributed networks. Memmott postulates the co-originary status of subjectivity and electronic technologies (page 49) and describes a process of ‘cyborganization’ in which ‘human subjects [are transformed] into hybrid entities that cannot be thought without the digital inscription apparatus that produces them.’ (page 49) For Hayles, such technotexts ‘play a special role in transforming literary criticism into a material practice, for they make vividly clear that the issue at stake is nothing less than a full-bodied understanding of literature’ (page 26).

lexia hack

Hayles suggests that a difference between Lexia to Perplexia and the print texts she discusses (A Humument; House of Leaves) is the amount of noise that resides on their surface. However, her chosen print texts are also very noisy. Meanwhile, Hayles does not mention the experiential distinctions between reading at a computer versus reading print which Bernstein among others mentions. By downplaying differences between the material nature of electronic and print texts, Hayles can put them on the same literary continuum. She thus implicitly refutes those who warned ‘Kaye’ of the equivocal relationship between electronic writing and literature.

MSA combines structuralist insights with a more traditional close reading and political, phenomenological, and epistemological tensions result. Meanwhle, as Koskimaa points out, Hayles alludes to the ‘highly contested field where allies and enemies sometimes count more than arguments’. Hayles seems to be situating herself above this field, but although she has the grace to avoid the rancour of some others in the field, she is deeply within it. The creation of territory is in general to be mourned; as we are seeing in Iraq, territories require borders, and borders seem to demand either expansionist or defensive strategies.

three areas

Prior to her discussion of Lexia to Perplexia, Hayles describes 3 eras of electronic writing analysis.

1. 1980s-1990’s - focused on the link, but the revolutionary claims for such works now appear inflated ‘for they were only beginning to tap into the extraordinary resources offered by electronic environments’ (pages 27-29).

2. 2nd generation electronic literature used other software and interfaces than Storyspace which ‘experimented with ways to incorporate narrative with sound, motion, animation, and other software functionalities’ and works that use combinatorial strategies, a computational perspective which ‘kills the literary priest’ (pages 27-28). These are different than technotexts because they don’t pay ‘particular attention to interactions between the materiality of inscription technologies and the inscription they produce’ (page 28).

Thus she introduces:

3. Media-specific analysis (MSA) ‘insists that texts must always be embodied to exist in the world. The materiality of those EMBODIMENTS interacts dynamically with linguistic, rhetorical, and literary practices to create the effect we call literature’ (page 31). All this talk of embodiments reveals one body that is greatly lacking: the human (reader’s) body. Although Hayles acknowledges it elsewhere, it seems we have yet to thoroughly reconceptualize propriocentric relations with revised notions of textuality like MSA in the wake of postmodernism. (Espen Aarseth’s text/machine [Cybertext: perspectives on ergodic literature. Baltimore: Johns Hopkins University Press, page 21] places the operator within the machine, but I suspect that a close analysis of this diagram would throw up conundrums.)

A Humument

A Humument by Tom Phillips is an artist’s book inspired by Burroughs’ cutups and applied to an obscure Victorian novel. The original hypertextuality of the novel is compounded by Phillips’ grafting of a second-level hypertextuality. According to Hayles, by re-presenting another book, A Humument reflects on the material production of the novel and in particular on the editorial decisions that novelists make.

House of Leaves

House of Leaves by Mark Z Danielewski concerns media and textual mediation. House of Leaves ultimately becomes a ‘a metaphysical inquiry’ which ‘instantiates the crisis characteristic of post-modernism, in which representation is short-circuited by the realization that there is no reality independent of mediation’ (page 110). In ‘a frenzy of remediation’ House of Leaves consumes other media, but traces of these other media remain which result ‘in a transformed physical and narrative corpus’ (page 112).

MIT Press

Hayles responded to MIT Press’s series of small format books with extensive visual content ‘with excitement, seeing in… [Peter Lunenfeld’s] offer an opportunity to explore the interactions of words with images that had already attracted my attention in electronic media’ (page 143). Several areas needed to be conceptualised: the design, the autobiography, and the relationship to the website.


According to the designer, Anne Burdick, ‘In order to create a book that embodies its own critical concepts - a technotext - it is imperative that the design evolves in tandem with the text.’ Hayles wanted to create a book that re-presents itself, over and over. Burdick’s response was to create ‘material metaphors’ that ‘work and rework the body of the codex, amplifying the book’s status as a book.’ Burdick concludes that this approach reveals ‘much about the complex relationship between showing and telling, from the role of context in quotations to the ways in which we read’ (page 140). As Koskimaa comments the combination of quotes from the text and images (for example, page 90, from A Humument) is inspired.

design hack

Burdick wants to facilitate affordance. Underlined capitalised words are ‘amplified entry points into the text’ and graphical elements offer ‘an alternative view of the conceptual terrain of the book’ (page 140). As Thomas points out however, the hypertextual-style underlinings in the book are semantically confusing; highlighting the difference between hardcopy and electronic texts may work against Hayles’ argument for continuity between electronic and print texts. With Zervos, I wondered whether various aspects of this distributed type of materiality foregrounded differences in materiality which Hayles actually seeks to suppress, as part of her project to create a discipline that encompasses them all.


One regret of Hayles is that footnotes and bibliography are forced onto the MIT Press website (page 143). The website contains essential academic aspects of the publication. Requires the Flash 6 plugin. Also worth a look is Eric Loyer’s Flash response, commented upon by Zervos.

website hack

The website material, in particular the extensive bibliography, is so significant that I had to download and print it. Indeed MIT Press designed it to be printed and inserted in the book. This doesn’t work, because the codex book is not designed for extra material to be slotted in. The alternative is to simultaneously read in two different media, which is not enjoyable. Although Zervos is rather harsh on which is an undeniably interesting experiment in distributed textuality, reading becomes unduly complex, and the reader is confronted by material difference to such an extent that it may indeed work against Hayles’ project


Hayles struggled with the autobiographical content that was part of the brief until she hit on the invention of an autobiographical persona, Kaye, which allowed her ‘a means of exuberant engagement’ (page 142).

autobiography hack

As a reader, I found the Kaye persona difficult to form a relationship with. I was constantly trying to judge how literally I should accept Kaye as a description of Hayles. If Kaye is Hayles, then what is the point of Kaye? However, Kaye’s existence makes an eminently readable narrative.


Hayles experiments at the edges of language and criticism to situate technotexts in relation to ‘the materiality of the literary artifact’. Issues raised include the role of reader-response (phenomenological) approaches to textual materiality, and the nature of theory itself.


Materiality…emerges from interactions between physical properties and a work’s artistic strategies’ (page 33). It ‘depends on how the work mobilizes its resources as a physical artifact as well as on the user’s interactions with the work and the interpretative strategies she develops…’ (page 33). We hear about Kaye’s trouble in interpreting these texts because she had been concentrating too hard on reading words: ‘Finally it hit her: the work embedded the verbal narrative in a topographic environment in which word was interwoven with world’ (page 41). Kaye’s physical relationship to these textual topographies remains under-explored, but the introduction of spatial metaphors begs this question.


Kaye’s scientific training may predispose Hayles to treat the text as an objective artefact. MSA pivots around the materiality of the text, and implicit in MSA seems to be the existence of an ideal MSA-inspired reader/critic (that is, Hayles herself). To my more phenomenological mind, the materiality of the text is largely determined by a user’s perception of the material parameters of its production, which cannot be objectively proven.


In one interlude, Hayles ponders the different definitions of theory for the sciences (distillation) and literature (interpretive frameworks) (page 104). Literature can’t renew itself by relying on new phenomena to explore like science can, because it depends on an established canon of lit texts. But lit scholars need noise unlike sciences (page 105). As ‘Kaye’s’ intellectual development took her away from science towards literature, she faced sharp differences in approach, from her own predisposition to ‘solving problems’ to that of ‘investigating problematics’ and increasingly desired to develop the latter approach (page 15).

However, at the same time ‘she never abandoned her comitment to precise explanation, feeling that if she really understood something she should be able to explain it to others so it was clear to them’ (page 14). Thus Hayles’ ideal MSA-informed reader, seems to tie in with a pseudo-empiricist world-view: texts are like molecules, available for analysis; if you can’t analyze texts and molecules you undermine the validity of rigorous research in science and in humanities. Epistemology pervades Writing Machines; it remains to be seen whether an MSA can be developed which is more reader/operator-response centered, in which the extent to which material analysis is foregrounded can be a facet of a reader’s interpretation.


Foregrounding materiality is a timely addition to contemporary criticism, but Hayles seeks to do too much - I want ambiguity, a liminal / emergent condition that I don’t want to resolve. My uneasiness with Hayles’ predeliction to over-determine the text is reflected in Aarseth’s belief that ‘cybertexts’ simply are (enjoyably) messy (Aarseth 1997, page 3) (although Aarseth also seems intent on taking some the messiness out with his textonomy).


Hayles’ ‘cultural studies’ style arguments are incomplete in Writing Machines. A cultural studies perspective would discuss the territorialization of critical inquiry itself. For example, is Hayles predisposed to talk about materiality because her meta-project is to develop territory in which ‘electronic writing’ is on a continuum with print literature? A successful synthesis of this putative continuum would be in the interests of a Professor of English and Design / Media Arts.

Hayles includes visual media and programmed environments in MSA. Their relationship to ‘literature’ needs much further exploration, and whether this approach can be reconciled with existing disciplines is yet to be demonstrated. Aarseth (1997), to whom Hayles pays scant attention (as both Zervos and Koskimaa point out), suggests an over-arching analytic concept of ‘cybertext’: this seems at least logical. As Eskelinen says ‘Cybertext theory can justify the study of digital and electronic textualities in their own terms, instead of submitting or committing to the traditions of print literature(s)…’ Hayles’ ‘technotext’ is currently too tied to print literature to achieve this level of conceptual independence, although I hesitate in joining Zervos, who believes that one of Hayles’ core problems is her refusal to acknowledge that there is ‘some immaterial quality that is contained in, or produced by, all the actualisations [of poetry]’ beyond a poem’s materiality. This vaguely Romantic transcendentalism assists to create all the aspects of ‘literature’ that I have personally been struggling to escape: the concept of the literary work that justifies copyright (page 31) and the notion of author as indidivual genuis (page 32) being two that Hayles herself mentions.