John Cayley

© 1997
PMC 7.3

 

Book Unbound*

 

Abstract: “Book Unbound” is a “collocational cybertext,” a self-assembling poetic collage that can be read in two ways: either automatically in the “bound” mode, or in an “unbound” mode that allows readers to extract and recycle words from its recombinant text stream. The present version is a HyperCard stack (Mac only, HyperCard program not required) available for downloading. –Editor

 

---

 
Editor’s Note: “Book Unbound” is a HyperCard stack. The present version runs only on Apple Macintosh computers. Activate one of the links below to download a compressed, binary version of the stack (a self-extracting StuffIt archive). If you are using a correctly-configured graphical browser, the file should be converted and decompressed file automatically. If it does not, save the file, convert it with BinHex then double-click to launch the self-expanding archive. If you own a copy of HyperCard, download the stack only (120k). If you do not own HyperCard, download the stand-alone application (676k).
 

---

 

Author’s Remarks

 

“Book Unbound” is a literary object which, I believe, exemplifies certain potentialities of cybertext. ¹

 

The first version of “Book Unbound” was produced in 1995 using HyperCard for the Apple Macintosh, although I had made earlier collocational pieces in 1993. From one point of view–especially when read in its “bound” mode–“Book Unbound” is simply a text generator. As such, it employs a fairly simple algorithm, which I call “collocational” but which might also be recognized as a simple stochastic transformation. The transformation can proceed beginning with any word of the given (or source) text. Any other word–occurring at any point in the given text–which follows (collocates with) the word last chosen may then follow it and so become in turn the current word. Clearly, in this type of transformation, at the very least, each pair of successive words are two-word segments of natural English. However, the text will wander within itself, potentially branching at any point when a word that is repeated in the given text is picked, and this will most often occur when common, grammatical words are encountered.

 

The given (or source) text of “Book Unbound” is a brief, closely-written piece concerned with the book, so-called new media, and varieties of textuality which they underwrite. In “Book Unbound” I have chosen to keep the source text hidden. Anyone who is interested to see it can either hack into the Hypercard stack or track down the catalogue of the exhibition for which it was written, “Mapping Knowledge” (curated and edited by Les Bicknell), an exhibition of book art held at The Minories Gallery, Colchester, UK, November 1994. Even when read in its “bound” mode, and starting from its initial state, I am more or less satisfied with the performance of the generative text. This performance is the “due form” of what “I” have (pre-)written or programmed, by composing the source text and applying these quasi-aleatory collocational procedures.

 

However, “Book Unbound” has other, more interesting, potential which could be taken much further in the future work of this field. ² Once read in its “unbound” mode and altered by a reader it becomes, in part, the work of that reader (or, if you prefer, that reading).

 

When you open “Book Unbound” and read it in this “unbound” mode, you change it. New collocations of words and phases are generated from the given text according to the algorithm described briefly above. After each screen fills, the reader is invited to select a phrase from the generated text by clicking successively on the first and the last words from a continuous string of text. These selections are collected on the page of the book named “leaf,” where they are accessible to copying or editing. But they also become a part of the store of potential collocations from which the book goes on to generate new text. The selections feed back into the process and change it irreversibly. If the reader continues to read and select over many sessions, his or her preferred collocations may eventually come to dominate the process. The work may then reach a state of chaotic stability, strangely attracted to one particular modulated reading of its original seed text. Each reader’s copy of the work thus becomes unique: non-trivially different from every other copy.

 

“Book Unbound” is (inevitably) an experiment with limitations. ³ For example, it is not possible to select discontinous text from one screen or from previous screens of generated language. Neither does the present work allow you to add your own new words to the total vocabulary of the piece (except to the freely editable “leaf” page). You can only select new or preferred collocations from those which are generated within its textual microverse. However, even within this defined and bounded space some interesting experiments can be performed, selecting, for example, only noun phrases, or only linguistic function words. Readers might make several copies of “Book Unbound” and train them in different ways, watching for divergences in the nature of the texts they generate, as for example, one version becomes a gallery for a small group of objects and another becomes a Steinian obsessive, possessed, let’s say, by pronouns.

 

John Cayley
May, 1997

 

---

 

Addendum

 

Remarks on “Book Unbound” from Espen Aarseth’s Cybertext: Perspectives on Ergodic Literature (PhD dissertation, University of Bergen, 1995; forthcoming from Johns Hopkins University Press):

 

John Cayley's "Book Unbound" is a literary work not easily classified by traditional aesthetics... it takes over the screen and spits forth short suggestive sentences one word at a time ...

 

The program is assembling these lines from its "hidden texts" according to certain algorithms. As the process goes on, the hidden text is changed by what is displayed, and the user can select passages for inclusion in the regenerative process. Thus the text output is influenced, and will be different for each copy of the text. Is it still the same text? Cayley calls the produced output "hologograms," fragments that contain holographic versions of the initial material....

 

This text is an impurity, a site of struggle between medium, sign and operator. The fragments produced are clearly not authored by anyone, they are pulverized and reconnected echoes of meaning, and the meaning that can be made from them is not the meaning that once existed. "Book Unbound" is an extreme paragon of cyborg aesthetics, an illustration of the issue of communicative control. The pleasure of this text is far from accidental; it belongs not to the illusion of control, but to the suggestive reality of unique and unrepeatable signification. It would be a grave mistake to see this text as a metaphor of the "impossibility of perfect communication" or as the embodiment of the gap between sign and meaning in texts. Instead, it shows how meaning struggles to produce itself, through the cyborg activity of writing.

 

Notes

 

*. “Book Unbound” was also included in the CD-ROM issue, number 3, of the arts magazine, Engaged (London, 1995).

 

1. Here, I distinguish cybertext from hypertext according to criteria first set out by Espen Aarseth in his essay, “Nonlinearity and Literary Theory” (in George Landow, ed. Hyper/Text/Theory, Baltimore: John Hopkins University Press, 1994) and refined in Cybertext: Perspectives on Ergodic Literature (forthcoming from John Hopkins University Press), and an article, “Text, Hypertext or Cybertext: a Typology of Textual Modes Using Corresponding Analysis,” forthcoming in Research in Humanities Computing, 6, ed. Susan Hockey, Nancy Ide and Giorgio Perissinotto, Oxford: Oxford University Press). Aarseth uses cybertext as a more inclusive term embracing, for example, generative or “ergodic” textuality (the latter explicitly demands work from the reader), while reserving “hypertext” for the (chiefly passive) link-node structures which are now familiar to over 50 million of (scare quotes) “us.”

 

2. In its HyperCard version, my later work “Pressing the Code Key” employs a somewhat elaborated version of the “Book Unbound” form, although it does not address the limitations outlined below. As an essay, “Pressing the Code Key” was published in ejournal v.6, n. 1. See: http://www.hanover.edu/philos/ejournal/archive/ej-6-1.txt.

 

3. There are also software limits to the HyperCard version, which cannot cope with text fields containing more than 32 kilobytes (about 32,000 characters or 5,000 words). Eventually, these software limitations would also be reached.