I then came upon theories of intertextuality which provided the basis for Methodology Number Two. Starting with a recent book by Graham Allen (Allen, 2000), I found that a post-structuralist version of intertextuality had all the makings of a good methodology through which I could speak. Intertextuality examines relationships between texts. An intertextual reading seeks to explore how meaning is produced through a text's relationship to other texts. Meaning can never be stable since a text's meaning depends on the context in which it is read. I was enthusiastic about intertextuality for a number of reasons. First and foremost theories of intertextuality look at the construction of meaning, although primarily from a textual perspective. To interpret a text and to discover its meaning, required looking at its relation to other texts. Barthes encourages us to remember the origins of the word ‘text’, “a tissue, a woven fabric” (Barthes, 1977: 159). The woven fabric constructs a metaphor for relations between texts themselves and represents intertextual approaches.

The foundations and development of intertextuality can be traced through some of the most important thinkers and movements of the 20^th century. Early in the century, Saussure’s work on the relational nature of meaning in linguistics points the way for subsequent thinkers to consider the relational nature of meaning in many contexts (Saussure, 1983). In his first principle, Saussure argues that the linguistic sign is arbitrary. Saussure describes the linguistic sign as a two-sided entity, signifier and signified, whose two elements are intimately linked, each element triggering the other. It is the link between the two that is arbitrary.

Saussure cites the example of the sign ‘sister’ and the French word (signifier) soeur stating that the same idea might just as well be represented by any other word for example, geragog. The relationship between signifier and signified is not, however, random since at any time language belongs to those whose use it. Language is a social institution that inherits past usage that anchors it in a community and in a time.

As a strategy for investigating the way in which relationships, on which linguistics depends, work, Saussure suggests two types: syntagmatic and associative (paradigmatic). The syntagmatic relation is the way in which words, phrases, sentences and other complex units are strung together in a language. Thus syntagmatic relationships are based on formal contiguity. “A word depends on its value, and hence its significance, both on its immediate neighbors and on its location within the linear organization of the sentence in which it appears” (Silverman, 1983: 105).

Silverman notes that syntagmatic analysis can be used in other areas, citing cinematic sequences in film and spatial disposition of objects in painting. Metz uses syntagmatic analysis to examine the image track of films (Metz, 1974). Fiske uses syntagmatic relationships to help understand how meaning is generated in communication studies (Fiske, 1990). Tolson describes the structure of media texts by examining their syntagmatic dimension (Tolson, 1996).

Syntagmatic analysis provides examination techniques useful at both the level of overall interactive structure and the level of mouse/cursor interaction. A syntagmatic analysis of the interactive structure of a work would involve examining the paths or choices one makes in a new media work, where those paths lead, the number of paths/choices available at any given moment and the forms that following a number of paths create. The results of such an analysis would lead to a relatively objective assessment of the complexity of interaction within the bounds of the analysis.

The mental association of one word or unit in language to others is called the associative relation. Saussure describes two such series of the many associations possible: the association based on similarity of signification where terms share a common element, or the association based on similarity of sound patterns. In the context of new media art, a paradigmatic analysis would help uncover the significance of interface design choices. For example take the choice of a cursor icon. Many new media art works change the cursor in the context of interactive events. What are those cursor changes? Why was that particular choice made rather than an alternative choice? Fiske sums up the paradigmatic choice by stating that the meaning of what was chosen is determined by the meaning of what was not. (Fiske, 1990: 58).

Saussure provides part of the foundation for intertextual theories. Allen writes: “the linguistic sign is, after Saussure, a non-unitary, non-stable, relational unit, the understanding of which leads us out into the vast network of relations, of similarity and difference…”. Saussure’s contribution alone does not provide the foundation for intertextual approaches. What is needed is a human-centered, socially specific contribution, which is provided by Bahktin.

Bahktin, with Medvedev and Volosinov, use the term utterance to understand the social significance of communication. An utterance is monologic if it produces one meaning. An utterance is dialogic if it produces many meanings. Working against the monologic utterance, seen as a purely individual act, Bahktin promotes the dialogic utterance constructed between two socially organized persons. Of the dialogic utterance Bahktin/Volosinov state:

In examining the novel’s diagoloic character, Bahktin suggests the notion of double-voiced discourse that “stem[s] from a recognition that language is never our own, that there is no single human subject which could possibly be the object of psychological investigation, that no interpretation is ever complete because every word is a response to previous words and elicits further responses.”

The first to articulate a theory of intertextuality explicitly was Julia Kristeva in her essay “The Bounded Text”. “The text is therefore a productivity, and this means: first, that its relationship to the language in which it is situated is redistributive (destructive-constructive), and hence can be better approached through logical categories rather than linguistic ones; and second, that it is a permutation of texts, an intertextuality: in the space of a given text, several utterances, taken from other texts, intersect and neutralize one another” (Kristeva, 1980: 36). In this articulation we find notions of a text produced not from a single author but as part of a culture. This notion results in Barthes pronouncement of “the death of the Author” (Barthes, 1988).

Initially I found Barthes’ S/Z a promising model for the examination of an interactive new media work as it gave an extremely interesting and thorough examination of a narrative work. In the end I found the approach too specific and confining. While the post-structuralist versions of intertextuality were very interesting, they did not fully suit the ideas I had knew I had percolating in the back of my mind.