Barthes conception of the (ideal) writerly text, that functions as a 'galaxy of signifiers ... reaching as far as the eye can see' on the surface seems to contain some interesting analogies with actual computer-mediated hypertext. Barthes divides various texts in the rather generic and classic 'readerly text' that provides for only one interpretation that leaves the reader a passive consumer of the text, and the far more open and multi-interpretable 'writerly text'. According to Barthes, this writerly text should provide for 'multiple entrances', therefore rendering the 'reader' as a more active producer of the text. Barthes provides us with a powerful critique on the implicit power relations within and between the various lexias of a text and the related hierarchy in the standard academic text. S/Z appears to be proto-hypertextual not only in its content, but only partly in its format: by placing Balzac's story at the end of the book rather than at the beginning, and by dividing S/Z into a multiplicity of small chapters where the various de- and connotations appear as proto-hypertextual remarks inserted into his text (rather than classic academic annotations), Barthes shows how 'hypertextualising' subverts several power relations at work in a text.
However, immediately a remark can be made here: Barthes' creation of the opposition between the lowly-valued readerly and the highly-valued writerly seems to be rather hierarchical in itself, as well as the very formalist categorisation of the codes in a text into a system of strictly five ones, reminds very much of traditional academic writing. One could argue though that Barthes' inserted these concepts to perform an ironical twist in S/Z: the over-stating of the writerly and the five codes could actually be read as a persiflage upon traditional academic writing, showing its workings in full glory...
In the appropriation of Barthes' concepts to computer-mediated hypertext as especially taken up by George Landow in Hypertext 2.0, a couple of critical remarks could be made. First, it appears that the concept proposed by Barthes of the lexias seems to mean something rather different from the lexias suggested in hypertext. In S/Z, lexias are small portions of text that convey, or de- and connotate towards (multiple) meanings that, when selected and woven together, potentially crystalize into certain discourses and ideologies. The lexia in hypertext rather flatly seems to be any paragraph, word or sentence (that can be connected by a software-driven link). The analogy appears to be definitely there on the formal level, but not necessarily on the level of content. Barthes actually acknowledges that he chooses the lexias for the purposefulness of his argument so as to deconstruct and show certain workings of Balzac's text (and this choosing is of course closely related to his situatedness as an academic in France some many decades ago); but computer-mediated lexias in hypertext usually rarely seem to aim at deconstruction on the level of the content of the text that the lexias are part of. The problem here is that any interconnected bunch of hypertext documents is not necessarily always inherently more 'writerly' than any printed text. The format may seems less hierarchical, and I acknowledge that form definitely has its implications on the content, but the content within lexias itself, the choice by whom of which lexias to link to what and why, and the overall discourse(s) of the group of texts themselves may still be very mono-ideologically inclined.

home | introduction | paradigms | terms | hypertext | remediation | critiques | memex | bibliography