Intertextuality, as defined by Michael Riffaterre, "depends on [a system of] limitations in our freedom of choice, of exclusions, since it is by renouncing incompatible associations within the text that we come to identify in the intertext their compatible counterparts." He further states that this intertextuality is the complete opposite of hypertextuality because the former builds a "structured network" of limits that will keep the reader on track (towards the "correct" interpretation), the latter is a "loose web of free association." 6
This comparison forces me to question Riffaterre's understanding of hypertext. The quote comes from a 1994 article, when hypertext was somewhat different from today's (1997) version, but certainly not an amorphous, unstructured mass of material arbitrarily selected. Two distinct types of information linking in hypertext refute Riffaterre's argument. First, embedded links are placed in a text by the author. They are very rarely random. A second form, "searches", are dependent on the programming of the search engine (program). Currently, different search engines give different "hits" to the same inquiry, which indicates that someone has decided how the search will be limited because computers can not make such decisions without instructions.
Riffaterre ultimately sees the intertext from the Aristotilean perspective of certifiable truth. He even goes so far as to imagine that the "Institutions of Interpretation" have not changed since Aristotle. Perhaps some in academia can maintain that illusion, but those who have grown up as "other" would argue the point. At any rate, he embraces an artificial standard when he states,
[I]ntertextuality is made manifest either by syllepsis or by a gap, or by an ungrammaticality. . . Each of these is immediately perceptible to readers, who need no more, to respond to the text, than the senses nature gave them. 7
He then goes on to discuss the intertext of Dickens' Bleak House, a French sonnet by "the mad poet Gerard de Nerval, who is famous for his arcane writing", and Goethe's poem Mignon. I believe that reading, let alone responding to the syllepses, gaps and ungrammaticalities which Riffaterre deconstructs, required much more that the senses nature gave me. The connections are not obvious and only a person with time, education, inclination and interest in the subject could or would make an attempt to find the conclusions Riffaterre suggests as valid.
The idea that only specific, validated conclusions are correct does conflict with hypertextuality, but intertextuality does not per se. Riffaterre seems to think that computers will read and interpret text. This may be true, but a computer could only do this based on instructions from a human which parallels the professor/student relationship where students are instructed to sort through vast amounts of data and come up with an edification of some idea while making sure to chronicle all important links to sources.
Landow views hypertext intertextuality as a liberation from the "structured network" of Riffaterre. While the opportunity to to read "in terms of author and tradition" is available, the reader can also bring into play her own group of intertextual resources. Paralogic hermeneutics would insist that intertextuality must be negotiated between two parties during the act of communication which would also argue Riffaterre's view of intertextuality.