In discussing hypertext in the classroom there are at least two issues which should be considered separately. The first, and more technical, question is what hypertext systems already exist or might be devised to support learning. Second, how does the use of hypertext by students and teachers change the traditional patterns of authority in the classroom. In the main account on offer in Part III by Dryden, these questions tend to be confused, with the result that powers are attributed to technology that rightly belong with the agents that are using it -- students and teachers. The rhetoric of hypertext advocates (Landow, Lanham, etc.) urges that the hypertext medium is inherently liberating and democratizing, but I will argue here, with Dryden's chapter as evidence, that this is a fallacy. Hypertext is no more liberating than paper or stone tablets: it depends who is using it, and for what purpose.

A serious consideration of hypertext and learning (which I don't pretend to offer here) would begin with the issue raised by Dobrin (p. 306): what are the conventions that govern hypertext, and should we undertake to teach them. Why do we teach students how to read reference books or Shakespeare plays, but not tabloid newspapers or magazines like LA Style? An inexperienced person may need some instruction in how to read a hypertext. Note that this is an empirical question: if hypertext is, as its advocates claim, closer to how text "wants to be" then presumably readers will need less time being instructed in its use. In comparison with the clunky, physical world of books, readers will recognize that text has gone native and will feel at home.

Hypertext for instructional purposes, however, seems likely to be carefully designed and constructed. The best known examples, such as Perseus or Intermedia, were both designed with extensive navigational help and pre-structured documents and pathways to assist student learning. They are far from the free, associative carnivals of intertextuality celebrated by postmodern theory. Landow, for example, is manifestly schizoid in his writing about hypertext, claiming extensive support for his view of hypertext from Barthes or Derrida in one place (e.g., in Hypertext 2.0, 1997), while sternly laying down rules for careful hypertext linking in another (Landow, 1991). But serious proposals for instructional use of hypertext largely bypass the postmodern scenario (cf. Lanham's multimedia textbook, or Siemens's proposed hypertext-based research tool). Moreover, students contributing to the kind of interpretive hypertext envisaged by Johnson-Eilola (p. 214) are finding a voice, a place in a community of interpreters: they are unlikely to feel encouraged by having their voices appropriated by a deconstructive account claiming the presence of voice to be an illusion, an artifact of the scene of writing, or by being told that the coherence of their writing is to be disassembled into its morceaux, cited as a sign of something else, the always-already.

Hypertext in the classroom is said to open the possibility of displacing the teacher from the position of "exclusive authority" that has been difficult to achieve within print culture. In hypertext, say Moulthrop and Kaplan, "there is no way to resist multiplicity" (Johnson-Eilola, p. 213). There are two distinct questions at issue here, however: first, whether displacement of authority cannot also be achieved independently from electronic media; and second, whether hypertext could not be used for authoritative instructional purposes just as well as a textbook. Obviously, the answer to both questions should be a yes. Experimenting with student centered classrooms has been underway for much longer than the use of hypertext in schools (e.g., Boud, 1981). And while I cannot point to examples of hypertext used for authoritarian purposes, this is hardly surprising: it is not what publishers of books and journal articles about hypertext are likely to encourage. A teacher faced with a set curriculum to deliver, however, and students with an examination looming, are just as likely to "resist multiplicity" in a hypertext as in a textbook.

If hypertext is to enable new classroom practices and displace "conventional literary education" (Johnson-Eilola, p. 217), this will come about not because of some inherent virtue in the medium but because teachers and students have rethought their roles. Dryden's wish to see more democratic, student-centered learning (p. 284) is unquestionably an admirable goal, but he doesn't need hypermedia to realize it (p. 292). It is the project-based methods that his students were enabled to take on that are empowering, not the technology. As Dobrin puts it, it is a mistake to pin your hopes on technology for social change (p. 310): the collaborative society Dryden facilitated in the classroom began with inviting students to formulate their own research topics and then standing aside while they used the new media productively to carry out their assignments. What is different here is the change in learning, not the medium used to implement it.

Let me be more specific on this. What was replaced in Dryden's classroom was the conventional transmission model of learning, which begins with the teacher in possession of all the knowledge and the students having none. The teacher then transmits this knowledge to students by requiring them to take notes from her lectures, to complete worksheets, to memorize, and to take tests. The alternative adopted by Dryden can be conceptualized through the framework of the project approach, as described by Katz and Chard (1989) or Miall (1989). This suggests that learning takes place most effectively if it unfolds in three phases: a first mapping phase during which students elicit what they already know and formulate questions about the topic; a second phase during which students actively search for information about the topic; and a third phase in which they bring together and report on the new information they have gathered. A common feature of projects is collaborative work, since students often work better with the incentive of each others' example findings, questions, and the discussions that collaboration makes possible.

An examination of Dryden's report shows him focusing first on the dramatic outcome of students presenting their multimedia reports (p. 289), a third phase event. He focuses next on the use of electronic resources during the second phase collaborative, research activity of the students (pp. 290-1). These offer excellent examples of the exploratory (phase 2) and constructive (phase 3) uses of hypermedia, and show that in the presence of a teacher as facilitating as Dryden students are indeed liberated to follow their own initiatives and to be inventive in all aspects of their learning. This comes primarily from changing the politics of the classroom, however, not the fact that students have multimedia tools at their disposal (although this may help). It is quite possible to carry out projects with no multimedia resources, as students in my own courses regularly demonstrate (see their project reports on the course web pages for The Shelleys, Romanticism).

Absent from Dryden's account, however, is phase one: there is no mention of the initial aesthetic, experiential response of students to the texts and films they undertook to study. What did they already know? How did they choose the issues on which they based their subsequent work? But the key issue for our purpose is this: what role did hypermedia resources play, if any, in those first responses at phase 1? Did students read Cinderella or Macbeth on screen? And if so, how did this influence their first responses? Dryden doesn't tell us anything about this phase.

The question of whether to read literary texts on computer or not is perhaps a vexed one (not least because postmodern theories now refuse to distinguish literary from non-literary texts: Landow, for example, regularly conflates primary and secondary reading in his accounts of Intermedia). My own position (see Miall, 1990) is that the initial reading of a literary text is best done away from the computer, but that re-reading for information, research, or study could be done effectively through a hypertext system (perhaps more effectively than in any other way, but powerful tools for this have yet to be devised). The state of absorption appropriate to the first reading of literature is less easy to sustain in a hypertext environment, with the cursor blinking, the distraction of embedded links, and the white noise of the computer fan deadening the aural play of sound in the inner ear. Dryden's chapter, in conclusion, offers a remarkable example of project work in action. But his main claim is unfounded: multimedia has no necessary relation to the student-centered classroom.