While it may be argued that hypertext is not inherently a collaboration between reader and writer, it can be certainly be constructed as such: the writer provides the lexia and the original set of multiple links which the reader may follow--by choosing certain links and not others, the reader constructs a meaning which is different from the meaning that any other reader will construct, and which may be different from any meaning the writer intended. This form of "collaboration," however, is superficial; it does not require an active engagement between reader and writer. Constructive hypertext systems which allow the reader to add new links and new lexia to the original work constitute a more concrete form of collaboration, as the original writer may attack, defend, or question the links or lexia that the reader has added, and all links and lexia remain accessible to successive readers. Because a constructive hypertext requires multiple collaboration as opposed to the internal individual act of collaboration performed in the reading of an exploratory hypertext, a dialogue may arise and any number of new voices may join the discussion. Because "in the space of hypertextual writing, anything that arises will be merged, gathered into the network of polyvalent discourses" (Moulthrop and Kaplan, 1994, p. 235), contradictory or resistant discourse can exist alongside those voices which converge upon a point of consensus, thus allowing the coexistence of consenting and dissenting voices, or constructing a kind of consensus that "incorporates a multiplicity of independent voices rather than constructing a final, authoritative position" (Taylor, 1992, p. 140).
Multiply-authored hypertexts allow the student to write individually and then link his or her individual writing to a larger text; therefore, the student can be evaluated as an individual and as part of a larger group without necessarily having to depend on others. Also, the dialogue-oriented nature of hypertext can encourage students to participate in the authorship of the hypertext document by providing multiple points of intersection for the individual discourses.
At the same time, hypertext does not fall into the trap which Karen Burke LeFevre describes in Invention as a Social Act(1987):
The teacher who adds a few group activities to the composition classroom does not automatically have a dialectical view of invention. Reconceiving invention as a social act does not mean that we assemble a group of atomistic individuals--"add people and stir"--who later resume their private search for knowledge (p. 49).Moreover, Anne Ruggles Gere (1987) cautions that while the democratic give-and-take of collaboration is essential, it does not by itself guarantee that any learning will take place. Participants in collaborative groups learn when they challenge one another with questions, when they use the evidence and information available to them, when they develop relationships among issues, when they evaluate their own thinking. In other words, they learn when they assume that knowledge is something they can help create rather than something to be received whole from someone else. (69) By focusing on the creation of knowledge in the form of linked lexia, hypertext can be used to explicitly show how students can work together to negotiate meaning with the discourse communities of their classes without necessarily coming to any kind of totalizing consensus.
Once the writer engages the hypertextual discourse, it becomes difficult to disengage from the dialogic process which occurs as the varied voices of the class create an ongoing textual conversation. The ability to add commentary to constructive hypertexts in the form of new lexia and to expand existing hypertexts by linking them together enables the emergence of a multivocal dialogue which does not privilege a dominant discourse, but encourages hypertext readers to hear and engage all of the voices which are present.