It reminds me of the responses I got from my multilingual respondents, when I was writing The Language Imperative, to the question "When you switch from one of the languages you speak to another, do you also switch from being one person to being a different person?" There were a few "I don't know" responses. But almost all were either "Yes, of course -- what a ridiculous question!" and "No, of course not -- what a ridiculous question!" There appear to be certain questions ["Is the linguistic relativity hypothesis valid?" is another one] for which the meta-answer is "Yes, for one subset of human beings; no, for a different subset of human beings." I think this "hypertexting in spoken language" question may be yet another example.
We're not going to settle this issue here -- which will be a relief to those of you who disagree even with the definitions of terms that I extracted from the comments of those who claim to "hypertext" as they speak. I want to wind up this discussion, therefore, by mentioning just one more thing that might be relevant, and that has been on my mind as I've been reading what you have to say: the possibility that our disagreement about what goes on when you talk is a function of chronological age. That is, it may be that most people over 50 -- maybe most people over 40, or even over 30 -- can't hypertext as they talk, but most people younger than that can..
A moderately large body of literature has accumulated over the past several decades (starting, roughly, with Marshall McLuhan) presenting a hypothesis that the brains of human beings who've grown up with everpresent electronic communication -- especially music videos, television equipped for "surfing," and (later) Internet conversational media -- have been significantly altered by that perceptual experience, so that those human beings process information differently from their elders. For example, it's hypothesized that most people my age (having grown up with print media) are accustomed to the idea that narratives have a beginning, a middle, and an end, and have trouble following plots that aren't structured that way. (That's true for me; I have an awful time reading William Gibson.) The e-generation, on the other hand, is used to flipping from one narrative to another, seeing only bits of one story and pieces of another, and is used to the quite different and nonlinear structure of music videos. The claim is that these different perceptual histories have literal, not metaphorical, effects on the human brain.
If that's true, then it may be that those of you who claim to be able to hypertext as you talk (with varying degrees of reported ability to listen attentively to others in the conversation at the same time) are correct, simply because your brains are different from my brain. I wouldn't be surprised.
Yesterday I hunted for one or two reasonably brief articles representative of the literature I mentioned above, and couldn't find anything really suitable. Either the articles were huge pdf files full of technical jargon, or they weren't available online, or they went -- in my opinion -- Way Too Far, given the slimness of the available evidence. I did find two items that were something like the general-interest piece I was looking for, and I'll give you the references here just for completeness' sake, but they're not what I wanted. It may be that someone among you has better examples to offer, in which case I'll be pleased to know about them.
One is titled "Brain Candy"; it's by Malcolm Gladwell, was a New Yorker piece, and is at http://www.newyorker.com/printables/critics/050516crbo_books . The other is Andy Oram's "Marshall McLuhan vs. Marshalling Regular Expressions," at http://www.oreillynet.com/lpt/a/2504 .
Thank you for the discussion; you've given me a great deal to think about and to work through.