It's hard to know where to begin, so I guess I'll just dive in.
First, I agree with Rob, Tom, and Carol regarding the need to develop semiotic models and techniques that will be practical in the real world of instructional design. I got interested in semiotics originally because I thought it held some real promise for re-conceiving our current models and techniques, and I still believe that. What it mostly offers right now is a different way of thinking, and I am trying myself to figure out how that different way of thinking affects how we do design. Potentially, semiotic thinking could affect nearly every aspect of design, which suggests to me that efforts ought to proceed in a number of directions simultaneously.
Interface design, as Rob [Phillips, 10 Nov 94] suggested, provides an excellent example. How do we decide what is an "intuitive interface design?" Semiotically speaking, it would be one in which the user easily interprets the signs used in the interface for what they are supposed to stand for. I am reminded of an example given, at one of our seminars last year, by a designer of multimedia instruction who described a program in which an icon of a parachute was used for the button enabling users to quit the program. I remember being quite mystified as to why that symbol had been chosen, because it certainly didn't make any sense to me. However, when the speaker said the parachute stood for the metaphor of "bailing out," it made perfect sense. Evidently, however, the sign was not interpreted correctly by most users, because it was changed to something more obvious. It seems to me, then, that semiotics offers us a means of determining what signs should be used in interface design to create ones that will be "intuitive" to the intended users. I believe Ward Cates has been investigating metaphors as a basis for interface design, and this seems to me a fruitful way to proceed.
In response to Tom Reeve's [8 Nov 94] question regarding the possibility of semiotic design leading to constructivist learning environments, I think the answer is yes. In fact, asking learners to express what they know or how they interpret the world in different types of media is a way to facilitate understanding of how signs function in the signification process. In other words, students can become aware of how various sign systems communicate meaning. Different aspects of the same object are highlighted when different signs are used to communicate its meaning. For example, a book and a movie highlight different aspects of the same story. A formula and a graph provide different views of the same relationship. Becoming aware of such differences will enable learners to use signs critically, as Gary [Shank, 9 Nov 94] said, to "suggest various forms of intended and unambiguous meaning."
I'm not sure I agree with Rob [Phillips, 10 Nov 94] that semiotic design could lead to behaviorist teaching environments. A strictly behavioral approach doesn't put any stock in what instructional strategies mean to learners, only whether they produce certain desired outcomes, usually expressed in terms of some behavior being exhibited more or less frequently. When I read Holland and Skinner's programmed text as an undergraduate in psychology, I was certainly successful in learning the concepts being taught and achieving the defined goals, but I was also bored out of my skull and about decided to change majors if that was what psychology was all about. Behavioral approaches simply don't deal with affective responses very well, but they are as much a part of the meaning we derive from experience as are cognitive or psychomotor responses.