I've been kicking around some ideas about the differences between science and technology. A number of authors have noted the importance of the interconnection between the two (Zilzel, Smith, Merton, etc.), and a few have explored differences in methodology (Vincenti, Henderson) but we still lack a comprehensive explanatory model.
I'd like to present my own idea: The Bingo Card Model [PDF]. My thought is that the primary difference between science and engineering lies in ontology. Science in inherently ontologically incomplete; existing theories dictate a priori what should be explored and articulated. A corollary to this idea is that only those things that should theoretically exist are recognized (cf. Kuhn). Engineering, in contrast, is always ontologically complete. Only those things that have been created have any ontological significance. Furthermore, additions to the world of things come from adaptations and expansions of those things that already exist (i.e., affordances).
Of course, there are some problems with the model. I conflate engineering, technology, and many aspects of material culture. I will also run into a problem with validation. I'm not sure how to provide a rhetorical grounding for presentation. There may be something in my work on machine books (i.e., they show machines that may or may not exist).