It's been interesting to read the truly substantive comments that have been made in response to Clark's paper (despite a slow start). I'd like to add my two cents worth.
First, I think the word "game" is one of those loaded terms into which people quickly pour their biases as soon as they hear it. All adults, it seems, have formulated an opinion on the subjects of games in education. I have found it can be hard to get a discussion centered on the proper issues underlying game design in education once the word is "let loose" because of all the baggage that people seem to carry along with it. Of course, it's not hard to understand why. The word game invokes a feeling of childishness which can quickly insult an adult. I, too, would be reasonably suspicious of any instructor who started gushing about "our goal is to have fun and play games."
Having said that, I have come to the conclusion that we in education have unfortunately undersold the value and power of gaming. Perhaps we have just come to take it for granted or no longer know how to take it seriously. Just like we have "Killer Apps" in the computing industry, so too I feel that gaming is one of those "Big Strategies" that just permeates human learning (particularly if you take a Piagetian perspective on things). A similar construct we also have never taken seriously in our field is that of human play. Again, the word itself invokes quick comments and feelings, many of which are not justified when you take time to look a little deeper: "Children play, adults do not." "Work is respectable, play is not."
As I have studied and thought about how to promote and support the best and most important kinds of learning (i.e., "intrinsically motivating, meaningful learning environments in which individuals self-regulate their own learning," to use the vernacular), I find that gaming and play attributes quickly bubble to the top of ways to put the theory into practice. These attributes include clear goals and expectations, a clear role or purpose for the learner, a meaningful context for a learner to judge whether or not their actions/responses are appropriate, feedback that makes sense, motivation built on individual curiosity and optimal challenges, etc. For these reasons, I think Roger Schank and his colleagues at Northwestern University are right in referring to games as "interactive stories." Consequently, I think that games should share the special status held by stories within cognitive psychology.
Another thing that draws me to game design is the inherent blending of both artistic and analytic elements, an issue that has been raised this week. The best games have a clear structure and purpose requiring careful analytical design, yet have those artistic elements that no one can sufficiently explain why they are needed or how they work--you know it when you see it. My graduate training in instructional technology was rooted in disavowing any artistic elements in the design process (aka "the proper study of IT"). I've long been annoyed by that because I know that the very best instruction is also among the most creative and innovative. Despite our best efforts to the contrary, we clearly do not understand the instructional design process very well, at least when carried out at its very best. We rely on professional designers to bring their own unique creative genius to bear on a problem in the context of careful and continual field-testing (call it formative evaluation or rapid prototyping if you want), even though they undoubtedly also use the best analytical tools to support their work (and to be sure they don't overlook the obvious).
As others have noted, there is a long and important educational history to concepts such as gaming and play that we in instructional technology have failed to recognize. (Few probably know just how massive their respective literature bases are.) I think, though, that there is a unique opportunity afoot due to the important advances in interactive (processing), visual/aural (artistic?), and networkable (social?) features of desktop computers. This interesting convergence of computing capability means that designers of "interactive stories" have much to look forward to in the next few years. I hope we can take advantage of it.