Well. This week has absolutely sucked. When did I last update this thing?
Oh, right. Anyway.
The dirty little secret about the videogame medium is that it'd probably be more accurate to say the videogame media. Granted, no universally agreed-upon definition for medium/media exists--I seem to recall a good, functional one about communication technologies and the social protocols that surround them--and given that HoMM5 runs on the same physical hardware as Blogger, the way we think about them and use them certainly has to come into it. But the trouble with coming up with a definition of "videogame" is that the commonalities between, say, Resident Evil and Second Life are not all that much stronger than the connection between Resident Evil and, well, Blogger. Part of why it's important to be able to identify different texts as belonging to different media is that it allows for the construction of critical theories appropriate to the medium in general.
I wonder if Eagleton's tripartite division in pre-structuralist lit theory might be useful in helping us see some of these distinctions. Authorial intent is not a sexy concept, of course, and it's unclear where some of my own perspectives fall...to what extent does it make sense to say that a game "says" or "does" something? Are we talking about the author? Generally not. If we do talk about the author, it's usually because someone fucked up. Part of the thing that makes mediocre games so compelling as objects of study is looking at the pieces and not being able to resist coming up with explanations about how they were supposed to fit together, before the dev team ran out of time or money. Other games, like Black & White and Frasca's still-fictional Strikeman, more or less demand to be looked at in terms of authorial intent, at least in terms of vision. Even emergent systems would seem to have a vague intent of their own, if only an intent to allow players to play with these rules over here but not those over there. But art doesn't generally work out the way we plan, and the engine we see is the product of several different intentional actors, along with mistakes, quick fixes and changes in direction, even before our perspectives as players come into play.
That said, as we go into the realm of multiplayer and user-generated content, reception theory does seem like it'd be the closest analogue to what would be most effective. There's certainly a lot to be said about authorial intent in Second Life, just as fan cultures have done some interesting things with the thoroughly authorial and linear Resident Evil, but in general, certain theoretical approaches will work better for some genres, and pinning some of those down might be more important than coming to a complete understanding of what videogames are.
After all, the bar is pretty low here. According to Eagleton, nobody knows what the hell Literature is anymore.