Maybe it's the Canadian in me (sorry, eh?), but prudence is a very important aspect of "innovation". Perhaps that is included or implicit in what you wrote, Danielle, but for folks like me it needs making explicit. There is nothing particularly virtuous or even necessary about constant change. Recognizing when change is needed, now there's a skill I wish we all had (myself included).
Indeed, for some 80 years or more, now, cognitive psychology has emphasized that the first, and perhaps most critical step, to problem-solving (and that's what innovation is, or ought to be) is defining what the problem is. (see Karl Duncker, and "functional fixedness": http://en.wikipedia.org/wiki/Functional_fixedness ). And within organizations, defining the problem commences with knowing that there is one which requires a solution. Innovation should not be solutions in search of a problem, but the recognition of important-enough problems, and their creative solution.
So, encouraging innovation requires a certain propensity to reflect, and to adopt "the big picture" when doing so, in order to both know when it is prudent to innovate and to avoid the conceptual traps that Duncker noted several generations back.
Both Peter's and Henry's posts allude to a few other critical aspects:
1) encouraging such reflection, perspective-taking, and intervention, by documenting and actively remembering past instances of innovation, and failures to innovate;
2) giving tacit and explicit permission to try something new.
Encouragement without permission is like saying "My door is always open" and leaving it open just a crack, is if to say "My door is always open, but I'm really busy right now, so please don't bug me". And permission without encouragement tends to result in very little off-line cogitation, and insufficient preparation.