Just to play devil's advocate, wouldn't any list of actions greater than one create "an illusion of parallelism"? Even if I limit myself to only one next action per project, I'm still gonna have a list of multiple actions and I'm still gonna be limited to only being able to do one thing at a time.
The reason I've always liked David Allen's suggestion to go ahead and list multiple actions for a given project if you've identified more than one, and none of them are dependent on anything else happening first, is that it seems more efficient to me. Why go back to project support repeatedly if I already know I can make three phone calls, send two emails, run two errands, create a spreadsheet for that project now?
Also, when I'm working in a given context it sometimes makes sense to stay in that context. I can start with one phone call. Maybe the one I think is most important in the moment. But when I'm in "phone mode," sometimes it makes sense to keep making calls. Sometimes it's not much harder to make ten calls than it is one (especially since there's a good chance some will go to voicemail). If I'm working on @home tasks, sometimes it's actually easier to keep working on that list than it is to switch contexts to remain in a given project. If I'm running @errands, it's nice to know all of the errands I could run without having to go back to project support. Especially since I wouldn't be looking at project support while I'm out and about; if I ran a bunch of errands, went home and identified another I could have run, I'd have to go back out. I'd be kicking myself.
I'm not saying that's always the case for me. Sometimes I stay in "project mode" and switch contexts. Sometimes I stay in "context mode" and switch projects. I know some people in this forum have expressed the view that it should be one or the other. For me, it's situational.
Anyway, I'm not trying to convince you you're wrong or start an argument. Like I said, playing devil's advocate. Or think of it as food for thought. Or something.
