Projects are my contexts

The thing that struck me as genius when I first read David Allen’s Getting Things Done was the idea of contexts. The next actions concept was a clever way to reframe something most of us already did—put our to-dos in chronological order—but it was the recognition that you could only do certain things under certain conditions, and the organization your task lists according to those conditions, that sold me on GTD.

But many are realizing that Allen’s original contexts of @work, @computer, @phone, and so on, don’t make as much sense today as they did in 2002. I don’t know about you, but I’m never not @phone, and I’m seldom not @computer. Merlin mentioned this problem on Back to Work a few weeks ago.1 The technology we now carry around in our pockets and our backpacks have made the traditional GTD context taxonomy much less useful than it was a decade ago.

Certainly some of the old contexts still make sense: I can’t mow the lawn unless I’m home, and I can’t buy Pop-Tarts unless I’m at the grocery store. But a lot of the work related contexts are always “on” for knowledge workers.

You may have seen this article by Sven Fechner, which addresses the problem and develops a new set of contexts based on time and attention—it is, in a sense, a categorization of tasks based on the state of mind you have to be in to do them. If Client A is direct and to the point, but Client B is longwinded, you would not, under Fechner’s system, put calls to them in the same context because they require a different sort of effort and attention from you.

(Do you remember when Baysian spam filters first got popular? One of Paul Graham’s selling points in “A Plan for Spam” was that these filters became personalized over time. They learned what you consider as spam and filtered accordingly. Fechner’s context system reminds me of this because the context categorization of a task is a personal thing. You may enjoy talking to Client B and see it as a refreshing break in the middle of your day. I, not being what you’d call a “people person,” would see it as a horrible burden that required a great effort. We’d assign different contexts to the same task.)

I’m intrigued by Fechner’s system, but I doubt I could stick with it—too much time spent deciding on the context for each task. I’ve switched to a system whose main feature was given away in the title of this post: I treat my projects, especially my work projects, as contexts.

This system has the advantage of simplicity. I never have to think about the appropriate context for a task. It does, however, share with Fechner’s system a regard for my state of mind.

When I’m working on, say, a project at work that involves the analysis of a roof structure, I want to give that analysis as much of my continuous attention as outside distractions will allow. Some of the work will be programming, some will be drawing, reading, laboratory testing, analyzing photographs, discussing the problem with my coworkers, etc. In a traditional GTD system, these tasks are all in different contexts because I’m moving from one place to another, using different tools and equipment. These are also in different contexts in Fechner’s context taxonomy because they require different levels of time and attention. But within my brain they’re all the same context: the context of analyzing that roof structure. And it’s keeping my brain in the same context—as it sees context—for as long as possible that makes me most productive and most satisfied with my work.

So there are no @s in my task lists anymore, just projects and next actions. Will this work for you? I have no idea. But so far it’s the system that makes the most sense for me.


  1. I can’t remember which episode it was, but I think it was either “Corner Casey Kasem” or “Muscle of Failure.” Listen to them both; it won’t hurt you.