I finally managed to organize all my printed articles into folders. Quite a tedious task, but maybe worth sharing my experience in more detail.
First the background and the objective: I had probably around 400 printed journal articles, kind of sorted into piles and lying around my office threatening slowly to engulf me. The articles had accumulated over the last few years and featured both rather extensive collections on well-defined topics (like policy responsiveness) and scattered individual pieces that I liked for some reason on topics I mostly don’t keep track on (like regime collapse). Obviously, I would want the articles organized into folders so that 1) they look neat, 2) I have quick access when I need them, and 3) I am able to do quick surveys of particular topics.
The solution I opted for is organizing the articles into approximately 30 topics and, within each topic, alphabetically. The more common way of using only alphabetical ordering doesn’t work well for libraries without a catalog because you need to remember the author of an article in order to find it. And making and keeping a catalog would be too tedious.
Unfortunately, one cannot rely on tags to organize physical objects like printed texts. Discovering tags (as used in blogs for example) has been a real revelation for me and my efforts to put order to the world around me. Tags out-compete hierarchical classification any time. But for my folders, I had to settle for non-overlapping classification into a small number of categories (topics). This mostly works fine except for the rare cases where an article can go into more than one category. Usually this happens with articles that are interesting both for their substance and their exemplary application of a method. My solution for these cases is to put the articles where I will most likely benefit from finding it in the future.
What I discovered during the process is that the system of academic topics is fractal. Each topic can be decomposed into subtopics and each subtopic can be decomposed into others, ad infinitum. A few years ago, all I had to have is one topic/folder for organizational design and change; later, I had to take the literature on co-ordination out of it; later, the articles on EU co-ordination were taken out of their parent topic; finally, after writing a paper on EU co-ordination in Central and Eastern Europe, even this sub-sub-topic had enough entries to make a separate folder necessary. So there isn’t much point into imposing hierarchy on your topics, and probably nobody else’s classification can work for you. In all likelihood, you don’t need an ‘EU co-ordination in Central and Eastern Europe‘ entry in your classification scheme. The fractal nature also means that you have to accept that the scheme will change over time as you go deeper into sub-topics and sub-sub-topics. You need to be pragmatic rather than dogmatic about your classification: no need for hierarchy, and no worries that your topics might cover issues differing greatly in their generality. My topics include side by side International Relations (which I mostly don’t care about) and Case studies of EU compliance in the new member states. You know when an issue is mature enough to get a separate topic when it makes finding anything else in the parent topic cumbersome.
In general – print as little as you can since reading articles in on a screen not only saves trees but makes for an easier organization of your library as well.
I will be extremely curious to read about your ideas about organizing your personal library in the comments.