Now back to philosophy. Or at least, to philosophical taxonomy. David Bourget and I are finalizing a new project for access to online work in philosophy. To a very rough first approximation it will be like MindPapers generalized to all of philosophy, although there will be many significant differences (it will be less ambitious in some respects, more ambitious in others). More on that when it goes public, hopefully within the next month or so.
One part of the project is a classification scheme, under which any paper in philosophy can be classified in up to three areas. The idea is that at least eventually, the classification scheme should be about as fine-grained as the MindPapers scheme. Philosophy is divided up into five clusters (Metaphysics and Epistemology, Value Theory, Science Logic and Mathematics, History of Western Philosophy, Other Philosophical Traditions). Each cluster is divided into six or more fields (in M&E, for example, these are Epistemology, Metaphilosophy, Metaphysics, Philosophy of Action, Philosophy of Mind, Philosophy of Language, Philosophy of Religion). Each field is divided into 5-10 areas (in Philosophy of Mind, for example, these are Consciousness, Intentionality, Perception, Metaphysics of Mind, Epistemology of Mind, Mental States and Processes, and Misc). Then each area can be divided into up to ten topics, and up to ten subtopics of each in turn (in roughly the way that the main areas of philosophy of mind are divided into topics and subtopics in Mindpapers).
With help from a number of others, including consultants in a wide range of areas, we've made a very rough first attempt at a taxonomy. Unsurprisingly, this taxonomy is better-developed in some areas than others. And even in areas where it is reasonably well-developed, many tricky decisions have to be made. Eventually, we will have ongoing projects for the refinement and development of these categories, with systematic consultation. For now, however, we're trying to get things up to scratch for a first draft.
At the moment, the various areas of M&E (apart from philosophy of religion) are developed to a fair degree of detail, though there is still work to be done. Value Theory and Science/Logic/Mathematics are more patchy. Some fields in these clusters (such as Philosophy of Gender, Race, and Sexuality, Philosophy of Education, and Philosophy of Social Sciences) are hardly subdivided at all, while other areas (such as General Philosophy of Science, Social and Political Philosophy, and many other areas in these clusters) are extremely patchy and incomplete. As for History of Philosophy, we've decided not to subdivide this for now beyond a few obvious groupings, and then categories for a few individuals in each period (these were settled by mechanically choosing those with more than n entries with their name in the title in the database, followed by a small amount of tweaking), although we may well subdivide these further eventually. And we haven't made any attempt to subdivide the Other Philosophical Traditions.
For now, I'm calling for feedback from the philosophical community, either via e-mail or via comments on this blog. Especially valuable will be thoughts on categories that we've missed, on ways to structure categories that don't yet have much structure, and on better ways of structuring things in tricky cases. The field structure is largely set (though perhaps not irrevocably), but the area structure is still a work in progress in many cases, and topics and subtopics are still very much in progress. In many cases what we have has been compiled partly from our own sense of the fields, partly from Internet sources (the Stanford Encyclopedia has been valuable), and partly from feedback from others, but all three of these things have been more extensive and useful in some fields than in others. So further feedback will be very useful at this stage. For now, our aim is to fill in structure in the first three main clusters, though thoughts about the other two clusters are also welcome. Of course we probably won't be able to follow every suggestion, but we'll at least give every reasonable suggestion some consideration.
Some methodology: we'd like to stick to 5-10 subcategories per category where possible, as this makes the system much more usable (and there are also relevant technological constraints). Of course sometimes there will be fewer (especially at the topic/subtopic level), and some topics may not have subtopics at all, especially in smaller areas. In a few cases it seems unavoidable to have more subcategories, but this shouldn't happen too often. Typically, finest-grained categories will have around 15-100 papers in them. More than this calls for further division, while less than this calls for less division. For now, we don't want to go beyond five levels. A given category can be crosslisted under multiple parent categories, marked withn asterisk in the taxonomy. Each crosslisted category has a primary parent category under which it is listed without an asterisk, though in some cases this selection is fairly arbitrary. For many purposes the choice of primary parent category won't matter much, as the relevant papers can show up in multiple places.
Of course there are many ways to produce a taxonomy like this, and this is just one way. For a start, the system is produced by analytic philosophers and has a bias toward carving things as an analytic philosopher would. Still, we'd like the system to be useful to continental philosophers and those in other traditions. Continental philosophy is covered to a considerable extent under 19th and 20th century philosophy, and there's a separate field under "Other Philosophical Traditions" both to capture contemporary work that doesn't fit well elsewhere, and to provide a way to mark continental papers for those who are looking for them. There's no reason why papers from continental and other traditions can't be included under the first three clusters too, though, and we're open to using categories that will help make that possible. Even from within an analytic perspective, of course many taxonomies are possible, but as always some choices need to be made. We're not suggesting that this is a definitive taxonomy (it's just something to make a website more useful), and we'd look forward to seeing other taxonomic attempts by others. And again, keep in mind that what we have currently is extremely preliminary and is highly sketchy and inexpert in some fields.
With that in mind, all constructive feedback by e-mail or in the comments is welcome. Feedback from those with expertise in relevant areas is especially welcome, and non-anonymous comments are especially appreciated. Again, here's the draft taxonomy. Thanks in advance,