Sunday, August 17, 2008

Literature Maps

Academics write 'literature reviews', and talk about 'logical space'. I wonder if there's any way to literally map out the existing literature, with nodes and connections, to provide an explicit representation of our communal knowledge? Two kinds of map would be especially valuable: one displaying the chronological unfolding of the literature, in terms of which papers are responding to which other papers (this should be relatively easy to construct); the other would look deeper and attempt to map logical space itself, and the way various arguments and positions are related, regardless of who put them forward or when (though this data should also be included, as a reference aid).

It would be a Herculean task, and no individual philosopher is in a position to codify the sum knowledge of the entire academic community. But it seems susceptible to chunking, and thus possible to construct gradually through wiki-style peer production. (Any Ph.D. could convert the 'literature review' from their dissertation, and authors could fill in the details for their own papers as they are published.)

It's the sort of thing which would be difficult to get off the ground, but once the 'tipping point' is reached, it could be extraordinarily valuable. (Or so it seems to me. What do you think?)

Perhaps the main challenge (in principle) is to work out exactly how to standardize the represented information. How, exactly, does one go about converting a literature review into a formal map? (Presumably not any old citation is substantial enough to qualify as a connection, but where do you draw the line? Is there any principled way to do this?)



  1. This would be an awesome project; and there are all sorts of variations on it that could be developed if the original projects went well.

    For the actual mapping something like Exhibit project would seem ideal.

  2. There are bibliography projects already (I'm thinking of the Chalmers/Bourget Mind Papers effort) which use a certain degree of automation. An automated citation-scraping project, leveraged off projects like that perhaps, could provide a skeleton for a human editing effort.

    The advantage of that would be an immediate map of connections showing what paper cites what. Editors could then dive in wiki-style and edit the connections (and add to them), and tag them as responses, comments, corrections or retractions to the cited paper.

    I suspect that's the sort of thing that would galvanise a broad human work-effort: because even seeing what the automated system generates would be interesting for people. It might be a better 'business model' for the project then a "from ground up" effort by experts, though obviously less authoritative from the get-go.

    That of course assumes that automated scraping and/or existing list extraction is even feasible to build a connection database. There's limited data presented on Blackwell-Synergy pages for example and I hear there are some other tricks that can be used, but it would take some serious IT-noused effort over and above academic expertise.

  3. For a pretty extensive example of this, see

  4. Franco Moretti of Stanford writes about mapping literature; you may find it interesting in the least.

    And I'm sure others are reminded of Borges' story, The Library of Babel.

  5. Check out for artist Ashley Hunt's version of this idea. Spectacular.


Visitors: check my comments policy first.
Non-Blogger users: If the comment form isn't working for you, email me your comment and I can post it on your behalf. (If your comment is too long, first try breaking it into two parts.)