Would you do it on the web? Would you get stuck in a rut and find that you were looking at the same things in the same ways?

What techniques would you use? Google scholar's pretty good. And of course you would have bibliographic references within those. Maybe you have paid access to a subscription database so that you can actually get to see more of those papers, not just their abstracts.

How do you find out, though, those authors who come sliding in out of left field into your subject? How do you get access to other people's ideas of what might be relevant? How do you break out of your own personal viewpoint on what relevance is?

Google wants to help you, heck, everything on the web wants to give you nicely-personalized, relevant results.

But what counts as relevant isn't just those things that most closely chime with the resources or the theories you're getting comfortable with. It's those things that challenge you, those things that are uncomfortable, and those things that take you off across into other disciplines and areas where analogous research or insights are waiting to be brought into contact with the ideas you brought with you.

I still think that libraries are good places to get this to happen. They can be good places to be surprised and challenged by information that supplements and disrupts the material you can easily gather online.

This isn't an argument for books, per se – though I do think that there are cognitive and critical reasons to access a range of media that includes them – as much as it is an argument for the importance of knowledge structures that cover a range of subjects as a public commons, rather than as a personalised niche.