It was like they were standing at far end of a room full of books and periodicals, with no walls separating the room from the outside, standing on the edge of an unfinished floor and observing poorly resolved things floating around before them that might or might not be useful data or other constructs. Shapeless forms of possible knowledge floating in the dark and cold unknown. Every now and then the scientist is able (using some tool or another) to grab on to one of these poorly defined forms in an attempt to wrestle it into a place where it could be understood. Sometimes, the thing they would grab would be reformed into books and articles to add to the shelves in their proper place. Sometimes (often) putting the new item on the shelf required tossing what was already there out into the vague abyss at the edge of the room, sometimes the thing they grabbed would wriggle free and escape before sense was made of it. Sometimes it would be thrown back because it was crap.
Definitely worth a read.
Aside from its own merits, it reminded me of a conversation I had with him recently. I don't know how the subject came up, but I do remember telling him how I wanted books to change. It was a recent conversation, so I was probably a bit out of it, but I still agree with what I said then.
Blogs, particularly science blogs, have spoiled me. There are books out there I want to read, weighty little things like Jared Diamond's tomes and 1491 that are broad overviews of topics with which I just don't have enough familiarity. But I'm not reading them. I'm not hesitating because it's too much reading, mind you. Blogs haven't spoiled my attention span. I'm hesitating because the books are wrong.
They have to be. These books cover so much that their authors can't be experts in all the details (where they are experts in any of them). There is information that has gone through at least two rounds of Telephone--simplification and translation to popular or vernacular that will change some of the meaning. There is information that is outdated. I don't want to learn and rely on this stuff only to have to redo all my work later.
I want what I have when I read a science article online and see a trackback, or what I get simply by keeping up with several science blogs. I want a notation that someone else has something to say on the subject. Not unedited, of course, because that extra shaping is part of the point of a book, but I want flags and the option to see when an expert says, "Well, this matter shouldn't be presented as settled," or "Stating this in that way, while not incorrect, is misleading. The more complex truth is...."
When I'm reading for information, even if I'm reading something aimed at the lay audience that I am, I want information, and I'm slowly becoming dissatisfied with anything else.