The Battle Of The Books

I can't say enough good things about the "Battle Of The Books" public debate last night at the NYPL. (For background, see here.) It was substantive, focused, occasionally hilarious, and always deeply interesting. As I suggested here a few weeks ago, I'm not a supporter of the Author's Guild action against Google, despite being a member of the Guild myself. But last night muddied the waters for me, in the best of ways. Two quick observations:

1. Google has amassed a serious image problem -- at least among the Manhattan intelligentsia -- in a shockingly short amount of time. There were some real boos at a few choice moments last night, and a few audience questions that were positively white-hot with rage at Google. The hotter they were, the more confused they seemed to be about what Google was actually doing. But still. There was a subculture of animosity in the room that was stronger than anything Microsoft would conjure up today. That's something Google needs to take very seriously. "Don't be evil" in a way is the easy part. The trickier one is: "Don't be dismissive when people end up thinking you're evil anyway."

2. I offered this up as an open question to Lessig, and he didn't seem to buy it, but I came away from the session with one new variable suddenly floating around my head. The fair use doctrine ultimately has to have a stable definition of what it means to "use" a piece of copywritten material for it to be at all useful as a legal category. And so it occurred to me that part of the problem here is that "use" changes dramatically when you move from "read" to "search." When you're reading something, access to the entire work is a defining part of the experience. So if you eliminate 95% of the work, it seems logical to file that as a critical subtraction under fair use. If you only quote three paragraphs in your literary analysis, that's cool. There are still 2,000 paragraphs of the original work sitting out the shelf waiting to be read. Effectively going unused.

But what do you "use" when you search? Google wants you to think that what you use is the result of the query. So if they send you back a few sentence fragments and a link encouraging you to buy the whole thing, then they've more than done their part, fair-use wise. They've only "used" a small portion of the book, and they've tried to make a sale. But the use-value of books when you're searching is different from the use-value when you're reading. If you were only searching snippets, the search would be, literally, useless. The value comes from knowing that your software representative is, on some basic level, reading the entire work and coming up with matches based on that comprehensive survey. Imagine doing a search and having Google report back "No matches, based on looking at 2% of the available documents." That would be supremely unhelpful, precisely because the use of search depends on the full range of the database being queried, and not just the sliver of positive matches culled from that database.

Google agrees that it would be a flagrant abuse of copyright if they scanned every page of Freakonomics and then served up the whole book to anyone willing to suffer through GoogleAds while reading it. And that's because the whole work is at stake. But if the value of searching comes from querying the whole work, not reading it, aren't we really talking about something different?

I'm sure Battelle has long ago wrestled with this somewhere, but it seemed like news to me.

Comments

[ognjen]

You ask:

"But if the value of searching comes from querying the whole work, not reading it, aren't we really talking about something different?"

and I say yes, we are. But the "fair" use of copyrighted material in this context is really one that doesn't diminish the value of the work as offered by the copyright-holder.

When we're talking about printed books v. Google Prints (or G. Books, or whatever), Google's use doesn't really compete with the publisher's offering -- when you buy the book, you normally don't get any login&password which would give you, the buyer of the book, the opportunity to search the material. The two products/services aren't exactly competing. (You could scan and OCR the whole book, of course.)

I can think of one publisher which *does* offer this searching service, it is O'Reilly, but notice that you *don't* get to use Safari for the price of the book you purchased.

In my opinion, Google would infringe the copyright if either one of these two conditions were met:
1) for each book purchased, you are offered a reasonably convenient way to query the text at no additional cost; or
2) a third-party search provider allows you to extract full text of the books, or its significant part

[Felix]

I'm troubled at the way the bigger debate seems to ever be reduced to a much smaller debate about fair use, and whether or not Google Book Search constitutes fair use of copyrighted product. I think that this product is so new and different that trying to squeeze it under the confines of the fair use doctrine is always going to be uncomfortable.

So you might be right that there's a way of reading the fair use doctrine which is more favourable to the Author's Guild and less favourable to Google. But at the same time I think we should be asking ourselves whether the fair use doctrine, which clearly wasn't designed for this situation, is really the thing we should be looking to here. I would rather look at the intention behind the fair use doctrine, and whether that is being violated.

My gut feeling, a bit like yours, is with Google on this one: Google Book Search is a win-win product for everybody concerned -- Google, authors, and publishers. But you also make a good point that the gut feeling of the Author's Guild pooh-bahs and of a large part of the Manhattan chattering classes in general is quite the opposite. I suspect a lot of the opposition is a knee-jerk reaction to a bazillion-dollar company muscling in to something which a lot of people anachronistically consider still to be a relatively genteel business. But you're quite right that Google's spinmeisters have done a dreadful job here.

All the same, the Author's Guild has done an atrocious job of really explaining what its issue is, and why/how its members would be harmed by Google Book Search. There will always be a Change Is Bad contingent whenever something new comes along, and the book world has more of them average. I suspect that most of the opposition really is as unthinking as you seem to think it is -- which means the Guild is really letting down its members by fanning the fires rather than approaching this issue in a more constructive and intelligent manner.

[ginsu]

Imagine Google hired 10 million people, and assigned each one a single book to read. Then imagine that Google distributes each search query to all of 10 million of them, and each person reports back the few lines, if any, of relevance in the assigned book. And then Google prints the few lines.

Would you have any problem with this? If so, on what basis? If not, why would it be different if Google uses technology to enable something that would otherwise be so clearly impractical?

[Steven Johnson]

Some excellent thoughts here, but a quick reply to ognjen's post. Yes, it's true that traditional publishers don't offer a digital search feature with their books. (Annoyingly, I believe -- you should get a free ebook with each real book you buy.) But they do come with a whole host of meta information that helps you scan the information inside the book: chapter titles, introductions, indexes, footnotes, etc. I buy a lot of books to be able to scan them for specific information, not read them all the way through. So I think there is a kind of crude search technology built into print books that Google is competing with. They're doing it a thousand times better, of course, which is why I'm largely in favor of the project. But there's more competition than I think you suggest.

Obviously, people don't read novels or more traditional narrative books this way. But people won't be using Google for the most part to find interesting novels. It'll be people like me, looking for information on a specific subject, who don't need an entire book's worth of results, just a few pages.

Now, I think probably Google's defense here is that the snippets they're returning are so small that you're forced to track down the book. Which makes the service simultaneously safer for them legally, and less interesting to me as a consumer. I like Amazon's idea of being able to buy a few pages or a chapter a la carte, not unlike buying songs from iTunes....

[Jason Schultz]

Further to you point, though, is the important distinction between copyrightable content and uncopyrightable factual information about a book. Copyright law does not protect indicies or "card catalogue"-type descriptions of a book because they are informational, not expressive in an artist sense.

The fact that these services or searches may be valuable doesn't change that. So while the AG may wish they had the right to control queries and metadata about their works, the law has never given them that control -- and for good reason. Control over factual information is a monopoly on science and knowledge which is inherently against progress toward education and freedom of information.