Epstein on Google Print

Richard Epstien is a hero of mine, so this essay on the Google Print controversy was frustrating to read. His point of departure is the recent tiff between James DeLong and Larry Lessig over Lessig’s airplane analogy.

Now, I agree with Epstein and DeLong that Lessig’s exposition of the Causby case was sloppy. When I first looked up the case while writing this article, I was surprised to find that the point he highlights was in fact an aside to the case.

Nevertheless, I think it’s pretty clear that Lessig was drawing an analogy to the Google Print case. He wasn’t citing the case as a precedent, nor was he claiming that the issues presented were precisely the same. What he was saying was that mindless applications of old rules to new technological situations can lead to absurd results. Clearly, requiring airplanes to get permission from every landowner whose property he crosses would be absurd.

So that begs the question: is the result sought by the publishers in the Google Print case absurd? I agree with Lessig that it is, but I think it’s not as obvious as in the airplane case. Nevertheless, I think that a careful review of how Google Print actually works demonstrates that the analogy between the airplane example and this case is quite strong.

The problem with Epstein’s article (and most of the writings of Google’s critics) is that they don’t seem to pay much attention to what Google is actually doing. As a result, they end up unwittingly attacking straw men that are different in vital ways from Google Print as it actually exists. For example:

Here the value in use of the air rights to places you can’t reach from the ground is trivial, relative to the costs of running the system, so we take the rights, save the administrative costs, and know (as well as these things can be known) that the all landowners are better off than before because of the gains that they derive from access to the transportation system or to the goods and services that it provides. This last point is critical because Lessig, in his answer to DeLong, challenges him with this question: “Did anyone receive compensation for the taking? No.” But once implicit-in-kind compensation is taken into account, the right answer is yes. All landowners are left better off than before. There is no implicit wealth switch as with the Google project.

What’s frustrating is that Epstein never explains what he means by an “implicit wealth switch.” Google is certainly hoping to profit from Google Print, just as the pilots of airplanes surely hoped to profit from their flights. But Google’s profit isn’t coming at the expense of any revenues that would otherwise accrue to publishers.

This becomes obvious when we examine how Google Print actually works. For in-copyright books for which publishers have not opted in, Google displays no ads at all, so arguably there’s no revenue to distribute at all. Once publishers opt-in, Google does display ads, but it splits the revenues with publishers. And if publishers feel the terms Google offers are unacceptable, the publishers are free to opt out entirely.

In short, Google has done everything it reasonably can to avoid any “wealth switches” that might run afoul of copyright law. Yet Epstein and company persist in asserting otherwise without explanation.

Epstein does raise a point that deserves a response: why do I and other Google supporters make such a big deal out of this case? The reason is that copying is the fundamental activity of a computer. The kind of copying Google is engaged in is fundamentally different than any copying anyone ever did in the analog world. No one will ever see the full copies in Google’s index, and those copies are, in and of themselves, completely useless to the average consumer. (they become valuable only when combined with Google’s search software) But while such copies were nowhere to be found in the pre-digital world, it’s kind of copying that’s unavoidable for any digital application seeking to manipulate copyrighted content. It is, therefore, vital that the copyright law distinguish this kind of copying from the ordinary kind that was the intended subject of copyright law for the first 200 years of our nation.

If the courts decide that all copying–including intermediate, purely electronic, and non-user-accessible copies like those in Google Print–is presumptively subject to the heavy hand of copyright regulation, it will have a dramatic effect on the development of future technologies, because broad classes of technological innovations will require their inventors hire a lawyer before they can even get their product off the ground. Such a requirement can be a very high barrier to entry for firms that might not have Google’s deep pockets.

This is where the analogy to Google itself, which Epstein poo-poos, comes in. Had a regime similar to that sought by book publishers been in place for web sites in the 1990s, the inventors of the original Google search engine–two grad students at Stanford–would have needed to obtain millions of “opt in” agreements with webmasters before the Google search engine would have been large enough to be useful. They would have faced an impossible chicken-and-egg situation: hiring lawyers takes capital, attracting capital requires a proof-of-product, but a proof-of-product is illegal until you’ve hired a team of lawyers to clear the rights to several million web pages. Fortunately for all of us, there was no Association of Web Site Publishers to stimie Google’s development. They created unauthorized copies without anyone’s permission, and the result is the search engine we all know and love.

Here again, Epstein pronounces this comparison “silly” without explanation. It’s impossible to know what he’s thinking. There certainly are some legitimate differences between the two cases, and perhaps he has some of those in mind. But I think that the more likely explanation is that, not being immersed in technology, he doesn’t grasp how fundamental and pervasive copying is to the operation of digital systems. He probably doesn’t grasp the fact that the copying in Google Print is not in any sense brazen and unprecedented as the publishers suggest. To the contrary, such copying occurs in every full-text search engine, and it is likely to occur in dozens of future digital applications that haven’t even been invented yet. The rules we create now will help to determine whether our copyright law nurtures those new innovations or smothers them before they can get off the ground.

SHARE: