November 7, 2002  ·  Lessig

Ted’s latest (and his patience with me is wearing, so perhaps sadly, his last), makes a passionate argument against my source code escrow idea, based on the nature of software and the creativity that builds it. I realize I must have somewhere inspired this debate about “nature.” I renounce it. No more talk about nature, or the philosophy of creativity. My argument is simple (maybe simplistic, maybe naive) pragmatism.

Ted had asked me why software should be treated differently from other “copyright property.” My simple response was that giving software 95 years of protection was to give the public nothing of value in return. Ted seems to agree with that: Some modification of the existing system is in order. That is all that motivates my argument. But Ted still doesn’t find that argument persuasive.

Yet his response reminds me of my years studying philosophy. He argues that there is a “means of production” that goes into every creative work. For the Godfather, for example, it “was the mind of Mario Puzo.” For software, it is the source code. And obviously, as the government should not demand Mario Puzo make his mind public in exchange for the copyright(s) supporting the Godfather, so too should the government not demand of the software author that he make his mind (the source code) public in exchange for a copyright on code.

That’s a useful point for making a distinction, but it is not an argument against escrowed code. Obviously, every creative activity includes lots of stuff that went into the creation that is not properly thought to be part of the product. Dave pointed to this point earlier: In writing a novel, there’s tons of creativity that is opaque in the final product. We don’t know, and can’t see, what the author had for breakfast, what arguments she had with her spouse, who he flirted with before writing chapter 10, what threat the publisher made that made her end at chapter 15. All this is hidden in the final product, and hence all creativity is, in this sense, “opaque.”

True enough. But I approach this from a pragmatic perspective, not a philosopher’s perspective. I pick out source code as the sort of thing that the world would benefit from having revealed because, as Seth Schoen reminded me and as the GPL defines it, source code is “the preferred form of a work for making modifications to it.” So if a copyright system wanted to encourage others to build on earlier code, providing the source code is at least one very good way to enable that creativity.

Of course, it is not the only way to enable others to build on earlier code. Providing clear APIs, etc., would enable a bunch of building. Just not as much. And reverse engineering object code would enable a bunch of building. Just not as cheaply. My point is just that we should design the system to support the “preferred form … for making modifications.”

Unless there’s a great reason not to, and this is what I’ve not seen yet. I find Ted’s claim that “I do not want to live in a society which compels me to register the contents of my brain with the federal government” colorful, but not persuasive. Poets know their poetry by heart. Does anyone think that poets are being forced to “resister the contents of [their] brain[s]” with the government when the copyright law requires a deposit of their work? This is source code, Ted. No doubt it is the product of lots of sweat and blood. But let’s not get carried away.

Siva Vaidhyanathan wrote me that my distinction may make sense today, but how long will it survive. As convergence means that all content is increasingly software, will the source code for that have to be deposited as well?

The answer to that depends, of course, on whether the source code for that too is “the preferred form of a work for making modifications to it.” Maybe, and if so, then yes. But let’s see first.

(Nick Coghlan also wrote to me that it is a mistake to call the “source code” the product; the source code, he nicely points out, is the “design – we just happen to have tools for automatically turning the design into a usable product.” Nicely put.)

Ted’s map of the different ways in which one might depend upon the right to copy is an excellent and helpful clarification of this debate. In light of that, I take it the crux of any real disagreement between Ted and me is precisely this question about modification. There is a long history of depositing copyrighted works with the government as a condition, or requirement, of copyright. Authors of books had to do it; movie studios (for a time) too with movies. These deposits were in part to enable the works to be copied once the copyright expired. The only thing my proposal adds to this tradition is the ability not just to copy, but to modify and build on the earlier work. That ability is enabled more easily with the source code than without it.

So why not enable it if we can? After an author earns the return promised by the copyright system, why not make the creative work available in the form others could most easily build upon? What reason (beyond mind-control rhetoric) is there for providing only handicapped access to the creative work?

(Compare: The Second Circuit Court of Appeals in the Corley case said that the fair use right didn’t include the right to access content in the technically more convenient manner. Thus, there was no violation of fair use if the only way to cut a snippet from a DVD was to film the snippet from the display on your TV. That position is consistent with Ted’s: the law need not give you the best, or easiest, means of modifying a copyrighted work. But in both cases, I don’t understand why. Why not push towards the easiest method for building on the work of others, once the copyright itself has been respected?)