September 22, 2020

Book Club Discussion: Code, Chapter 2

This week in Book Club, we read Chapter 2 of Lessig’s Code, and Other Laws of Cyberspace. Rather than kick off the discussion with an essay, I’ll just open the floor for discussion. (To say something, add a comment below.) I’ll chime in with my opinions as the discussion develops.

For next Friday, we’ll read Chapters 3 and 4.


  1. off topic says:

    off topic

    dont know, if its in the blog software, or a custom template: but the link down on the front page to the rss feed is broken, see source fragment:
    a href=”feed:http://www.fre[…]


  2. Please note: this is my first time reading this book, so perhaps I will be missing some points that will be clarified later, but…

    The stories in this chapter certainly capture many of the dilemmas we face
    in cyberspace. However, I don’t know that I see how some of Lessig’s conclusions follow from the stories.

    The first story on Avatar space is intended to illustrate a solution to the issue of regulability.
    The possibilities in Avatar space are determined by the code—the software, or architecture, that makes the Avatar space what it is.

    …there is regulation of behavior in cyberspace, but that regulation is imposed primarily through code. What distinguishes different parts of cyberspace are the differences in the regulations effected through code. In some places life is fairly free, in other places controlled, and the difference between them is simply a difference in the architectures of control—that is,
    a difference in code.

    Strictly speaking, I’d have to say this is correct. De facto, code enforces fairly rigidly what I may and may not do. But that is a limitation of code, partly due to the fact that the developers had limited time/resources/vision to anticipate what else I might choose or be permitted to do. Equating Law with Code presents three problems, in my mind:

    Who gets to select the regulation that is implemented in the code?
    Does circumvention of “the” code (e.g. tinkering) automatically make one a “law”-breaker?’
    Who’s in charge when the code hits an exception?

  3. In case anyone’s interested, Orin Kerr has started a very apropos discussion of the whole “code is law” concept over at the Volokh Conspiracy blog. I’m having a fine time continuing my Lessig-shredding there…

  4. I think Joe Liu got it right over at Volokh. The buzzphrase “code is law”, by itself, isn’t helpful. It’s not literally true, and it doesn’t even capture the valuable part of Lessig’s argument. But Lessig has a point nonetheless. Many people use “code is law” as a shorthand for what they interpret Lessig as saying. To foster a useful debate, it’s probably more useful to engage Lessig’s argument than his slogan.

  5. …And I think Orin Kerr (and I, of course) got it right in response. The phrase, “code is a regulator”, from chapter 2, may be a better, less aggressive statement of Lessig’s thesis. But if that’s his thesis, then it’s a very uninteresting one, for the reasons Orin and I stated over at the Volokh Conspiracy. After all, everything is a regulator in that sense, including every other product of human engineering, each of which ends up constraining our behavior in various ways. And law, politics and custom deal with them all, all the time, without any breathless law professors having to warn us time and again about these rather obvious effects.

    As I stated over at the other blog, I think what motivates Lessig and his supporters is not the unremarkable observation that code design constrains the way people behave, but rather their vision of the remarkable potential of code to remove constraints on behavior, which they see as unique and enthralling in the same way as any other enthusiast does about his or her passion. Hence the constraints imposed by code design, which the rest of us take in stride just as we do the constraints imposed by car design, house design, or kitchen design, strike Lessig et al. as an evil obstacle to their heavenly vision of unlimited, well, “freedom to tinker”.

    To re-use my analogy from the other blog, they’re like car enthusiasts railing against the constraints imposed by modern automotive technology–on-board computers and the rest–which make it much more difficult for them, as amateurs, to fiddle endlessly with their engines the way they used to. The rest of us may sympathize, and we might even rally to their defense–if the cost to us of doing so isn’t too great. But we’re not likely to consider the issue nearly as important as they do, because viewed from outside their community, it really isn’t.

  6. I see your point Simon, though as a EE/Coder-type and computer-tinkerer I sympathize with LL’s sentiments. I would offer a rebutle by pointing out that the internet is becoming more and more interlinked with important concepts like … free speech, freedom of assembly, etc. This makes the constraints mentioned by Lessig in connection with Code relevant in a broader context than the design constraints you mention regarding automobile hobbyists — your analogy breaks down (IMHO) when the internet becomes more than a hobby (which i predict it will be for basically everyone on the planet eventually). Imagine if newspapers or TVs or radio, back in the day, had certain corporate-controlled constraints that colored the way news was presented and distributed; I think this clearly would be a concern for everyone.


  7. edit: rebuttal
    darn, I wish my keyboard would auto-spellcheck.

  8. Will, I agree with you that the Internet is more than just a hobby, and that the constraints imposed by its architecture might therefore turn out to be quite important sometimes. But exactly the same thing could be said about cars. The fact that the Internet is in some sense made of code, while cars are made of steel, plastic and glass, just doesn’t make that much difference, as far as I can tell.

    Now, we do regulate car design and manufacture, in all sorts of ways, for all sorts of reasons–safety, for instance, or pollution control. But we don’t think of this as freeing us from the constraints surreptitiously imposed on us by the corporate designers of unsafe, smog-belching cars. Rather, we look at each aspect of car design, and ask, “is it worth the costs and burdens of imposing legal restrictions on the choices of car designers and manufacturers, in order to achieve some societal goal that might not otherwise be met?”. Usually, the answer is “no”, but occasionally, it’s “yes”.

    The same process applies to software or the Internet. We can ask, as a society, “what do we want from software and the Internet? Is it feasible to obtain? If so, what would it cost–in the general sense–to force its architects to give it to us? Can we get it anyway, by other means–the market, for instance–without having to use force?”

    I have no compunction about criticizing the current Internet architectrue, as anybody who reads this blog knows. But I don’t understand why that’s different from criticizing current car design, or house design, or anything-else-design. Lessig is entitled to lambaste various current or future proposed features of Internet or software architecure. (And I am free, I hope, to ridicule his complaints mercilessly.) But I think his decision to harp on the supposed special nature of “code”, as if it were fundamentally different from any other product or service, was deeply misguided, based on a fundamental misunderstanding, and responsible for creating a red herring the size of a humpback whale.

  9. Ellen M. Rigsby says:

    Dan wrote:
    The same process applies to software or the Internet. We can ask, as a society, “what do we want from software and the Internet? Is it feasible to obtain? If so, what would it cost–in the general sense–to force its architects to give it to us? Can we get it anyway, by other means–the market, for instance–without having to use force?”

    I thought Lessig’s point is that code is now getting written in a way that elminates the process that would ask this question. One might ask, Ala Dan’s comments last week, whether such a question would ever really reach more than an aristocratic few, but the ways that Congress has pursued intellectual property law (I’m thinking DMCA for example) has eliminated the possibility to question how DeCSS might allow me to play dvd’s on my linux machine, even when I bought them. Perhaps not the biggest free speech issue, but not the smallest either. I would say that US Law in its current form isn’t very good at distinguishing between types of content without effectively censoring. The interpretation of the First Amendment that protects all speech allows for the kind of questioning that we want.

  10. Ellen, are you arguing that “the ways that Congress has pursued intellectual property law” that are unsatisfactory to you are somehow a consequence of the nature of software? I can think of lots of reasons why the battle over intellectual property law might be going the way it is, but none of them have anything to do with the properties of software. Indeed, strikingly similar battles were fought over cassette recorders and VCRs–items which have little or no software in them.

  11. I had written a very nice, relatively lengthy reply. However, I didn’t enter the proper verification code and it was lost.

  12. Code can be a crappy law.

  13. Exactly my point.

  14. To me, the most interesting of Lessig’s four stories is the one about the dog and the poisonous plant in a virtual world. In a virtual world, code really is (physical) law, in the sense that code defines what is possible. This raises some interesting issues.

    There is plenty more to be said about this story. Here is one example. Rather than asking why Dank doesn’t make a pain-free dog, we can ask why he doesn’t make a dog that is immune to poison. That should be possible, right? If Dank makes a poison-free dog, why doesn’t Martha make a meta-poison that kills even poison-free dogs? And so on.

    Who is going to win this arms race? It all depends on how the virtual space is designed — on the detailed design of the interface that controls interactions between plants and dogs. The interface must inherently limit at least one of the parties, in the sense that Martha cannot have a perfect dog-killing plant at the same time that Dank has a perfect plant-proof dog.

    Part of the trouble with the simplest understanding of “code is law” is that it assumes that code is written by some distant person. When the participants in the story can write code themselves, we can’t just ask everybody to code their way out of the constraints on them. When goals conflict, everybody can’t win, no matter how much code gets written.