Professor Larry Solum has an interesting article up called, “The Layers Principle: Internet Architecture and the Law.” This post does not cover the entire article. I’ve described some bits (in outline form) and then pointed to one particularly interesting point. I will probably come back to the other parts of the article later (I might also clean up this post a little, because it’s kind of fragmented).
The Layers Principle – respect the layers
What are the layers? A little different than Benkler’s, made for the narrow context. These are the content, application, transport (TCP), internet protocol (IP), link (between user’s computer and physical layer), and physical layers. Communications move from highest to lowest layers, then back up again.
What does it mean to respect the layers?
1. Keep them separate.
2. If you’re making policy to affect one layer but you are directly altering the functions of another layer, minimize the distance between the two layers. (Ex. Say you’re blocking amazon.com. You’re trying to affect the content layer, but you go after the IP layer. You can block amazon.com in many different ways, but you’re impairing the functioning of IP.)
Briefly: Fit between means and ends (otherwise, overbreadth and underinclusion), and transparency on the Internet
Reviewing features of Internet
Lessig’s code thesis
Tranparency – network does not “associate data packets with application file types”
Solum is positing that layers, not E2E, is the key feature that brings about transparency.
E2E does follow from layer separation, but layer separation does not flow from E2E.
Transparency is a direct result of how the layers treat each other – if layers are separated, accepting and not modifying the info, then it will be transparent.
“Layers principle and its corollaries should be viewed as establishing a framework for analysis of alternative policies for Internet regulation that incorporates a strong presumption against regulations that violate the layers principle.”
Wants to stop incrementally violating principle more and more (“incrementalism”).
But, doesn’t want this to be an absolute rule.
There’s a lot of justification here, and a lot of interesting ideas about how this principle could work into legislative and judicial analysis. There’re also some great case studies. I won’t go into that right now – the article is quite long, and I’ve got other stuff to describe for you.
Solum moves back to Benkler’s formulation at the very end, because there are some regulations aimed at Benkler’s layers that need to be compared to regulations in Solum’s framework.
Take the DMCA, for example. This is a regulation targeting the content layer by going after a lower layer (code). Solum says the transparency requirement doesn’t apply in the same way, because Benkler’s layers are more descriptive categories rather than “an implementation of design principles.”
Then Solum moves into “A Layers Approach to the Substantial Non-Infringing Uses Doctrine.” He notes that the Sony principle can be associated with the idea of ensuring a good fit between means and ends that applies when using his layers principle. You can also think of as e2e, Solum argues, because “unless there is near identity of the problem and the technology, the law should not target the lower technology layer in order to regulate the problems at the upper layer.” But, he says that these principles should be distinguished from the layers principles because, unlike with TCP/IP, substantial innocent use is not an inherent design feature of all of Benkler’s communications layers. Additionally, he repeats that there’s no transparency requirement.
The point Solum then makes is pretty interesting. Problems of fit between means and ends are “even more profound than the problems of fit that undergird the Sony rule. To these problems of fit are added the even more compelling problems identified by the transparency thesis. In other words, the underlying rationale of Sony applies more strongly to layer-violating regulations in the TCP/IP layers than it does on the facts of Sony itself.”
Thus, you can find this layers principle already embedded in case law – that’s a neat idea. It’s particularly interesting if it actually has weight in other areas, outside the copyright context. As noted, Solum does suggest ways that the layers principle could become part of judicial reasoning, and this seems like a good in.
This point is particularly interesting to me because it comes from the same professor who has helped me understand the importance of distinguishing between a case’s dicta and its holding. When you try to glean precedent from a case, your assessment will radically change depending on whether you look at just the narrow holding, a broader version of the holding, or the general reasoning itself.
Whether or not this reasoning would actually work, it’s interesting to think about how “the normative basis for the layers principle is already anchored in the deep structure of American law.” I wonder: where else, outside of Sony, can we find it?