Labratorium

The Cyberscholar Working Group, initiated by Berkman Fellow Urs Gasser and ISP Fellows Eddan Katz and Nimrod Kozlovski, is a monthly forum for fellows and affiliates of the Yale Law School Information Society Project and the Berkman Center for Internet & Society at Harvard Law School to discuss their ongoing research. Each session is focused on the peer review and discussion of current projects submitted by a presenter. Meeting alternatively at Harvard and Yale, the working group aims to expand the shared knowledge of young scholars by bringing together these preeminent centers of thought on issues confronting the information age. Discussion sessions are designed to facilitate advancements in the individual research of presenters and in turn encourage exposure among the participants to the multi-disciplinary features of the issues addressed by their own work. At the first meeting of the working group on November 20th, 2003 at Yale Law School Mr. James Grimmelmann http://www.laboratorium.net/ will present his current research.

Regulation by Software

Software, like law and physical architecture, can regulate behavior, but not all forms of regulation are created equal. I will present a theory of software as a "modality of regulation" with an emphasis on understanding what we gain and what we give up when we make the choice to use software instead of some other form of control.

Three key features determine software's nature. Software is automatic: it requires no human intervention. Software is immediate: it prevents rather than punishes. And software is plastic: it can do almost anything a programmer wants. From these three features, we can derive a number of predictions about the effects of using software in a regulatory context. These predictions are borne out when tested against cases in which software _has_ been so used, including multiplayer online games, which represent a field in which software has been used as a pervasive system of control.

We can predict that software favors the consistent application of rules over the exercise of discretion. We can predict that software will, under many circumstances, create higher transaction costs than legal regulation (which can be ignored where it is inefficient). We can predict that software systems will be comparatively fragile. And we can predict that it may be difficult to establish that software is a legitimate source of authority, since every disagreement with software's choices can be rephrased as a complaint that the software is buggy.