The discussion in Brown's paper about these computer generated "personal" messages, and the way they can be attributed to software, which was written by people, reminds me of Roundtree's concept of rhetorical agency. This example is somewhat analogous to the simulation anecdote Roundtree uses. It stands to reason that, since the programs that spit out these messages were ultimately written by people, those people -- along with the computers and software and all other participants, human or non-human -- share rhetorical agency in the "personalized" message that a potential voter would encounter. And it doesn't stop there. The people who share rhetorical agency in this event can also be those who contributed to developing the voter database that resulted in a potential voter receiving one of these personalized messages (which could include door-knockers, phone bank workers, and many others), as well as the designers of all respective technologies which made those things possible. There is essentially no end to the chain of people, places, and things that share rhetorical agency in any given situation, it seems.

Furthermore, the stated situation also seems to apply procedural rhetoric to a very broad and old concept like knowing one's audience, which was emphasized by Aristotle, Longinus, et al.