In this project I assess the feasibility of using signature lists from internet petitions to characterize acquaintance networks.  Such petitions circulate widely over email (although they seem to have begun to wane as more people are scolded for forwarding them) and often include hundreds or even thousands of sequential signatures. While they are inefficient as petitions (since they include more duplicate than unique signatures if/when they are returned), by assuming that adjacent signatories know one another (with appropriate corrections for those who sign multiple names), each petition can be treated as a tracer sent through a social network.  Multiple petitions can be sequenced matched to produce a large tree structure which is a sample of a diffusion network.  This structure can be analyzed statistically and visualized graphically.  Comparing this data to population and geographic mobility data yields some insight into how the internet changes and recapitulates geographical separation and physical mobility patterns.

This paper first describes the methods of data collection, cleaning, maintenance and analysis. This is followed by a review of the mathematics of petitions and networks relevant to analyzing this data.  The results of the analyses of several families of petitions are presented both statistically and graphically using molecular modeling software.  Comparisons are made between the insights produced by this data and studies of virus propagation.  Finally, I suggest the limitations on inferences that can be drawn from such analysis and some thoughts on lessons it offers for understanding the small world effects offered by the internet.