Research:Non-finite Processes in Human Social Phenomena

This page documents a completed research project.


Key Personnel edit

  • Simon DeDeo / simon[at]santafe.edu
  • (postgraduate research students, as recruited)

Project Summary edit

Two papers have been published in the peer-reviewed literature: Group Minds and the Case of Wikipedia; Human Computation (2014) 1:1:5-29. Collective Phenomena and Non-Finite State Computation in a Human Social System; PLoS ONE 8(10): e75818 (2013).

An open-access pre-print on this work is now available at the arXiv.

Below is the abstract of a recent talk delivered at UCDavis to the Complexity Sciences Center.

Non-finite Processes in Human Social Phenomena

Simon DeDeo, Santa Fe Institute

10:30 AM, Wednesday, October 17th

Physics 195

Computational models of social and biological phenomena have often focused on the finite-state case. In this work, I present a mathematical result, the Probabilistic Pumping Lemma, that constrains the behavior of all finite state processes. I then present evidence for violation of this Lemma, and thus statistical evidence for the emergence of non-finite computation, in cooperative phenomena associated with the collectively-edited encyclopedia Wikipedia. I distinguish between psychological, or individual-level, and collective aspects of this violation, and present evidence for the existence of separate Universality Classes for the two cases.

Methods edit

Statistical analysis of public data (revisions, revision timing, revert behavior, linguistic structure of comment fields). We do not do subject recruitment.

Dissemination edit

Public lectures; technical talks; publication in Open Access journals including archiving of versions at [1].

Wikimedia Policies, Ethics, and Human Subjects Protection edit

Per discussions with our local IRB contact, IRB approval is not required for this work.

Benefits for the Wikimedia community edit

Why does Wikipedia work so well? What are the signatures of a "healthy" community? How does the nature of the editing process change as one goes from single-user to collaborative endeavours?

Timeline edit

October 2012 -- October 2013: talks delivered; recruitment of research student; publication of first results. October 2013 -- October 2014: follow-up research; additional release of results as found.

Funding edit

  • Omidyar Postdoctoral Fellowship (supporting PI). Funded by the Pierre and Pam Omidyar Foundation.
  • Research also relevant to National Science Foundation Emerging Frontiers Grant EF-1137929 (Information Theory in Biological Processes); not directly funded on this grant.

References edit

  • Early results and discussion presented at SFI Complex Systems Summer School 2012. See Emergence Lectures.
  • Early results presented at University of California, Davis, Department of Physics (October, 2012).

External links edit

PI Webpage.

Contacts edit

simon[at]santafe.edu