Title: One path or many? Policy development and diffusion across Wikipedia language editions
Benjamin Mako Hill
In my day job, I teach and do research at the University of Washington. I am also an active editor on several wikis and work with the Wikimedia community and the Wikimedia Foundation community to support academic research related to wikis and Wikimedia in a number of ways. Please see my user page on meta for more information.
I, Seth Frey, am a cognitive scientist and computational social scientist who studies human decision behavior in complex social environments. My expertise is in computational approaches to self-governance and the cognitive science of the commons. I am a professor in Communication at the University of California Davis and an affiliate of the Ostrom Workshop at Indiana University. I was a behavioral economist at Disney Research Zurich in Walt Disney Imagineering, a Neukom Fellow at Dartmouth College’s Neukom Institute, and a student at the New England Complex Systems Institute. I earned a Ph.D. in Cognitive Science and Informatics at Indiana University in 2013, and a B.A. in Cognitive Science from UC Berkeley. My work has appeared in New York Times Magazine, TEDx, BBC Radio, Hacker News, and Nautil.us Magazine. It has been funded by the NSF, NASA, and the JSPS.
Room: Room 324
Start time: Fri, 18 Aug 2023 10:20:00 +0800
End time: Fri, 18 Aug 2023 10:30:00 +0800
Type: Lightning talk
Track: Research, Science, and Medicine
Submission state: confirmed
Duration: 10 minutes
Do not record: false
Presentation language: en
Abstract & description[edit source]
Do comparable self-governing collective action institutions converge on comparable policy systems? Do they do so via comparable developmental paths? We test both theories using data on 60 policies shared by 245 Wikipedia language editions. We find the most popular policies are the first adopted. We find more evidence for a common shared path than several.
The Wikipedias are a petri dish for the diffusion of policy in collective action institutions more broadly. How similar are their policy systems, and how similar are the paths they took to find them?
By collecting all cross-linked papers in the project namespace and vetting at least one language version, we are able build a human-vetted, computer legible indication of cross-edition equivalence of policy structures, despite our ignorance of 326 of the 332 languages that Wikipedia is written in. With this, we extracted the order of adoption of 61 policies that we shared across the editions.
We find that policies that are shared tend to be shared widely, that nearly every shared policy can be found in the English edition, and that the clearest predictor of policy adoption order is policy popularity across editions.
Although we do not definitively eliminate the possibility that language editions follow multiple paths in converging on their policy systems, the evidence suggests that editions follow a single noisy developmental path, potentially suggesting strong influence across editions and a stronger role of common structural constraints than diverse cultural constraints in determining patterns of policy adoption.
Our results point to a strong culture of sharing among the most common policies within Wikipedia language editions. Encouraging editions to draw from these highly shared policies, rather than developing their own, may economize on the finite time and energy of each project’s volunteers. This seems to be particularly likely to be true in small projects.
Further details[edit source]
Qn. How does your session relate to the event themes: Diversity, Collaboration Future?
The 300+ Wikipedias are 300 attempts to satisfy one goal. In what way are their policy systems driven by differences in their cultural source and similarities in their common mission?
Qn. What is the experience level needed for the audience for your session?
Everyone can participate in this session
Qn. What is the most appropriate format for this session?