As part of the Guardian’s ongoing Battle for the Internet series, looking “challenges facing the dream of an open internet”, it published an interactive map purporting to rank the level of government interference over Internet access around the world, with data sourced from the OpenNet Initiative (ONI). Bizarrely, on Sri Lanka it notes that “The ONI has given Sri Lanka a clean rating in terms of internet filtering. According to the World Bank, 12% of the country was online in 2010.” Emphasis mine. What follows is an excerpt from an email penned to the principal investigators at ONI, in addition to Ethan Zuckerman.
Ronald J. Deibert
Director, The Citizen Lab, Munk Centre for International Studies, University of Toronto
John Palfrey
Faculty Co-Director, Berkman Center for Internet & Society, Harvard University
Rafal Rohozinski
CEO of the SecDev Group and Psiphon, Inc., and Senior Research Advisor to the Citizen Lab
Jonathan Zittrain
Professor, Harvard Law School
Co-Founder, Berkman Center for Internet & Society
###
I read with interest the Guardian’s Battle for the Internet feature, which runs an infographic with data sourced from ONI, noting explicitly that Sri Lanka “has a clean rating in terms of internet filtering”. This is in fact very far from ground realities.
- http://bundlr.com/b/web-censorship-in-sri-lanka is an overview of the degree to which the ‘net is monitored and controlled in Sri Lanka.
- http://www.scribd.com/doc/88102850/Arbitrary-Blocking-and-Registration-of-Websites-The-Continuing-Violation-of-Freedom-of-Expression-on-the-Internet is a statement issued by the Centre for Policy Alternatives (CPA) on the arbitrary blocking of websites by the government.
- http://www.scribd.com/doc/83503200/Freedom-of-Expression-on-the-Internet-in-Sri-Lanka is a detailed report, also by CPA, on the degree to which FOE is curtailed in online spaces in Sri Lanka.
We have flagged our concern via @groundviews on Twitter, because it is important for us that vital indexes like ONI, when republished via a respected masthead like the Guardian, reflect far more accurately the challenges we face in the real world, and through erroneous analysis arising from significant gaps in data collection, don’t go on to strengthen a government in Sri Lanka that is demonstrably repressive, and indeed, violent.
###
As Nigel Nugawela, the co-editor of Groundviews and lead researcher of CPA’s report looking at FOE online noted over email,
“Also very problematic. I hope they re-score or issue a statement. If the government was competent enough, this is the sort mistake that they could use to buttress their claims that there is FOE on the Internet in the country or highlight the unreliability of indices produced by international organisations.”
Three tweets on ONI’s absurd analysis published via @groundviews follow. Going by comments on the Guardian’s website, it also appears that it isn’t just Sri Lanka ONI has got horribly wrong.
BIZARRE! @opennet says no 'net filtering in #srilanka in @guardian guardian.co.uk/technology/dat… But see record bundlr.com/b/web-censorsh… @EthanZ #lka—
Groundviews (@groundviews) April 17, 2012
1. @opennet analysis of #srilanka in @guardian wrong guardian.co.uk/technology/dat… See @CPASL statement scribd.com/doc/88102850/A… on 'net filtering #lka—
Groundviews (@groundviews) April 17, 2012
2. @opennet analysis of #srilanka in @guardian wrong guardian.co.uk/technology/dat… See @CPASL report on 'net censorship scribd.com/doc/83503200/F… #lka—
Groundviews (@groundviews) April 17, 2012
