Human rights is the global currency of modern politics. Whenever the United States attacks a country, diplomatically or physically, it cites human rights claims. And by a not-so-surprising irony, the critics of the United States and its allies complain of human rights violations as well.
So human rights workers should be universally feted and supported. Instead, however, they are chronically underfunded, goaded to justify every detail of their work, and threatened with dire harm.
For these reasons, human rights work requires free software.
I heard this unusual call for free software (and I think it's obvious in this context that the proper term to use is "free" and not "open source") in a speech on October 5th by Dr. Patrick Ball, the spirited and plain-spoken deputy director of the Science and Human Rights Project of the American Association for the Advancement of Science.
Patrick is best known for the eight hours of testimony he gave before an international criminal tribunal at The Hague to show that Serbian atrocities, and not NATO bombing or Kosovo Liberation Army aggression, were responsible for the mass deaths and displacement of Kosovo Albanians. His evidence was drawn from a statistical correlation of many sources, including thousands of interviews made by three human rights organizations of fleeing Albanians, sometimes within hours of the killings and destruction they witnessed.
Patrick's venues have ranged from Haiti to Sri Lanka; at his talk, he discussed recent visits to Guatemala and Sierra Leone. His gruesome specialty lies in accumulating many individual stories of death, torture, and terror; correlating them to determine their degree of consistency and reliability; and running statistics that show patterns over time and geography.
One is struck by the incongruity between this horrific material and Patrick's affable, down-home manner, but his dedication to ripping away the masks of the world's evil and vindicating the memories of the victims comes through clearly. As for his believability, an audience member with a doctorate in statistics told me later, "This talk was one of the most compact, yet clearest, presentations that I have ever heard on what statistics can and cannot do."
There is a good deal of overlap between Patrick's work and the mission of Computer Professionals for Social Responsibility, which has recently started a project called Privaterra, which provides software support to human rights workers. The creator of Privaterra, Robert Guerra, knows Patrick and invited him to speak at the 2002 CPSR annual meeting. I recently wrote a Weblog on the conference.
So I was familiar with the subject of Patrick's work when he started his lunchtime keynote, and was too busy stuffing my face at the start to even bother taking notes. But my thoughts really started churning when he unexpectedly started giving accolades to free software. They brought a new urgency to old debates. Here are his observations.
Accountability and Verification
Human rights workers, who hold powerful forces accountable for their behavior, need to be accountable themselves for all of the data and opinions they offer. This accountability extends to the software they use. And only free software can meet that requirement.
Imagine an American scientist bringing a closed, proprietary encryption program or statistical package to political activists in a foreign country and saying, "Just use this; take my word that it works right." That's a non-starter. If the software is open source, even though the human rights staff might not be able to personally verify that it's accurate and free of bias, they can take the source to a university or other expert and have it vetted.
The same challenges arise when a human rights organization publicly presents its results. The politicians, generals, and other power-holders will dispute every step in reasoning. A lot of an organization's credibility lies in its process for collecting data and its use of statistics, but the software has to be certified to be trustworthy, as well. An open package whose source can be checked by any technically qualified person removes a potential area of dispute.
(As an aside, this consideration shows why it's a good idea to use free software for any public or governmental functions -- most of all for elections, where the reliability of any software solution is questionable in the first place.)
Pages: 1, 2