Freedom and Restrictions in Open Data

Very interesting situation developed last month within the Russian community of OpenStreetMap. It is illegal to map the location of military facilities in Russia. In fact, a permit is required to do any mapping. The Russian community started a vote (well, in practice and discussion, a non-binding poll) on whether military installations should be removed from the Russian map, to protect the safety of the local community.

Response from the international community was that any mapper has the choice to map whatever they like, especially if it physically exists. Deleting someone else’s data would be against core principals of OSM, of open data.

As Frederik states, this puts us in an ideological paradox

With Russia we now seem to have a case where it is possible that a
majority of the local community decides to work differently from the
principles that we have grown to cherish, namely letting everyone map
what the hell they want. We now have people who not only want to keep
certain things out of the map, but even delete these objects where they
have been mapped by others. It is even possible that they form an OSMF
chapter supporting these actions.

If this happens, then we have a clash of two principles. Either we say
that the principle of subsidiarity overrides the “freedom to tag”,
allowing the Russians to restrict that freedom in their area because it
is, after all, their area; or we say that the “freedom to tag” is the
overriding principle and we will not tolerate any local community to
tell us outsiders how they would like to be mapped.

This could have repercussions for mapping in many restrictive authoritarian regimes. It’s my position that for the vast majority of cases, open data is better and the default, and in fact, makes us more secure and informed. This was one critique of our mapping in Gaza, that showing details of Gaza would make Gazans more vulnerable. Fortunately, Gazans also agree that information is all about changing perception.

In general, I view these edge cases as a question of power. Hiding information protects those already in power, but not those that are already marginalized. Legitimate cases to me is only information that puts dis-empowered people at risk, such as refugee routes along the Burmese-Indian border. But does this mean I would remove those tracks if someone added to OpenStreetMap? This and other questions remain to be answered.

4 thoughts on “Freedom and Restrictions in Open Data”

  1. You also have the Chinese planning to start enforcing their mapping regulations, and they don’t stop at military locations, although as the blog author of this post he correctly points out that if civilians have access to this information multinational corporations and foreign governments had it long ago, so the crack downs are more likely to be protectionism of local industry and government owned entities selling satnav data and equipment.

  2. A few years ago, I had a similar discussion about mapping petroglyphs in the southwest USA. Aboriginal American sites are frequently vandalized, especially in very remote locations. Many of those sites are sacred places. It one of the reasons many tribes maintain their own GIS databases under tight security.

    I guess it depends on which side you are on. If you are on the Russian Military side, then you are concerned about protecting your bases much the same as the American tribes are concerned about protecting their sacred places. Because we are not on the side of the Russian Military (and we view the Military as oppressive in nature), we feel mapping the bases is an activist expression.

    Maybe we need anonymous accounts that can add contentious data which can be easily filtered.

  3. @Eric unless you are in Russia you most likely aren’t breaking any local laws, and users can filter out any information they like from the database, as for culturally sensitive sites, I’m not sure what to do there, although in terms of vandalism this is no different than people going into cemeteries and knocking over headstones, which also happens, so I guess we shouldn’t map cemeteries for the same reason?

Comments are closed.