Emacsen's Blog

Edit Wars in OpenStreetMap

Reading the feedback from my Why The World Needs OSM post has been enlightening. One criticism of OSM that has come up multiple times is the concern about edit wars like those seen in Wikipedia. It's a topic that I know something about, both as a long time OSM contributor, and also as a DWG member. This post explores how common edit wars are and the circumstances that surround them.

Before I go any further on this topic, I want to state for the record that the opinions expressed in this post are my own and don't reflect the DWG or the OpenStreetMap Foundation in any way.

OpenStreetMap DWG members operate similarly to Wikipedia Administrators - we handle issue of users breaking community standards and intervene in issues of conflict resolution. We don't enforce any particular mapping conventions, but we do enforce community guidelines.

What are the kinds of editing conflicts seen in OSM?

The most common conflict is around the issue of data imports. OSM is not a collection of information sources, but a single unified dataset where geographic objects interlock. Because of this, and because of OSM's emphasis on primary sources and community mapping, imports are very difficult to do right. In response, OSM has a set of import guidelines which must be followed. When they are not, the DWG steps in and enforces that the guidelines are met.

The second most common place that the DWG steps in is vandalism. Vandalism is purposefully editing the map in a way that doesn't reflect the reality. OSM community members often fix small mistakes by newbies or small scale vandalism. The DWG often gets called in to fix larger scale vandalism, usually by drive-by troublemakers.

The third most common type of case that the DWG handles are around tagging conflicts. Tagging is the process of classifying objects. OpenStreetMap uses a free form folksonony and sometimes disputes arise about how to classify certain features. I hesitate to call this an edit war because it is almost always a single user whose views differ from those in the rest of the community.

What about actual edit wars?

Actual edit wars do occasionally come up, but they almost always involve political disputes. One group believes that an area of land belongs to them, and or has a certain name, while another group believe that they own the land, and or it has a different name and the two sides cannot come to consensus. In this special case, the DWG has special policies laid out from the Foundation in how these delicate situations are handled.

When an edit war happens, the DWG gets involved, but the frequency of a genuine edit war is very low, on the order of one edit war every six months or longer. It's far less frequent than what someone outside the project may imagine.

Why are edit wars so common in Wikipedia, but so uncommon in OpenStreetMap?

Wikipedia edit wars concern two issues, facts and presentation. In particular, Wikipedia relies on secondary sources. When two facts collide, an editor must decide which source is better. OpenStreetMap relies instead on primary sources of contributors visiting an area and editing OSM to reflect what they see. Secondary sources are considered inferior and many OSMers use the derogatory term armchair mapping in reference to using these sources. Because of this, when a conflict arises, the person who visits the location is going to be given deference over any other source, even officially published government data.

As for presentation, Wikipedians may argue over the best way to write a paragraph, or where to list a particular fact. In OSM the general rule is to update wrong information, and to add detail when you can.

I think that our policies around ground observable truth are the main reason why edit wars are infrequent and almost always concern political boundaries, which are the one feature that OSM maps that is not ground verifiable.

The other reason that I believe edit wars are less common in OSM is that OSM has a higher bar required for editing. OSM does not allow anonymous edits, and unlike Wikipedia. If a particular user is causing problems, they can be easily identified.

Additionally, OSM is quite a bit smaller than Wikipedia. Even though OSM has over 1.2 million user accounts, on a daily basis only about 2000 users ever edit the map. While 2000 is certainly not a small number, it's relatively easy to spot a troublemaker. But even then, a single user cannot start a war- someone who changes large parts of the map on their own is an isolated vandal, while actual edit wars would require at least two parties. This virtually never happens.

When another editor makes a change to something I've created, in every case I've seen, they've enhanced the map. Either they've expanded on the data that I entered or they corrected a mistake I made. This is the norm in OpenStreetMap- additive and corrective edits. Edit conflicts are so infrequent that most users will never encounter them, and even to the group that oversees them, they're a special case. It's possible that as OSM grows that the number of edit wars will increase, but I don't think that's likely, as we haven't seen the number of edit wars grow despite the project's growing success over the years. Instead, I think that the source of edit wars, that is the number of geopolitical conflicts is going to be a better predictor of the trend.