[OSM-talk] Just facts?

Christoph Hormann chris_hormann at gmx.de
Thu Jun 19 17:46:46 UTC 2014


On Thursday 19 June 2014, Frederik Ramm wrote:
>
> "Ultimately, map data is pretty much fact and whether it exists or
> not is a binary statement. [...]

This is IMO what mapping should aim at, as outlined by the verifiability 
rule - practical mapping and actual data is however often quite far 
away from this goal.

More importantly though i think the most significant way to take 
influence for a mapper is through selection of what to map and what 
not.  Since there is neither a policy nor a mechanism to enforce 
completeness of mapping and it is unrealistic to create such what is 
mapped and what is not represents a huge element of subjectivity and 
possibly bias in the data even if all the data itself is strictly 
factual and verifiable.

Imagine for example someone mapping all the buildings in a town but 
deliberately leaving one building unmapped.  It is much more likely 
then this building stays missing from the database than if the 
buildings had not been mapped systematically but over time by community 
efforts.  And there is nothing factually wrong with such mappping, it 
is just incomplete.

But even without deliberate attempts to influence the mapping like this 
it seems clear to me that paid mapping will focus on different things 
than normal community mapping.  Kind of 'where the money is' vs. 'where 
the people and their interests are'.  If suddenly half of the mapping 
in OSM would be paid mapping you can be certain that the thematic and 
regional focus of OSM would change - and again without any non-factual 
or non-verifiable elements in the data itself.

Getting back to the comparison with Wikipedia i think there are a number 
of important differences to keep in mind when comparing these two.

- In contrast to the data in Wikipedia the OSM data is normally not 
viewed directly, it is used in maps and other services which provide an 
additional layer of abstraction and interpretation and due to the 
diversity of map styles and services available the possibilities for 
someone editing OSM data to take influence in some form are much more 
indirect than in case of Wikipedia.

- One of the major mechanisms leading to bias in Wikipedia is the 
removal of data.  You can only effectively push a certain POV if you 
can actually remove information unfavourable to it.  For OSM this would 
mean simple additions of new data by paid mappers would probably be 
less critical than deletions and modifications like changing tags.

-- 
Christoph Hormann
http://www.imagico.de/



More information about the talk mailing list