I once listend to a Creative Commons Salon titled: What Does it Mean to Be Open in a Data-Driven World? and in that discussion there was a great discussion on what it means to have data which flows and is openMinute 50 has a really interesting comment about sharing scientific data.

http://blip.tv/creative-commons/creative-commons-salon-mountain-view-what-does-it-mean-to-be-open-in-a-data-driven-world-4725230

Somewhere in that hour one of the speakers made a remark about when data is missing from a dataset, the missing data is more evident, or noticeable than if the data had actually been left in the dataset. This got me thinking over the last few moths since I saw this Salon, is it really the data which is important? My conclusion is that, no the data is not really what is important in dataset. The data has some value but what is really valuable is how the data is connected to itself or to other datasets. It is the relationships around the data which makes the data valuable. This sort of relationship between value and objects reminded me about the example of the fingers in movie Patch Adams.

One of the implications I have been thinking about has been the implications on Language Documentation Data. In a way this kind of data has many if not all the challenges of being scientific datasets.

But if it is not the data which is important, but rather the relationships to and among the data, then it would seem that simply charging people to be able to access data would be an example of data stewards selling themselves short of their true potential. The question that data stewards should be asking is not How safe can we keep this data, or How much can we charge for access to this data, but rather What relationships can I find out about because I have this data and am making it progressively more useful to various kinds of people?