Today I learned a bit about minor mistakes that one can make with social software. One important thing is language. I just learned that the basic assumption that all members of a community share the language for their contributions does not hold if the community contributions are assembled through private blogs or other web2.0 sources. I hoped that within team, or for that matter a very small community, there is some kind of interest of a shared language to exchange information. However, although there might be shared interests, not all members share the same language in their possible contributions. In order to achieve valuable contributions that are accessible for all members of a community, a software must indentify the language of the contributions, and decide which information should be used.

This is not a very new insight, but it raises some implications for experimental settings which use web2.0 techniques. First, the experimental group has to be carefully selected, so the web2.0 concepts and the social concepts hold. Second, the used tools have to be carefully investigated with regard to these factors, too. Finally, the experiment can not be run on its own but requires constant monitoring in order to take care, if user actions are within the experimantal constraints.