What if there is no clear set publisher? Especially if the original data could be considered to be "Common Domain" in the first place, such as a thread in a bulletin board or discussion group, such as perlmonks. Look at thepen.com, he is republishing it, but he is no way claiming ownership or and rights pertaining to it, he is merely providing a different interface to the same data. I agree this arguement starts to fall apart when you start to consider semi-commercial websites, ones that pay authors to write articles for them, in which case you would need to make more of an effort to attempt to contact the original author.