The epistemology(s) of volunteered geographic information: a critique