We conduct a case study of a laboratory experiment involving a group support system and explain how it went awry. We take the perspectives of the experiment's human subjects and the researchers themselves as the basis on which to interpret what happened in the experiment. We interpret the researchers as imputing, to the human subjects, the conduit model of communication and the calculator model of human information processing, which together constitute an instance of Ricoeur's hermeneutic world behind the text. We interpret the human subjects as importing, into the laboratory, their socially constructed world of personal friends, their histories and even their popular culture a world that is an instance of Ricoeur's hermeneutic world in front of the text. We explain the experiment's going awry as following from the researchers' not accounting for, much less being aware of, the disparity between the two worlds. In taking the human subjects and the researchers seriously as human beings, we make recommendations about how such experiments might be better conducted, particularly in information systems research.
