If on a winter’s night two researchers…: a challenge to assumptions of soundness of interpretation

This paper reports on a quasi-experiment into triangulation, which is increasingly frequently cited as a guarantor of validity and reliability of findings. The methodology that we are exploring is the increasingly widely employed combination of corpus linguistics and (critical) discourse analysis. It has been argued that corpus approaches can offer greater objectivity because they are data-driven (or at least data-supported), more generalisable as they are based on larger samples, and more transparent given the research may be replicated on the same data. In order to explore the extent to which integrating corpus approaches may contribute to the stability of interpretations the authors set up an exploratory experiment. We attempt to answer the question: would two researchers starting with the same corpus and research question and (broadly) theoretical / methodological framework come to the same/similar conclusions