ABSTRACT
In this chapter I want to pursue the argument that many of the assertions made about semantic prosody in the literature are strongly infl uenced by the way corpus data, and more specifi cally concordances, are arranged for the user by the attendant software.
