Abstract/Summary

This paper takes four frameworks for understanding linguistic politeness (Brown and Levinson, Watts, Terkourafi, Hall) and tests each on the same corpus to see whether they yield results that are useful and/or in keeping with the other information we have about the material. The corpus used consists of 661 polite requests made in letters by a single Roman author, Cicero. The results demonstrate first that politeness theories are helpful as explanatory tools even in dealing with very well-known material, and second that no one theory is best: different theories are more and less useful in answering different questions about the data. It is therefore suggested that the use of multiple frameworks will provide the best understanding of the data.