This research paper focuses on the "bad data" problem that exists in historical pragmatics and discusses whether epistolary discourse might help to overcome it. The "bad data" problem is defined as a problem encountered by researchers in historical pragmatics that have to rely on written data in order to research spoken language in past eras. The goal of this research is to examine if epistolary discourse is adequate in overcoming this problem. The corpus is based on the Lettere di Levante (Letters from the East), a collection of 88 letters written in the first half of the fifteenth century. In this research, we focus on one letter dated July 14, 1428. In order to determine the aspects of spoken language found in the letter, the frameworks devised by Jucker (2008), Koch and Oesterreicher (1985), and Culpeper and Kytö (2010) were applied. The results seem to suggest that this particular corpus contains examples of spoken language, albeit a particular subtype -diplomatic discourse.