On the Linguistic Properties of Formulaic Speech

In the study of oral-formulaic performance the formula has always been understood as playing a pivotal role (Lord 1960, Foley 1995). But although it is clearly a linguistic unit, the technical study of formulae by linguists has been slight. This paper intends to remedy this lack by proposing some linguistic theories as to the nature of the formula. They will take the form of formal and testable proposals. By formal I mean mathematically modelled. Why place such a study before the readers of Oral Tradition? Two preliminary observations are in order. First, it has become clear over the last twenty years that the use of formulae is not restricted to the performance of oral literature (Edwards and Sienkewicz 1990). The theories developed below draw on the study of vernacular oral traditions and specifically on the author’s work on some of these traditions (Kuiper 1996). It is not intended that this should be a slight on the work of others. It is just that mine is the only work I know of that looks at formulae from a technical perspective (cp. Pawley 1991 and 1992). However, the theories proposed are intended to generalize to formulaic speech used for the performance of oral literature. Second, the work being attempted here is essentially crossdisciplinary between the post-Lord study of formulaic literature and the linguistic study of formulaic speech. As such it attempts to explain the properties of formulae via formalisms used in linguistics. The cost of this is that such formalisms may not be familiar to a number of the scholars working on formulaic speech within a more literary or ethnographic tradition. However, it should be clear that formulae are linguistic units and as such the formalisms of linguistics are appropriate, not to say, the appropriate tools with which to study the linguistic properties of formulae. While there are some costs for scholars who may be unfamiliar with contemporary linguistics in understanding what such tools have to offer, the