| Summary: | In this paper I will explore the limitations of using a single methodology
in language description and documentation. Specifically, I will argue as others have
(Chelliah 2001; Sherzer 1987; Urban 1991), that multiple methods of data
collection are necessary in order to adequately describe a language. To illustrate my
point I will examine the distribution of ergative case marking in Darma. Described
in early sketches as having a split-ergative system, I demonstrate that the pattern of
distribution for the ergative morpheme is not obligatory and that it can appear in a
variety of contexts that extend beyond the limits described in the split-system.
Using data obtained through multiple methodologies, I show that the use of the
ergative is not only optional, but its use may serve a pragmatic function that is not
yet fully understood.
|
|---|