The Navajo Sound System (review)

Let me begin this review by reminding readers of the distinction between documentation and description:

The aim of a language documentation is to provide acomprehensive record of the linguistic practices characteristic of a given speech community … This … differs fundamentally from … language description [which] aims at the record of a language … as a system of abstract elements, constructions, and rules …

(Himmelmann 1998:166)

Accepting this distinction, The Navajo sound system (NSS), although mainly descriptive, is also documentary in effect, because it is based on digital recordings (which one assumes are securely stored and archived) and because, to my mind, its many spectrograms serve as both documentation and description simultaneously, inasmuch as they provide a record of the language that bypasses the researcher’s ‘impressions’ in a useful way. NSS is an excellent description of aspects of Navajo sound structure. Although it could have gone further as a documentary work, providing, say, an accompanying CD with the sound files used in the description or a web address from which they could be downloaded, this is by no means an omission unique to NSS. It is a fact about most research reports on endangered languages. The answer to the problem is not to chastise fieldworkers (or the author of this important study) for failing to add documentation to their past goals, but to change the way we linguists generally think about field research and to underscore the importance of teaching new generations of field researchers in field methods classes about the need to make both documentation and description goals of their field research. Let me say first, however, why I think that NSS fills an important gap as a descriptive study, before getting into the specifics of the book. A bit of personal history might help to make the point.

When I began my own field research, the cassette tape recorder was cutting-edge technology (at least to me). As a poor student I tried to spend as little as possible on my recorder and I used recycled cassettes. My first publications as a linguist were on the sound system of the Pirahã language, describing a sui generis segmental inventory and stress pattern (Everett 1982, Everett & Everett 1984). Today if you asked for my original tapes, I am not sure I could find them. But if I did, they would be muffled and of little use. And in this I believe I am not alone. My articles on Pirahã were accepted for publication without any request by any reviewer to actually hear or see spectrograms of my data. Years later, Peter Ladefoged came to Brazil and recorded data related to my phonological analyses of Banawá, Wari’, Pirahã, and Oro Win. Those recordings, all high-quality digital recordings, are safely and carefully preserved in UCLA’s linguistics department. Ladefoged’s visit to Brazil was part of a National Science Foundation sponsored research program that he and Ian Maddieson undertook to document the sound systems of human languages and to teach their fellow linguists more about the importance of careful recording and phonetic analysis as part of the core tasks of field linguistics (see e.g. Ladefoged & Maddieson 1996 and also Ladefoged 2003). They raised the bar in descriptive work, something I would almost label a paradigm shift.

Joyce McDonough, during her time as a postdoctoral researcher at UCLA, was influenced by Ladefoged. This influence shows in the conception and execution of NSS. And just as the general linguistics community owes Ladefoged and Maddieson a tremendous debt for their contribution to the improvement of standards of language documentation, the community of Americanists owes adebt to M for NSS. I do not think that any field researcher on American Indian languages should now claim to have seriously documented or described a language unless they have provided a record of the spoken speech of that language through high-quality recordings, in a long-term storage format, and careful phonetic support for the sound-based analyses in the...