Please use this identifier to cite or link to this item:

Beyond the text: Diversifying documentary corpora

File Size Format  
42012.pdf 11.86 MB Adobe PDF View/Open
42012.mp3 28.75 MB MP3 View/Open

Item Summary

Title:Beyond the text: Diversifying documentary corpora
Authors:Heston, Tyler
Contributors:Heston, Tyler (speaker)
Date Issued:03 Mar 2017
Description:The past two decades have seen an explosion in the quantity of documentary materials available on minority languages from around the world, including especially documentation of naturally occurring speech events, or texts. Texts have proven themselves a valuable and versatile basis for research, though like any type of data, they are not equally suited to all types of research questions. In this presentation, I draw on my own experiences of fieldwork in the Pacific to articulate a case for the value of experimental data, and in particular, the crucial importance of controlled phonetic data as part of any well-rounded corpus.
One of the key challenges to phonetic research is the sensitivity of phonetic variables to subtle differences in environmental factors (Gordon 2003, Maddieson 2001). Controlling for all possible confounding effects arising from adjacent segments, syllable position, prosodic position, and speech rate in a corpus composed of naturally-occurring data alone requires a prohibitively large corpus (Maddieson, 2001), though through careful elicitation design, data of equal or better quality for phonetic research can be obtained with a fraction of the time and effort.
A key methodological tool for studying phonetics is the use of frames—a phrase that is relatively neutral semantically, allowing a large range of lexical items to occur in the same position (e.g., “Say ____ again”). Recording minimal pairs in the same frame is a simple but powerful technique allowing one to control for both segmental and prosodic organization simultaneously, allowing a true comparison of the sounds in question. The use of controlled data is perhaps even more crucial for studies of prosody than segmental research, given the proliferation of segmental, morphosyntactic, discourse, and even affective factors that influence the prosodic contour of each utterance. Elicitation of short, phonetically-controlled sentences has proven itself as particularly useful strategy for prosodic research, both in studies of word prosody (Yakup & Sereno 2016, Garellek & White 2015) and research on intonation (Himmelmann & Ladd 2008, Jun & Fletcher 2015).
While naturally occurring texts are an immensely useful and important aspect of any documentary project, they do not invalidate the benefits of other types of data. Phonetic elicitation contributes directly to the goals of documentary linguistics to create diverse and lasting language records, allowing exploration of aspects of a language that remain otherwise intractable. For these reasons, the inclusion of controlled phonetic data is a necessary and profitable component of any documentary corpus.
Garellek, Marc & James White. 2015. Phonetics of Tongan stress. Journal of the International Phonetic Association 45(1). 13–34.
Gordon, Matthew. 2003. Collecting phonetic data on endangered languages. Proceedings of the 15th International Congress of Phonetic Sciences, 207–210.
Himmelmann, Nikolaus P. & D. Robert Ladd. 2008. Prosodic description: An introduction for fieldworkers.
Appears in Collections: 5th International Conference on Language Documentation and Conservation (ICLDC)

Please email if you need this content in ADA-compliant format.

Items in ScholarSpace are protected by copyright, with all rights reserved, unless otherwise indicated.