A large corpus of recorded oral tradition can be created using two recording machines, one playing back the spoken texts and the other used to capture an oral annotation. Recording speakers who are commenting on earlier recordings is a method for providing annotations that bypasses literacy.
This method was discussed by Will Reimans at the 1st ICLDC last year, and also in an article to appear in the journal Language Documentation and Conservation later in 2010. Tony Woodbury (2003: 45) suggests that our time as linguists is better spent not interlienarising texts, but instead asking “elders to ‘respeak’ them to a second tape slowly so that anyone with training in hearing the language can make the transcription if they wish.”
BOLD PNG
Recently, Steven Bird launched a project using this method, called BOLD: PNG in which he has provided small digital recorders (100 were donated by Olympus) and training in the method of “Basic Oral Language Documentation” (BOLD) for speakers of a number of PNG languages. These recorders are not ‘best practice’ for recording what may be the first and perhaps only recording of texts in a language. They are easy to use, but record in a proprietary format that needs to be converted and have no external microphone. There is a suggestion that they could provide a vector by which later, higher quality recorders may be available for use by those now experienced in the practice of recording, naming and annotating oral tradition recorded in this way. The BOLD project is asking for expressions of support to help motivate the large team of speakers involved in the work, so, if you feel like contributing, go to their guest-book and drop them a line.
Woodbury, Anthony C. 2003. Defining language documentation. In Peter K. Austin (ed.), Language Documentation and Description 1:35-51. London: SOAS.
Follow
I have undertaken similar efforts with Tlingit recordings that I have a difficult time understanding or deciphering. (I have recordings from one deceased individual with a pronounced stutter and high pressure to speak, so I often need help with him.) What I do is get a running ‘respeaking’ of the original, with a fluent speaker listening to a few seconds of the original and then repeating the phrases. I always give them time to comment on each segment, and also ask questions myself. Most Tlingit speakers are illiterate in their language anyway, so this is a very effective way to work with them.
I disagree however with the argument that our time is not well spent interlinearizing. A linguist could fairly claim this if they are discussing a language with relatively simple morphology. But Tlingit and its Athabaskan relatives have insanely complex morphology, and picking apart the meaning of individual words is an extraordinarily difficult endeavour. Having watched language learners try to do so, it’s apparent that they desperately need the assistance that detailed interlinear glosses can give them.
In addition, if the language in question has a good bit of dialectal diversity, or even if the particular speaker(s) recorded have unusual idiolects, then it’s essential to provide a transcription that clarifies what the speakers are saying versus what the ‘standard’ language would have. This is another issue that easily confounds language learners, and quickly leaves them feeling disheartened and discouraged about recovering their ancestral language.
The role of interlinearized texts is being downplayed by a number of vocal linguists nowadays, claiming that other areas of work are more important. I disagree fundamentally, and feel instead that interlinearizing and translating are essential components of what we do both for our own analytical work and for our services to the language community.
(As an aside, the tiny size of the comment window on this site absolutely sucks.)
Myself and some Babar Islander friends in Indonesia are recording narratives using a Zoom H2, retelling and re-recording in slow mode, and also splicing in LWC free translation breath-group by breathgroup for language learners. Anyway, we use Audacity for the splicing and stuff. I would love it if we could get away from the computer though because it is a big bottle neck to sustainability and growth of the documentation/mobilisation program. So I would like to get some hand held recording devices which can also do these kind of edits on the fly. I am looking at the iPod or iPhone ruinning the FiRe app. Have you guys heard of anyone using this configuration?