Using video in language documentation: a LIP discussion

This is a recap of Linguistics in the Pub held at Prince Alfred Hotel, Carlton on Tuesday the 6th of September written by Lauren Gawne. From now on this will be a regular feature here at Endangered Languages and Cultures.

For the topic of video in language documentation we were lucky to be joined by Joe Blythe (Max Planck Institute, Nijmegen) and Jenny Green (ELDP funded Postdoc at The University of Melbourne), who have both worked extensively with video and both recently returned from fieldwork. Joe started off the session by talking us through some of his data. Joe has just returned from a field trip in Wadeye where he is continuing to collect conversational data. On this trip Joe tried working with some new speakers and some of his regular speakers but in different environments. He found it interesting that a shift in location for people he worked with regularly, for example into a house instead of out bush, would lead to very different behaviour towards the camera. He was very kind to show us not only some of his excellent (and often quite scenic) data but also some of these less successful attempts. Even less successful recordings are interesting in their own way.

Joe uses JVC cameras with XLR inputs for his Sennheiser K6 modular mics, which can either be attached directly to the camera or detached and moved closer to the speakers. Joe talked about how it’s important to be concerned about sound quality, and how the quality of the equipment was important but so was the environment for recording. Joe prefers working out bush where the audio is less affected by echo than it is when working near buildings and concrete.

We also learned some valuable recording techniques from Joe. Using a GPS with a compass allows you to locate the speakers in absolute space – useful for orienting their gestures in later transcriptions. Using a bag of rice as a convenience and multifunctional tripod. Also, exporting the initial transcription from ELAN into a subtitle file for checking the transcription with (literate) speakers.

Jenny’s work provided a nice counterpoint to Joe’s. She uses several cameras at once – a Panasonic HS900 and a Sony CX-550, and a range of mics, including a Sennheiser lapel mic and a Rode NTG-2. Both researchers work with Australian Indigenous languages, and both are interested in recording video, but while Joe is interested in naturalistic conversational data, Jenny’s work focuses on recording verbal art, including sign language (in collaboration with Batchelor College linguists),  and sand drawing in Central Australia. Some of this data, as far as Jenny is concerned, is never likely to be completely natural, as it takes so long to set up the rig.

Jenny showed us the elaborate equipment required for her work – the overhead camera for the sand drawing, the ladder and microphone stand that now replaces the ladder, and the signadome tent or mobile ‘studio’ used for the sign project. She talked about the logistics of collecting high quality video data – and compared the (sometimes) lone linguist to an entire film crew; director, producer, camera-person, sound recordist, driver, set-designer, caterer. She talked about the difficulties in trying to obtain good recordings when both sound and image are given equal status, and about the need to think about the quality of light when filming with as much attention as is paid to worrying about the sound environment.

For Jenny, one of the major issues with working with video is thinking about what the final products will be. As researchers should we be concerned about creating a product that could be returned to the community? With video editing becoming so easy should we be thinking about turning the data into other products? How much importance should we give to how the speakers and the community see the data? How much control should they have over the way their own images look. Often linguists find themselves in situations in the field where the records they make of important cultural activities will be unique and potentially multi-purpose – used for analyses of speech, gesture, drawing, or sign, but also for making films for community and wider consumption.

Opening out to a more general discussion after the presentations lead to some interesting discussion. One of the first points to be raised was what should be recorded? Is it now best practice to video record everything in case it is one day useful to future researchers? How do we deal with that much data on a practical level? In the field such practice eats through storage space, and universities are only just starting to offer staff a sufficient amount of server space to make such high volume data storage viable. There is also the problem of archiving; many archives are still not fully equipped to deal with high volumes of video data.

The discussion also came back to Jenny’s point about what the final product of video recording is. It is becoming more popular to turn video recordings into products for the community and general public, but this may lead to a conflict in collection styles – for example, for linguistic research the consensus is that more naturalistic is better, but for consumable products a different aesthetic kicks in. For example is direct eye contact of an interlocutor with the camera preferred or to be avoided? (Joe thought that one of his sessions was a ‘failure’ because the men he was filming looked at the camera, whereas Jenny deliberately sets the camera up a surrogate interlocutor). Do linguists care about film conventions, like the ‘Law’ of thirds? (in sign language filming Jenny tends to position the signer right in the middle of the frame, and this gives film people the horrors). For naturalist documentation the camera is fixed in position with a view wide enough to capture all people present, and the filmer goes away; if there is an objective of making other types of films a range of different shots is preferred – close-ups, wide shots etc. If several cameras are running at once it is possible to do all of this!

Another point was that video is evolving quickly and it’s hard to keep track of what the benchmark for good video recording is. For example, it was very rare to be shooting in HD only a couple of years ago but now HD cameras are ubiquitous and affordable. In terms of extracting video and storing video several researchers present felt like every new project had to pioneer their own process, unlike audio where the benchmarks and processes have remained rather stable.

Overall it was a good discussion, a great mix of the practicalities of recording video data and the best practices to which we should aspire.

One Comment

  1. I’m really sorry I missed these presentations, as video documentation is a core part of the business of a sign language researcher like myself. I have been particularly inspired by Jenny Green’s interest in the aesthetic of the finished digital video product, and my discussions with her have got me rethinking my approach to this issue. The aesthetics of filming has been a low priority for work on Auslan and British Sign Language (BSL) documentation projects in the past, where we’ve focussed on making the participants feel as comfortable as possible by not having too much of the equipment and personnel associated with film production present at data collection sessions (not too many cameras or too much lighting equipment, no camera operators, no researchers -only deaf community fieldworkers – present during filming etc). This was partly motivated by a desire to avoid style shifting in a more self-conscious form of signing, or towards a variety more influenced by English. You can see some of the results on the BSL Corpus Project website, where the video data has gone live just recently (ELAN annotations and translation files will follow in the future): http://www.bslcorpusproject.org/data. Although we’ve got some great data, I must say that some of it doesn’t look very professionally filmed, and some feedback reaching us from the British deaf community has expressed disappointment about this issue. After all, some feel strongly that, if the language is going to be documented in the form of an open-access online archive, then the data really ought to look visually appealing as well, because it will create a positive impression of the language and the community among visitors to the BSL Corpus Project website.

Leave a Reply