Today on the ELAR blog, we celebrate UNESCO’s World Day for Audiovisual Heritage by highlighting a few ELAR collections that not only create new audiovisual documents but also incorporate heritage materials in meaningful ways.
By Sydney Rey
This year’s theme for UNESCO’s World Day for Audiovisual Heritage is “Window to the world.”
“Audiovisual materials as documentary heritage objects provide a window to the world as we observe events we cannot attend, we hear voices from the past who can no longer speak, and we craft stories that inform and entertain.”
The following collections and ongoing projects incorporate legacy materials in such a way that ensures the voices of the past are not lost and that access to these records of cultural heritage and memory are made available both to the communities they are born out of and to the scientific community at large. I was fortunate enough to interview some of the collectors and ask their perspectives on how revisiting and improving heritage documents promotes accessibility and re-ignites cultural memory.
A comprehensive documentation of Panare, a Cariban language of Venezuela by Natalia Cáceres and Marie-Claude Mattéi-Müller
The original goals that Natalia Cáceres and Marie-Claude Mattéi-Müller set out with in planning their ELDP project included training Panare community members, collecting new data, making a digital dictionary based on a paper dictionary and new data from texts, as well as adding transcriptions to existing legacy materials. However, with the advent of COVID-19, their primary focus shifted necessarily to the tasks which could be done in isolation. Namely, the digitizing and re-transcription of those existing legacy materials.
These materials consist of cassette audio recordings and hand-written transcriptions made and collected by Mattéi-Müller in the 70s and 80s. The roughly 25 tapes document conversations between a local Shaman (now deceased) and his son-in-law throughout the son-in-law’s apprenticeship training. They capture a rare and naturalistic speech setting in which traditional knowledge is transmitted from one generation to the next but is now, a mere 40-50 years later, reportedly lost among community members. Of course, the original transcriptions and translations are incomplete and hand-written making the tapes in their present state of marginal use. This is where Cáceres and Mattéi-Müller are making their mark.
By revisiting these tapes, digitizing them, and providing time-aligned transcriptions and translations, they are able to salvage the record of an important aspect of Panare heritage; creating something of immense historical value for the community, as well as scientific value for future researchers. As Cáceres notes, “there are things we won’t know about the language without a better transcription.” Which is to say that, in addition to the cultural content of these materials, there may be linguistic content which is no longer available in the Panare speaker community.
I asked Cáceres about her perspective on the importance of audiovisual materials with respect to her work in language documentation. She said [paraphrased] that they allow others, as well as the original researcher (in the future), to interpret things that the researcher who made the recording might not have taken into account (which would be impossible to do if there was no recording). In oral communities, especially, if one doesn’t take the time to talk and pass on stories or history, then it is not transmitted to the next generation – this is how these things are lost. Therefore audio and visual recordings are essential if person-to-person transmission isn’t happening. Additionally, from a research perspective, thorough recordings give us better accountability in our work.
While digitizing previous audio recordings promotes data accessibility in its own right (especially when dealing with cassette tapes which have a shelf-life of 30 years – but only when properly stored and cared for), I also asked Cáceres about the importance of data accessibility for community members. To this she recalled a past project in which she was able to bring previously taken photographs of that community to present-day speakers and found that people were captivated by them. They would say, “People don’t dance like that anymore! People don’t do that anymore!” For them, these materials are a window into their history and making them accessible to the community reignites collective cultural memory.
When COVID-19 loosens its grip on the planet, Cáceres and Mattéi-Müller hope to continue their project as planned with field trips to the Panare community and work with Panare speakers. Specifically with the son-in-law of the Shaman from the original recordings.
Documentation of Negidal, a nearly extinct Northern Tungusic language of the Lower Amur by Brigitte Pakendorf and Natalia Aralova
With the primary aim of building an extensive corpus of interlinearized texts together with accompanying audio recordings and some video recordings of Negidal, Pakendorf and Aralova encountered and incorporated two types of previously recorded heritage materials to enrich their collection: 1) a small corpus of audio recordings made on cassette tapes sometime in the 1990s, some by a native speaker and others by an ethnologist and a linguist; 2) a set of more than 15 hours of recordings collected between 2005 and 2010 by a group of fellow researchers.
These recordings, though relatively recent, were of marginal value due to the fact that they were not fully transcribed, glossed, and translated. So Pakendorf and Aralova decided to do the painstaking work of transcribing, digitizing and making more readily interpretable all of the recordings that came before, before making new ones.
I asked the researchers what the specific value of doing this leant to their project and their corpus in general. “It allowed us to hit the field running,” said Aralova. Practically speaking, it saved them a lot of time recording long texts and working to transcribe them while in the field. Negidal had fewer than 10 speakers at the time they were working, two of whom passed away during the project so, needless to say, time was precious. They were also able to identify gaps in the corpus and texts that needed to be recorded differently in order to be better understood. For example, they recorded several procedural texts with video in order to fully capture deixis etc… that cannot be captured by audio only.
Moreover, the linguistic and socio-linguistic value that these annotated resources brought to the corpus was irreproducible in many ways. The recordings from the 90s include data from a male speaker who was a known and exceptional storyteller. Today, versions of his stories only exist in writing and there are no longer any living male speakers. Additionally, the recordings collected in the 2000s feature the (then) oldest speaker, Anna Nadeina, who was born in the early 1910s and was able to recall aspects of her early life from living nomadically to having windows made of fish skin.
“There’s so much in there that’s not just language – it’s memories. It’s a window into what life was like back then” – Brigitte Pakendorf
Today, there are only two fluent speakers of Negidal left. Had Pakendorf and Aralova chosen to ignore these previously recorded materials, it is possible that the language, voices, and knowledge encoded in their contents would have been lost forever. Instead, the son of the last male speaker now has a copy of his father’s recordings in a format that he can keep and understand indefinitely.
View the full Negidal collection HERE.
Documentation of Ingrian by Fedor Rozhanskiy and Elena Markus
In planning their documentation of Ingrian, Fedor Rozhanskiy and Elena Markus recognized from the onset that there were legacy materials associated with the language that were themselves endangered. There was linguistic data stored on tapes in private archives without backup copies and in a location that was inaccessible to the scientific community. So they accounted for this in their ELDP application and made half of their project’s main goal to collect and prepare these materials for modern, digital archiving.
They were able to achieve their goal and add new, high-quality recordings in order to build a corpus of digital, accessible, audio and visual materials which add up to 531 hours of data (15 hours of which are transcribed and annotated in ELAN). These recordings include data from six Finnic languages and data from as early as 1968. The content covers an immense amount of cultural content, traditional knowledge, and spontaneous speech, while also capturing the dialectology and language contact found in the region.
Rozhanskiy and Markus have made an important contribution to the field of linguistics through the building of this collection. But they have done so by honoring the work of those who came before them.
We often hear the urgent call to document endangered languages but not so much the call to re-document (or bring to present-day linguistic research standards) heritage materials. Yet we can see from these few examples that the value of this work is potentially very high and the time left to do it, potentially dwindling. The message that rings most clearly for me through the nuances of all these collections is that giving time, attention and patience to legacy materials can provide immense value to the work, progress, and understanding governing future documentation. In other words, when we choose to revive shared cultural memory through audiovisual heritage materials, we all benefit.
ELAR is currently involved with the development of quality standards and curation criteria for digital language data as part of the QUEST research project. The project aims to improve the reusability of culturally valuable audio and video recordings, and to ensure that their reuse is also possible in non-scientific sectors. The focus lies on digital data collected in language documentation and multilingualism research, whose reuse potential extends far beyond the boundaries of these disciplines. Accordingly, a sub-project of QUEST is dedicated to the development of criteria for the curation of language data, focusing on their accessibility by cultural institutions and more importantly speaker communities. By developing procedures that will allow corpora to be enriched with information, their long term findability and reusability will be improved. QUEST thus also contributes to the future-proofing of culturally significant audio and video recordings. More information on QUEST is available here.
Want to get involved with digitizing legacy materials? Check out the HELP for Endangered Legacy Collections initiative from the Committee on Endangered Languages and their Preservation (CELP)
Want to read more about legacy materials projects? Check out this recent post about recordings from the 60s and 70s in Valoc.