He Claimed My 7-Year Dying Language Archive for the UNESCO Prize — Then the Committee Asked Who Held the Village Consent

The village was called Ban Sao Han. It was two hours by road from the nearest town with reliable power, and the road turned to packed clay for the last twenty kilometers, which meant that in the rainy season, Dr. Yuki Tanaka drove the last section with both hands on the wheel and the Nagra wrapped in the padded case on the passenger seat.

The Nagra reel-to-reel recorder was a Swiss instrument, 1971, aluminum body, broadcast-grade. It weighed 4.2 kilograms and the take-up spool was slightly bent from a drop in year three — she had been crossing a bamboo bridge carrying the equipment case over one shoulder and the case had slipped, and the corner had hit the bridge rail before she caught it. She had opened the case. She had checked the transport. The spool was bent at approximately 3 degrees from vertical. The recorder still ran true. She had wrapped the recorder back in its foam and carried it across the bridge.

In Ban Sao Han, the recorder sat on a low table in the session room — a clean-swept space in Nai Khuon’s family compound, the windows open to the paddy field, the air at this hour dense and still. Nai Khuon was 84. She was one of eleven fluent speakers of Mekong Tau still living. The language had no alphabet. It had no dictionary. It had the recordings Yuki had made over seven years, and the annotation framework she had built to make those recordings navigable.

Yuki set the microphone at the height of Nai Khuon’s natural speaking posture — 34 centimeters from the speaker face, which she had determined in year one after testing three distances and comparing the harmonic content of the tonal peaks. Mekong Tau was a four-tone language. The tones were the grammar. The wrong microphone distance blurred the onset of tone 3b — a high-falling contour with a voiced onset that had no equivalent in any neighboring language — into the shape of tone 2, which was unvoiced and level. Yuki had written a four-page analysis of this problem. The analysis was session 7 of the archive annotation file.

Linh was positioned to the side with the secondary microphone. She was 24, a community member who had assisted the field sessions for three years. She did not need instructions now. She knew the room geometry.

Yuki pressed record.

Nai Khuon began to speak.

She was reciting the water-calling text — a ritual speech that Nai Khuon had said she learned at twelve from her grandmother, and that her grandmother had also learned at twelve. Yuki had recorded it four times over seven years. Each recording was different — the tone contours varied slightly with Nai Khuon’s health and time of day and the temperature in the room, which affected the speed of her articulation. The variation was data. Yuki annotated each version in the tonal markup system she had developed in year two when the existing field linguistics notation failed to capture the micro-pitch movements that distinguished Mekong Tau’s tones in connected speech. The notation system was now used by two other research teams working in the same language family.

She had not published that either.

She annotated. The reel turned. The bent spool turned at the same rate as the other. The tape moved.

Linh made a small sign with two fingers — take-up speed. Yuki adjusted one notch. Linh nodded.

This was how the session worked. This was how 91 sessions had worked.

She read the UNESCO grant announcement on a Tuesday, from her desk at the university, which was the third desk on the right in the shared historical linguistics section.

The university website had posted the announcement that morning: Department of Anthropology Receives UNESCO Intangible Cultural Heritage Documentation Grant. The announcement led with the project name — Mekong Tau Language Archive and Documentation Initiative — and with the PI’s name: Prof. A. Voss, Principal Investigator.

She scrolled to the project personnel section.

Field researchers: Dr. Yuki Tanaka.

She read “field researcher.”

She looked at the Nagra on the shelf behind her desk. The bent take-up spool caught the light from the east window. The reels from the last session with Nai Khuon — session 91, the water-calling text, the fourth recording — were still loaded. She had not rewound them. She would annotate session 91 this week.

She turned back to the computer. She closed the announcement tab.

Eight months ago:

Voss had called her to his office on a Thursday. He had shown her the grant draft — she had not been consulted during the drafting. The cover page: University Endangered Language Documentation Program. Prof. A. Voss, Principal Investigator.

He had said: “I’ve used the Mekong Tau archive as the primary evidence. Your fieldwork is the backbone of this application.”

She had read the personnel section. Field researchers: Dr. Yuki Tanaka.

She had said: “The consent agreements are in my name.”

He had said: “Yes, we’ll address that in the evaluation phase. The important thing is getting the application in.”

She had said: “Okay.”

She had left his office.

She had gone back to the archive and spent three hours annotating session 84.

She opened the consent agreement folder.

Seven communities. Seven signed agreements. Her name on each one. The agreements were personal — she had spent seven years building the relationships with the speaker communities that made the consent possible. Nai Khuon had signed the first agreement in year one, and she had signed it because Yuki had been there, in the room, asking in the right language, in the right way, at the right time of day when Nai Khuon’s family was home. The consent was given to Yuki. Not to the university. Not to the department. To the researcher who had been in the room.

She closed the folder.

She did not look at the Nagra.

She opened session 91 annotation.

The university research showcase was held in the faculty building atrium, tables set up in a horseshoe with project display boards. Voss presented the Mekong Tau project from the center position.

He had a 90-second audio clip loaded on his laptop — session 62, the tonal inventory passage, which Yuki had selected in year four as the clearest single demonstration of all four tones in continuous speech. She had selected it because the speaker in session 62 — Nai Khuon, at 81, in the spring session — had been particularly clear that day, her articulation unhurried, the room temperature exactly right. The clip had been processed through the noise-reduction filter Yuki had calibrated specifically for the Nagra’s ambient hiss profile.

Voss had the Nagra on the table because the showcase required an artifact display. Yuki operated the playback.

He said: “Our documentation program has captured the final fluent speakers of this language before they are gone. What you are hearing is a tonal system that has no written form and no other complete audio record.”

Yuki turned the reel.

The clip played. The four tones of Mekong Tau were audible — tone 1 level, tone 2 unvoiced rising, tone 3b high-falling with the voiced onset that she had spent four months learning to hear correctly. The room was quiet.

A faculty member from another department said: “How do you know which tone is which? They sound the same to me.”

Voss said: “That’s the beauty of the project — the annotation framework allows non-specialists to navigate the archive.”

He did not say whose framework. Yuki went back to the recording level monitor.

The email from Dr. Marie Petit arrived the following Thursday.

Subject: UNESCO Intangible Cultural Heritage Assessment — Mekong Tau Documentation Initiative — Consent Documentation Request.

Yuki read it at 8:30 AM.

The UNESCO evaluation committee was conducting the formal assessment of the documentation initiative. The evaluation protocol required:

(a) the original field recording logs, signed by the researcher who holds the community research consent agreements;

(b) copies of the signed consent documents from each speaker community;

(c) a letter from the community consent holder confirming the archive’s availability for UNESCO’s intangible heritage record.

The email was addressed to her directly. Dr. Petit had looked up who the consent agreements named.

She opened the consent folder. Seven agreements. Her signature on each one — and the community elder’s signature, and sometimes a witness’s. She had folded each agreement carefully after signing and put it in the archive envelope herself, in the same room where the consent was given.

She looked at the Nagra on the shelf.

The bent spool was visible from her desk.

She closed the folder. She did not email Voss. She opened session 91 annotation and kept working. She had three more hours on the tonal contour analysis for the water-calling text.

Voss read the UNESCO evaluation notice in his email queue at 11 AM. He was between meetings. He forwarded it to his research administrator with a note: “Please coordinate with the department’s documentation on this — standard evaluation process.” He assumed the documentation request meant institutional letters, project summaries, the kind of administrative materials his research administrator handled routinely.

He went to his next meeting.

He had not read the part of the UNESCO protocol that specified: “Community research consent documentation is personal and non-transferable — UNESCO requires that the consent holder be identified by name and sign the evaluation documentation directly.”

He had not read this because he had not read past the summary paragraph.

She had been annotating session 91 for six days.

The water-calling text had a structure she had not fully understood until the fourth recording. In the first three recordings, she had treated the opening phrase as a tonal preamble — a set of syllables that established the key before the main text began. In the fourth recording, the one from the current season, she had noticed that the tonal preamble was not identical across versions. It varied in a way that appeared systematic. The opening phrase in summer was different from the opening phrase in winter. The difference was two micro-pitch movements in the second syllable of the first tone, which her annotation markup system could record but which no existing linguistic analysis framework had a category for.

She was building the category.

This was what the archive was. Not just the recordings — the analysis that made the recordings speak. A recording of Mekong Tau without the annotation framework was a sound file. A sound file that a non-speaker could not parse, could not cite, could not use to reconstruct the tonal system if the speakers were gone. The annotation framework was the difference between a record and an archive.

She had built the annotation framework. She had revised it four times over seven years. The current version was 47 pages, with a notation index and a sample annotated text.

The university’s research showcase had described the archive. No one had described the framework.

She opened session 91 and kept working.

The research administrator’s name was Paul. He had been in the department for eight years. He found the answer in forty minutes.

The UNESCO Intangible Cultural Heritage Convention ethical protocol — he had pulled the document from the UNESCO website — was explicit: community research consent agreements entered into between a researcher and an indigenous or minority language community were personal instruments. They expressed the community’s consent to be recorded and for the recordings to be used in the ways specified in the agreement. The consent holder was the researcher named in the agreement. An institution could not hold the consent on behalf of that researcher. An institution could receive the recordings as an archive asset, but the consent to make and use those recordings belonged to the researcher who had negotiated it.

Paul had looked at the consent agreements, which were on file in the department’s research records.

They were signed by Dr. Yuki Tanaka.

Every one of them.

He went to Voss’s office.

He told him.

Voss said: “Can the department sign as an institutional partner?”

Paul said: “The protocol says the consent holder must sign directly. The consent holder is named in the agreements.”

Voss said: “And the agreements name Dr. Tanaka.”

Paul said: “Yes.”

Voss looked at the UNESCO protocol summary on the desk. He had told Yuki eight months ago that they would address the consent question “in the evaluation phase.” He had said it because it had seemed like a solvable problem — a documentation formality that his research administrator would handle by producing the appropriate institutional letter.

He now understood that the community consent was not a documentation formality. It was not institutional. It had never been institutional. The seven speaker communities in the Mekong Delta had given their consent to Yuki because Yuki had been the one who came, every year, for seven years, and sat in their session rooms, and listened, and annotated, and asked them in the appropriate way. They had not given their consent to the Department of Anthropology. The Department of Anthropology had not asked them.

He called Yuki’s extension.

She was annotating session 91 when the call came.

She had the playback running on the headphones — Nai Khuon’s voice, the water-calling text, the high-falling tone 3b onset at minute four of the recording. She had been marking the onset boundaries in the annotation file. The onset was cleaner in this recording than in the three previous ones because the room temperature had been lower, which had the effect of reducing the muscle tension in Nai Khuon’s jaw and allowing the voiced onset to develop more fully.

She took off the headphones.

Voss said: “UNESCO need the consent documentation and the letter. The agreements are in your name.”

She said: “I know.”

He said: “Can you send them what they need?”

She said: “Yes.”

There was a pause.

He said: “Yuki, I should have — the evaluation phase question. I said we would address it.”

She said: “The evaluation phase is now.”

He said: “Yes.”

She said: “I’ll send the documentation.”

She hung up.

The Nagra was still on the shelf. The reels from session 91 were still loaded. The bent spool caught the light from the window. She did not look at it. She opened the consent folder and began to prepare the documentation package.

She prepared the documentation package at her desk in the evening.

She had the consent folder open, the session logs open, and a draft of the consent-holder letter in a separate document. The work was methodical. Each consent agreement required a certified copy — she had scanned them years ago for the digital archive, so the files existed; she needed to add the certification note to each one. Each session log entry was already complete — she had maintained the logs in real time, entering each session’s details on the day of recording. She did not need to reconstruct anything.

The consent-holder letter took the most thought. She drafted it three times.

The first draft was too long — she had included a full description of the consent negotiation process for each community, which ran to four pages. She deleted it. The UNESCO evaluation did not need her to explain how the consent had been obtained. The agreement documents explained that. She needed to confirm that she held the consent and that the archive’s use was within its scope.

The second draft was too formal — she had written it in the style of an institutional letter, with department letterhead language and passive construction. She deleted it. The consent was personal. The letter should be personal.

The third draft was two paragraphs. She sent it.

She understood, as she sent it, that this was the moment the question was answered. Not when the UNESCO record was updated. Not when Voss issued the amended grant record. Now, at her desk, assembling seven years of community relationships into a documentation package, each page signed with her name because no one else could sign it. The communities had given their consent to the researcher who came, not to the institution that funded the travel.

She sent the package at 9:14 PM.

She reopened session 91 annotation.

The documentation package was a three-part submission.

Part one: seven consent agreements, each signed by the community elder, by Yuki, and in four cases by a witness. She scanned each one, front and back, and certified the copies as accurate to the originals. The originals were in the archive envelope in her desk drawer. She had kept them there since year one.

Part two: the field recording logs — session 1 through session 91, each log entry listing the date, the location, the speaker name, the session type, the duration, and the equipment used. Her name was on every log as the recording researcher. The Nagra’s serial number was in every entry.

Part three: the consent-holder letter. She drafted it in two paragraphs. The first paragraph confirmed that she held the community research consent agreements for the Mekong Tau language documentation archive, that the agreements were personal instruments issued to her by the speaker communities, and that the archive’s use in UNESCO’s intangible heritage record was consistent with the permissions granted in those agreements. The second paragraph confirmed that she was the sole researcher responsible for the recording and annotation work in the archive.

She sent the package to Dr. Petit.

The UNESCO evaluation record was updated within the week. She received the confirmation email on a Friday morning.

Dr. Petit: “Dr. Tanaka — your consent documentation is exemplary. The community agreements provide the fullest ethical coverage for an archive submission we have reviewed this cycle. You will be listed as Lead Researcher and Consent Holder in the UNESCO nomination record. The archive will be designated as the primary evidentiary foundation for the Mekong Tau Intangible Heritage nomination.”

She read “Lead Researcher and Consent Holder.”

She put the email in the Mekong Tau project folder. She had session 91 annotation to finish.

Linh was on the next field trip three weeks later. Yuki told her about the UNESCO confirmation on the drive to Ban Sao Han — not the full account, just the result: the evaluation had gone through, the archive was accepted, the nomination was moving forward.

Linh said: “Your name in the nomination.”

Yuki said: “Yes.”

Linh said: “Nai Khuon will be happy.”

Yuki said: “Yes.”

Linh said: “She always said — if the language lives anywhere after us, it should live with someone who learned to listen.”

Yuki kept driving. The road was clay for the last twenty kilometers.

Voss stopped by her office the following week.

He stood in the doorway. He said: “Excellent outcome. The department is grateful for your fieldwork.”

She said: “The consent holders were generous.”

He said: “Yes.” He paused. “I’ve updated the research register. You’re PI on all subsequent Mekong Tau work. Any community-facing activity, any new grant application — that goes through you.”

She said: “Okay.”

He said: “I’ve also submitted the reattribution to the UNESCO records office. The nomination document will list you as Lead Researcher from the outset.”

She said: “Good.”

He said: “Good work, Yuki.”

She said: “Thank you.”

He left.

She picked up the Nagra from the shelf. She checked the reel. The session 91 tape was still loaded. The bent spool turned one click when she tested the transport. She set the recorder back on the shelf.

She had session 91 annotation to finish. The tone 3b onset boundary at minute four still needed three more timestamps marked.

The original UNESCO grant announcement — Prof. A. Voss, Principal Investigator — had been shared by the university’s social media accounts on the day of the announcement and reshared 340 times before the university updated its own post. The reshares were on accounts the university did not control. They could not be deleted. The university’s post now read: “Documentation work led by Dr. Yuki Tanaka.” The 340 reshares still showed the original text.

She had seen them.

She did not look at them anymore.

The UNESCO nomination record was a permanent document.

Dr. Petit’s office had sent a formal copy of the nomination record by post — a laminated certificate-weight sheet, the UNESCO logo at the top, and below it: Mekong Tau Language Documentation Archive. Lead Researcher and Consent Holder: Dr. Yuki Tanaka. The archive’s designation as evidentiary foundation for the Intangible Heritage nomination meant that the record would be cited in the official nomination file, which was a publicly accessible document held by UNESCO’s cultural heritage division.

The nomination file would be searchable. Her name was in the searchable field.

The seven speaker communities had been notified by the UNESCO evaluation office — Dr. Petit had sent letters to each community elder, translated by Linh into the local language. Nai Khuon had received her letter in Ban Sao Han. Linh had read it to her. Yuki did not know exactly what Nai Khuon had said in response, because Linh had summarized it: “She said she was glad it went to the right place.”

The Nagra was on the shelf.

The session 91 annotation was complete. She had marked the tonal boundary at minute four — the tone 3b onset, the high-falling voiced contour. She had identified the systematic variation in the opening phrase across the four recordings. She had started a 12-page analysis document on the variation pattern. She would present it at the field linguistics conference in November.

The annotation framework version 5 was half-drafted. She expected to finish it by the end of the year.

There were eleven fluent speakers. Malee was learning.

She set the Nagra reel recorder on the low table, the same way she always did — flat surface, microphone positioned toward the speaker’s natural head height, the power cable behind the table leg and out of the way. The take-up spool was still bent from the drop three seasons ago. She had never had it repaired. The reel she had loaded was fresh. The reels from the last session with Nai Khuon were in the archive box beside her field bag, labeled in her hand.

Nai Khuon’s granddaughter, Malee, was seated across the table. She was sixteen. She had been attending the sessions for two years. Today she would speak a full tonal sequence for the first time on tape. The UNESCO nomination record was in the office at the university, Yuki’s name in the lead researcher field. Linh was adjusting the room microphone. Yuki checked the record level.

She pressed play-record. The bent spool turned. The tape began to roll.

Malee looked at the recorder. She looked at Yuki.

Yuki nodded once.

Malee began.

She was reciting the same water-calling text Nai Khuon had recorded in session 1, seven years ago. Malee had learned it from her grandmother over three years of listening to the playback. The tonal contours were close — not identical, because Malee was sixteen and her grandmother was 84 and the language moved differently in a younger voice, but the structure was there. The four tones were distinguishable. Tone 3b had the voiced onset.

Yuki annotated.

The research register update had arrived that morning: Mekong Tau Documentation Initiative — Principal Investigator: Dr. Yuki Tanaka. She had read it. She had filed it in the project folder.

The tape rolled.

She marked the tonal boundary at the opening phrase — the same boundary she had first marked in session 1, in the same room, in the same light, with Nai Khuon’s voice on the tape instead of her granddaughter’s. The notation was the same notation she had developed in year two. Malee’s voice was in the archive now. The archive had a second generation.

The 340 reshares of the original announcement were still there — she knew without checking. She did not check anymore. The UNESCO nomination record showed her name. The research register showed her name. Malee’s voice was on the tape.

The bent spool turned. The tape rolled. Yuki annotated.

There were eleven fluent speakers of Mekong Tau. Malee was learning. She was not fluent yet. She was seventeen sessions away from fluent, by Yuki’s estimate, if the sessions continued at the current pace and if Nai Khuon’s health held through the next dry season.

Yuki had a field schedule for the next eighteen months.

She kept annotating.

She had not told Voss about the tonal preamble variation she had found in session 91.

It was not a secret. It was not finished. She was still building the analysis. When it was finished she would publish it — a paper in the Journal of Linguistic Field Methods, in the section on tonal annotation methodology. She had a target length of 5,000 words and a working title. She had a list of three peer reviewers she planned to suggest to the editors.

The paper would not mention Voss. It would not need to. It would be her name on the paper, with the archive sessions cited as primary sources. Session 91, Dr. Yuki Tanaka, Ban Sao Han, 2025. The UNESCO nomination record would be in the bibliography.

The research register notice was in the project folder: Mekong Tau Documentation Initiative — PI: Dr. Yuki Tanaka. She had filed it beside the UNESCO nomination copy.

Malee was at minute six of the recording.

The water-calling text had sixteen phrases. Malee had completed nine. The tonal contours in the first nine phrases were accurate — Yuki had been checking each one in real time against the session 91 annotation. In phrase seven, Malee had produced the tone 3b onset without hesitation. The voiced onset was clear. Yuki had marked it in the annotation notebook.

There was no equivalent for tone 3b in any neighboring language. The analysis of it was 47 pages in the annotation framework and 12 pages in the new paper. It was a sound that had existed in this valley for longer than the university had existed, and it was in Malee’s voice now, on the tape, the bent spool turning.

Linh caught Yuki’s eye from across the room. She gave the two-finger sign — record level.

Yuki checked. The level was right.

She nodded.

Malee continued.