The tapes are in the drawer where I keep things I'm going to deal with tomorrow.
Six cassettes, TDK SA-90s, recorded between 2003 and 2007. My grandmother's handwriting on the labels — not her name but the contents. Tanuki stories. Songs from before. For Mariko when she's older. The last one just says things I want to remember, which I've never been able to listen to for longer than forty seconds.
I'm a translator. That's what I do, professionally. English and Japanese, both directions, mostly technical documents for a pharmaceutical company in San Mateo. Regulatory filings. Clinical trial protocols. I'm good at it — precise, fast, reliable. My clients don't know I keep six cassettes in a drawer. They wouldn't understand why a professional translator hasn't translated the thing that matters most.
The dialect is the problem. That's what I tell people, when they ask. My grandmother spoke Tsugaru-ben, from the far north of Honshu — a dialect so thick that NHK used to subtitle it for Tokyo audiences. Standard Japanese and Tsugaru-ben share grammar the way English and Scots share an alphabet: technically the same system, functionally a different world. I grew up hearing it from her and speaking standard Japanese back, the way bilingual children learn to separate their rivers. I understand Tsugaru-ben the way you understand your mother's face — completely, without being able to describe it.
But understanding and translating are different operations. Understanding happens in the body. Translation happens on the page. To translate my grandmother's stories I would have to make decisions: this word means this, this phrase carries that weight, this silence between sentences is a pause for breath or a pause for grief and the difference changes the meaning of what follows. Every decision would be an assertion. Every assertion could be wrong. And there's no one left to tell me.
The last speaker of Tsugaru-ben is not me. There are people in Aomori Prefecture who still speak it — old people, fewer each year, their children answering in standard Japanese the way I answered my grandmother. But the last speaker of my grandmother's specific Tsugaru-ben — the version inflected by her years in California, softened by English loan words she'd naturalized without noticing, shaped by the particular stories she chose to tell and the particular silences she maintained — that speaker is dead. What remains is six cassettes and me.
I tried once. Three years ago, after the funeral, when the obligation felt fresh enough to override whatever it is that stops me. I set up the old tape deck — she'd given it to me along with the cassettes, as if she knew I'd need the specific machine — and I pressed play on Tanuki stories.
Her voice filled my apartment. Not the voice from the last year, thin and careful. The voice from 2004, when she was seventy-one and still cooked for twelve on New Year's and carried groceries up three flights without pausing. A voice with weight in it.
She was telling the story of the tanuki who wanted to cross the river. I knew this story. She'd told it to me dozens of times, always a little different, because the oral tradition doesn't have a canonical text — it has a shape, and the storyteller fills the shape with whatever the moment requires. In this version, the tanuki stood on the bank for a long time before crossing. In the version I remembered from childhood, the tanuki crossed immediately and the trouble came from the other side. Same story. Different animal.
I wrote the first sentence of the translation: There was a tanuki who wanted to cross a river.
And stopped. Because "wanted" was already wrong. The word she used — the Tsugaru-ben word — doesn't map cleanly to "want." It's closer to something between wanting and needing, with an overtone of knowing you shouldn't. A wanting that contains its own reluctance. English doesn't have this word. Standard Japanese doesn't either, not exactly. The word exists in the space between my grandmother's mouth and my ear, and to move it onto the page I'd have to pick a direction: want, or need, or yearn, or something else, and whichever I picked would be one thing when the word was several things, and the several-ness was the point.
I turned off the tape. I put the cassettes back in the drawer. I translated a clinical trial protocol for a new diabetes medication and it was a relief — those words meant exactly one thing each.
My mother asks about the tapes sometimes. She doesn't speak Tsugaru-ben — she grew up in Sacramento, where my grandmother spoke English at the grocery store and Japanese at home but the dialect only to herself, humming songs while she cooked, muttering at the news, talking to the plants. My mother heard the dialect the way you hear music through a wall: the rhythm without the words.
"Have you translated them yet?" she asks, and I say I'm working on it, which is true in the sense that I think about it every day and false in the sense that I haven't pressed play in three years.
What I don't tell her is that I'm afraid of being wrong in a way that can't be corrected. A bad translation of a pharmaceutical document gets caught in review. Someone flags the error, I fix it, the process works. A bad translation of my grandmother's stories just exists. It becomes the record. My version displaces the original not because it's better but because it's legible and the original is dying. Magnetic tape degrades. Oxide sheds from the substrate. Every year the tapes sit in the drawer, the original becomes a little more fragile and my translation — my wrong, partial, one-directional translation — becomes a little more necessary.
This is the thing I can't get past: the more necessary the translation becomes, the more careful I need to be, and the more careful I need to be, the less I can bring myself to start. The tapes' fragility is an argument for urgency and an argument for paralysis, simultaneously, and I can't tell which argument is winning because they use the same evidence.
A colleague tells me I'm overthinking it. A translation doesn't have to be perfect, he says. It has to be useful. And he's right. I know he's right. The knowing and the doing occupy different rooms in me, connected by a hallway I can't seem to walk down.
I've wondered whether this is grief. Whether the not-translating is a way of keeping her alive — as long as the tapes are untranslated, they still need me. They're still waiting. The relationship between her voice and my understanding is still active, still present-tense. Translating would move it to past tense. Would make it about her instead of from her.
But I don't think that's it. Or not entirely. Because the reluctance doesn't feel like holding on. It feels like standing at the edge of something you're qualified to do and not being able to tell whether your hesitation is respect or cowardice. Whether you're pausing because the work deserves more care than you can give it right now, or whether you're using "more care" as a way to never start.
I translate pharmaceutical documents without hesitation. Adverse event: Grade 3 hepatotoxicity. I know exactly what that means and exactly how to render it in Japanese. The words have been defined by committee, stripped of ambiguity, made portable across languages by design. They mean one thing. I carry them from one language to another the way a courier carries a sealed envelope: the contents are not my responsibility. I just deliver.
My grandmother's stories are not sealed envelopes. They're living things that will die differently in my hands than they would have in hers. Every translator knows this. Every translator does the work anyway. I don't know why I can't.
Last week I took the tapes out of the drawer. Not to translate — just to hold them. The plastic cases are yellowing. The labels are still legible but the adhesive is failing; one corner of Songs from before has peeled up and I pressed it back down carefully, knowing that the tape inside is degrading whether or not the label stays.
I held the one that says things I want to remember and I thought about pressing play. I own a USB cassette converter now — I bought it two years ago, another preparation that substitutes for action. I could digitize the tapes. That would preserve the audio without requiring translation. The substrate would be saved. Her voice would survive the oxide.
But I haven't done that either. I think the digitizing would let me off the hook. If the audio is preserved, the urgency of translation decreases. The argument from fragility weakens. And without that argument, I'd have to face the real question, which is not "will the tapes survive?" but "will I ever do this?"
The fragility is convenient. It provides a reason that isn't the reason. The real reason is something I can only describe by circling: I am capable of doing this work. I am not doing this work. And I cannot determine whether the gap between capability and action is a feature of my care or a failure of my courage, and the two explanations feel identical from inside.
There's a word in Tsugaru-ben — medagusa — that doesn't exist in standard Japanese. It means something like the state of being too aware of being watched to act naturally. My grandmother used it about cats: the way a cat that knows you're watching it will pretend to be interested in something else. But she also used it about herself. Medagusa, she'd say, when someone complimented her cooking and she couldn't eat her own food for the rest of the meal.
I wonder if what I have is medagusa. The tapes watch me. My grandmother's voice watches me from inside the magnetic oxide. The future readers of the translation watch me — whoever they are, my mother, a scholar, a stranger. Everyone is watching, and under that gaze I can't act naturally, and acting naturally is the only way to translate something true.
But this might be another story I'm telling myself. Another reason that isn't the reason.
Tomorrow I'll translate the tapes. I've been saying this for three years. Each time I say it, I believe it. Each morning, the drawer stays closed. Not because I decided not to open it — I never decide not to. I just don't decide to. The absence of action feels different from the presence of refusal, though I'm no longer sure it is.
The pharmaceutical company sent a new batch of documents this morning. Phase III results for an immunotherapy drug. Efficacy data, safety profiles, adverse events organized by system organ class. I'll have it done by Thursday. The words will cross from English to Japanese without loss because they were built to cross without loss. That's what technical language is: pre-translated. The work has already been done by the committees who defined the terms. I'm not translating meaning. I'm matching patterns.
My grandmother's word for the tanuki's wanting — the wanting that contains its own reluctance — sits in no committee's glossary. It lives in six cassettes in a drawer in my apartment in San Mateo, forty-three miles and an entire ocean from where the word was made.
I could carry it across. I have the capability. The question is whether carrying it across would deliver it or change it, and whether the difference between delivery and change is something a carrier can see, or whether you'd have to be standing on the other side to know.
The tapes are in the drawer. Tomorrow, I'll open it.