This is part of Transom’s occasional series of advice columns on how to handle language translation on the radio. You can find the entire Thoughts On Translation series here.
Who Speaks for the Speaker?
I was talking shop with the excellent Steve Wadhams of CBC Radio at a Third Coast International Audio Festival a few years ago, and I asked him, in my rudimentary Canadian, how he handled non-English voices in his radio stories. He told me he mainly did what everyone else did — run the voice in the clear for a few seconds, then dip it under a voiceover.
But there was something he’d like to do, he said, if he could get away with it. That was to let the foreign-language actuality play in its entirety, no matter how long, then follow with a complete translation. That way listeners would hear the speaker’s voice, and the music and arc of her statement, even if they couldn’t understand the words.
Radio is as much about voices as it is about words. Arguably more. A voice conveys attitude, personality, and mood. It provides information about age, education, ethnicity, and social class. It has timbre, rhythm, tempo, dynamics, and pitch. Most importantly, it connects speakers and listeners in ways that words on a page — or words read later by someone else — rarely can.
You don’t want to lose all that just because your subject is one of the 6.6 billion people whose first language isn’t English.
Including foreign-language voices in a radio story creates all sorts of creative, technical, ethical, and semantic challenges. Here I want to focus on a subset of those, surrounding the choice of the voices that listeners hear in the finished piece.
But First… Why Bother?
Sure, English may not be everyone’s mother tongue, but it’s still the most widely spoken language in the world. Wherever you go somebody speaks it. Why not just do stories about them?
Everyone who reports from abroad does this to some extent. I may be guiltier than most. For the WORKING series on Marketplace, I profiled two Nigerian men, a French woman, a Kenyan woman, and an Italian woman, all in English. And in every instance, I ended up wondering what the hell I’d been thinking.
Raffaella was very smart and her written English was excellent. I talked to her on the phone and had no trouble understanding her. But even after four days of hanging out with her, I had precious little tape I could use. In the opening sequence, I repeat every iffy word immediately after she says it. I sell this as a clever device but in fact it’s an act of desperation. Can you hear the regret in my voice?
So yes, lots of people speak English. But most people don’t, and in many places, those who do are hardly representative of the general population. Insisting that a non-native speaker speak English can also be cruel. How would you like to go on the radio in your second or third or fourth language? (All that said, Raffaella claimed to be thrilled with the way her profile turned out.)
There are two common approaches to scripted voiceovers. The first is to find someone who (a) speaks English with the appropriate accent, (b) isn’t too hard for American listeners to understand, (c) matches the subject in gender and more or less in age, and (d) can read from a script without sounding like he or she is reading.
I think the voiceover works beautifully. But there are risks in having a non-native speaker try to mimic, in English, someone’s emotion and speaking style. My main gripe? Too often, the voiceover person sounds less articulate than the original speaker. Which isn’t surprising, since she’s (a) reading and (b) speaking a foreign language. To avoid this, you need to find someone who isn’t just a good actor, but who also speaks excellent (rather than just passable) English.
Once you’ve found her, it’s essential that she listen to the original tape, and even better if she translates it herself. You may want to ask her to listen to the tape and record a translation without looking at a script.
The second common approach to voiceover is to grab a native English-speaking colleague or friend and have him or her read the lines as neutrally as possible. In a way this can be more honest than trying to imitate the original — it’s like a film producer using subtitles rather than dubbing. But a flat (or overly jaunty) read or a badly mismatched voice can really take a listener out of a scene.
Here’s an example of a non-theatrical voiceover from “Breaking the Ice,” a terrific piece produced by Michael May for This American Life in 2013.
Michael chooses a woman to speak for the man, and she does it in an affectless American accent. Crucially, for the first few sentences, we hear every word the man says. I think it works because the original and the voiceover never compete for our emotional attention. We know that the guy is speaking from the heart and the woman is only translating. We care about him, not her.
There’s another voiceover option you rarely hear: Using a native English speaker whose situation is somehow analogous to that of your subject. If you’re profiling a kid in a Brazilian favela, for instance, you might use an American kid from the projects. If you’re talking with a farmer in rural Tanzania, you might use an American farmer.
Listen to the opening of “Saving Jungle Souls,” another Vanishing Homelands piece by Sandy Tolan and Nancy Postero. It begins by recounting the moment, years earlier, when Chief Ataiba, leader of a band of Yuqui Indians, decided to leave the Bolivian Amazon and live in an Evangelical mission.
First we hear the jungle, then Ataiba speaks in the clear for several seconds. It’s not a huge amount of time, but it’s enough to give us a feel for the place, the language, and Ataiba’s voice. Then comes the voiceover, which I love.
“When I first heard Ataiba I started thinking about who could voice it and Danny McCabe immediately came to mind,” Sandy told me recently. “There was something in the timbre of both men’s voices.”
I suspect there was also something about the way McCabe identified with Ataiba’s dilemma. McCabe was a Navajo elder from Arizona whose family had lost its home in a land dispute. My Homelands colleagues have often used Native Americans to do voiceovers for indigenous people in other countries, and it can be remarkably effective. The recording sessions can be moving, too.
No matter whom you choose to do your voiceovers, getting them right takes lots of direction. It helps if you pay the people — they may have to be there for a while.
No matter how good the read, a scripted voiceover almost always sounds, um, scripted. A real-time translation from a good on-site interpreter can sound much more spontaneous and alive.
But getting usable tape from a three-person, two-language conversation can be a culturally, linguistically, emotionally, and ergonomically complicated business. You have to be extremely clear about what you want. Generally that means waiting for the subject to complete a thought (but not letting her ramble on and on), speaking in the first person (not saying, “She says…” at the beginning of every sentence), and translating word for word rather than paraphrasing.
No need to paraphrase this.
Help Transom get new work and voices to public radio by donating now.
Here’s a clip from a piece I did for NPR’s Morning Edition as part of Homelands’ Worlds of Difference series. The story is about how Kenya’s Maasai people are rethinking their traditional opposition to public education. In this scene, we’ve just stepped out of a dark hut, where we witnessed a pre-wedding ceremony for a teenage couple.
I actually recorded the translation a few minutes after the interview. I knew it was an important cut, but the interpreter (a young Maasai journalist) stumbled when translating it. So after the interview was over I took him aside, played it back to him, and he translated it again. I like the way he searches for words. To me it sounds much more in-the-moment than if I’d written it out and asked him to read it.
Sometimes it makes sense to be transparent about the fact that you’re using an interpreter. Kelly McEvers is great at this — you feel like you’re with her, interpreter at her side, hearing the world through the same filters she does. I like the honesty of this approach; you just need to be careful not to push your subjects’ voices to the background.
The final clip is from a story I did in Peru in 2001. The piece is about an indigenous man, Espiritu Bautista, and an American anthropologist, Richard Chase Smith, who are working together to create a digital map of the ancestral territory of Bautista’s people. In this scene, we’ve just driven several hours to an overgrown field; Bautista has never been there, but Smith has a hunch the site is culturally significant.
I let Bautista speak for a few seconds before I describe what I’m seeing. Then Smith explains what Bautista is saying, but not word for word. I’m always wary of having people speak for other people, but here’s a situation where I think a paraphrase works better than a straight translation. The scene documents a moment of discovery, and I want my listeners to experience it along with us. If I’d gone with a voiceover, especially one recorded later, I think it would have pulled us away.
It May Be Greek To You…
Okay, so you’ve done your interviews, you’ve sent the tape to a translator, you’ve gotten back transcripts, you’ve used them to write your voiceovers. All you’ve got to do now is mix. Here’s your chance to let your listeners hear your subjects’ voices! Just be sure that they’re saying what you think they’re saying. What sounds to you like “man speaking ______” may in fact be “Why are you sticking that mic in my face?” Someone out there is bound to know.