Leelo Keevallik on UR Sverige Forskar

Leelo Keevallik was featured on UR’s Sverige Forskar program, available here: https://urplay.se/program/229034-sverige-forskar-de-ordlosa-ljudens-sprak

For a (rough! personally done!) transcript in English, see below:

De ordlösa ljudens språk. 2022.

Producent: Jan Axelsson. UR: Sverige Forskar.

New research is mapping out the sounds we make, those that are rarely found in dictionaries. Extending from this work, we can improve how we teach languages and make our digital devices smarter.

We are constantly producing sounds. Not just words, but sounds.

And traditionally, these sounds haven’t had a place in linguistic research. But the sounds between the words can have more meaning than we thought.

Leelo: Yes, for example, one can use the word ‘mmm’ when one likes food, ‘brr’ when it’s cold, or one can say ‘ahh’ [Oh] when one hears something new, and so forth. The big question is: where does language begin and end?

Like language detectives, researchers decipher these non-lexical sounds. A research group is Linköping has for several years documented the sounds we humans make.

Leelo: To capture these sounds, we record people in real life, since we can’t come across these sounds in the lab. Then I try to find which sounds go with which movements

In the research, the sounds and movements are all recorded. Then the materials are transcribed and analyzed.

Leelo: Ah-oh-ah, ah-oh-ah are short, simple steps. <ah: ah: ah: ah:> are longer steps. Hey-oh are kicks. Then we have flashy steps and those are oOp ah!.

Leelo: One thing we have looked at is how parents use lipsmacks when babies have the first meals.

The researchers have been able to show that ape mothers use the same sounds when they feed their own young. So we are more like apes than we think.

Those sounds we give off when we strain can also function as language.

Leelo: One might think that someone acts strong when they grunt, but I have looked at when people move pianos or when people lift dung together in a sheep stable, and I’ve been able to show that people grunt when they are about to lose their grip. One is strong when you have air in your lungs and hold your breath. When the breath seeps out, that’s when the grunts are produced. In this way, grunts emerge more as a call for help.

In the future, both AI and language students learn to understand the sounds between the words.

Leelo: Voice assistants will assist us more and more, not least in clinical contexts. And then it would be perfect to have a voice assistant could recognize when someone is groaning in pain. In that case we wouldn’t need to always give them instructions to call for staff help.

Leelo: A less narrow perspective on language would help us to communicate better.

%d bloggers like this: