Name: Do Sounds Carry Their Own Meanings? Onomatopoeia and Arbitrariness of the Sign
Uploaded: 2016-11-10T06:13:33.000Z
Duration: 9 min 45 s
Description: Thousands of YouTube videos with English-Chinese subtitles! Now you can learn to understand native speakers, expand your vocabulary, and improve your pronunciation...

So let’s talk about matching. Sometimes, when we look at the world around us, it's

Adverbs of degree

really easy to tell what things go together. But when it comes to working out which sounds

go with which meanings, a lot of the time, it's not so straightforward. If you go in

thinking there's some natural order that gives your words their inherent wordiness, you're

going to have a hard time. I’m Moti Lieberman, and this is the Ling Space.

Words in our languages can really feel like there's something just naturally that-thing

about them. Like, take a word like “horse”. To an English speaker, that set of sounds

just obviously goes with that four-legged, long-faced, maned mammal. What else could it be?

But in other languages, the word for “horse” can sound totally different – it's

馬 [ɯma] in Japanese, kabayo in Tagalog, pferd in German, סוס [sus] in Hebrew, and more. These

don't really sound alike at all, and you could find similar lists for most words if you look

It’s not like this is a revolutionary idea – over a hundred years ago, the linguist

Ferdinand de Saussure was already wondering about what this meant for how language works.

He realized that there wasn't something fundamental about anything that ties sound and meaning

together, an idea which he termed the “arbitrariness of the linguistic sign”. By this, he meant

that a language could pick potentially any combination of sounds to point out any object

or idea in the world. Horses don't have to be called horses.

This idea lies at the heart of our modern conception of words. We can't just tap into

some collective vocabulary unconscious; we have to learn each and every thing in every

language we pick up. It's a large part of why we assume that kids have such a challenge

in front of them when it comes to language acquisition, as we've talked about before.

And this applies to nearly all concepts in language! Pretty much everything you can think of

in the world has no deep-rooted connection between what it is and the words we use to describe it.

One exception to this is onomatopoeia, or words that imitate the sounds they're associated

with. So English speakers decided a long time ago that when you hear a bee fly by your ear,

and it makes that [bzzzz] sound, that's a buzz. The word is pretty much the same as

the noise coming from the insect. It's not arbitrary – there's a clear link between the two.

Except… how true is this, really? If we look across different languages, we can often

find things that sound really different showing up as onomatopoeia for the same thing. Like,

take our horse from before, and picture it prancing down a nice cobblestone street. What

sound does it make? Well, in English, it goes clip-clop. But in Turkish, the same horse

running would go digidik-digidik, or in Punjabi it would go tun-tuk, or in Japanese, paka-paka.

In each case, you can imagine how the sound got turned into the onomatopoeic word, but

the results aren't close to being the same. And you can find this for lots of similar

onomatopoeic activities – roosters crowing, people sneezing, dogs barking, and so forth.

But even if there are differences, we still have this intuition that underneath, these

aren't quite as arbitrary as calling a horse a horse instead of an uma. And we can show

that through psycholinguistic experiments. A 2016 study dug into this, using Japanese

ideophones, or onomatopoeic words. See, Japanese has this whole class of these ideophones that

get used to express not just things like quacking or laughing, but lots of other adjectives

And they usually come in the form of two pairs of two syllables each. Like, excited is わくわく [wakɯwakɯ],

or exhausted is へとへと [hetoheto]. There are a ton of these pairs, but they should be less

immediately associated with the concepts underneath them, right? I mean, there's nothing particularly

fluffy about a word like ふわふわ [ɸɯwaɸɯwa]. A fluffy horse doesn't go ふわふわ as it fluffs on by.

To check into this, the researchers in the study took 38 of these words, and gave

them to native Dutch speakers, who didn't know any Japanese. The participants got shown

pairings of each Japanese word with a Dutch translation twice. But not all of them had

the right Dutch word: half of them got matched with the correct translation, and half of

them with the opposite translation. So, like, ぼろぼろ [boɾoboɾo] for worn out would

get the Dutch “worn out” versleten, but うきうき [ɯkiɯki] for happy would get paired with

After that, the participants were shown another set of word pairs: one of the Japanese words

with either the Dutch translation they'd already seen, or with some other random Dutch word.

They were asked to hit a button as quickly as they could to show whether this pairing

was one they'd seen already. And even after only having two exposures, something really

cool already turns up: it was a lot easier for the Dutch participants to remember a pairing

if the translation was correct in the first place. So they memorized the pairs where both

the Japanese and the Dutch meant the same thing 86.1% of the time, versus only 71.1% for the

ones where the words had opposite meanings.

And that wasn't the end of their evidence. The researchers also included a follow-up

part, where they confessed to the participants that they'd been devious, and that half the

Japanese words they'd learned were actually the opposite of the Dutch words. But they

didn’t tell them which half. Instead, participants were given all the Japanese words, and this

time, they had to choose between two opposites for the Dutch translation.

One opposite was always the Dutch word that they'd originally seen, and the other was something

new: so, like, if they got the word for energetic, きびきび [kibikibi], they'd have to choose between

the Dutch words for energetic vs. tired. If they were just picking randomly, knowing that

half of the translations they’d learned before were wrong, their success rate should

be at chance, or about 50%. But here, even after being told to disregard everything they'd

heard, participants still picked right 72.3% of the time.

But maybe this is just something about Japanese and Dutch – maybe the word matching between

the languages is better than we'd think. But ah, those careful researchers. They also

ran a study where participants went through the exact same steps, except with regular

Japanese adjectives. You know, the kind that are supposed to be totally arbitrary, with

no deep connections between the sounds, like 高い [takai] being tall or 安い [yasɯi] being cheap.

And with these words, there was less of an effect. They didn't remember the real

vs. opposite words any better, and while they were still above chance for the guessing between

two opposites part, their performance wasn't as high as with the ideophones.

So it looks like there's some meaning gain to the ideophones themselves – there really

is something fluffy about ふわふわ. And it's not just here where this looks like it

can be helpful – it may also ease word learning in kids, as we'll talk about back on our website.

But why should this be the case? Well, the power of these words probably stems at least

partly from the sounds themselves, which seem to come with certain associations. You probably

don’t think of [k] or [f] as having anything particular meaning associated with them. They’re just sounds.

But let’s try something that’s a bit more intuitive, based on an experiment first done

in 1929. Take a minute and look at these two shapes. All right? Now, if I told you that

one of them is a molmo and the other is an ikitik, which would you think is which?

The vast majority of people will associate the blobby shape with lower, rounder vowels

like [ɑ] or [o], and voiced labial consonants like [m] or [b], and smooth liquid [l]s. The

sharp, pointy shape will probably get voiceless stops like [k] or [t], high vowels like [i]

and [ɪ], and other sharp, pointy sounds. And this effect has been shown in languages from

English to Spanish to Tamil! This suggests that there might be a deeper association between

sounds and shapes, which has led some people to believe that the words we associate with concepts

are somewhat less than random, and that may connect into the differences in learning ideophones.

And it also explains why so much effort goes into picking out new product names. When a

new brand or product is getting ready to go on the market, a really important part of

the process is coming up with a good name. You want people to get the right idea about

Subtitles ListPlay Video

Do Sounds Carry Their Own Meanings? Onomatopoeia and Arbitrariness of the Sign

perceive

imitate

associate

random