Bouba/Kiki Effect • Hacker News Board

Re: Bouba/Kiki Effect

#71

Post by hyperpape » Tue, Jul 20, 2021, 2:23 AM UTC

beaconstudios wrote: post #68

twelvechairs wrote: post #45

Earlier quoted context omitted.

I don't understand what point you are trying to make. You seem to have made exactly mine. Symbolism is a choice. The Bouba/Kiki thing is not.

they're arguing that symbols are not arbitrary, which is true but I think they inferred that you were making that argument as opposed to the softer "symbols are conventional" one that you did make. Convention does not necessarily imply arbitrariness. If symbols were arbitrary then maybe a rabbit could symbolise ferocity, but they are instead more like metaphors that need to tie into something concrete.

Thanks, this is really helpful. I did in fact think that the use of convention implied more arbitrariness. The convention that we drive on the right can be reversed. But I suppose there can be conventions which are partially constrained by some underlying phenomenon.

Re: Bouba/Kiki Effect

#72

Post by elihu » Tue, Jul 20, 2021, 2:42 AM UTC

wyager wrote: post #8

Bouba/Kiki are onomatopoeia for the sounds objects of those shapes are liable to make if you drop them or otherwise physically interact with them. Bouba is probably soft/flexible and Kiki is probably hard/rigid.

That was my initial reaction. "kiki" sounds like hard, sharp things clattering against each other like rocks or broken glass, and "bouba" sounds like something flexible and bouncy.

Re: Bouba/Kiki Effect

#73

Post by dr_dshiv » Tue, Jul 20, 2021, 3:03 AM UTC

twelvechairs wrote: post #36

gfaure wrote: post #4

There's a Tom Scott video about this with some further academic references: https://www.youtube.com/watch?v=1TDIAObsqcs Bouba/kiki is most likely a specific instance of the more general notion of https://en.wikipedia.org/wiki/Sound_symbolism . Another well-known cross-linguistic association is the one between front vowels and small objects on one hand, and back vowels and large objects on the other.

'Symbolism' is a terrible word to describe something like this. Symbols are convention and can change. This is the opposite - an innate cross-sensory connection that will not change.

Agreed. The presymbolic term people might be looking for is "resonance." Read early JJ Gibson (who introduced affordances). He talks about this dichotomy.

Re: Bouba/Kiki Effect

#74

Post by dcow » Tue, Jul 20, 2021, 4:44 AM UTC

So now we’re going to have to check our internal shape bias and correct algorithms that encode as much into classification of arbitrary objects. Great.

Re: Bouba/Kiki Effect

#75

Post by 310260 » Tue, Jul 20, 2021, 4:54 AM UTC

dcow wrote: post #74

So now we’re going to have to check our internal shape bias and correct algorithms that encode as much into classification of arbitrary objects. Great.

I'm gonna assume you're kidding.

Re: Bouba/Kiki Effect

#76

Post by qalmakka » Tue, Jul 20, 2021, 5:27 AM UTC

If I think about it I've always instinctively associated unvoiced sounds like /k/ and /t/ to sharper shapes, and sounds like /b/ or /v/ to "rounder", softer shapes. I think that voicing makes everything "smoother", in some way?

Re: Bouba/Kiki Effect

#77

Post by dcow » Tue, Jul 20, 2021, 6:07 AM UTC

310260 wrote: post #75

dcow wrote: post #74

So now we’re going to have to check our internal shape bias and correct algorithms that encode as much into classification of arbitrary objects. Great.

I'm gonna assume you're kidding.

I wish I were 100% kidding.

Re: Bouba/Kiki Effect

#78

Post by wodenokoto » Tue, Jul 20, 2021, 6:28 AM UTC

antiterra wrote: post #66

kazinator wrote: post #59

The primary author of the Dictionary of Japanese Grammar book series, Makino Seiichi, presents a related hypothesis about sound symbolism in Japanese. He identifies a pattern that "hard" sounds like K often appear with words associated with distance between the speaker and the subject, whereas sounds like N are associated with less distance. For instance, the clause nominalizing particle KOTO expresses a more objecti…

But koko/kore/kono vs soko/sore/sono doesn’t seem to match that hypothesis. How do you make sure you’ve got enough examples to prove the argument?

The structure of the argument as presented in the book is

“we see this sound symbolism in onomatopoeia” and gives a very large list of examples, followed by “this can be used to explain other phenomenons such as [insert parents statement]”

You are right that it is interesting how and if ko-so-a-do fits this pattern, but I don’t think it’s a “gotcha” against sound symbolism in Japanese.

Re: Bouba/Kiki Effect

#79

Post by jzshen » Tue, Jul 20, 2021, 6:44 AM UTC

This totally works for other things. "Mamy", "sira", "sakai", are words for various tastes (eg spicy) in Malagasy and I thought that they just sounded 'right' when I first heard them.

(mamy:sweet, sira:salty, sakai:spicy)

Re: Bouba/Kiki Effect

#80

Post by mohn » Tue, Jul 20, 2021, 6:54 AM UTC

horse666 wrote: post #46

jxramos wrote: post #17

Earlier quoted context omitted.

ps it's a joy to watch VS Ramachandran get excited about synesthesia, definitely worth hunting down videos of him on YouTube.

Yes, absolutely! I loved listening to the Reith lectures many, many years ago, when Vilayanur Ramachandran gave them. You can still listen to them here, http://www.bbc.co.uk/radio4/reith2003/ In Lecture 4 (Purple Numbers and Sharp Cheese) he talks about the Kiki / Bouba effect. His enunciation is amazing.

I was not able to play the RealMedia streams from BBC, but here's an alternative link to lecture 4 at 15m01s:

https://www.youtube.com/watch?v=Xttp5GpiVLs&list=PLez3PPtnpn...