Most words in most languages are arbitrary, which means that there is no connection between the sounds of the words and the meanings that they express. For example, in the word dog, it's not like d means pet, o means four legs, and g means likes rolling in muddy puddles. However, all languages have sets of words where there is a direct connection between sounds and meaning. This ranges from simple onomatopoeia to ideophones, which are distinctive sound-symbolic words which depict all kinds of sensory imagery.

Lockwood and colleagues showed that Dutch people with absolutely no knowledge of Japanese are able to accurately guess the meanings of Japanese ideophones. When asked whether the ideophone fuwafuwa means pluizig (fluffy) or houtig (wooden), over 90% of people will correctly say that it means pluizig, even though they have never heard the word before. This sensitivity also affects word learning. The researchers taught their participants 38 Japanese ideophones... except that half the ideophones were learned with their real Dutch meanings (e.g. fuwafuwa and pluizig), and half the ideophones were learned with their opposite Dutch meanings (e.g. kirakira and dof, whereas it actually means fonkelend). In a test, the participants learned the ideophones with the real translations much better than the ideophones with the opposite translations, scoring 86.7% for the real translations and only 71.3% for the opposite translations.

This study also used EEG to measure participants' brain activity. The researchers found that there was a difference during the test phase; the ideophones learned with their real translations elicited a bigger P3 response than the ideophones learned with their opposite translations. The most interesting result was when looking at the individual participants' brain activity and correlating it with their ideophone sensitivity scores. The P3 in response to the ideophones in the real condition was similar for all participants, but the P3 in response to the ideophones in the opposite condition varied a lot. People who were more sensitive to sound symbolism had lower P3s in response to the ideophones in the opposite condition, while people who were less sensitive to sound symbolism had higher P3s in response to the ideophones in the opposite condition. In fact, people who were least sensitive to sound symbolism showed the same brain activity to all ideophones.

What this means is that everybody can identify and exploit sound symbolism to help word learning when the sounds and the meanings match (like fuwafuwa and pluizig). However, people who are more sensitive to sound symbolism are also more aware of when things don't match (like kirakira and dof), and they find it harder to ignore these clashes during word learning.

More information:

Lockwood, G, Hagoort, P and Dingemanse, M 2016 How Iconicity Helps People Learn New Words: Neural Correlates and Individual Differences in Sound-Symbolic Bootstrapping. Collabra, 2(1): 7, pp. 1–15, DOI: http://dx.doi.org/10.1525/collabra.42

Data and analyses scripts:

https://osf.io/ema3t/