w/ Mark Dingemanse and Mits Ota


Japanese ideophones are often used as a model system for exploring the purported learnability benefit for motivated associations between words and meanings: Japanese mothers preferentially use ideophones when communicating with their young children, and both English and Japanese children find ideophones easier to learn than normal adjectives. This suggests a special status for ideophones in aiding acquisition- one which is most often traced to the motivatedness of the associations between sound and meaning. We might suggest, for example, that ‘gorogoro’ is an appropriate label for a heavy object rolling for a number of reasons:

  1. It contains the rounded vowel /o/, which is known to be sound-symbolically associated with size
  2. It contains the voiced consonant /g/, which is known to be sound-symbolically associated with size. This association is further highlighted when contrasted with ‘korokoro‘, which means a light object rolling.
  3. The reduplication of ‘goro’ is itself iconic of the process of rolling, which is an event that repeats both temporally and spatially.

Previous researcher have suggested that these reasons all enhance the learnability of words like ‘gorogoro’, and experimental work in artificial language learning contexts has broadly supported this possibility: broadly speaking, motivated associations between words and meanings enhance learning. ‘Gorogoro’ and other ideophones should not, however, be considered entirely in isolation. Previous research exploring sound-symbolic phenomena like the Bouba-Kiki effect, for example, often makes use of sets of words that are systematic, in addition to being motivated. As suggested in my PhD Thesis (here), some previously reported results suggesting a learnability benefit for motivatedness can be partially explained as a learnability benefit due to systematicity.

Might the same be true for ideophones like ‘gorogoro’? Ideophones commonly make use of reduplication at rates much higher than the rest of the lexicon, after all, so why is this the case? For an ideophone like ‘gorogoro‘ the answer seems straightforward- reduplication is itself iconic, but what about ‘iraira‘ (angry). Here the segment ‘ira’ is not reduplicated as an intensifier or in any straightforwardly iconic way (‘ira‘ is not a morpheme meaning angry, with ‘iraira‘ meaning very angry). This project seeks to explore the role of reduplication in ideophones and its relationship to language learning (see this project for a related exploration).

Broadly, we see three possible (non-mutually exclusive) roles for the function of reduplication

  1. Reduplication is, itself, iconic
  2. Reduplication is a disambiguating cue
  3. Reduplication aids in memorability (both not because of iconicity or systematicity)

Hypothesis 2 above is the central hypothesis that we are interested in exploring in this project. We suggest that the use of reduplication (and other phonotactic markedness common to ideophones) might serve as a signpost for language learners, one that says look here for iconcity, and that thus might influence expectations about associations between features of the word and its meaning (pointing, for example to the motivated use of rounded vowels and voiced consonants in a word like ‘gorogoro’).

We explore this possibility using an experimental approach modified from Dingemanse et al. (2016), which allows us to disambiguate between support for the above three hypotheses using native speakers of both English and Dutch.