Exploring meaning-sound systematicity in Korean

Studies of word-level meaning-sound systematicity in English and four other European languages have shown that words that sound similar tend to have similar meanings. The term ‘systematicity’ in this research tradition is defined as statistically non-arbitrary relations between sub-domains of language, in contrast to the traditionally assumed Saussurian arbitrariness. We explore such systematicity in a typologically distinct language, Korean. We find a relatively high level of systematicity, which we attribute to the method of analysis where we applied Latent Semantic Analysis based on eo-jeols—sequences of syllable-blocks bounded by spaces in an internet corpus of written Korean. Eo-jeols embody a psychologically realistic spectrum of linguistic structure and influence, compared with previous purely lexically based studies of systematicity. Systematicity was pervasive in our sample of the Korean lexicon—partitioned by word frequency, etymological origin, syllabic constituents (onset, vowel, coda, rhyme), syntactic categories, homonyms, onomatopoeia, and loanwords—suggesting a fundamental basis for systematicity. We explain meaning-sound systematicity in terms of related degrees of cognitive effort in speaking and listening.