Iconicity in English and Spanish and Its Relation to Lexical Category and Age of Acquisition

Abstract

Signed languages exhibit iconicity (resemblance between form and meaning) across their vocabulary, and many non-Indo-European spoken languages feature sizable classes of iconic words known as ideophones. In comparison, Indo-European languages like English and Spanish are believed to be arbitrary outside of a small number of onomatopoeic words. In three experiments with English and two with Spanish, we asked native speakers to rate the iconicity of ~600 words from the English and Spanish MacArthur-Bates Communicative Developmental Inventories. We found that iconicity in the words of both languages varied in a theoretically meaningful way with lexical category. In both languages, adjectives were rated as more iconic than nouns and function words, and corresponding to typological differences between English and Spanish in verb semantics, English verbs were rated as relatively iconic compared to Spanish verbs. We also found that both languages exhibited a negative relationship between iconicity ratings and age of acquisition. Words learned earlier tended to be more iconic, suggesting that iconicity in early vocabulary may aid word learning. Altogether these findings show that iconicity is a graded quality that pervades vocabularies of even the most "arbitrary" spoken languages. The findings provide compelling evidence that iconicity is an important property of all languages, signed and spoken, including Indo-European languages.

Fig 1. Relationship between words’ lexical category and iconicity ratings based on written and auditory English stimuli.

estimates of an alien’s accuracy in guessing an English word’s meaning from its sound, iconicity ratings of written Spanish stimuli with verbs in infinitive form, and in 3^rd person singular form. The measure of guessing accuracy was rescaled to match the scale used in the other experiments for graphing purposes only. Error bars represent standard error of means.

Fig 2. Relationship between AoA and iconicity ratings as measured by coefficients from each mixed effects regression model from each experiment.

The coefficient represents increases in iconicity per each 1 percent increase in proportion of children that produce a given word at 30 months. The left-most bars depict coefficients from analyses controlling for effects of the basic set of covariates: frequency, number of phonemes, number of morphemes, concreteness, and association with babies for English language Experiments 1–3; frequency, number of phonemes and number of morphemes for Spanish language Experiments 4–5. The middle group of bars depict coefficients from analyses controlling for the basic set of covariates plus systematicity. The right-most bars depict coefficients from analyses controlling for the basic set of covariates and excluding all onomatopoeia and interjections. Estimates of alien accuracy were rescaled to match the scale used in the other experiments for graphing purposes only. Error bars represent standard error of means. *** indicates p < .0001, ** indicates p < .01, * indicates p < .05.

Fig 3. Relationship between AoA in each language and iconicity ratings in each language for translation pairs.

The bars show the coefficient from a multiple regression predicting iconicity from AoA in each language. English iconicity ratings are based on the average judgments obtained from Experiment 1 (written words). Spanish iconicity ratings are based on the average judgments across Experiments 4 (infinitive verbs) and 5 (conjugated verbs). Error bars represent standard error of means.