2 December 2020

Measuring phonological systematicity in infants’ early words

Catherine Laing

Early in development, infants’ words are often very simple in structure: Early productions are phonologically systematic (Laing, 2019; Vihman, 2016), and reliant on phonological structures that are simple to produce, such as consonant harmony (/beibi/ to represent baby) and open CV syllables (/da/ to represent dog). Furthermore, many of infants’ earliest words are phonologically similar: in data from their bilingual (English-Spanish) daughter’s early word acquisition, Deuchar and Quay (2000) show that 13 of her first 20 words are produced with a CV structure, and many words are phonologically identical: she produces car, clock, casa ‘house’ and cat as /ka/, and papa ‘daddy’, pájaro ‘bird’ and panda as /pa/. This points to continuity in the transition from babble to words; Vihman (2017) proposes that infants draw on what is most accessible to them in early word production, with earliest words matching the simple structures and stop consonants produced in babble.

In this talk, I analyse the phonological similarity of infants’ developing lexicon to observe development of systematic phonological structures over the first three years. Using data from the Providence corpus (Demuth et al., 2006), I use network analysis to model phonological similarity of over 140,000 word productions from 5 infants (2 males) between ages 0;11 to 3;10.