A corpus of consonant-vowel-consonant real words and nonwords: comparison of phonotactic probability, neighborhood density, and consonant age of acquisition

Abstract

A corpus of 5,765 consonant-vowel-consonant sequences (CVCs) was compiled, and phonotactic probability and neighborhood density were computed for both child and adult corpora. This corpus of CVCs, provided as supplementary materials, was analyzed to address the following questions: (1) Do computations based on a child corpus differ from those based on an adult corpus? (2) Do the phonotactic probability and/or the neighborhood density of real words differ from those of nonwords? (3) Do phonotactic probability and/or neighborhood density differ across CVCs that vary in consonant age of acquisition? The results showed significant differences in phonotactic probability and neighborhood density for the child versus adult corpora, replicating prior findings. The impact of this difference on future studies will depend on the level of precision needed when specifying probability and density. In addition, significant and large differences in phonotactic probability and neighborhood density were detected between real words and nonwords, which may present methodological challenges for future research. Finally, CVCs composed of earlier-acquired sounds differed significantly in probability and density from those composed of later-acquired sounds, although this effect was relatively small and is less likely to present significant methodological challenges to future studies.

Figure 1

Scatter plots of child versus adult positional segment sum (top), biphone sum (middle), and neighborhood density (bottom). Solid line indicates the linear regression fit line. Dashed line is a reference line indicating a perfect correlation.

Figure 2

Normalized (i.e., z-score) positional segment sum (top), biphone sum (middle), and neighborhood density (bottom) by consonant acquisition class for real words based on the adult (open bar) or child corpus (vertical line bar) and nonwords based on the adult (filled bar) or child corpus (dotted bar). Error bars indicate standard errors.