Is learning /r/ the same as learning to speak more clearly?

The English alveolar approximant /r/ is a late-acquired sound, and many children with speech sound disorder (SSD) require intervention for misarticulated /r/. However, it is unclear whether improvement in /r/ production seen in typical development or in children with SSD receiving therapy is a consequence of—or concurrent with—overall increases in the clarity of their speech. It is possible that /r/ learning reflects developmental increase in speech clarity. To explore this relationship, we examined changes in the degree of distinctiveness between /r/ and /w/ and a measure of clear speech, vowel space dispersion, in 145 typically developing children who participated in a longitudinal study when they were four and five years old. The distinctiveness of children’s /r/ and /w/ productions was measured as the difference in F3–F2 distance between /r/ and /w/, and vowel space dispersion was measured as mean of Euclidean distances between vowel tokens and center of F1–-F2 vowel space. Preliminary results show that increases in vowel-space size across time-points did not correlate with increases in /r/-/w/ distinctiveness. However, /r/-/w/ distinctiveness at age four negatively predicted vowel space dispersion at age five. Together, these results suggest that /r/-/w/ learning does not reflect changes in overall speech clarity.