A measure for cluster proximity (MCP) in child speech.

The proximity of consonant clusters in child speech is measured in the literature by the proportion of adult-like produced clusters to targeted clusters. With this measure, all clusters that are produced adult-like score 100%, while all others score 0%. Consonant clusters in child speech go through three main developmental stages: omission, reduction, two-member production (as targeted or substituted). Several authors claim that two-member clusters produced as two members should be considered acquired whether realised as targeted or not. In view of all these, a measure is needed that quantitatively differentiates all possible realisations of two-member clusters: omission, reduction, vowel epenthesis, two-member production, distinguishing substituted consonants from adult-like ones. Such a measure is proposed in the present article. To demonstrate its validity, the measure is applied to cluster productions of typically and atypically developing monolingual and bilingual children cross-linguistically. Moreover, it is applied to a bilingual Greek-English child's consonant clusters at age 2;7, when all cluster realisations are present. The aim is for the measure to be used in establishing norms for each stage in cluster development which will, in turn, guide assessment and intervention in phonologically delayed/disordered children.