This paper presents a cross‐sectional and longitudinal investigation into the development of a contrast in voicing production in simple and complex consonant clusters for 21 children. The results showed that, although the group results indicated a progression that may have been caused either by development in articulatory proficiency or by an growing underlying representation, individual children ... [Show full abstract] develop in a manner that is not consistent with a developmental model postulating only articulatory development. Abstract This paper examines variation in language development based on the case study of a Greek‐ speaking child with Specific Language Impairment in the phonological component. To account for variation, we extend the empirical coverage of partial ordering (Anttila 1997a,b) to language acquisition along the lines of Tzakosta's (2004) Multiple Parallel Grammars theory. The theoretical gain of the proposed model is that it provides a principled basis to define developmental paths and also draw a distinction between smart and non‐smart paths. This is stated as the Grammar Inclusion Hypothesis: advanced grammars consist of total orders which are proper subsets of early grammars; a smart path is one that subsumes the total order of the target grammar.