Comparative Markedness and Induced Opacity.

Results are reported from a descriptive and experimental study that was intended to evaluate comparative markedness (McCarthy 2002, 2003) as an amendment to optimality theory. Two children (aged 4;3 and 4;11) with strikingly similar, delayed phonologies presented with two independent, interacting error patterns of special interest, i.e., Deaffrication ([tɪn] 'chin') and Consonant Harmony ([ɡɔɡ] 'dog') in a feeding interaction ([kik] 'cheek'). Both children were enrolled in a counterbalanced treatment study employing a multiple base-line single-subject experimental design, which was intended to induce a grandfather effect in one case ([dɔɡ] 'dog' and [kik] 'cheek') and a counterfeeding interaction in the other ([ɡɔɡ] 'dog' and [tik] 'cheek'). The results were largely supportive of comparative markedness, although some anomalies were observed. The clinical implications of these results are also explored.