Shape and function cues have been effective in eliciting the names of taught words (Capone & McGregor, 2005; Capone Singleton, 2012). However, shape cues provide the additional benefit of extending those names to untaught exemplars. Here, we compare shape and indicator cues (a new control), but test children repeatedly over time and task, to discern the effect of shape cues. We use error analysis, naming performance and word extension tasks. In particular, error analysis and word extension to never-before-seen-exemplars help disambiguate what aspects of the semantic-, lexical- and linking of the two are strengthened by shape cues.