This chapter reports on a cross-linguistic developmental study comparing speech and gesture produced in narratives by adults and children speaking either French (a non-pro-drop Romance language) or Zulu (a pro-drop Bantu language). We asked 72 participants (French: 12 adults and 24 children; Zulu: 12 adults and 24 children) to narrate a short silent cartoon. Zulu narratives were more detailed and contained fewer comments than the French. Zulu-speaking participants produced more representational and fewer pragmatic gestures than their French-speaking counterparts. Language differences do not explain this result. Rather, the findings support the gesture-speech co-expressivity framework and suggest an impact of literacy practice norms on multimodal narrative performance during later language acquisition.