Previous papers in these pages have dealt empirically with the child's first words, the child's first imitations, and the use of yes/no and wh-questions with infants. In this study, we touch on all these issues, but attempt to place them in a systemic-functional language framework and a cultural-historical learning one. First, we deal with some of Ruqaiya Hasan's acute criticisms of Vygotsky, using one of Vygotsky's untranslated last lectures to show that he was cognizant of some of these objections and moving to meet them. We then trace a child's responses to bilingual questions up to and including the moment when the child can answer. We find that the child doesn't learn by repeating words but instead by construing and re-enacting a whole interpersonal exchange. This appears to validate both the Halliday-Hasan critique of Vygotsky and Vygotsky's later approach.