In science classrooms, the epistemic practices of explanation building and argumentation often extend over multiple episodes of talk during a single lesson or across several lessons. Analyzing this kind of discourse requires a way to identify patterns that emerge over time to better understand student participation and how teachers support students' disciplinary work. In this paper, we share the development of a unique graphic representation of classroom talk which we call barcodes. These barcodes assisted our analysis of when and how, over multiple points in a school year, three elementary science teachers facilitated students' science sensemaking during whole-class discussions in ways that ended up promoting, sustaining, or constraining students' collective development of ideas. Barcodes allowed us to see that each teacher regularly engaged students in rigorous whole-class talk over a school year, yet each classroom had distinct patterns of teacher involvement and activity sequences that preceded or co-occurred with these conversations. Paired with transcripts, barcodes illuminated a relationship between teacher responsiveness to specific student ideas and higher discursive rigor. Finally, iterative cross-referencing between barcodes and transcripts sparked further inquiries into supportive conditions for talk that were not as apparent using transcripts alone. In this way, the barcode functioned both as an analytical tool and a final visualization of discourse events in a series of lessons from grades 5 and 6 science classrooms.