Opportunities for computational modeling bring about new and productive uncertainties in students' explanations of the relationship of phenomena “in the world” to their models of those phenomena. Computational abstractions that become useful in the lifeworld of the classroom are steeped in the history of how those abstractions came to be used and understood. In this paper, we describe how two elementary classrooms used a programming and modeling language in ways that link their material enactments, measurements, and paper-based forms of representations to productive computational abstractions. The heterogeneity of computational abstractions––and how they came to enter the shared language within these classrooms––have implications for our understanding of how learners perceive their shared production of scientific explanations and the computational and non-computational tools they use. Findings have implications for the design of programming environments and for K12 computational modeling, both in regard to science education and computing education.