At present, quality earth science education in grade school is rare, increasing the importance of post-secondary courses. Observations of post-secondary geoscience indicate students often maintain errant ideas about the earth, even after direct instruction. This qualitative case study documents model-building activities of students as they experienced classroom instruction that braids history, inquiry, and model-based-learning within the context of earth dynamics. Transcripts of students' conversations, and their written work indicate students primarily employed model accretion to enhance their mental models. Instances of accretion were descriptive, pertaining to what their model consisted of, as opposed to how it explained the target phenomenon. Participants also conflated "continent" with "tectonic plate" and had difficulty attributing elastic properties--the mechanism for earthquakes--to rocks or "plates." We assert that the documented learning difficulties resulted from use of the metaphor "tectonic 'plate,'" reinforced by other everyday experiences and meanings. We suggest students need time with new models or concepts to develop strong descriptions before developing explanations. They need concrete experiences and explicit discussions concerning mapping those experiences to concepts. Lastly, because students often apply common meanings to scientific terms, we should not ask if they understand, but ask how they understand the concept.