There's a debate that flares up every few years in math education. One side says we've gone too far toward "understanding" — students don't know their facts, can't solve equations cleanly, can't compute. The other side says we've gone too far toward procedures — students get answers without knowing what they mean, and the minute a problem looks different from the homework, they're stuck.
Both sides are describing real classrooms. The question is what to do about it.
It helps to be precise about what these two things actually are. Procedural fluency is the ability to carry out a method — solve the equation, find the slope, compute the mean — accurately and without unnecessary effort. Conceptual understanding is harder to pin down, but here's a working version: it's the connected mental structure that lets a student handle a problem they haven't seen before.
Robert Skemp made this distinction fifty years ago with an image that still earns its keep.1 Imagine two people who both know how to get from their house to the grocery store. The first has memorized the route — left at the corner, three blocks, right at the light. They can do that trip every time. But ask them to get to the grocery store starting from the library, and they're stuck. They don't have a route for that. The second person has a mental map of the neighborhood. They've also walked from home to the store many times — they know that route cold — but they could also tell you how to get there from the library, or from anywhere else, because they understand how the streets fit together.
Both people can get to the store. One of them can also do everything else.
This is the case for conceptual understanding, but it's also the case against a false choice. The person with the mental map didn't develop it instead of learning routes — they developed it through walking the routes, paying attention, noticing how this street connects to that one. The relationship runs both ways. Procedural work builds conceptual understanding when students are paying attention to structure; conceptual understanding makes procedures stick because students can reconstruct a method they've half-forgotten instead of guessing.
The research on this is clearer than the public debate suggests. Procedural and conceptual knowledge develop together, each supporting the other, and the strongest instruction builds both at the same time.2 The classroom studies that compared traditional and reform approaches found the same thing from a different angle: the highest-leverage move wasn't a curriculum or a method, it was sustained attention to meaning *while* students were doing the math.3
What does this look like on a Tuesday?
It looks like a teacher who, when a student solves 3(x - 2) = 15 correctly, doesn't just say "good" and move on. They ask: Why did you start by distributing? Could you have divided both sides by 3 first? Would that have worked? The student practiced a procedure, and the question opened a small window into structure — that the equation has a multiplicative grouping, that there's more than one entry point, that the choice has consequences.
It looks like comparing two student solutions to the same problem and asking what's the same and what's different about the moves they made.4
It looks like asking students to predict what a graph will look like before drawing it, then noticing where the prediction missed.
None of these moves abandon procedural work. They sit on top of it. The student still has to solve the equation, draw the graph, find the slope. The difference is whether the procedural work builds a map or just adds another memorized route.
That's the work. Not picking a side. Building both at once, on purpose, with the kind of small instructional moves that fit into a regular lesson on a regular day.
Endnotes
1 Skemp's original 1976 article distinguished "instrumental understanding" (knowing how to do something) from "relational understanding" (knowing how and why, and how things connect). His town-and-map analogy is the source of the neighborhood image used here. Skemp, R. R. (1976). Relational understanding and instrumental understanding. Mathematics Teaching, 77, 20–26.
2 Rittle-Johnson and Star's research on equation solving found that procedural and conceptual knowledge are mutually reinforcing — gains in one tend to produce gains in the other. Rittle-Johnson, B., & Star, J. R. (2009). Compared to what? The effects of different comparisons on conceptual knowledge and procedural flexibility for equation solving. Journal of Educational Psychology, 101(3), 529–544.
3 A synthesis of research across very different teaching traditions identified two features that consistently predicted student learning: explicit attention to concepts during instruction, and giving students room to struggle with important mathematics. Hiebert, J., & Grouws, D. A. (2007). The effects of classroom mathematics teaching on students' learning. In F. K. Lester (Ed.), Second handbook of research on mathematics teaching and learning (pp. 371–404). NCTM.
4 The instructional move of comparing solutions — same problem, different approaches — has a strong evidence base for building both procedural flexibility and conceptual understanding. See Rittle-Johnson & Star (2009) above; also Star, J. R., & Rittle-Johnson, B. (2009). It pays to compare: An experimental study on computational estimation. Journal of Experimental Child Psychology, 102, 408–426.