The reverse case gets less attention, but I think it is just as important. To make it, I want to borrow a distinction from organizational theory between first-order and second-order learning. My wife is a cardiologist. As such, she often has conversations with patients about the relationship between their cardiovascular health and their lifestyle. After seeing her, patients often leave with both a prescription for a statin to help treat their cholesterol or heart condition, and a list of things they must consider changing about their lifestyle. Some patients will take the statin and, after seeing improvement in their cholesterol numbers, conclude there is absolutely nothing wrong with their lifestyle. Other patients will take the statin and make systematic changes to their diet, exercise, sleep, and the management of stress.
Both categories of patients engage in learning and change, but these are different categories of learning. The first category of patients, who only took the medication, engage in what organizational theorists would refer to as first-order learning. While there are different types of first-order learning, what they all have in common is that they leave the prevailing system intact. The medication makes the existing pattern marginally more sustainable; thus, they conclude there really is no need for system change. The second category of patients, who also change their lifestyle, engage in second-order learning, which fundamentally transforms the system (their lifestyle) that produced poor cardiovascular health to begin with.
Deeper learning and continuous improvement both have an orientation toward transformation. But they hold that orientation differently. In deeper learning, the need for transformation is an assumption you argue from, not towards. The starting premise is that our schools, as currently designed, are not built to consistently produce the relational, adaptive, identity-rich outcomes we say we want—and that no amount of tuning the existing system will get us there. In continuous improvement, by contrast, transformation tends to be an assumption you argue towards, not from. The implicit bet is that if you incrementally improve enough of the surrounding pieces, a transformed system will eventually emerge.
That distinction matters greatly. Improvement, by its nature, reduces dissonance. It makes the prevailing system feel a little better, a little more humane, a little more functional. The metaphor I find useful is a pot of water. Water does not transform into steam unless the heat stays high enough to release it as a gas. Improvement, applied within a system that is not having an explicit conversation about the need for transformation, can quietly turn down the heat. The system feels less broken. The pressure for fundamental change recedes. And improvement begins to function as validation—proof that there was nothing fundamentally wrong with the system in the first place.
This is what deeper learning offers continuous improvement: a way to front-load the conversation about transformation, so that the discipline of incremental work does not collapse into a quiet defense of the status quo. Continuous improvement can tell us whether things are getting better. It cannot, by itself, tell us what “better” means. Deeper learning can. Without clarity about the transformational intent, the rigor of improvement methods can be turned, quite efficiently, toward improvement goals that are not worthy of our efforts.
Deeper learning also offers a counter-argument to an oft-cited critique of continuous improvement: that improvement science is just incrementalism dressed up in better clothes. I take that critique seriously, and I believe the improvement community should too. But I think it confuses two different things: small steps and small ideas. Continuous improvement, properly understood, is not about pursuing modest ambitions. It is about testing big ideas at a small scale, so that we do not burn community will and trust by spreading something that does not yet work. Or as Tony Bryk advises, “learn fast so you can implement better.” In the presence of a transformational vision, increments are how a transformed future actually arrives.
Transformation is the result of disciplined, rigorous, incremental work, sustained over time, in service of an aspirational marker on the horizon. Deeper learning gives us that navigational bearing on the horizon.
