Preface

AI hasn't created a teaching problem — it has exposed it. Modern schooling teaches for completion, not for cultivating thinkers. This creates a gap between students completing assignments and those assignments successfully developing the thinker. This also leads to a practice gap for teachers. This book is an attempt to close that gap. It couldn't be more timely: current use of AI in education threatens to deepen exactly the compliance our schooling inadvertently produces, and yet it can support solving it.

The practice gap is visceral. You feel when a room has gone quietly compliant, when a student who is "participating" has actually checked out and is instead copying their neighbor's or an AI's answers — there's no thinking there. We need to cultivate the thinking of students without making our jobs harder — and doing so makes teaching more satisfying.

By some estimates, teachers already have to make a decision every three seconds.¹ We need teaching decisions framed before lessons start. By the end of the book you should be able to name which kind of thinking students present with, and what kind of response it calls for. In fact, you will be able to support student agency, which I define as capabilities + decision-making + origination + answerability. You will also come to see which forms schools currently make visible and which they keep hidden.

I've worked under many different educational conditions: as a homeschooling parent, a Montessori cooperative founder, a researcher in democratic schools, an adjunct in university classrooms, and a high school science teacher. This has given me a broad perspective on what students can do and how teachers can support them.

A note on method, so you know what kind of book you are holding. This is a case study of one teacher: my classrooms, and the sites I have taught and researched in, observed over years and set in dialogue with the theorists I have read along the way. It is a dialogue with them, not a meta-analysis of the literature. I credit where I can, and I have tried not to get lost in historical trails or esoteric side-quarrels. Where researchers have tested something I have only watched, I cite them; where I make a claim from my own room, I say so.

And one thing worth saying plainly, because readers will wonder where I stand. I do not think race or class is the engine of what this book describes; the engine is economic. Nothing inherent in race determines school performance, and class as such confers no protection: wealth without intellectual mentorship fails too. Where race enters, and it does enter, I will show it entering through this book's own mechanism. The full argument is in Chapter 4.

Years of chronic illness forced me to think iteratively: notice carefully, change one thing, observe, revise my understanding. Pain keeps you honest; you cannot prescribe a solution before you have observed the problem, nor can you invoke a solution because it's ideal or has been sold to you. That habit of attention — holding the tension between what you want to happen and what is actually happening — is something teachers do daily. At the same time we do so while surveilled (pacing guides, walkthroughs, standardized tests), which deepen — and, I'll argue, often create — the practice gap. The constraints are real.

Answerable Teaching argues that the crisis commonly called student disengagement is often a crisis of recognition. Students make different kinds of bids to act as knowers — through partial explanations, reframings, refusals, social claims, and self-authored pursuits — but schools tend to recognize only those that produce legible, curriculum-aligned answers. Drawing on classroom cases across public, progressive, and democratic schools, Katherine von Duyke offers a taxonomy of agency and a new ethic of teaching: teachers must become answerable to the forms of thinking students actually bring.

A note on the students in these pages. Every name is a pseudonym, and identifying details have been changed. Cases drawn from research were collected under appropriate consent arrangements; cases from my own classrooms are described with details altered to prevent identification. I have tried throughout to render what each student was doing accurately, even where accuracy is less flattering to me than a smoother version would be. They are not here to illustrate my successes. They are here because they illustrate a form of thinking that taught me how to respond better.

Whether you are in a graduate seminar, in a for-credit professional development course, founding a school, or reading to better support your own children, you will find yourself nodding your head at familiar and frustrating scenarios. The first chapters give you the framework. The middle chapters give you the teaching moves (and I'll frame how AI can make these doable and offer you tools). The final chapters consider structural changes to school contexts making this work more possible for teachers.

— Katherine von Duyke, Wilmington, Delaware

Preface note

¹ Jackson (1968) actually reports a teaching decision roughly every twenty seconds; estimates vary. Jackson, P. W. (1968). Life in Classrooms. Holt, Rinehart and Winston. That count didn't have to account, back then, for every time a teacher has to put their password somewhere.