For the better part of a decade, Natalie Wexler has helped reshape the conversation about American education. At a time when the field was (and in many ways still is) preoccupied with the naïve belief that reading comprehension can be taught with a content-neutral junk drawer of “skills and strategies” lessons, she has made a clear and necessary case that knowledge matters and that curriculum, education’s long-neglected, red-headed stepchild, belongs at the center of efforts to improve student outcomes. Like the rest of us who have made this case, she walks in the footsteps of E.D. Hirsch Jr., who turned 98 this week and has spent more than four decades arguing that literacy and equity, meaningfully understood, depend on shared knowledge and a coherent curriculum.

In many precincts, that argument has largely carried the day. “High-quality instructional materials” (HQIM) are now a central reform strategy in states and districts across the country. This is real progress. As Hirsch himself and Dan Willingham wrote this week in Education Next, “knowledge is having a moment in education fashion,” a phrase they use advisedly, given the field’s long history of faddish enthusiasms.

But reading Wexler’s recent Substack on how to tell whether “a knowledge-building curriculum…actually deserves that label,” I found myself returning to a chronic concern of mine that’s only grown sharper over time. This concern is not about whether curriculum matters. It does. The question is whether the gratifying enthusiasm for knowledge-rich curriculum represents a durable shift in practice or simply the latest turn in a familiar cycle. And whether fine-grained curriculum evaluation matters in a system where curriculum—any curriculum—is not yet central to the student experience.

Wexler’s excellent dissection of the widely used Benchmark Advance ELA curriculum echoes and advances a pair of Substack posts by Olivia Mullins, an elementary science teacher with a Ph.D. in neuroscience. Mullins points out that the units in Benchmark Advance are organized by abstract “themes,” rather than a coherent set of topics, and sometimes fail to clarify important vocabulary or facts needed to understand readings. Wexler affirms Mullins’s argument that Benchmark “lacks both the depth and the coherence to do an effective job of building knowledge.”

Wexler and Mullins’s analysis is both careful and correct. You can’t judge a curriculum by surface signals—test scores, adoption lists, or ratings alone. And claims from publishers that their wares are knowledge-building should be taken with a grain of salt.

To know whether a knowledge-building curriculum actually deserves that label, you have to look at what it asks students to do all day. Are students reading rich, content-heavy texts and grappling with ideas that build over time? Are lessons organized around coherent topics that accumulate knowledge, rather than disconnected skills practice? Do classroom tasks reflect what we know about how learning works—retrieval, attention to meaning, and sustained engagement with content—rather than superficial activities?

The trouble is that even if the answer to all of those questions is “yes,” you still don’t know whether students will, in fact, get a “high quality” curriculum. You don’t know if instruction, as students experience it, systematically builds knowledge over time.

“As students experience it” is the rub. It is more revealing to watch what people do than what they say. And what educators do with curriculum suggests that, for all our talk about its importance, it does not yet command real authority. It is still treated as optional—something to be reshaped or replaced at will—rather than the backbone of instruction. This in turn suggests it may be premature to fetishize fine-grained differences among curricula: Happily, we are getting better at distinguishing strong curricula from weak ones on paper. But we have not yet built a system that reliably delivers any curriculum as intended.

A curriculum is a designed experience. It reflects deliberate decisions about what to teach, in what order, using which texts, and toward what ends. Done well, it reflects expertise and promotes coherence over time. But that is curriculum as design. Curriculum as practice in American schools is something else entirely.

Much education analysis can leave the impression that classroom practice is largely settled and that the only meaningful differences lie in implementation quality. Talk of a “Mississippi Miracle” or “Southern Surge,” for example, suggests a level of uniformity that exists only in our heads. From state to state, within districts, even across the hallway in a single school, you can encounter entirely different instructional experiences. In one, students may be reading a complex text and grappling with ideas. In the other, they may be completing activities only loosely connected to the core content. Both teachers are, technically, using the same curriculum, but in practice they have diverged.

Understand this is not an anomaly, it is our system. Teachers routinely adapt, supplement, skip, and replace lessons. This is not to suggest that teachers should simply “shut up and teach the script.” Thoughtful adaptation—grounded in evidence, aligned to the curriculum’s sequence, and responsive to student needs—is part of the craft. But the default assumption too often runs in the opposite direction: that the curriculum is suspect and that professional judgment necessitates modifying or replacing it. A healthier norm would reverse that presumption—treating the curriculum as sound unless there is a clear, evidence-based reason to depart from it.

Remarkably, even as more educators embrace “HQIM” and districts invest heavily in it, this habit of customization appears to be intensifying, not receding. Teachers are now using more supplemental resources than in prior years, not fewer. On average, teachers report drawing on multiple curriculum programs and roughly five additional resources, such as Share My Lesson and Teachers Pay Teachers, a number that has grown over time. In other words, the rise of “HQIM” has not displaced the long-standing norm of customization. It has been layered on top of it.

The result is a wide and consequential gap between what is intended, what is implemented, and what students actually experience. And it is that last category—the experienced curriculum—that ultimately determines what students learn.

Not incidentally, there is also a hidden cost to this culture of constant customization that’s rarely acknowledged: teacher time. I spent 20 to 30 hours a week early in my career cobbling together lessons from scratch or searching for materials to “engage” students, differentiate instruction, or teach content-neutral “skills and strategies.” Those hours would have been far better spent studying student work, refining my delivery, or communicating with families. One of the underappreciated benefits of a coherent, high-quality curriculum is that it frees teachers from reinventing the wheel, allowing them to focus on the parts of the job that matter most and that, unlike curriculum creation, only they can do.

To be clear, this state of affairs is not entirely the fault of teachers. David Steiner, director of the Johns Hopkins Institute for Education Policy, has described how even rigorous, thoughtfully designed curricula often remain, quite literally, on the shelf. At the core, he writes, is a mindset: Teachers “don’t believe their students can manage the rigor of grade-level HQIM instruction—thus, the avoidance and watering down.” But that’s not irrational or merely low expectations. Teachers are responding to classrooms where many students are below grade level and where the supports to bridge that gap are weak or incoherent. Steiner is persuasive that the onus is on publishers to include pre-unit diagnostic quizzes throughout the curriculum and on schools to use those results to design effective academic interventions.

Without these sorts of shifts in professional practice and norms, attempts to evaluate a curriculum, whether through test scores, reviews, or earnest deep-dive analyses like Wexler’s, are of limited use. We are observing what is printed, not what is delivered. It’s a bit like trying to infer children’s diet and nutrition by walking the aisles of a grocery store. The availability of healthy food is encouraging and tells you something about what is possible. It tells you very little about what children are actually eating.

This helps clarify what it means—or ought to mean—when we say a curriculum “works.” If a curriculum works, materials are used with coherence and purpose. Leadership is aligned. Expectations are shared. Professional development reinforces the curriculum rather than introducing competing priorities or initiatives. Teachers treat curriculum as a foundation rather than a suggestion.

In sum, we are facing not one challenge but two. The first and fundamental one is to establish the expectation that curriculum is central to the student experience and something to be used as designed. The second challenge is how to evaluate and valorize knowledge-rich curriculum as an essential tool of literacy and equity. Wexler is right to insist that ELA curricula should build knowledge in a careful and coherent way. But before we can confidently judge whether a curriculum is any good, we may need to ask a simpler—and more elusive—question: How can we tell what students are actually experiencing?

Until we can untangle the knot between what is intended, what is designed, what is taught, and what students actually experience, our claims about curriculum effectiveness will remain necessarily provisional—not because curriculum does not matter, but because, too often, it is not what students actually get.

The bottom line is that we are winning the intellectual argument about curriculum—but we have not yet won the operational one. We’ve barely tried.

Editor’s note: This was first published on the author’s Substack, The Next 30 Years.