Coaching Letter #225

On Practice Architectures

My well-thumbed copy of Kemmis et al.

Hello, I hope you are doing well. Thank you for subscribing to The Coaching Letter—I really appreciate it. Here in the northeast United States we are under about a foot and a half of snow—I didn’t realize until I saw the news reports how long it has been since we’ve had so much snow. Our house is on 2 acres and is surrounded by woods, so it’s very beautiful when the sun shines on the snow. But it’s very cold, and forecasted to get down to minus 5 Fahrenheit in the next few days. As I’ve said before, I have this weird compulsion to start off by talking about the weather—I’m sure it’s a habit I picked up from decades of almost daily phone calls with my dad, in which he always started by either asking me or telling me about our respective weather conditions. And it’s certainly easier than thinking about the mad things that are going on in the world. I’m a big believer in focusing on the part I can control, and what I can control is sharing with you ideas and tools we can all use to make our sliver of the world better and fairer for the adults, and more importantly, the children, who are in it. Anyway, wherever you are, I hope you are equipped for your weather.

This Coaching Letter is about Practice Architectures. I doubt you have heard this term, since it has never come up in any of the classes I have taken, any of the books and articles I have read, nor any of the conversations I’ve had with colleagues and clients, ever. I learned about it from ChatGPT when I asked “What concept or construct is adjacent to the work I do that we haven’t talked about yet?” and got back the ironic professorial “Here’s one I’m a little surprised we haven’t really unpacked yet, because it sits right next to your work and quietly explains a lot of what you’re already seeing: practice architectures.” I find ChatGPT’s self-anthropomorphizing mildly irritating. Nevertheless, I ordered the book it suggested: Changing Practices, Changing Education, by Kemmis et al., and it is indeed super useful.

This Coaching Letter is intended to bridge from the last Coaching Letter on the psychology of change (#224) to a larger emphasis on systems as key determinants of practice.

Here’s my thesis, stated in what I will admit are rather extreme terms, so as better to make my point. In educational organizations, AKA school districts, there is a powerful tendency to push implementation of strategic plans (and other guiding documents and policies) downward through the organizational hierarchy. This move allows those higher up the chain, AKA “Central Office” and principals, to offload responsibility for successful implementation and results onto those lower down, AKA teachers, justifying this as the morally upstanding practice of “holding teachers accountable.”

This practice was certainly true in two of the school districts I worked in. In one, the assistant superintendent, whom I liked as a person, would explicitly state, at the beginning of the year, that the board had given the superintendent goals, and the superintendent had said to the assistant superintendent, “My goals are your goals”, and now the assistant superintendent was telling us, the principals, “My goals are your goals”, and the implication was that we were going to go back to our buildings and say to teachers, “My goals are your goals”, and they were somehow going to figure out how to raise student achievement with little guidance from Central Office as to how they were supposed to do that. It was a strategic donut.

Teachers, at the same time, do themselves no favors by resisting this downward push through invoking rationales that just make them look resistant or defensive, such as: I don’t have time for this; Is this just another thing we have to do?; My kids aren’t ready for this; I tried that before but it didn’t work; This might work in theory, but our context is different; We’re already doing this; We tried this a few years ago and it fizzled out; This feels like it’s taking away my professional judgment/autonomy/creativity; I’m worried this will confuse students; Am I going to get dinged for this if I try it and it doesn’t go well? Is this going to be another walkthrough “look-for”?

I could go on. Teachers’ responses are justifiable, but do not address the root causes of behavior, which tend to have much more to do with systems and situations than they have to do with the traits of the individuals expected to enact change. Specifically, in the last Coaching Letter (#224), I wrote about the Fundamental Attribution Error, which is the general human tendency to overestimate the role of the individual as an explanatory factor relative to the situation or system in which the individual is operating. (I also used the term trait inflation, which I particularly love.)

Therefore, if the situation is more powerful than the individual, we should be looking to how systems are set up to explain why people are behaving the way they are, and the concept of Practice Architectures (PA) gives us a way of doing that. So, here’s what I’ve learned about PA.

What people can do in practice is prefigured by the architectures of the situation they’re in.

OK, so, the first thing here is the use of the word “can”. And not, for example, choose, decide, allowed, or even just do. This implies that there is a range of possibilities that people can operate with—I think there are technical terms for this, like permission structure or Overton window that do similar work of outlining what is acceptable thinking and doing in a given context. It also implies that people make choices according to what they know how to do, what is sustainable over time (because heroes burn out pretty quickly), what they think is valued, what they think they might be “dinged” for (which is code for some kind of risk or judgment), and so on.

Use of the word “can” also has other implications. It means that what people think they can do—in any sense—precedes what they decide to do. Which pokes holes in the autonomy argument, because it makes it obvious that choice is already constrained. It also means that we are looking at the situation more than the word “capacity” connotes, because capacity is about the person, and can is more ambiguous—it exists in some combination of personal and organizational factors. In other words, can may mean has the capacity to and chooses not to actually do it (because, for example, they don’t believe their kids are capable of rigorous learning tasks), or cannot may mean wants to do it but doesn’t believe that it’s an acceptable choice (because, for example, they think that their supervisor won’t understand it).

And finally, can places the responsibility for action more squarely on the system, because it raises the question: What must change in the system so that people can act differently? And the obverse: What is maintaining current practice as the most desirable choice? Not to put too fine a point on it: Teachers don’t fail to enact equitable, ambitious instruction because they lack virtue; they are operating inside architectures that make some moves easy, some moves costly, and some unthinkable.

OK, enough about can. Let’s make a brief digression into coherence. As I’ve written before (CL #218), coherence is not an end in itself, or a mechanism in itself, but results from, and contributes to, the desirable situation where all the parts of a system are aligned to support equitable instruction. So PA as a construct is very much aligned with coherence; it’s the degree to which the practice architectures across classrooms, teams, and leadership roles are mutually reinforcing rather than contradictory. So let’s actually describe the components of Practice Architectures.

When I said above that what people can do in practice is prefigured by the architectures of the situation they’re in, those architectures come in three intertwined forms.

1. Cultural–discursive: the language, concepts, and stories that make certain actions intelligible.

2. Material–economic: time, schedules, tasks, tools, pacing guides, room layouts, curriculum materials.

3. Social–political: power relations, norms, who gets to decide, who speaks, who is accountable to whom.

OK, so here’s a more detailed breakdown, framed in terms of high quality, equitable instruction:

1. Cultural–discursive practice architecture is all the language, concepts, mental models, and explanatory frames that bring clarity to challenging, grade-level instruction so that teachers understand what they’re expected to accomplish, and how. This entails shared definitions of terms such as rigor, struggle, support, and success. In districts where equitable instruction is possible, there is a shared way of talking about students, learning, and teaching that rejects deficit explanations and centers opportunity to learn, task quality, and instructional support as causal mechanisms. Struggle is named as evidence of thinking, not incapacity; rigor is defined in terms of cognitive demand rather than pacing or coverage; differentiation is understood as responsive teaching during students’ engagement with shared, grade-level tasks rather than pre-emptive simplification. These discourses matter because they shape what teachers notice, how they interpret student behavior, and which instructional moves feel professionally appropriate.

2. Material–economic practice architecture refers to the concrete conditions that shape what teachers can actually do in classrooms and in their professional learning: time, schedules, tasks, tools, curriculum materials, pacing guides, and the physical and organizational structures of schools. In districts where high-quality, equitable instruction is possible, these material conditions are designed to make ambitious instruction feasible rather than heroic. Teachers have regular, protected time to plan, study tasks, examine student work, and learn from one another; they have access to grade-level, cognitively demanding tasks so that ambitious instruction is the default rather than an act of individual invention; and they work within schedules and pacing expectations that prioritize depth of understanding over coverage. Classroom tools and routines—such as discussion structures, representations, and formative assessment practices—are treated as instructional supports rather than compliance mechanisms. These material arrangements matter because they determine whether teachers can realistically enact what the district claims to value, or whether those values remain aspirational slogans disconnected from daily practice.

3. Social–political practice architecture refers to the patterns of power, authority, norms, and accountability that shape who gets to make instructional decisions, whose judgment counts, and what happens when instruction does not immediately “work.” In systems that support equitable instruction, authority over instructional improvement is not concentrated solely at the top but is distributed across roles in deliberate ways: teachers are positioned as learners and contributors to collective knowledge, coaches are positioned as partners rather than enforcers, and leaders are responsible for creating the conditions, and removing the barriers, so that learning can occur. Norms make it safe for teachers to surface uncertainty, examine their practice publicly, and take instructional risks without fear of blame or punishment. Accountability is oriented toward learning and improvement rather than surveillance, with clear distinctions between spaces for growth and spaces for evaluation. These social–political arrangements matter because they shape the risk landscape of instruction: they determine whether teachers experience ambitious teaching as professionally responsible or professionally dangerous.

And here’s another wrinkle. PAs are stacked: One practice becomes the architecture for another. For example, the practice of leadership creates the practice architecture for teaching. Because leaders are the ones making decisions about how systems are designed. Here is a non-comprehensive but representative list of how that plays out, using PAs as a frame:

1. One of the ideas that I have become more and more clear about in the last few years is that teachers have a mental model of teaching that entails being maximally helpful--answering questions, explaining, precluding confusion, premature scaffolding, differentiating the task demand. While this ensures that students are maximally successful in the short term (they get the assignment done quicker, better, and with less friction), it doesn’t do them any favors in the long run, because the cognitive demand of the task is reduced and therefore opportunity to learn is constrained and therefore students are likely to learn less over time. So the normative mental model of teaching that the practice architecture of the system should support is that the job of teachers is to maintain cognitive demand (for more on this see CL #213 and Core of the Matter #9) not to be maximally helpful. And that cultural–discursive architecture is disproportionately influenced by specific leadership action, i.e. how principals and curriculum specialists talk about teaching during workshops, team meetings, and any other conversations. And, the work of those leaders in creating that architecture is influenced by the work of more senior leaders and the language they use and the mental models they promote.

2. Leaders who expect teachers to change without attention to material–economic architecture are, as my dad would say, away with the fairies. We hear from teachers all the time that schedules don’t work; there are multiples interruptions during instruction; teacher collaboration is not baked into the schedule and, in some cases, the teacher contract actively gets in the way of time for collaboration; they do not have easy access to high quality tasks (if I could wave a magic wand and fix just one thing…). I know very well that districts don’t have limitless resources, but I also know that we are not necessarily talking about huge amounts of money, but coordination, organization, and prioritization. For example, I bet very few districts have high quality instructional tasks in, say, World History, embedded in their curriculum (although they frequently have common assessments and performance tasks). Creating those tasks feels like a huge undertaking, they don’t have anyone at the district who can take that on, and they don’t have the money to hire anyone, so the problem persists--and is another example of where the burden of carrying out strategic priorities gets pushed down to teachers in the form of “lesson planning”, when it is really “task design”. But even if they could pay a small stipend to a department chair to start compiling these tasks as they are created by teachers teaching the course at different schools, then over time a task bank would be created that all teachers would have access to. This is long-term work that does not require oodles of money, just coordination, organization, and prioritization.

3. The social–political practice architecture of leadership is often the hardest to change, in part because it rests on deeply ingrained and largely unquestioned assumptions about authority, expertise, and judgment. In many school systems, leadership is implicitly defined as the right—and the responsibility—to evaluate the quality of teachers’ instructional practice, even when leaders are working outside their own areas of subject-matter or grade-level expertise. This reflects a widely accepted mental model of leadership in which positional authority is treated as a proxy for instructional judgment. That model is so normalized that it rarely invites scrutiny, despite the fact that it is difficult to defend on logical or empirical grounds. Research on learning, expertise, and organizational improvement consistently shows that judgment rendered from outside the practice itself is a weak lever for change, particularly when it is coupled with power over evaluation or employment. As I argued in the last Coaching Letter on the psychology of change (#224), people reliably become defensive when they experience threats to their competence, identity, or standing; and judgment—especially when delivered by someone with formal authority—is one of the most reliable ways to trigger those defenses. From a social–political perspective, then, leadership practices that center evaluation and judgment inadvertently create conditions that make learning less likely: they increase self-protection, reduce openness, and narrow the range of instructional risks teachers are willing to take. Creating psychological safety does not mean avoiding challenge or lowering expectations; it means deliberately reducing the role of judgment in learning spaces so that teachers can surface uncertainty, examine their practice honestly, and engage in improvement without having to defend their professional worth. When leaders shift from being judges of practice to designers of the conditions under which practice improves, they fundamentally alter the social–political architecture of the system—and, in doing so, change the practice architecture for teaching and the improvement of teaching. In this way, practice architectures and culture are mutually reinforcing: the conditions leaders create shape what people believe the system values, and those beliefs in turn stabilize or resist further change. As my colleague Andrew would point out, this is a reinforcing loop.

We’ve written a lot about this lately, which we have workshopped with partner districts and networks—you may wish to check out From Direct to Indirect Instructional Leadership, or Building Leaders’ Role in High-Leverage Instructional Improvement (which I’m now thinking we should re-organize using PA as a framework), or System Conditions that make the work of facilitators easier or harder (which was generated by the Thinking Classrooms NIC and which I organized into PA).

And, of course, teaching creates the practice architecture for learning.

The point of Practice Architectures is not to excuse anyone from responsibility, nor to suggest that individual skill and judgment don’t matter. It is to relocate responsibility to where it actually belongs: in the design of systems that make certain kinds of practice possible, sustainable, and sensible. If we want different teaching, different learning, and more equitable outcomes, then we have to stop asking whether people will change and start asking what must change in the system so that they can.

One tiny little geeky point? I realize that I’ve been using the term “capacity” fast and loose for years. We should stop thinking about capacity as a trait people possess (“we need to build the capacity of our leaders to…”) and start thinking about it as a property that emerges from the interaction of all three practice architectures. If the cultural–discursive architecture is weak, people may have time, tools, and permission, but they don’t know what counts as good work or how to interpret what’s happening. Everyone does what they think is best, but there is unproductive (or worse) variation in the system. If the material–economic architecture is weak, people may understand exactly what high-quality practice looks like and feel safe trying it, but they cannot sustain it. Action requires heroism which leads to cynicism. If the social–political architecture is weak, people may understand the work and have the time and tools, but they won’t risk acting on that understanding. People don’t act out of self-protection.

As you may recall (some of you are incredibly close readers of the Coaching Letter), CL #223 was trying to tee up a series on creating systems and structures for equity-focused on improvement. This was the second one. And really hard to write—it took me a long time to get my head around Practice Architectures. My plan is to keep going with big ideas that have important implications for practice. If you have any feedback, I would love to hear from you. And if there is anything else I can do for you, please let me know. One other thing—I mentioned at the top that I cope with all that’s going on by focusing on what I can control, and writing the Coaching Letter is a big part of that, and so I am very grateful to recipients of the Coaching Letter who came up to me at the Maintaining Cognitive Demand workshop last week to let me know how much they appreciate reading it. It meant a lot to me. Best, Isobel