Chapter 3 — The Three School Models

Each model polices its teachers as much as its students.

You met Angie and Nancy in two very different schools, and that difference matters. Angie was in a democratic school; Nancy was in a progressive one. That difference is the reason their agencies could surface where I could see them at all. In a traditional classroom, their thinking would have stayed underground.

Many of us have heard of these alternatives: the Montessori school across town, the free school a colleague's child attends, the project-based charter from a conference presentation. What we are rarely taught to do, in teacher preparation or anywhere after, is analyze what each model values: which forms of agency it surfaces, which it buries, and what each choice costs the learner and the teacher.

I present the models as pure forms for the sake of analysis, but no real school is pure. We all borrow pieces of these models, and anyone can mistakenly import a model's failure mode along with its strengths, assuming the strengths will land evenly on all students and all kinds of coursework. They will not. The primary model you teach in also shapes what you can see, because a model's language supports you in defining what you see. So this chapter maps the three main school models: what each surfaces and buries, the founding paradox each runs on, and the way each one polices the teacher as much as the student. Rarely does any school follow one model faithfully, whatever the mission statement promises; your school's mileage may vary from the extracted forms I present here.

Traditional schooling: agency as a capacity to be delivered

The traditional model optimizes for what Scott (1998) called legibility:[1] rendering learning into standardized, countable, comparable outputs. A fixed curriculum, segmented time, standardized assessments; a teacher who delivers the curriculum and a student who receives it. Its roots are in Enlightenment-era values: every child, regardless of background, gets access to the same body of knowledge in order to participate in democratic society. Those values are part of why traditional schooling persists. Corporations would also like to offload training onto state funding, and parents want their children to have a leg up in life via diplomas, but the founding aspiration was the cultivated citizen.

Classical liberal education, the artes liberales, was the education befitting a free person.[2] It aimed at a flexible, independent intellect with the capacity to think for oneself; content was grist for that cultivation, not the endpoint. It is close to Freire's stance that all students begin as knowers, not empty vessels to fill, and it understood that knowledge has no endpoint. Thinking was held to be foundational to a democratic society. But it was, from the start, an education for the wealthy leisured class: for those with the standing and free time to be cultivated, at a time when democracy transparently meant a slice of society rather than all of it.

When that ideal was universalized into mass public schooling, the rhetoric of cultivation survived; for most students, the cultivation of the self did not. The curriculum, not the person, became the center. Standardized transmission now stands as a proxy both for education and for equity, while intellectual cultivation stays concentrated where cultural privilege already is. School tax money is never distributed evenly.

Nussbaum's capabilities approach attempts to revive the classical intent: liberty is not rights on paper; individuals must be positioned to exercise their rights, which means sufficient standing and income to have a voice.[3] Equity support becomes internal to becoming a rational thinker. But in the shell game of educational terms, the cultivated individual gets traded for future economic capability. I take this up in Chapter 11.

It helps to see public traditional education as contested ground between three modern goals. Labaree (1997) argues that public schooling is pulled in three incompatible directions: democratic equality (preparing citizens for shared civic life through equal access to a common curriculum), social efficiency (producing the workers an economy needs), and social mobility (giving individuals a credential to get ahead).[6] The first wants a shared, equalizing experience. The second wants service to the GDP. The third wants sortable, competitive difference.

Democratic equality has motivated a school system meant to be not only public but equitable, a goal highly contested in the Trump era. Special-needs students are provided more services; public schools are held to a common set of standards. Outcomes-based learning and No Child Left Behind draw on this goal, and it was the argument behind Brown v. Board of Education, support for LGBTQ+ youth, and bilingual education. As Anyon (1980) showed, though, the outcomes of social efficiency are not evenly distributed: some schools prepare students to be executives, others for middle management, still others for service jobs, and the neighborhood school you attend predicts your future employment. This has prompted parents to find ways, or the political means, to choice their children into schools in better neighborhoods; to establish charters that allow greater exclusion (equal in the front door, unequal through the back door of expulsion); and to flee to private schools offering tutorial and Socratic environments. To misquote Animal Farm: some schools are more equal than others. The third goal echoes the Enlightenment aspiration of rational, autonomous, self-actualized adults, and in public schools it often decays toward credentialism and career-pathway options.

Private, alternative, religious, and charter schools pull these goals toward one value or another: to afford the decay, to avoid equity, or to change the method of delivery. Some are founded in opposition to the values themselves, as instruments of racial, immigrant, or identity segregation; some to promote methods that are project-based, progressive, Socratic, or real-world in nature.

Kant named the contradiction the traditional model runs on: how shall I cultivate freedom under conditions of compulsion?[7] Compulsion first, freedom after seems to make sense; anyone who has had a two-year-old understands that children do not arrive as rational planners. It seems to make sense, that is, until you think of agency as developmental.

Here is the model's deeper cost. Focused on outcomes, it lacks language for origination: a student's effort to understand from their own perception of the world, their own thinking as a knower, their own differences with the teacher, their own curricular ends. Without the language, these things become unintelligible, or at least unimportant. And that lack of language effectively polices teachers into an orientation toward outcomes. The traditional model sees student agency as a capacity to be built for later: first you receive the standardized curriculum; then, as an adult, you get economic power with it. This is the orientation that unseated Angie's knowing, and its characteristic pitfall is what Bakhtin called monologism:[5] knowledge arrives finished, and a student's working model is an error to be corrected rather than a beginning to be explored.

Marcus's story is one of the reasons I felt I needed to write this book. In my biology class, Marcus was well-behaved; if quiet disconnection counts as good behavior. When he wasn't asleep, he at least wasn't disruptive. He simply had little interest, if not zero, in my biology assignments. So one day I approached him in a friendly way and asked plainly, "Why aren't you interested in any of my assignments?"

"I don't care about biology," he said. I flinched, recovered, and asked, "What do you care about?" As he talked, a different Marcus came into view: not a disengaged student but a purposeful young man. He was getting up at 4 a.m. to work out. He was researching stocks, making significant money at it, and building small business ventures. He had a fierce drive for mastery, just not the kind his school wanted him to show. By any psychological measure Marcus was motivated, disciplined, purposeful. Within school, his motivation did not align with institutional goals, so he appeared disengaged or even resistant. Motivation can be strong and still be institutionally illegible. The traditional regime reads output, not source; it has no routine use for the question "What do you care about?" when there is no official artifact to read. Marcus displayed what Chapter 2 defined as authorial agency.

When I suggested to administrators that an independent study could put Marcus's accomplishments on his high school transcript, because on his GPA alone a college would not see him as an attractive candidate, I was told I was trying to save him. The phrase accused me of a personal crusade, a missionary stance toward this student. I felt gaslit: I was being labeled as someone who had lost professional distance, while the administrator labeled him lazy, unmotivated, and immature, finalizing him as a student not quite worth anyone's time. Marcus did eventually graduate with a good GPA, because he later realized college was something he wanted. It was not hard for him to achieve once he had a personal reason.

Unlike Marcus, Priya, whom you met in the last chapter, performed school without actually learning. This model often rewards the faking and erases real origination at the same time.

To be fair, the traditional model persists because it holds real goods: relatively low cost, sufficient structure for safety, teachers who often care, direct teaching through curated content. These have their place for some students and for some content; even Marcus thought his English and math courses had value to him. Had Marcus gone to a democratic school, he might have gained in origination but lost on academic scaffolding. Hold that trade-off; it is the hinge of this chapter.

Progressive schooling: agency as constructed experience

The progressive model, historically most associated with Dewey and Montessori, optimizes for experience meant to support origination and autonomy: active learning, inquiry, projects, choice-making, the classroom as a laboratory of democracy.[8] It is built to cultivate engaged agency, students making sense of real phenomena rather than receiving inert facts. Just as Angie wanted to work out the division of fractions without the jingle, Dewey argued that learning had to begin where knowledge itself begins. In The Child and the Curriculum (1902) he wrote that the curriculum "must be restored to the experience from which it has been abstracted. It needs to be psychologized; turned over, translated into the immediate and individual experiencing within which it has its origin and significance." Dewey understood that the procedural form of dividing fractions is shorthand for getting the right answer, and that the shorthand does not illustrate what is meant by it.

What Dewey saw is that students need the opportunity for sense-making, the substrate of engaged agency. This is why John's students stayed engaged even on a day the lesson wasn't doing that work: most days, their sense-making is satisfied, so one opaque lesson did not tip them into disengagement. Meaning-making goes further. It begins from one's own perceptual standpoint, taking a position toward the idea while staying answerable to the subject under study.[4] The student is still focused on the teacher's content but has another way of seeing it; this is the ground on which epistemic agency can begin. Nancy is the clean case. Her classmates asking who gets the extra penny? were sense-making; Nancy, giving the problem an aesthetic order, had crossed into meaning-making.

That is where the progressive model finds its own paradox. Not compulsion-now-for-freedom-later, but individually constructed knowledge aimed at standardized endpoints. Students build, manipulate, and explore, but toward the answer the teacher already holds; their construction funnels to a preset destination. John felt something important in Nancy's work, as he told me later, but he had no language to articulate to himself what it might be. So he corrected her and moved on. Because progressivism values sense-making, it lacks the language for epistemic bids, even though these teachers, often more attuned to their students, regularly sense them.

The progressive model is the one I align with most, inspired by Dewey, Holt, and Montessori, so my sharpest example of progressive policing is one I did to myself, and to my own fourth-grade child. I kept pressing my oldest son, Brian, to use manipulatives in math, certain the hands-on way was the only true sign of understanding. He finally had to tell me I was frustrating him: the manipulatives slowed him down. The orthodoxy was so much mine that I had stopped seeing the learner in front of me, even though he was my own child, and served the method instead. The progressive police lived in my own head, coupled with a traditional education of my own that had taught me knowledge has known endpoints.

There is another blind spot in progressivism, and it is an equity issue. Progressive education assumes the kind of cultural capital Lareau calls concerted cultivation, and even among children who have it, some are simply more academically interested during their school years than others. Because the model values autonomy while holding known endpoints, the expectation (mine included, once) is that given sufficient freedom and material, students will become academically interested and proficient at reaching those endpoints. On the axis running from origination to academic scaffolding, the students who need direct instruction or other scaffolds can be left to themselves. Autonomy without scaffolding can produce inequity. We will later look at Patall's work, and at a study of Philadelphia library branches, to see how autonomy and resources alone are insufficient for developing engagement even in what students choose to learn.

I was discussing these matters recently with a family friend, Mike, who saw this from the inside of the progressive school he attended. "Only a slice of kids really succeeded in that environment," he said: the ones who came in with a strong bent toward academic learning. He and many classmates floundered. The positive, he added, is that all of them became far better at exercising autonomy than the public-school students he later observed.

Yong Zhao, in World Class Learners (2012), argues the mirror image about the traditional model: a standards-and-testing regime can teach only what is prescribed, so students talented or curious outside it never get the chance to discover those capacities. Countries whose schools are less authoritarian, he notes, show greater creativity and entrepreneurship. Right diagnosis; but as we will see, "more freedom" is not by itself the cure.[15]

So, like the traditional model, the progressive model holds real goods: the development of autonomy and confidence, the psychologizing of learning. Its downside is that the endpoint disposition gets smuggled in as an entry condition. The traditional model suppresses origination; naive progressivism presupposes it; neither cultivates it. For Marcus, a progressive school would probably have left him still sleeping at his desk, with administrators sympathetic to my concerns rather than gaslighting me, and the same outcome for Marcus.

Democratic schooling: agency as self-governance

The democratic model, of which the Sudbury schools are the clearest American form, optimizes for individual freedom through self-governance: students set the rules, choose what to learn, and share real authority over the conditions of school life. Its philosophical anchor is that students are already motivated, and the adult's job is to protect the conditions under which that motivation can run its own course. Greenberg distinguishes democratic from progressive education by observing that the progressive teacher who spots a child's first glimmer of interest and rushes over with books is a "turn-off to curiosity" (Greenberg, 1985). It is also the kind of place where Angie's thinking could surface and her friends could gather around a chalkboard, because they were not obliged to be somewhere else.

The American democratic form developed from two intertwined strands. A libertarian branch ran through the UK: Homer Lane's Little Commonwealth (1913), then A. S. Neill's Summerhill (1921), with emotional well-being first and optional class attendance, but no student-authored curriculum. An anarchist branch began with Francesc Ferrer's Modern School (Barcelona, 1901), which crossed to the US as the Ferrer Modern School and closed in 1953; it had a directed curriculum that was non-coercive in method. The free-school wave of the 1960s and 70s, catalyzed by John Holt's Growing Without Schooling newsletter, led to the decisive American turn, both non-directive and student-governed: Sudbury Valley School (1968, Daniel Greenberg and colleagues), with no classes or set subjects, town-meeting governance, formal law with a student judiciary, and "staff" rather than "teachers." The student is treated as the auctor, literally the originator, of their own learning; what I called authorial autonomy: control over what to learn, for how long, and to what standard (von Duyke, 2013). In this model learning is, as it so often is in life, largely incidental and emergent: through the school meeting, judicial committees, requested classes, tutorials, watching peers. They like to quote Dewey: "education is life."

The democratic critique of the other models runs like this: they fail to distinguish autonomy as a psychological good in one's life from autonomy in the classroom, and real classroom autonomy must include the choice not to partake of any of the options. Autonomy cannot be said to exist without the freedom to decline. But educational models in general, the democratic one included, fail at a further distinction: autonomy-as-disposition (real, transferable, valuable regardless of its object) versus its academic outworking, which may need scaffolding. In practice the democratic model tends not to include the freedom to ask for extrinsic structure within its account of autonomy.

That leaves Sudbury-model schooling with its own paradox: maximum freedom can undermine autonomy. Withholding structure in the name of freedom leaves the student to direct their own learning without a curator, and the same self-direction that would liberate a Marcus can leave other students feeling bereft. At the Sudbury-model school I researched, middle schoolers would leave the school for a year, giving up their freedom to acquire more scaffolding for their learning, and then return to regain their autonomy and use the new scaffolding at a direction and pace they found valuable. What does it tell us when children must exit a model of freedom in order to get what their learning needs? Self-direction works best for students who arrive with unusual focus, or already fluent in the cultural capital the model assumes.[9] It is not a universal good you can pour into any classroom.[10] Autonomy support, as Patall's work shows, is conditional; it most benefits those who already have footing. Freedom without sufficient support can reproduce the very inequity it was meant to level. And Nancy's epistemic bid would probably have gone unread here too: the Sudbury school in my research held that staff have no duty to teach, and her flower arrangement would have been read as play.

I learned how tightly the democratic model polices its own adults in person. The chalkboard conversation with Angie and her friends, the one I opened this book with, leaned hard into the borders of the model's value system. Staff took me aside and said I had skated dangerously close to coercive teaching. I had assigned nothing, graded nothing, required no one to stay; the students had drifted over on their own and stayed because the conversation interested them. But in a school organized around non-interference, an adult who draws students into an inquiry they did not initiate is suspect by definition. The very moment a student was showing me how her engaged agency had been squashed was, inside that model's value system, a transgression.

What the three have in common

Watch what happened in each story. A lack of language can hide the fact that alternatives are possible rather than foreclosed, and social policing then relocates the structural absence into a pathology of the person:[11] I was coercive in the democratic school; I was trying to save him in the traditional one; and in my progressive homeschool room I policed myself with the ideology of a method. Each model polices its teachers as much as its students. That is why the models must be understood as structural: teachers act inside constraints that reward some moves and punish others, and a missing vocabulary is itself a form of policing.

Notice, too, the identical relocation onto the person.[12] The administration did not admit the institution had no path for Marcus; I was accused of trying to save him. The democratic staff did not confess the model has no room for drawing a student into inquiry; I had "skated close to coercion." And I did not tell myself this method is not serving Brian; I concluded I was a bad teacher for wanting to teach directly. In each case a structural absence got rewritten as a personal defect. And if the defect is always personal, who is left to question the structure? Each model, in its own way, forces a choice between origination and educational scaffolding, and what goes missing is the intellectual apprenticeship that results when origination and rigor coexist.

Language fails teachers in two distinct ways, and it is worth pulling them apart.

The word does not exist (absence). The concept is unavailable, so the phenomenon cannot become an object of attention. John sensed something important in Nancy's penny arrangement, but he had no term for meaning-making versus sense-making, no notion of an epistemic bid, so he could not respond to what he intuited. He corrected her and moved on. The missing word did not just leave her pattern unnamed; it made it unseeable as a bid for a different way of understanding the problem.

The word exists but is ruled illegitimate (suppression).[13] The Sudbury staff had ruled an entire register out of bounds: "teach," "curriculum," "intervene," interest-as-a-hook. When Angie and I began our chalkboard inquiry, the only concept available for what I had done was the delegitimizing one, coercive teaching. Off-doctrine views got the same treatment applied to students: ideas that differed from staff's were relabeled "absurdist," implying the students lacked rational thought, and some of those students disengaged from every curricular activity for the rest of their enrollment. Nancy, in her school, was merely "off topic."

Language is powerful because it can make a set of behaviors into a comprehensible concept, and a comprehensible concept is diagnostic and answerable.[14] Naming reorganizes perception: "wait time" let teachers notice the one-second gap they were leaving after questions; "funds of knowledge" let a child's home knowledge register as an asset instead of a deficit. The absent word and the forbidden word produce the same incapacity to act, and both then charge the person, student or teacher, with the gap.

Not every binary is false. But we have to ask of each one: is the trade-off intrinsic, a true either/or, or is it produced by a missing support?[15] It may be that public education cannot afford the teaching power required to support student origination; money is a real limiter, though only in the sense of what political budgets prioritize. Technology is sometimes offered as the answer, but I suspect our sense of what is pedagogically possible is largely independent of it: the chalkboard has been with us since the 1800s, and we have not used two centuries of technical advancement in remarkably different pedagogical ways since. Our understanding of how students learn is still shaped by educational outcomes, and the research literature still mostly describes autonomy and agency as binaries a classroom or a student either has or lacks, their cultivation assigned to home life (Lareau) rather than counted among the responsibilities of schooling.

Chapter 2 gave us names for more classroom thinking like Nancy's and Angie's: engaged, epistemic, democratic, and authorial agency, and the shadow forms that appear when agency is chronically unseated. Teachers who have language for what they intuit in student actions, and for the structures that confine them, can then decide honestly whether they want to teach differently or need to teach somewhere else. What comes next is the why: how the daily machinery of schooling manufactures the shadow forms, and why that pressure lands hardest in high-poverty schools.

Endnotes — Chapter 3

[1] "Legibility" is used here in James Scott's sense — the rendering of complex social reality into standardized, countable forms a central authority can read. Scott, J. C. (1998). Seeing Like a State. Yale University Press.

[2] The classical liberal-arts tradition — the artes liberales as the education of the free, leisured citizen, organized as the trivium and quadrivium. (A history-of-liberal-education source — e.g. Kimball, Orators and Philosophers, 1986 — can be added if the historical claim needs support.)

[3] Nussbaum, M. C. (2011). Creating Capabilities: The Human Development Approach. Belknap Press of Harvard University Press. Developed in Chapter 11.

[4] "Effortful agency" is my own term. Matusov, E., von Duyke, K., & Kayumova, S. (2016). Mapping concepts of agency in educational contexts. Integrative Psychological and Behavioral Science, 50(3), 420–446.

[5] Monologism is Bakhtin's, developed in Problems of Dostoevsky's Poetics (C. Emerson, Ed. & Trans.; University of Minnesota Press, 1984). His The Dialogic Imagination (1981, University of Texas Press) carries the companion concept of dialogism.

[6] Labaree, D. F. (1997). Public goods, private goods: The American struggle over educational goals. American Educational Research Journal, 34(1), 39–81.

[7] The paradox of cultivating freedom under conditions of compulsion is Kant's. Kant, I. (1900). On Education (Über Pädagogik) (A. Churton, Trans.). D. C. Heath. (Original work published 1803.) [verify translation/edition]

[8] Dewey, J. (1916). Democracy and Education. Macmillan; and Dewey, J. (1902). The Child and the Curriculum. University of Chicago Press. On Montessori's method, Montessori, M. (1912). The Montessori Method (A. E. George, Trans.). Frederick A. Stokes (see also Lillard, A. S., 2005). The "laboratory" image is Dewey's.

[9] Bourdieu, P. (1986). The forms of capital. In J. G. Richardson (Ed.), Handbook of Theory and Research for the Sociology of Education (pp. 241–258). Greenwood Press.

[10] On the conditional benefits of autonomy support and choice: Patall, E. A., Cooper, H., & Robinson, J. C. (2008). The effects of choice on intrinsic motivation and related outcomes: A meta-analysis. Psychological Bulletin, 134(2), 270–300; see also Patall, Cooper, & Wynn (2010) and Patall et al. (2022).

[11] Mills, C. W. (1959). The Sociological Imagination. Oxford University Press. — public issues vs. private troubles; the structural-to-personal relocation.

[12] The conflict between women loving their home and children vs. loving a career is a false dichotomy largely manufactured by missing supports — childcare, leave, flexible work, an equal partner in the domestic labor. Supply those and much of the "incompatibility" dissolves. The social policing fills in where resources are lacking — you can't have it all, she's neglecting the kids, she's not really committed at work — presenting the either/or as both a fact of nature and a fact of character, so the woman who wants both is read as greedy or naive. What's missing gets hidden twice: once as "just how it is," and once by pathologizing the person.

[13] Language that is suppressed: Bourdieu names the everyday version — legitimate language and symbolic violence — one group's register is consecrated as "correct" and the hierarchy is registered as natural, so the dominated police themselves (Bourdieu, Language and Symbolic Power, 1991). Freire stresses that to be denied the word is to be denied the power to name — and so to transform — one's world (Freire, Pedagogy of the Oppressed, 1970). When the regime rules the word illegitimate, the teacher self-censors, and the phenomenon goes unaddressed.

[14] Naming reorganizes perception: "wait time" (Rowe, 1986) helped teachers notice the on-average one-second gap between asking a question and providing the answer; naming it made it changeable. "Funds of knowledge" (Moll, Amanti, Neff, & González, 1992) — before the term, a child's home knowledge didn't register as academic ability, so its presence was read as deficit rather than asset.

[15] Right diagnosis, wrong cure. Zhao's implied remedy to the traditional model — more autonomy, creativity, choice — diagnoses the disease but mistakes the cure: autonomy support is conditional (Patall, 2008), and the progressive and democratic models presuppose pre-existing capital, so "more freedom" is in danger of reproducing advantage (the library study: Neuman & Celano, 2012).