Legibility, Local Knowledge, and Institutional Power

How Systems of Control Destroy the Knowledge They Need

States, institutions, and administrative systems need to see the things they govern. They need populations to be countable, land to be surveyable, organizations to be auditable, and activities to be reportable. This need to see — to render the world legible — is not a pathology. It is a prerequisite for governance. You cannot tax what you cannot count. You cannot regulate what you cannot observe. You cannot fund what you cannot measure.

The problem is not legibility itself. The problem is what legibility destroys in the process of creating itself. Every act of administrative simplification — every form, every metric, every reporting requirement, every compliance framework — imposes a grid of categories on a reality that does not naturally organize itself into those categories. The grid captures some things and misses others. What it misses is often the knowledge, the relationships, the informal practices, and the situated judgment that make the system actually work. The grid makes the system visible to the administrator. It makes the system’s actual functioning invisible.

This is the central insight of James Scott’s Seeing Like a State (1998), and it is the intellectual foundation of the People layer’s analysis of institutional power.

I. Scott and the Failure of Administrative Ordering

Seeing Like a State

James Scott’s argument, developed through case studies ranging from scientific forestry in 18th-century Germany to Soviet collectivization to Tanzanian villagization to Brasília’s planned urban design, is that large-scale administrative ordering projects fail in predictable ways when they combine four elements:

1. The administrative ordering of nature and society. The state (or institution) imposes a simplified, standardized grid on a complex reality. German scientific forestry replaced diverse natural forests with monoculture plantations arranged in straight rows — legible, measurable, harvestable on schedule. The diverse forest was illegible to the state. The plantation was legible. The plantation was also ecologically fragile, vulnerable to disease, and ultimately less productive than the diverse forest it replaced.

2. A high-modernist ideology. The belief that scientific and technical knowledge is sufficient to organize human activity — that the expert’s model of the system is adequate to the system’s actual complexity. High modernism is not science. It is scientism: the unwarranted confidence that the model captures everything that matters.

3. An authoritarian state (or institution) willing and able to impose the plan. The administrative grid is not adopted voluntarily. It is imposed — through regulation, through funding conditions, through compliance requirements, through the market power of the institution that controls the resources.

4. A prostrate civil society that cannot resist. The populations affected by the administrative ordering lack the political power or organizational capacity to push back, modify, or subvert the plan.

When all four elements are present, Scott argues, the result is reliably catastrophic: the administrative plan destroys the local knowledge and informal practices that the planners did not see, could not see, and therefore did not value, and the simplified system that replaces them is less functional, less resilient, and less adaptive than what it replaced.

The Healthcare Application

The American healthcare system exhibits all four of Scott’s elements in its administrative structure:

Administrative ordering. The compliance framework — 2 CFR 200 Uniform Guidance, Medicare Conditions of Participation, HRSA program requirements, state licensure regulations, accreditation standards — imposes a standardized grid on organizations that vary enormously in size, capacity, context, and community need. The grid requires uniform reporting, uniform documentation, uniform organizational structures. A 15-person Rural Health Clinic in Montana and a 500-person FQHC in Chicago must comply with the same HRSA program requirements. The grid does not see the difference. It is not designed to.

High-modernist ideology. The belief that compliance metrics adequately represent organizational performance. That quality scores reflect quality. That documentation reflects care. That the dashboard is the organization. This is not a conspiracy. It is the occupational ideology of administrators, regulators, and funders who must govern at a distance and who therefore rely on the legible representations (reports, metrics, audit findings) because they cannot access the illegible reality (the care, the relationships, the situated judgment).

Institutional power to impose. The organizations subject to the compliance framework do not choose it. They comply because compliance is the condition of funding. The power asymmetry is total: the funder defines the terms, the auditor enforces them, and the organization conforms or loses the grant. The grant agreement is a contract of adhesion in the legal sense — take it or leave it, with no meaningful negotiation of terms.

Limited capacity to resist. The organizations most affected — small rural providers, tribal health programs, community behavioral health centers — are the organizations with the least political power and the least organizational capacity to challenge or modify the administrative framework. They comply, absorb the cost, and develop workarounds.

II. Local Knowledge: Mētis and Situated Practice

Scott’s Concept of Mētis

Scott’s Seeing Like a State introduces the Greek concept of mētis — practical knowledge, cunning intelligence, the kind of know-how that comes from intimate, sustained engagement with a specific domain. Mētis is the fisherman’s knowledge of where the fish run in his particular stretch of river. It is the farmer’s knowledge of how this specific field responds to different planting strategies across seasons. It is tacit, contextual, embodied, and often inarticulable — Polanyi’s tacit knowledge in a social and practical register.

In healthcare, mētis is everywhere:

The triage nurse who can tell — from the way a patient walks through the door, from the sound of their breathing, from a quality she cannot name but recognizes from a thousand prior encounters — that this patient is sicker than their vital signs suggest. The knowledge is real. It is clinically significant. It cannot be captured in a triage algorithm.

The community health worker who knows that Mr. Torres will not take his medication unless his daughter is involved in the care plan, that the daughter works at the cannery until 3 PM, and that the best time to do a home visit is Thursday afternoon when the daughter has a half day. This knowledge is not in the EHR. It is not in the care plan template. It is in the CHW’s relational understanding of a specific patient in a specific community context.

The grants manager who knows that the state program officer interprets “infrastructure expenditure” broadly enough to include telehealth equipment but will disallow the same equipment if it is categorized as “technology” — a distinction that is nowhere in the written guidance but is operationally consequential. This knowledge is institutional mētis: practical wisdom about how the system actually works, acquired through experience, and transmissible only through the kind of mentorship and community of practice that Lave and Wenger described.

The Destruction of Mētis

Scott’s central warning is that administrative ordering systems systematically destroy the mētis they cannot see:

Standardization eliminates contextual knowledge. When the compliance framework requires standardized documentation, the local variation in practice — the adapted, context-specific approaches that practitioners have developed through experience — is either suppressed (practitioners change their practice to match the documentation requirements) or hidden (practitioners continue their adapted practice but document something different). Either way, the institutional knowledge base is degraded.

Metrics replace judgment. When performance is measured by metrics, the metrics become the target (Goodhart’s Law: “When a measure becomes a target, it ceases to be a good measure”). The triage nurse’s clinical judgment is replaced by the triage algorithm’s score. The CHW’s relational knowledge is replaced by the contact log’s count of visits. The metric is legible. The judgment is not. The metric survives. The judgment atrophies.

Turnover erases institutional memory. Mētis is carried by people, not by systems. When experienced staff leave — and the workforce foundations documents explain why rural healthcare turnover is chronic and structural — they take their mētis with them. The EHR retains the data. The organization loses the knowledge of what the data means, how the system actually works, and why certain practices developed. New staff inherit the formal procedures. They do not inherit the situated knowledge that made the formal procedures functional.

III. Bourdieu: Habitus, Capital, and Institutional Reproduction

The Concept of Habitus

Pierre Bourdieu’s concept of habitus (Outline of a Theory of Practice, 1977; The Logic of Practice, 1990) describes the durable, transposable dispositions that individuals acquire through socialization within a specific social position. Habitus is not conscious strategy. It is the internalized sense of “how things are done here” — the practical mastery of the social game that allows individuals to act appropriately without consciously calculating each move.

In healthcare organizations, habitus is the set of dispositions that practitioners and administrators acquire through training, socialization, and sustained practice within a specific institutional environment:

The physician’s habitus includes: the disposition to take charge of clinical decisions, the expectation of professional autonomy, the internalized hierarchy of specialties, the assumptions about what constitutes quality care, and the practical sense of how to navigate the institutional environment (how to get things done, whom to talk to, what shortcuts are acceptable).

The compliance officer’s habitus includes: the disposition toward documentation and risk avoidance, the internalized standards of what constitutes “adequate” compliance, the practical sense of what auditors are looking for, and the assumptions about what the organization can and cannot do within the regulatory framework.

The community health worker’s habitus includes: the disposition toward relational engagement, the practical knowledge of the community’s social structure, the sense of what patients need that clinicians don’t see, and the often-uncomfortable awareness that their knowledge is not valued by the institutional hierarchy.

Capital and the Institutional Hierarchy

Bourdieu’s framework extends beyond habitus to include forms of capital — resources that individuals and groups can deploy within the social field:

Cultural capital: Knowledge, skills, credentials, and dispositions that are valued within the field. In healthcare, cultural capital is distributed along the credential hierarchy: the physician’s MD carries more institutional weight than the NP’s DNP, which carries more than the RN’s BSN, which carries infinitely more than the CHW’s lived experience. The credential hierarchy is not merely a quality assurance mechanism. It is a system for distributing cultural capital — and therefore institutional power — in ways that reproduce the existing hierarchy.

Social capital: Relationships and networks that provide access to resources and information. In healthcare, social capital includes professional networks (the physician’s referral relationships, the administrator’s connections to the state program office), community relationships (the CHW’s trust network, the nurse’s relationship with patients’ families), and institutional relationships (the grants manager’s relationship with the program officer). Social capital is distributed differently from cultural capital: the CHW may have more community social capital than the physician, but less institutional social capital.

Symbolic capital: Recognition, prestige, and legitimacy. The physician’s opinions carry symbolic weight in organizational decisions regardless of whether those opinions are better informed than those of the nurse or the CHW. The auditor’s findings carry symbolic weight regardless of whether they reflect the organization’s actual performance. Symbolic capital is the mechanism through which institutional hierarchies reproduce themselves — the powerful are heard because they are powerful, which reinforces their power.

Institutional Reproduction Under Regime Change

Bourdieu’s framework explains why institutional change is so difficult, even when the formal rules change: the habitus of the people within the institution was formed under the old rules, and habitus changes slowly. A compliance officer whose habitus was formed in the Medicaid compliance environment — whose dispositions, assumptions, and practical sense are all adapted to Medicaid rules — does not simply adopt a new habitus when the compliance framework shifts to 2 CFR 200 Uniform Guidance. She interprets the new rules through the lens of the old habitus. She sees Medicaid-shaped problems in a non-Medicaid regulatory environment. She applies Medicaid-era solutions to post-Medicaid challenges. The formal rules changed. The habitus did not.

This is a different explanation of institutional inertia than the one provided by organizational economics. North’s institutional economics explains inertia as a structural feature of the relationship between institutions and organizations. Bourdieu’s habitus explains inertia as an embodied feature of the people within the organization — a mismatch between internalized dispositions and the changed environment that those dispositions must navigate. Both explanations are true simultaneously. The structural inertia and the habitual inertia reinforce each other.

IV. Geertz Revisited: Local Knowledge as System Knowledge

The Knowledge That Organizations Run On

Clifford Geertz’s Local Knowledge: Further Essays in Interpretive Anthropology (1983) extended his earlier work on thick description to argue that every social system runs on local knowledge — knowledge that is specific to a place, a community, a practice, and a history, and that cannot be extracted from that context without losing its meaning and utility.

Healthcare organizations run on local knowledge in ways that their information systems do not capture:

Clinical local knowledge. The patterns of disease in this community. The drug interactions that are particularly relevant given this population’s medication profiles. The behavioral health conditions that present differently in this cultural context. The seasonal patterns of ED utilization that reflect this community’s agricultural cycle. This knowledge is carried by experienced clinicians and is lost when they leave.

Organizational local knowledge. How this organization actually processes a grant report (who does what, in what order, with what informal consultations). How budget adjustments get approved in practice (not through the formal approval chain but through a conversation with the CFO in the parking lot). Why the third-floor supply closet always runs out of IV start kits (because the night shift borrows from it for the ED and no one has formalized the resupply process). This knowledge is the operating system of the organization. It is nowhere in the documentation.

Community local knowledge. The history of this community’s relationship with government programs. The trust networks that determine which institutions community members will engage with. The informal leaders — not the elected officials or the org-chart executives but the people whose opinions actually shape community behavior. The cultural norms around health, illness, help-seeking, and institutional engagement. This knowledge determines whether any externally designed program will be used, modified, or ignored.

Why Local Knowledge Matters for the Operating Lens

The capabilitygraph’s operating lens works through four layers. Each layer’s analytical power depends, ultimately, on the local knowledge that the People layer provides:

Operations research can model patient flow, optimize scheduling, and predict utilization. But the model requires inputs that are locally determined: actual arrival patterns, actual service times, actual capacity constraints — including the informal constraints (the physician who always runs late, the radiology tech who leaves early on Fridays, the scheduling workaround that the front desk has developed) that no aggregate data set captures.

Public finance can analyze the institutional structure of funding, the compliance framework, and the political economy of the transition. But the institutional analysis becomes actionable only when it is grounded in local knowledge of how this specific organization interprets the rules, manages the tensions, and navigates the gap between what the funder requires and what the organization can deliver.

Workforce economics can model labor markets, predict shortages, and design incentive structures. But the labor market analysis becomes relevant only when it accounts for the local labor market conditions — the competing employers, the commuting patterns, the community ties, the informal reasons workers stay or leave — that determine whether the aggregate predictions hold in this specific place.

The People layer is not a fourth analytical framework parallel to the other three. It is the ground on which the other three stand. Without local knowledge, the models are maps of territory no one has visited. With it, they are instruments for understanding and intervening in specific human situations.

Intellectual Debts

• Bourdieu, P. (1977). Outline of a Theory of Practice. Habitus, capital, and the reproduction of social structures through embodied practice.
• Geertz, C. (1973). The Interpretation of Cultures. Thick description and the interpretive theory of culture.
• Geertz, C. (1983). Local Knowledge: Further Essays in Interpretive Anthropology. Knowledge as situated, contextual, and inseparable from practice.
• Bourdieu, P. (1990). The Logic of Practice. The practical sense of social action — habitus in operation.
• Scott, J. (1998). Seeing Like a State: How Certain Schemes to Improve the Human Condition Have Failed. Administrative legibility, mētis, and the systematic destruction of local knowledge by modernist planning.
• Polanyi, M. (1966). The Tacit Dimension. The structure of tacit knowledge and its irreducibility to explicit representation.
• Wenger, E. (1998). Communities of Practice: Learning, Meaning, and Identity. Knowledge enacted through shared practice in informal communities.
• Lave, J. & Wenger, E. (1991). Situated Learning: Legitimate Peripheral Participation. Knowledge transmission through practice.
• Perrow, C. (1984). Normal Accidents: Living with High-Risk Technologies. Systemic failure as a consequence of complexity and coupling.
• Weick, K. (1995). Sensemaking in Organizations. How people in organizations construct meaning from ambiguity.
• Goodhart, C. (1984). “Problems of Monetary Management: The U.K. Experience.” In Goodhart, Monetary Theory and Practice. The law: when a measure becomes a target, it ceases to be a good measure.