Polycentric Democracy — A Relational Map of Power, Networks, Institutions and Civic Accountability Citizen → local body → state → courts → market → platform/network → expert institution → community → media → data system → mutual checks → appeal/correction → shared accountability Central authority is not sufficient; fragmented authority is not automatically democratic; networks are not neutral; locality is valuable but must be free of exclusion; private platforms can have public effects; authority requires reasons, limits, appeal and audit; polycentricity is both cooperation and control. Problem In the traditional picture of democracy, power is sought chiefly within the state—the legislature, executive, judiciary, elections and administration. These indispensable centres remain decisive in modern society, but citizens’ lives are not shaped by government orders alone. Employment opportunities can be altered by company policy, credit by a bank score, public visibility by platform ranking, access to news by a recommendation system, educational opportunity by universities, the meaning of health by expert institutions, and social reputation by communities and digital networks. The question of democracy has therefore expanded beyond the state. “Network power” does not mean that some secret omnipotent network controls everything. Its meaning is simpler and analytical: many decisions arise within webs of relations where, instead of a central command, standards, protocols, access, reputation, data, contracts, interdependence and platform rules direct behaviour. Power resides here not only in commands but also in architecture, defaults, ranking, interoperability, visibility and exclusion. The first difficulty of networks is visibility. Government laws are generally published; but the rules of search ranking, content moderation, credit scoring, procurement platforms, payment gateways or logistics networks may remain obscure to citizens. Decisions that are not called “law” can nevertheless have law-like effects on life. Democratic scrutiny therefore cannot be confined to formal legal sources. The second difficulty is ambiguity of authority. When a decision is produced jointly by several institutions, it becomes easy after an error to say, “Not us—the system.” A vendor built the model, an institution supplied the data, an official accepted the output, a platform delivered the information, and a contractor acted upon it. Power may thus be joint while accountability is fragmented. A central test of polycentric democracy is that distributed decision-making must not become distributed irresponsibility. The third difficulty is scale. Local bodies are close to citizens and can therefore understand context; yet local majorities can also suppress minorities. Central institutions can secure equal minimum rights; but excessive centralisation can erase local knowledge, experimentation and autonomy. “Centre versus local” is therefore not a simple choice. The question is which function belongs at which level, within which rights-boundary, and under what structure of appeal. The fourth difficulty concerns the private–public distinction. A private company is not the state; but if its platform becomes basic infrastructure for public discussion, payments, transport, employment or identity, its private rules can generate extensive public effects. “It is private, so it may do whatever it wants” and “it has public effects, so turn it into the state” are both extremes. The middle path raises questions of proportional public obligations, transparency, contestability and interoperability. The fifth difficulty is expertise. In areas such as pandemics, river management, energy, cyber-security, financial stability or artificial intelligence, expert knowledge is necessary. But expertise is not a complete substitute for democratic legitimacy. Citizens cannot themselves perform every technical calculation, yet objectives, the distribution of risk, choices of value, acceptable harm and rules of appeal require public reasons. The sixth difficulty is informational inequality. In a network, actors with more data, computing power, legal resources, advertising reach, reputation or connections can exercise disproportionate influence. Even where formal voting rights are equal, agenda-setting, issue framing, visibility and lobbying may be unequal. Democracy is not only the moment of voting; the distribution of public attention is also power. The seventh difficulty is speed. Digital platforms and financial networks can alter decisions in seconds, while law, judicial review and public debate move slowly. Accountability that is too slow can become meaningless, but decisions that are too rapid can become unreflective. A polycentric order needs a proper combination of emergency action and mandatory subsequent review. The eighth difficulty is cross-border effect. Data, platforms, capital, environment, migration and supply chains move beyond administrative boundaries. A decision by one city can affect another region; a platform based in one country can shape public conversation among citizens of many countries. Polycentric democracy is therefore incomplete without nested and trans-local coordination. The ninth difficulty concerns identity and community. Local self-government can protect a group’s cultural dignity, language and tradition; but within a community, dissenting voices, women, the poor, migrants, minorities or young people can be suppressed. It is not enough to say “the community has spoken”: whose voice, through what process, with how much dissent, and with what opportunity for exit or appeal must also be asked. The problem of this chapter is therefore: when power is distributed across the state, market, platforms, communities, expert institutions, local bodies, data systems and communication networks, how can democracy become sufficiently polycentric for decisions to remain context-sensitive and flexible while protecting civic dignity from arbitrary domination, invisible network power, appeal-free classification, local oppression and private monopoly? Core Proposition First proposition: polycentric democracy does not mean merely many centres; it means relations of mutual limitation, cooperation, oversight and appeal among legitimate centres. If multiple centres of power collide without rules, that is not polycentric democracy but fragmented domination. Second proposition: power is not merely the capacity to issue commands; setting the agenda, opening or closing access, increasing or reducing visibility, making standards, controlling data, classifying people, determining protocols and changing the cost of alternatives are also forms of power. Third proposition: every centre of power should have a clear jurisdiction. It should be public which decisions fall within its authority, which do not, which rights are minimally non-derogable, and who conducts interim or final review in a dispute. Fourth proposition: subsidiarity is a useful principle. A decision that can be made effectively, fairly and with rights protected at a lower level should not be pulled upward unnecessarily. But where local capacity or rights protection is deficient, a higher-level role is legitimate. Fifth proposition: locality is not itself justice. Nearby institutions understand context, but local prejudice, dynastic control, crowd pressure or capture are possible. Local autonomy therefore requires external rights review. Sixth proposition: central equality is necessary, but total centralisation is not. National or constitutional levels can secure minimum rights, equal citizenship, fair procedure and broad redistribution; local levels can manage adaptation, experimentation and cultural specificity. Seventh proposition: network power should be made traceable. Major nodes, data sources, protocols, vendors, funding relations, ownership structures, moderation rules or standards that influence decisions should, where possible, be public or subject to oversight. गजेन्द्र ठाकु रक समानान्तर दर्शन — खण्ड २ Eighth proposition: private power should bear public obligations in proportion to its public impact. Not every private institution is a state; but the more its decisions shape citizens’ basic opportunities, communication or dependence on infrastructure, the stronger its duties of reason-giving, transparency, safety and appeal. Ninth proposition: exit is useful but insufficient. “If you do not like the service, leave it” is meaningful only when a genuine alternative exists. Under network effects, monopoly, employment dependence or essential-service conditions, voice, appeal and regulation are necessary. Tenth proposition: interoperability can be a democratic tool. If users can transfer data, social connections or service history with appropriate safeguards, arbitrary power within closed ecosystems can be reduced. Eleventh proposition: public reason should be the language of decision. Citizens are not all experts, but institutions should at least explain their objectives, principal evidence, risks, trade-offs, alternatives and avenues of appeal in an intelligible form. Twelfth proposition: expertise can be a source of authority, not a licence for domination. The method, uncertainty, conflicts of interest and value judgements behind expert conclusions should be made distinct and explicit. Thirteenth proposition: network robustness is a democratic value. If failure of one node shuts down the whole system, dependence is excessively centralised. Redundancy, fallback, local capacity and offline alternatives increase civic security. Fourteenth proposition: anti-capture rules are necessary. Electoral bodies, regulators, panchayats, universities, associations or platforms can be captured by money, caste or dynastic dominance, party control, corporate lobbying or bureaucratic monopoly. Rotation, disclosure, conflict rules and independent review are useful safeguards. Fifteenth proposition: democracy concerns knowledge production as well as decision-making. Citizens’ experience, local knowledge, expert studies, statistics, testimony from affected groups and public reasoning should all enter decisions through appropriate methods. Sixteenth proposition: the quality of participation matters, not merely its quantity. Giving everyone a chance to speak at a meeting is not enough when the agenda is predetermined, information unavailable, language inaccessible, timing unsuitable or retaliation feared. Seventeenth proposition: minority rights limit majority procedure. A decision by 51 per cent cannot abolish a person’s basic dignity, equal citizenship, life, bodily integrity, freedom of religion or non-religion, or right to a judicial hearing. Eighteenth proposition: a polycentric order can accept experimentation, but it must not dump the costs of harm on weaker groups. Pilots, sunset clauses, independent evaluation and reversible design make experimentation accountable. Nineteenth proposition: the audit trail is the memory of network decision-making. Records of when rules changed, which data were used, who approved a decision, what appeals arose and what corrections followed are essential for future accountability. Twentieth proposition: emergency power should be time-bound. Rapid centralisation may be necessary in a crisis, but without expiry, legislative or judicial review, public reporting and a restoration plan, emergency arrangements can become permanent normality. Twenty-first proposition: polycentric democracy is both adversarial and cooperative. Institutions constrain one another but also share information. A system of veto alone can become paralysed; a system of cooperation alone can become vulnerable to capture. Twenty-second proposition: complete network neutrality is not possible in every field, but discriminatory gatekeeping must be examined. Equal service, justified prioritisation and public-interest exceptions should be governed by clear rules. Twenty-third proposition: attention power is political power. Actors that control the ordering of public attention, trends, recommendations or advertising affect freedom of information and electoral fairness. Twenty-fourth proposition: polycentric democracy must be scale-sensitive. Village water distribution, metropolitan transport, river basins, national tax policy and global digital standards cannot be governed at the same institutional scale. Twenty-fifth proposition: the final test is non-domination. A citizen should not depend on the arbitrary mercy of any single state, market, platform, community or network actor; reasons, limits, alternatives, resistance and review should remain available. Principal Arguments The first argument concerns the functional basis of polycentricity. In complex societies information is dispersed: local actors may know more than central officials about particular water problems, the language of a school, irrigation on a field or urban transport. Lower-level decision capacity can reduce information costs, provided rights protection and coordination are available. The second argument concerns error correction. An error by a single centre can spread across an entire system; several independent or semi-independent centres can conduct alternative experiments and make comparison possible. Comparison is useful, however, only when results are public, standards comparable and the harms of failed experiments limited. The third argument concerns redundancy. A single chokepoint is dangerous during a cyberattack, natural disaster, political crisis or administrative failure. Local fallback, multiple communication channels, distributed data backups and alternative service providers can increase democratic resilience. The fourth argument concerns proximity of voice. Citizens can raise problems more directly in nearby institutions. Dialogue may be easier at the level of a panchayat, ward, school committee or cooperative. But if proximity turns into patronage or fear, an external appeal is necessary. The fifth argument concerns anti-monopoly. Competing centres can protect civic freedom against governmental, economic or informational monopoly. An independent judiciary, free media, associations, local government, cooperatives, universities and civic organisations can balance one another’s power. The sixth argument concerns specialisation. No single institution can understand every decision. Environmental boards, public- health bodies, data-protection authorities, local water committees or financial regulators can develop specialised knowledge. Polycentricity accepts specialisation, but links it to democratic oversight. The seventh argument concerns the reverse side of network effects. As more users join a platform, everyone has stronger incentives to remain there; convenience can thereby become monopoly-like power. Portability, interoperability and open standards can lower exit costs and increase democratic contestability. The eighth argument concerns agenda power. Decision-making is not merely a yes-or-no vote: which issue is discussable, which data are presented, which options appear on the menu and which news is placed on top are all prior decisions. Analysis of network power brings agenda-setting within democratic accountability. The ninth argument concerns procedural justice. Even a beneficial outcome can reduce citizens’ equal status if the process is appeal-free, reasonless, biased or opaque. Democracy treats citizens not as policy targets but as participants capable of asking and receiving reasons. The tenth argument concerns collective action. Air pollution, rivers, pandemics, cyber-security or traffic cannot be solved by one actor. Coordinated contributions from several centres are necessary. In common-pool problems, polycentric governance can combine local monitoring with higher-level coordination. The eleventh argument concerns fiscal accountability. When the institution that spends money is completely detached from revenue or outcomes, accountability may weaken. Yet complete local fiscal self-reliance can damage poorer regions. A combination of local discretion, transparent transfers and equalisation is therefore required. The twelfth argument concerns plural values. Citizens may hold different preferences in education, language, urban design, culture or welfare. Some decentralisation leaves room for diverse ways of life, while a constitutional floor ensures that diversity does not become a licence for discrimination. गजेन्द्र ठाकु र The thirteenth argument concerns epistemic diversity. Different institutions detect different failure modes. Courts may see rights, auditors expenditure, journalists corruption, local groups conditions on the ground, scientists causal risk, and civil society lived experience. Convergence among independent sources makes decisions more robust. The fourteenth argument concerns temporal scale. Some institutions may be designed for long-term thinking—a central bank, constitutional court or river authority; elected governments operate under shorter mandates; local bodies manage daily services. Different time horizons can compensate for one another’s weaknesses. The fifteenth argument concerns contestability. The democratic value is not merely choosing the right ruler but having institutional capacity to change a wrong decision. Appeals, review, elections, judicial challenge, ombudsmen, consumer switching, public criticism and policy sunsets are different forms of contestability. The sixteenth argument concerns the chain of legitimacy. The legitimacy of a delegated body can derive from democratic mandate, law, expertise, transparency and review. “We are experts” or “we were elected” alone does not fully justify every decision. The seventeenth argument concerns asymmetry. A platform knows much about its users; users know little about the platform’s decision logic. Administration sees citizens’ records; citizens do not see backend processes. Transparency, data access, reason- giving and independent audit can reduce this epistemic asymmetry. The eighteenth argument concerns continuity of dignity. The dignity-based ethics of the previous chapter gives polycentric democracy an ethical boundary: no centre, whether state or community, should reduce a person to something disposable, appeal-free or merely a fixed category. Pūrvapakṣa Pūrvapakṣa 1: polycentricity slows decisions. In crises, many veto points, consultations and appeals can make action impossible. Pūrvapakṣa 2: citizens need a clearly responsible government. Too many centres increase blame shifting and weaken electoral accountability. Pūrvapakṣa 3: local self-government increases inequality. Wealthy cities will provide excellent services while poor regions fall behind, breaking national equality. Pūrvapakṣa 4: expert bodies are anti-democratic because their officials are not elected. Pūrvapakṣa 5: imposing public obligations on private platforms will weaken innovation and freedom of expression. Pūrvapakṣa 6: interoperability, portability and transparency can damage security and privacy. Pūrvapakṣa 7: community-based governance is a romantic ideal; local elites, caste pressure, nepotism and violence are realities. Pūrvapakṣa 8: network power is so diffuse that regulating it is impossible. Regulatory efforts will always lag behind technology. Pūrvapakṣa 9: polycentric democracy breaks policy coherence; different bodies create contradictory rules. Pūrvapakṣa 10: ultimately rights cannot be enforced without the sovereign state, so the remaining centres are merely symbolic. Uttarapakṣa Reply 1: temporary command may be necessary in a crisis, but pre-defined emergency competences, time limits, reporting and ex post review can combine speed with accountability. Polycentricity does not demand endless consultation on every decision. Reply 2: responsibility mapping can reduce blame shifting. Every decision chain should identify in writing the decision owner, data provider, technical vendor, approving authority, appeal body and oversight body. Several parties may be responsible for one decision, but “no one” is not acceptable. Reply 3: decentralisation must be accompanied by fiscal equalisation, national minimum standards, targeted grants and capacity support. Local discretion does not mean complete dependence on the accident of local resources. Reply 4: the legitimacy of expert bodies arises from delegated law, competence, conflict rules, transparent methods, legislative or judicial review and public reason. Being unelected does not make such a body illegitimate, but unchecked expert rule is not justified either. Reply 5: obligations should be proportional. A small private forum and a dominant infrastructure platform should not bear identical burdens. Duties should increase with public impact, market power, dependence, risk and lack of substitutability. Reply 6: transparency does not mean exposing all source code to everyone. Tiered access, regulatory audit, privacy-preserving reporting, security exceptions and aggregated metrics can be used. Contestability and security are not necessarily opposed. Reply 7: local capture is a real reason for multi-level safeguards—secret ballots, anti-discrimination law, rotation, representation of women and minorities, an external ombudsman, judicial appeal and transparent finance. The benefits of local knowledge cannot be separated from criticism of elite domination. Reply 8: technology changes, but principles can endure—purpose, accountability, auditability, portability, non-discrimination, due process, safety and appeal. Regulation can focus on function and risk rather than on a particular interface. Reply 9: coherence requires coordination forums, shared standards, conflict-of-law rules, data interoperability, appellate hierarchies and periodic joint review. Polycentricity is not a celebration of contradiction; it is managed plural authority. Reply 10: the state has a distinctive power of coercive enforcement, but social authority, knowledge, markets, professional norms and digital infrastructure are other sources of real influence. To protect democracy, the state must regulate them while itself remaining limited by other centres. Indian Philosophical and Political Dialogue It would be historically misleading to search Indian traditions for the modern phrase “polycentric democracy” as such. Yet the histories of divided authority, assemblies, communities, associations, dharma, justice, rājadharma, local custom, debate and multi-level social institutions open comparative questions. The purpose is not to grant a modern concept an ancient certificate, but to understand both similarity and difference. The Vinaya tradition of the Buddhist Saṅgha raises questions of institutional decision, rules, discipline, community and dispute resolution. It is not modern democracy, but it can offer historical examples that provoke questions about collective rule formation and accountable procedure. Jain anekāntavāda is not a theory of political power, but epistemic many-sidedness teaches a caution: the viewpoint of one centre cannot capture the whole situation. This epistemic lesson can be useful in polycentric decision processes, although normative democracy does not follow automatically from it. The Nyāya tradition of debate teaches a discipline of public reason—proposition, reason, example, counterargument, defect and defeater. For modern institutions it is a useful reminder that claims to authority should not end in reasonless command. Kauṭilya’s Arthaśāstra presents a complex structure of multi-level administration, intelligence, taxation, cities, villages and state capacity; but its governmental perspective is not democratic polycentricity. Its comparative use lies in understanding the organisation of power, not in imitation. The history of gaṇa-saṅghas shows alternative forms of collective rule, but caution is necessary regarding membership, social boundaries and the historical sources. Modern universal citizenship should not be equated with ancient oligarchic or clan-based assemblies. The Constitution of India provides a more direct basis for modern polycentric structure through the relations among federalism, states, local bodies, an independent judiciary, the Election Commission, regulatory institutions, cooperative federalism and fundamental rights. गजेन्द्र ठाकु रक समानान्तर दर्शन — खण्ड २ Panchayati Raj can increase local participation and administrative proximity. Reservations intended to increase representation of formerly excluded groups also show the social-inclusion dimension of polycentric democracy. Yet proxy representation, family control, fiscal dependence and bureaucratic capture remain matters for scrutiny. Federalism is not merely a list of powers; it is also a mechanism for negotiating shared problems such as fiscal transfers, inter- state coordination, language, migration, rivers, disasters and infrastructure. Powers should be clear, but overlap requires forums for coordination. Fundamental rights constitute the constitutional floor of polycentricity. If a local majority, religious institution, company or the state itself arbitrarily violates a citizen’s basic dignity or equality, there must be a route to higher review. Public-interest litigation, judicial review, the right to information, audit, electoral competition, media and civil society can become separate nodes of accountability. Each is vulnerable to capture and excess; they are best understood as complementary and mutually checking institutions. India’s linguistic diversity offers an epistemic lesson for polycentric governance. If administrative information is available only in one high-status language, formal participation will not become real participation. Language accessibility is a structural condition of democratic access. Digital public infrastructure can widen access to services, but if identity, authentication, payment or welfare platforms become single points of failure, offline remedies, grievance redress, data protection and human fallback are necessary. In Indian cities, overlapping authority among state governments, municipal bodies, development authorities, utilities, traffic police, private contractors and resident groups often creates confusion about accountability. Polycentricity treats this overlap not merely as a coordination problem but as a question of clarity of authority. The comparative conclusion is this: Indian traditions and the modern Constitution both provide resources for thinking about multi-level institutions, but the standards of modern polycentric democracy are universal dignity, equal citizenship, public reason, contestability and rights protection. Mithila’s Parallel Perspective Historical regions such as Mithila, Vajji and Anga are connected through multiple networks of modern administrative boundaries, the Nepal–India border, language, river basins, migration, towns, villages, universities, religious sites and markets. Problems in such regions cannot be fully understood from a single capital-centred viewpoint. The water questions of the Kosi, Kamla, Bagmati, Gandak and other rivers do not end at village or district boundaries. Upstream and downstream flows, embankments, irrigation, roads, settlement, disaster, Nepal–India coordination, farmers and scientific institutions all require multi-level participation. River governance is a direct field of experimentation for polycentricity. Flood history shows that local experience and central engineering knowledge are both necessary. Local memory is not a substitute for technical modelling, and satellite data or models are not substitutes for village micro-topography, social vulnerability and historical knowledge of channels. A combined evidential system is required. In Maithili language policy, schools, universities, literary institutions, publishers, digital platforms, the state, institutions in Nepal, the diaspora and local communities are distinct centres. Language preservation does not mean giving one institution complete control; networks of open standards, Unicode, archives, teaching materials and community creation can be more durable. Panjis, genealogies, local records, temples and monasteries, private collections and public archives are separate custodial centres. Digital preservation requires common metadata, provenance, duplicate backups and rules of access; making an entire cultural memory depend on a single private holder is risky. Migration carries Mithila’s governance beyond territorial boundaries. People living in Delhi, Mumbai, Gujarat, Punjab, Gulf countries, Nepal and elsewhere affect local economies, families, education and the digital public sphere. Diaspora voices are useful, but should not automatically override the direct interests of local residents. Village-level institutions provide proximity, but local prestige, lineage, caste and wealth can suppress dissent. Secret ballots, representation of women and disadvantaged groups, external complaint forums, digital and offline transparency, and legal assistance can form a protective ring around local democracy. In Madhubani, Darbhanga, Janakpur, Purnea, Bhagalpur and surrounding urban areas, transport, waste, water, heritage, housing and the informal economy are often divided among many agencies. A single grievance map and an agency- responsibility chart can improve accountability for citizens. Within a cross-border cultural region, state sovereignty can be respected while non-sovereign cooperation continues in literature, research, environment, tourism, disaster information and linguistic resources. Polycentricity does not erase borders; it seeks the appropriate level of coordination. Local media and e-journals can become centres of public knowledge, but neither editor nor community is the final authority on truth. Source review, correction policy, archives, version history and multiple contributors can make knowledge power more accountable. In school governance, parents, teachers, students, local-language experts, education departments and examination bodies have distinct roles. If curriculum is decided in a closed way by a single actor, local language and history can disappear; complete localism, on the other hand, can weaken common educational minima. Mithila’s Parallel conclusion is that regional autonomy, local knowledge and cultural dignity are central values, but their sustainable form requires constitutional rights, evidence, cross-border coordination, anti-capture rules and multi-level appeal. Analytical Framework for Polycentric Democracy For democratic evaluation of any institution or network, the first question should not be “How much power?” but “What kind of power?” Coercive, economic, informational, classificatory, infrastructural, reputational, professional, agenda-setting and relational power operate through different means. The second question concerns jurisdiction: in what field is the actor a legitimate decision-maker? A hospital may set medical protocols, but punishing a citizen’s political views lies outside its jurisdiction. A platform may make service rules, but if an opaque score controls access to credit, additional scrutiny is required. The third question concerns dependence: can the affected person genuinely exit? Even where nominal alternatives exist, switching costs, loss of network connections, geographic monopoly or livelihood dependence can make exit meaningless. The fourth question concerns transparency: what level of the decision rule, eligibility criteria, evidence, ownership, funding, conflicts, data sources, error rates and change logs should be public? Legitimate limits arising from security or privacy should be stated separately. The fifth question concerns participation: who was consulted and who was not; was information given before or after the decision; was dissent from affected groups recorded; was the decision schedule accessible to participation? The sixth question concerns reason-giving: does the institution merely say “this is policy”, or does it offer relevant reasons? Reasons should be meaningful, specific and capable of being challenged on appeal. The seventh question concerns appeal: at what level, within what time limit and with what degree of independence is an appeal heard? If the original decision-maker hears its own appeal without a review standard, a formal remedy is not a real remedy. The eighth question concerns remedy: after an error is established, will only the record change, or can there also be compensation for harm, restoration of service, public correction, data deletion or systemic policy change? Remedies should be proportional. गजेन्द्र ठाकु र The ninth question concerns inclusion: are grievance mechanisms inaccessible because of language, disability, lack of internet, missing documents, poverty or geographic distance? A polycentric order requires multi-channel access. The tenth question concerns capture: on whose resources or relationships does a board, committee, regulator, association or local body depend? Conflict disclosures, cooling-off periods, procurement transparency and membership diversity should be examined. The eleventh question concerns scale-fit: is the problem local, regional, national, cross-border or global? Does the level of decision match the ecological or social scale, or merely administrative convenience? River basins do not obey district boundaries. The twelfth question concerns reversibility: if a policy is shown to be wrong, can it be changed, or will sunk costs and technical lock-in make it permanent? Modular design and sunset clauses increase democracy’s learning capacity. The thirteenth question concerns data power: who collects data about whom, how long are they retained, with whom are they shared, what inferences are drawn, and how far can the individual access, correct or delete them? Informational asymmetry is a central form of power in network governance. The fourteenth question concerns metric power: what behaviour does a performance score incentivise? If police are measured only by case counts, schools only by exam scores, or hospitals only by throughput, their missions can be distorted. The fifteenth question concerns the constitutional floor: however popular, local or expert an actor may be, which rights remain non-derogable? Without such a floor, polycentricity can become a federation of dominations. Forms of Network Power The first form is infrastructural power. An actor without whom others cannot function—a payment rail, identity system, cloud provider, telecom backbone, app store or search engine—can become a gatekeeper. Infrastructure neutrality, redundancy and due process are therefore political questions. The second form is protocol power. Technical standards determine who can connect, which formats count as valid and which features are possible. A protocol can have distributive consequences even when described as merely technical. The third form is ranking power. Search results, feeds, marketplace placement, school rankings, credit scores and risk-list visibility can change opportunities. The robustness, susceptibility to gaming, bias and appealability of ranking criteria must be examined. The fourth form is moderation power. Rules of online communities set limits on public discourse. Complete laissez-faire can increase harassment and manipulation, while excessively opaque moderation can suppress speech. Clear rules, notice, reasons, graduated sanctions and appeal are useful. The fifth form is identity power. When an authentication provider declares a person “valid” or “invalid”, access can be affected. Fallback procedures are essential for biometric mismatches, name variations, document errors or network outages. The sixth form is financial chokepoint power. Decisions by banks, payment processors or credit bureaux can cut a person or business off from markets. Fraud control is necessary, but unexplained deplatforming can cause severe harm in high-stakes settings. The seventh form is knowledge power. Journals, universities, accreditation bodies, expert panels and databases can determine which knowledge receives recognition. Peer review is necessary, but transparency in gatekeeping, conflict policies and plural inquiry are also necessary. The eighth form is reputational power. Rating systems, influencer networks, media framing or public blacklists can alter behaviour. Questions arise about correction of reputation, standards of evidence and disproportionate permanence. The ninth form is contractual power. Standardised terms can create wide-ranging rules without individual negotiation. Even where legal consent is formally present, substantive choice may be limited; review of unfair terms is part of democratic market governance. The tenth form is supply-chain power. A dominant buyer can alter suppliers’ wages, environmental practices or production standards. This can be positive leverage, but it can also shift unilateral burdens onto smaller producers. The eleventh form is algorithmic coordination. In ride-hailing, delivery, warehousing, content work or gig labour, task allocation can be governed by software. A worker’s “boss” may be a metric or algorithm rather than a person; explanation, contestation, data access and collective voice are therefore necessary. The twelfth form is emergency network power. Disaster communication, health tracing or relief-distribution networks can coordinate rapidly, but extraordinary data or surveillance authority should be rolled back when the crisis ends. The thirteenth form is social-graph power. Who is connected to whom, who is visible and who is trusted affect how information flows through a network. Bot manipulation, coordinated harassment and artificial amplification can distort democratic attention. The fourteenth form is standards-consortium power. Standards created by industry bodies or international technical groups may arise outside national law yet become the practical basis of conduct. Representation of affected publics, documentation and review are therefore necessary. The fifteenth form is cloud and AI dependence. If a public agency depends completely on a proprietary model or provider, policy capacity can become tied to private infrastructure. Procurement should therefore consider portability, audit rights, data location, termination assistance and open interfaces. Institutional Design: How Polycentricity Can Be Built One: create a responsibility register. Every public or high-impact system should clearly identify the owner, operator, vendor, regulator, auditor, grievance authority and final accountable officer. Two: publish a decision-rights matrix stating which actor recommends, who approves, who may override and who hears appeals. Ambiguous overlap increases corruption and blame shifting. Three: translate the constitutional or rights floor into operational policy through minimum checklists for non-discrimination, notice, hearing, privacy, accessibility and proportionality. Four: define the limits of local discretion. Local bodies may adapt to context, but they should not fall below protected rights or national minima. Five: provide fiscal equalisation so that local autonomy in poorer regions does not become merely nominal because of lack of resources. Transparent formulas, needs-based transfers and audit are required. Six: create multi-level ombudsmen. If a local grievance is not resolved, a higher independent forum should be available, with a simple, low-cost route for small claims. Seven: where participatory budgeting or deliberative forums are used, rules should govern participant selection, information packs, facilitation, conflict disclosure and minority reports. Eight: make public consultation meaningful—publish drafts before decisions, allow adequate time for response, use local languages, provide summaries of comments and issue a response matrix explaining why proposals were accepted or rejected. Nine: data-sharing agreements should be purpose-limited. Polycentric governance should not become a universal data pool; every act of sharing requires a legal basis, security, retention limits and audit. Ten: use interoperable records so citizens do not have to submit the same evidence repeatedly to different offices, while maintaining granular access control. Convenience is not the same as unlimited internal surveillance. Eleven: conduct regular independent technical audits of high-risk systems, ensuring auditor independence, adequate scope, transparent methodology and tracking of remediation. गजेन्द्र ठाकु रक समानान्तर दर्शन — खण्ड २ Twelve: require incident reporting for system failure, data breaches, mass wrongful denial, moderation errors and infrastructure outages, including public summaries. Accountability is incomplete without learning. Thirteen: apply sunset clauses to new high-risk policies or technologies; renewal should depend on evidence, not inertia. Fourteen: a regulatory sandbox should not be merely an industry convenience; it should include protection of affected users, capped exposure, transparent evaluation and an exit plan. Fifteen: common open standards can reduce vendor lock-in. Procurement should require export formats, APIs, documentation and migration support. Sixteen: collective representation can raise structural problems beyond individual grievances by workers, consumers, students, patients or platform users. Unions, associations, user councils and advisory boards can therefore have legitimate roles. Seventeen: institutionalise minority reports. Dissent should not disappear inside committee consensus; reasoned disagreement can become valuable evidence for future review. Eighteen: conduct periodic power reviews. Which actors have accumulated disproportionate influence? Examine market concentration, data concentration, lobbying, procurement dependence, political finance and ownership links. Polycentric Democracy in the Digital Age In the digital public sphere, platforms can simultaneously perform the roles of editor, market, infrastructure and community. This combination of roles increases conflicts of interest. If a content host is also advertiser, recommender, seller and rule- enforcer, functional separation or independent oversight may be appropriate. A universal single rule for content moderation is difficult because language, satire, contexts of violence, newsworthiness, political speech and local law differ. Yet publication of rules, notice, retention of evidence, appeal, emergency exceptions and transparency reports can provide a common procedural floor. Algorithmic feeds can indirectly influence democratic choice. “Engagement” is not a neutral metric; it can incentivise outrage, novelty or sensationalism. User control over ranking, chronological options, recommender explanations and research access can be useful safeguards. Political advertising can be governed through transparency about sponsor identity, expenditure, targeting criteria, audience size, archival access and disclosure of synthetic media. Microtargeting can turn persuasion into a wholly private act and make public rebuttal difficult. In the age of AI-generated content, authenticity infrastructure requires several independent centres—publisher signatures, platform labels, provenance metadata, fact-checking, media literacy and legal remedies. A single central authority of truth is dangerous, but so is complete anarchy. Federation can be an alternative to a single database for digital identity, but federation requires interoperability, security and recovery schemes. A person should not be completely cut off from basic services because of an identity-system failure. Open-source software can increase the capacity for public audit, but “the code is open” does not automatically mean that the system is accountable. Training data, governance, deployment context, security, funding and decision responsibility are separate questions. Open data can increase transparency, but it also carries risks of privacy loss, re-identification, low-quality data and unequal analytic capacity. Public data release should therefore include risk assessment, metadata and correction channels. Information sharing is necessary in cyber-security, but secrecy classifications should not become permanent loopholes. Sensitive details can remain confidential while governance, incident counts, accountability and independent oversight remain possible. The digital divide is a direct question for polycentric democracy. If online consultation alone excludes poor, elderly, rural, disabled or low-connectivity citizens, “digital participation” is not representative. Offline channels should remain available. AI policy can involve a central regulator, sector regulators, courts, standards bodies, public procurement, the research community, civil society, user groups and company internal governance. No single regulator can handle every technical and contextual problem. For cross-border platforms, national law is necessary, but enforcement remains limited without regulatory interoperability, treaties, mutual assistance, common safety standards and transnational research access. In children’s digital environments, parents, schools, platforms, regulators and the child’s evolving autonomy are distinct centres. Complete parental control and complete platform discretion are both inadequate; age-appropriate design and grievance routes are necessary. The central maxim of digital polycentricity is this: infrastructure should be distributed, authority clear, rights portable, decisions contestable, and human or social fallback available when systems fail. Contemporary Applications In municipal water supply, publish a role map for the municipality, utility, ward committee, laboratory, contractor and citizen group. Water quality should be tested by an independent laboratory, complaints should have both ward-level and central- helpline routes, and outage data should be public. In flood management, a basin-level authority, district administration, village committees, meteorological services, engineers and local-knowledge networks should share a dashboard. Evacuation decisions should respond to local context, while upstream forecasts are coordinated at a regional level. In school governance, clearly define the relationship among a national learning floor, state curriculum, local-language adaptation, school management committees and teachers’ professional autonomy. Examination marks alone should not be the performance metric. Universities require a multi-level balance among academic freedom, departmental autonomy, university governance, accreditation, funding agencies and student rights. No single political or managerial centre should control an entire discipline of knowledge. Health systems should distinguish the roles of national safety standards, hospital protocols, doctors’ judgement, patient consent, insurance rules and grievance bodies. An insurance algorithm should not be the final authority on medical necessity. In police technology procurement, elected governments should set policy, police should state operational needs, independent privacy and institutional audits should be conducted, court review should be available, and local community impact hearings should be held. For high-risk technologies such as face recognition, use limited pilots, documented purposes, watch-list governance, false-match audits, human confirmation, deletion rules and an independent complaint mechanism. In welfare distribution, there may be a central eligibility framework, but it should be accompanied by local handling of exceptions, offline verification, social audit, district appeal and automatic restoration after a proven system error. In banking and credit, a multi-centre framework can combine central regulation, credit-bureau standards, bank judgement, consumer-dispute bodies and court remedies. Specific reasons should be given for adverse actions. Gig-work platforms should combine company freedom over scheduling with duties concerning worker data access, pay-rule transparency, notice before deactivation, appeal, occupational safety and collective representation. In social-media moderation, a layered system of automated detection, trained human review, user appeal, an independent supervisory board and transparency reporting can reduce the errors of any single centre. In the electoral information ecosystem, clarify overlapping roles of election authorities, parties, media, platforms, fact-checkers, civil society and courts. Provenance, advertising transparency, spending rules and rapid legal remedies are better than a state “ministry of truth”. गजेन्द्र ठाकु र Heritage protection should involve joint governance by archaeology departments, local communities, municipalities, scholars, tourism agencies and owners. Neither tourist revenue nor ritual claims alone should determine outcomes. River-basin development should combine engineering decisions about dams and embankments with evidence about environment, displacement, livelihoods, sediment, downstream risk and local history. Consultation with affected communities should not be a formality after project approval. Diaspora funding can help local schools and heritage projects, but governance boards should balance local residents, teachers or users, donors and independent audit. Giving money should not automatically confer total control of the agenda. Where several institutions jointly operate a research-data repository, common metadata, access tiers, licences, provenance, correction logs and responsibility for long-term preservation should be explicit. In AI-assisted public administration, model vendors may make recommendations, but officials should remain decision owners, independent audit should be available, citizens should receive explanations and appeal rights, and procurement contracts should preserve audit rights. Public–private partnerships should not eliminate public accountability in the name of “private efficiency”. Service standards, tariff rules, performance data, appropriate contract disclosure, grievances, termination conditions and continuity plans should remain in place. In disaster relief, central resource mobilisation, district coordination, panchayat-level beneficiary verification, civil-society delivery support and public dashboards can complement one another; duplication and exclusion should be audited. In language technology, the state can support corpora, universities research, communities contribute, open-source developers build tools, publishers supply content and platforms provide distribution. Corpus access, attribution, licensing, privacy and quality standards should be shared rules. Chapter Conclusion The chief achievement of polycentric democracy is the recognition that democracy cannot be protected merely by finding the “right centre”. Modern power is distributed across many locations; networks of accountability must therefore also be built across many locations. Polycentricity is not a simple synonym for decentralisation. Local institutions, the state, courts, regulators, markets, platforms, communities and expert bodies may each possess legitimate domains; but each requires limits, reasons, transparency, appeal and higher or lateral review. Network power resides in material and technical structures—protocols, rankings, identity, payment, cloud systems, contracts, data and reputation. Such power cannot be removed from democratic scrutiny merely by calling it “private”, especially when citizens’ basic opportunities depend upon it. At the same time, it would be wrong to convert every form of private influence into state bureaucratic control. Proportional obligations, interoperability, independent oversight, competition, collective voice, judicial remedies and open standards provide multiple tools. The epistemic side of democracy is equally important. Local experience, scientific evidence, administrative data, citizen testimony, audit, journalism and historical memory are distinct sources. Multi-evidential rationalism combines them according to quality, independence and relevance. Dignity-based ethics supplies the floor of a polycentric structure: no person should be turned into a disposable input, a fixed label or an appeal-less subject of any centre. Non-domination is the political maxim of this chapter. In regions such as Mithila, Vajji and Anga, questions of rivers, language, cross-border culture, migration, local government, archives and the digital public sphere demand multi-level coordination. Here polycentricity is not only a theory but a practical necessity.