The kingdom needs the forest people. Their tracking. Their knowledge of which trees yield which timber. Their understanding of where the elephant herds move and how to capture them without losing the catchers. Their ability to fight on terrain other soldiers cannot move on. The kingdom does not control the forest people. They live outside the caste system. They organize themselves in tribes whose internal structure the Arthashastra never quite describes. Their loyalties to the king are less reliable than the loyalties of any other category of subject. The king uses them anyway, because he has no choice. They occupy the zones his administration cannot reach, and they are the only ones who know how those zones work.
Trautmann names the structural position directly: "forests contain forest people, who are very different from farmers, lying outside the caste system, organized in tribes. They are hard to control by the king but essential for the kingdom because they have valuable knowledge and skills pertaining to the forest that village-dwellers do not have."1
Three structural features in one sentence. Outside the caste system. Organized in tribes. Hard to control. The first feature is sociological — the chaturvarnya hierarchy that orders settled life does not order them. The second is political — they have their own internal authority structures, and those structures are not the king's. The third is operational — even when the king wants to control them, he cannot fully do so. The combination produces a population that is part of the kingdom in some respects (they live within its borders, fight in its armies, supply forest goods) and outside it in others (they are not taxed in the standard ways, not subject to the standard legal regime, not legible to the standard administrative apparatus).
The Arthashastra at 2.17.1-3 attaches forest people to specific economic zones: "The king is to establish forests for materials, one for each of the kinds of produce (kupya), and workshops for goods made from them, with forest people (atavi) attached to the forests."1 The atavi are not a free-floating tribal population. They are functionally embedded in the kingdom's productive geography — the materials forests need them, the elephant forests need them, the border zones need them.
But the embedding is one-way. The kingdom needs the forest people for what they produce and what they know. The forest people are not symmetrically dependent on the kingdom. They can survive without it; the kingdom cannot survive without their forests. The asymmetry is the source of their leverage.
The forest people's leverage derives from specialized knowledge that is hard to transfer. Trautmann is direct about this: their "valuable knowledge and skills pertaining to the forest that village-dwellers do not have."1 The knowledge includes which trees are the right age for timber, where the elephant herds are concentrating in any given season, which routes through the forest are passable in monsoon, which plants are medicinal and which are toxic, how to track a wounded animal across rocky ground, how to fight an enemy who knows the same tricks. None of this is written down. None of it is taught in any formal institution. It exists only in the heads and bodies of the forest people, transmitted across generations through practice.
Village-dwellers cannot acquire this knowledge by visiting the forest. Acquisition requires growing up in it. The king who tries to bypass the forest people by sending his own administrators into the materials forest finds his administrators returning with bad timber, missing the elephant census, getting lost on patrols, dying from poisoned plants. The knowledge is the strategic asset. The forest people are the only repository.
The Arthashastra recognizes this and engineers around it. The supervisors who direct the forest people are not themselves forest people. They are specialists who live in villages — elephant-trainers, physicians, the cowherds who handle the cattle herds at the forest edge. Trautmann names the structure: "the elephant trainer and physician are settled there (along with the cowherd) by grant of the king. They are not forest people, and live in villages, but they supervise and direct forest people who live where they work."1
Two-tier architecture. Village-dwelling specialists provide the interface between the kingdom's administration and the forest people. The forest people do the actual forest work. The specialists translate between the two worlds — converting the forest people's specialized knowledge into formal records (the elephant census), into kingdom-legible products (timber graded by quality), into military deployments (forest scouts assigned to specific army units).
The most strategically interesting use of forest people is military. The Arthashastra at 9.2.6-8 describes forest troops as "especially useful as guides and scouts, fighting on certain kinds of terrain, countering certain modes of fighting, and fighting the enemy's forest troops."1 Four use cases. Each names a domain where conventional armies fail and forest armies succeed.
Guides because they know the terrain. Scouts because they can move undetected through forest the enemy cannot search. Terrain combat because they can fight in conditions where infantry formations cannot deploy. And — the most interesting case — countering the enemy's forest troops, who present the same kind of advantage the kingdom's own forest people do. Forest warfare is its own specialty. Standard armies cannot fight it. Only other forest fighters can.
But Trautmann names the constraint: "their loyalties are less certain for kings than other kinds of soldiers."1 The forest troops are useful precisely because they are not fully integrated into the kingdom's apparatus — and the lack of integration is also their weakness as soldiers. They will fight for the king when the cause aligns with their interests. They may not fight for the king when the cause does not. They can desert into the forest in conditions that would be impossible for soldiers from the settled population. They can switch sides if a competing patron offers better terms. Their independence is the resource and the risk.
The Arthashastra does not solve this. It does not propose to make the forest people into reliable soldiers like the kshatriya warriors of the settled population. Doing so would require integrating them into the caste system, which would require breaking the tribal structures that produce their forest knowledge in the first place. The kingdom cannot have both — fully reliable forest troops and genuine forest knowledge. The Arthashastra accepts the trade-off and uses the forest people on terms that include their unreliability.
The Arthashastra's accommodation of the forest people is more sophisticated than it appears. Most pre-modern states tried to either fully integrate frontier populations or fully eliminate them. The Arthashastra does neither. It maintains a relationship of partial integration in which the forest people remain partly outside the kingdom's authority but are functionally part of its operations.
This is the architecture James C. Scott calls the "art of not being governed" applied from the state's side. Scott's account, focused on Southeast Asian highland populations, describes how frontier peoples maintain political autonomy by keeping themselves illegible to lowland states. The Arthashastra's account, from the lowland state's perspective, describes how a state that needs services from such populations operates without insisting on full legibility. The two accounts converge: the frontier is not where the state ends; the frontier is where the state operates differently.
The two-tier supervision architecture (village-dwelling specialists directing forest people) is the operational form of the partial integration. The kingdom does not need to govern the forest people directly. It needs to interface with them through people who can. The interface is the technology that makes the relationship work without requiring the integration the relationship cannot support.
The atavi attachment to materials forests at 2.17.1-3 (line 1147), the elephant-forest guarding by forest people at 2.2.6-9 (line 1158), the village-settled supervisor structure at line 1166, the four military use cases at 9.2.6-8, and the framing of forest people as "outside the caste system, organized in tribes" at lines 1241-1248 are attested in Kangle's translation. Trautmann's synthesis ("essential part of the royal enterprise... difficult to control because of it") is at line 1248.1
The Arthashastra's accommodation works because the kingdom does not need full governance over the forest people — it needs services. But the relationship requires that the forest people remain in the forest, with their knowledge, their tribes, and their independence. If the forest is destroyed (cleared for agriculture, logged out, converted to other uses), the forest people lose their knowledge base and their leverage simultaneously. The kingdom that grows by clearing forests undermines its own capacity to operate at the frontier. The text does not resolve this; in practice, it played out historically as the gradual displacement of forest populations by agricultural expansion, with consequent loss of frontier capacity.
A second tension: the loyalty problem is structural, not contingent. The forest people's reliability cannot be improved through the standard mechanisms (caste integration, formal employment, ideological training) without destroying the forest knowledge they are valued for. The Arthashastra implicitly accepts that some kingdom operations will fail because the forest troops desert at critical moments, in exchange for the operations that succeed because the forest troops are present. This is an honest trade-off the text names without trying to solve.
[Single source — Trautmann/Kangle. Olivelle 2013 priority second source for verification. The atavi designation, the materials-forest and elephant-forest attachments (2.17.1-3, 2.2.6-9), the village-settled supervisor structure, the military use cases at 9.2.6-8, and the "outside the caste system, organized in tribes" framing are attested in Kangle's translation. The reading of the partial-integration architecture as a deliberate trade-off is Trautmann's interpretive synthesis; the primary text describes the architecture without explicitly theorizing it as a trade-off.]
The plain version: any modern state with internal frontier populations — populations the formal apparatus cannot fully integrate but which provide essential services — runs into the same architectural problem. The Arthashastra's accommodation is a model. Modern states rarely manage it as cleanly.
History: The atavi/kingdom relationship maps directly onto modern equivalents — the Pashtun belt at the Pakistan-Afghanistan border, the Amazon basin's indigenous populations relative to Brazilian state authority, the Kurdish populations across Turkey/Syria/Iraq/Iran, the highland populations of Southeast Asia (Scott's case study). In each case, formal state authority is partial, the population provides services the state cannot do without (resource extraction, terrain knowledge, frontier security), and the relationship oscillates between partial integration and conflict. The Arthashastra's two-tier architecture (village-dwelling specialists supervising the frontier population) is the historical norm. Modern attempts to fully integrate these populations have generally failed; modern attempts to fully eliminate them have been catastrophic and incomplete; modern attempts to maintain the partial-integration accommodation tend to be unstable but sometimes work. The handshake reveals: the Arthashastra's structural insight applies cross-temporally and cross-civilizationally. The frontier-sovereignty problem is not a special case; it is the structural form of state authority at the edge of its administrative reach.
Behavioral Mechanics: Behavioral Mechanics Hub — The atavi are an early case study of what behavioral mechanics studies as embedded specialist populations. Modern equivalents are pervasive — the senior engineers at a tech company who maintain critical legacy systems no one else understands; the medical specialists who can diagnose conditions the rest of the field cannot; the relationship managers who hold the firm's most valuable client connections in their personal networks. Each is, structurally, an atavi. They have specialized knowledge that is hard to transfer. They have leverage proportional to the institution's dependence on the knowledge. Their loyalties are partly to the institution and partly to their own specialty community, and the loyalties can shift. The behavioral-mechanics insight is that knowledge asymmetry produces durable bargaining power, and the Arthashastra's two-tier supervisor architecture is the structural management response. Modern institutions that try to bypass the bargaining power (by replacing the specialist, by documenting the knowledge, by reducing dependency) often discover that the knowledge resists capture in exactly the way forest knowledge resisted capture by the king's administrators. The Arthashastra's accommodation — maintain the relationship without insisting on full integration — is the durable management strategy.
Pillai's Inside Chanakya's Mind names the broader sovereignty network the atavi belong to as a four-category schema: forest chiefs (atavika), frontier chiefs (anta-pala), city chiefs (paura), and countryside chiefs (janapada). Sutra 1.16.7 prescribes that the king's diplomatic and intelligence apparatus must independently maintain relations with all four — they are not collapsed into a single "rural population" or "outlying population" category but treated as four distinct sovereignty types, each with its own protocols and each capable of independent rebellion or alliance.P2 See Sutra 1.16.7: Forest/Frontier/City/Countryside Chiefs Network.
The four-category schema reframes the atavi page. The forest people are not the only internal frontier — they are one of four sovereignty types within the kingdom's borders, and the king's strategic-intelligence apparatus must maintain separate relationships with each. The two-tier supervisor architecture this page identifies is one specific instance of the more general principle: the king's authority operates through local intermediaries who hold their own authority within their constituency, and the kingdom's stability depends on maintaining good relations with the intermediaries rather than collapsing them into the formal bureaucracy. Pillai's reading puts the atavi into a fuller political ecology — they are part of a sovereignty network the kingdom maintains through the spy-establishment (mantra-shakti) and the four-upayas approach (sama-dana-bheda-danda — conciliation, gifts, division, force) calibrated to each chief-type. See Spy Establishment as Information Order for the network's information layer. The cross-tradition handshake produced: modern institutions running specialist-population strategies are typically managing one frontier-type without recognizing that the Arthashastra names four — and a competent management architecture builds out separate relations for each.
The Sharpest Implication
If the partial-integration architecture is the structural norm for any institution with embedded specialist populations, then most modern attempts to "fully integrate," "fully professionalize," or "fully document" specialist knowledge are fighting the underlying structure. The Arthashastra would say: stop trying. Maintain the partial relationship. Build the supervisor layer that interfaces between your formal apparatus and your specialist population. Accept that the specialists' loyalties will be partly outside your control, and engineer your operations around that fact rather than against it. Modern HR practices, knowledge-management systems, and "everyone is replaceable" rhetoric routinely violate this principle and pay the consequences in degraded specialist performance and high turnover. The architecture works when you accept what it is.
Generative Questions
The forest people's loyalty was always uncertain because they had alternatives — they could survive in the forest without the kingdom's protection. Modern specialists often have analogous alternatives (other employers, independent practice, retirement). Does the Arthashastra's framework — accept the uncertain loyalty, build the architecture around it — work when the alternatives are commercial rather than geographic?
The two-tier supervisor architecture (village-dwelling specialists directing forest people) is what the Arthashastra prescribes. Modern equivalents include middle-management roles that interface between executive direction and specialist execution. Are these roles being designed with the Arthashastra's structural insight, or are they more often being eliminated as "unnecessary layers" in flattening exercises that destroy the interface the operation actually needs?
The forest-people relationship was historically eroded by agricultural expansion that destroyed the forest. Modern equivalents — automation that replaces specialist judgment, codification that captures specialist knowledge in documents, AI systems that simulate specialist reasoning — may be doing structural damage to specialist populations whose long-term value the institution does not yet appreciate. What's the equivalent of the agricultural-expansion failure mode, and how is it playing out across modern industries?
[VERIFIED — source re-read 2026-04-30]