The granary overseer has to know what happens when paddy is milled. He has to know that one volume of paddy becomes a different volume of husked rice, that one volume of sesame becomes one-fourth its volume in oil, that legumes lose half their bulk when pounded. He has to know these ratios because his inventory accounting depends on them, and because the quality of the milling determines who eats which grade of rice. The king gets the finest milling — five adhakas from five dronas of paddy. The war elephant gets nine. The soldier gets eight. The slave gets the broken grains. The hierarchy of the kingdom is enacted, daily, in the differential weight and grade of rice issued from one storehouse to bowls scattered across the palace, the army, and the stables. The granary is the ledger where political theory becomes daily food.
The overseer of the granary (koshtagara-adhyaksha) is responsible for the technical operation of the storehouse — protection, movement, processing, ration issuance — and the accounting that holds the operation together. Trautmann lists the staff: "the sweeper, watchman, weigher, measurer, supervisor of measuring, dispenser, delivery supervisor, receiver of tallies, slaves and labourers."1 Eight functional roles plus generalized labor. The granary is not a building; it is a workplace with specialized division of labor.
The processing equipment is detailed: implements for weighing and measuring, grinding-stones, pestle and mortar, pounding machine, oil press, fan, winnowing basket, sieve, cane-basket, box, broom (2.16.62-63). The granary mills its own grain, presses its own oil, refines its own products. Raw paddy comes in; processed rice and oil go out. The overseer manages the transformation.
The transformation creates an accounting problem. Sutra 2.15.24 names the requirement: "He should personally observe the quantity by which grains increase or decrease when pounded, rubbed, ground or fired, moistened, dried or cooked."1 Volume changes during processing. The overseer who does not know the conversion ratios cannot tell whether the processing has been honest. The grain that came in as one volume should leave as a known fraction of that volume. Discrepancies are theft, spoilage, or accounting error — and the overseer must know enough to distinguish them.
Trautmann itemizes the ratios. Rice yields half its hulled volume when milled. Priyangu yields half plus one-ninth. Wheat and barley pounded yield the same. Sesame, barley, mudga, and masha beans when rubbed yield similar fractions. Legumes lose half their bulk; lentils lose a third. Linseed oil comes in at one-sixth the seed weight; sesame oil at one-fourth.1 The list goes on. Each ratio is an inventory anchor. Each anchor is a check against the people who handle the grain at every step.
This is bureaucratic-mathematical work, and it is also the substance of governance. The granary that fails its accounting fails the kingdom's reserve discipline. The half-granary famine rule — half held back for distress, half for use — only works if the overseer can verify that the use half is what is actually being drawn down. Without conversion-ratio knowledge, the half-granary rule degrades into an aspiration.
The Arthashastra's ration system applies two principles: finer quality for higher status, larger quantity for larger bodies. Trautmann names them as the "two implied criteria"1 underlying ration issuance.
The quality criterion runs through the milling. From five dronas of paddy, the granary can produce different yields depending on how finely the rice is milled. Better milling yields less rice but finer rice. The yield/quality tradeoff is operationalized as a ration ladder:
Twelve adhakas of coarse rice from five dronas of paddy goes to the young elephant — bulk, low quality, sufficient for an animal whose digestion is not particular. Eleven adhakas, slightly finer, goes to the mature elephant. Ten adhakas to the riding elephant. Nine adhakas to the war elephant — fewer adhakas, finer rice, the war animal whose performance the kingdom depends on. Then the human ladder begins. Eight adhakas to soldiers. Seven to chiefs. Six to queens and princes. And five adhakas — the highest quality, the smallest yield from the same five dronas — to the king. The king's ration is "or one prastha of rice-grains, unbroken and cleansed" (2.15.42).1
At the bottom of the system: broken grains. The grains that fragmented during milling are reserved for "slaves, workers and cook's helpers, and lesser animals such as ducks, geese, goats, sheep and cattle" (2.15.62, 52-56).1 Broken grains are still food. They are not contaminated. They are simply what remains after the unbroken grains have been allocated to the people whose social position warrants whole grain.
The hierarchy is not justified in the text. It is presented as operational fact. The king eats the finest rice. The slave eats the broken grains. The granary's job is to deliver each ration to its designated recipient on schedule.
The quantity criterion runs through proportional sizing. For Arya (upper-caste) males, the ration is one prastha of rice, one-fourth prastha of curry, salt at one-sixteenth of the curry, butter or oil at one-fourth of the curry. Lower castes get less: one-sixth prastha of curry, half the butter or oil. Women get less by a quarter. Children get less by a half. Trautmann is direct: "ration units are proportionate to the status of the person and the body size."1
The combination of the two criteria produces a fully specified ration matrix. Every recipient has a known quality grade and a known quantity. The granary's daily output is the sum across the matrix.
A modern reader looking at the ration ladder sees caste hierarchy implemented as differential nutrition. This reading is correct but incomplete. The system also does technical work that the hierarchical reading can obscure.
The processing efficiency varies. Better milling yields less rice. If the granary milled all rice to the king's standard, the kingdom would starve. If it milled all rice to the slave's standard, the king would eat coarse food and the social signal would collapse. The ladder distributes the milling cost across the population in a way that maximizes total caloric output while maintaining the hierarchical signaling the regime depends on. The system is calibrated, not arbitrary.
The proportional sizing also has technical justification. Soldiers do more physical labor than chiefs. War elephants need more food than riding elephants. Larger bodies need more calories than smaller ones. The Arthashastra's rule — quantity proportional to body size — is also nutritionally correct, even if it is also socially encoded. Children genuinely need fewer calories than adults; the half-ration for children is not arbitrary cruelty, it is approximate calorie-needs accounting.
The hierarchical encoding sits on top of the technical accounting. Where the two reinforce each other (large adult Arya males working hard need more calories, and the system gives them more), the architecture works smoothly. Where they conflict (a slave doing heavy labor needs the same calories as a free worker doing the same labor, but gets less because of caste position), the architecture imposes the hierarchy on the calorie reality. The Arthashastra does not flag this conflict. It treats the hierarchy as the master variable.
The granary staff list and equipment inventory are at 2.16.62-63 (line 757). The conversion-ratio observation requirement is at 2.15.24 (line 765). The specific ratios (rice/priyangu/legumes/lentils/oils) are at line 767. The quality ladder (12 down to 5 adhakas, plus broken grains) is at 2.15.42, 2.15.52-56, 2.15.62 (line 777). The quantity rules (Arya males, lower castes, women, children) and curry proportions are at lines 779-781. The horse-attendant monthly allowance is at 2.30.3 (line 781). All conversion ratios and ration specifications are direct from Kangle's translation.1
The ration system encodes caste hierarchy as a daily material practice. Trautmann's gloss — "ration units are proportionate to the status of the person and the body size" — equates two different things: technical body-size proportionality and social caste hierarchy. The Arthashastra does not separate them either. The result is a system in which it is impossible to tell whether a given ration differential is technical (smaller body needs less food) or hierarchical (lower caste gets less food). The conflation is part of how the hierarchy is naturalized — it presents itself as the natural extension of obvious technical rules.
A second tension: the ration matrix's complexity is also its fragility. With many recipient categories, many quality grades, many quantity adjustments, the system requires high-quality administration to operate as designed. In practice, ration corruption (substituting lower-quality grain, short-weighting, misallocating) was likely common. The Arthashastra's prescriptive precision is partly an attempt to constrain this corruption by specifying everything; the corruption likely happened anyway, and the prescriptive system functions partly as a baseline against which actual practice could be audited.
[Single source — Trautmann/Kangle. Olivelle 2013 priority second source for verification. The conversion ratios, quality grades, and quantity rules are attested in Kangle's translation. The "two implied criteria" framing (finer quality + larger quantity for higher-order recipients) is Trautmann's interpretive synthesis of the matrix's underlying logic; the primary text presents the rules without explicitly theorizing the unifying criteria.]
The plain version: the ration system is one of the clearest pre-modern examples of social hierarchy enforced through material practice rather than through ideology alone. The hierarchy is not preached. It is fed. Modern equivalents (institutional cafeterias, prison rations, military mess hierarchies, in-flight meal classes) reproduce versions of the same architecture. The principle — distribute differential material outputs to enforce hierarchical relationships — is durable across civilizational frames.
History: The ration matrix is the operational form of the caste system at the level of the royal household. The Arthashastra elsewhere prescribes caste relationships in legal and religious terms; the ration system is the same hierarchy enacted at the level of daily food. Modern parallels are striking. Military rations are tiered by rank in functionally identical ways (officers' mess vs. enlisted mess, with different food quality and presentation despite equivalent calorie content). Prison food is tiered by classification level. In-flight meals are tiered by ticket class. Each system encodes a hierarchy through the food. The Arthashastra makes the encoding explicit; modern systems often obscure it. The handshake reveals how durable the architecture is — and how much modern food-tiering systems are descendants of an ancient pattern that hierarchical societies repeatedly converge on.
Behavioral Mechanics: Behavioral Mechanics Hub — Differential rationing as a hierarchy-maintenance mechanism is structurally identical to what behavioral mechanics studies under the heading of status signaling through differential consumption. The king's unbroken whole-grain rice is a status signal. The slave's broken grains are also a status signal — the absence of the unbroken rice marks the slave as lower. The signaling works because the food is visible: in any communal meal, in any procession of supplies, in any ritual feast, the differential is on display. The behavioral-mechanics insight: hierarchical systems that depend on differential consumption are stable when the differentials are visible, and unstable when they are concealed. Modern equivalents — first-class lounges, executive dining rooms, premium product lines — preserve visibility deliberately. The Arthashastra's ration system was visible by necessity (everyone saw what everyone ate). The mechanism that maintained the hierarchy 2,300 years ago is the same mechanism that maintains hierarchies in modern consumer society. The ration matrix and the luxury-goods market are structurally related.
The Sharpest Implication
If hierarchical relationships are maintained through differential material practice — through who eats what, who sees what, who occupies which physical space — then political philosophies that treat hierarchy as primarily ideological are misdiagnosing the architecture. You do not undo a hierarchy by changing what people believe about it. You undo it by changing the material practices that enact it. The Arthashastra's ration matrix is a model: change the matrix, change the hierarchy. Modern equivalents work the same way. Desegregating institutional cafeterias did more for civil rights than any number of speeches. Equalizing access to first-class amenities undermines class distinctions more effectively than any number of editorials about equality. The implication: hierarchy is in the architecture. Reform requires architectural change.
Generative Questions
The Arthashastra's ration system encoded caste hierarchy through food differential. Modern bureaucratic systems often encode hierarchy through space differential (corner offices, executive floors, separate elevators). What other axes of hierarchical encoding are operating in any given institutional context — and which axis is most consequential?
The conversion-ratio knowledge required of the granary overseer was substantial — eight to ten ratios across multiple processing stages, applied to dozens of materials. Modern accounting has automated this kind of conversion tracking. What new failure modes does automation introduce that the ancient overseer's personal observation was protecting against? Possible answer: the automated system records what was reported, not what actually happened, returning the system to the embezzlement problem the Arthashastra addresses elsewhere.
The two-criterion architecture (quality + quantity, both calibrated to recipient status) treats body size and social status as proxies for each other in a way modern reasoning would separate. Was Kautilya wrong to conflate them, or is the conflation tracking a real correlation (large adult elite males in agrarian societies actually do more physical labor than young women or children, on average) that modern individualist accounting obscures?
[VERIFIED — source re-read 2026-04-30]