In Indic thought, the daṇḍa (“staff” or “scepter”) is a symbol of the capacity to inflict harm. It was associated in particular with the king, as defined by his use of violence in the practice of governance. The term is usually translated as “punishment,” although it has a somewhat broader semantic range. Within Dharmaśāstra, daṇḍa is the ultimate worldly means of enforcing the norms of dharma . The king’s role in the legal system, from this perspective, is shaped by his right and obligation as daṇḍadhara (“wielder of the staff”) to punish those deemed deserving of punishment under the law or who fail to perform their ascribed penances. The threat and actualization of his violence serves to put the law “in force.” At the same time, the king’s daṇḍa represents his raw ability to dominate others by force and is the fundamental source of his political power. By submitting it to considerations of expedience (artha ) and lawfulness (dharma ), daṇḍa is rendered a productive force for effective governance and further enhancement of the king’s power.
In this way, daṇḍa stands at the intersection of the political and the legal, the normalization (in a literal sense) of domination, where coercive violence becomes just punishment. An analysis of punishment, therefore, is an ideal place to explore how relations of domination (including, but not limited to, the king) were articulated and legitimized within the legal imagination of Dharmaśāstra. And in respect of this, there are two primary questions: Who could punish? And who was immune from punishment? Of these, the latter is far more complex, inseparable, as it is, from the dynamic contours of a penology characterized by differential punishment based on status. The application of punishment, whether harshly, lightly, or not at all, was shaped not only by the prejudices of a deeply inegalitarian society, but also by issues such as sovereignty, worth, and personhood. The former question—who may punish?—is better suited to the present context and addressing it will, at least, shine some light on the political community imagined within Dharmaśāstra as revealed by the recognition of the right to punish and its limits.
We read in the Nāradasmṛti :
gurur ātmavatāṃ śāstā śāstā rājā durātmanām |
atha pracchannapāpānāṃ śāstā vaivasvato yamaḥ || (NSm 19.57)
The guru is the punisher of the self-possessed; the king is the punisher of the wicked;
And Yama Vaivasvata is the punisher of those who conceal their sins.
This verse is probably an aphorism that circulated in legal circles and beyond (cf . VaDh 20.3). It identifies three “punishers” (śāstṛs , a term that can also mean “teacher”) of men: their guru, their king, and the god of the dead, Yama, himself a king. Now, this verse does not use the technical term daṇḍa (“punishment”), and the guru’s punishment likely refers to prāyaścitta , which forms an independent and restricted domain within the greater legal world of Dharmaśāstra. Yet, the verse is clearly conceptualizing punishment in the widest scope. So, while it should not be read as a constitution delegating the power to punish, it serves nevertheless as a useful entry point into the multipolar world of punishment in the legal imagination of the Dharmaśāstra literature, where two figures, the king and the teacher, predominate.
In Indic thought, the king was uniquely associated with the power of punishment. Manu goes so far as to equate punishment with the king himself (MDh 7.17). One of his epithets in this regard was daṇḍadhara , “the wielder of the staff” (e.g., NSm Mā 1.2), and among his sacred duties (rājadharmas ), along with waging war and protecting living beings, was his obligation to punish those deserving of punishment (e.g., ĀpDh 2.10.6; GDh 10.7–10.8). When the texts invoke punishment for offenses, often in the context of vyavahāra (litigation in state courts), it is to the king that they appeal. As Manu states:
anubandhaṃ parijñāya deśakālau ca tattvataḥ |
sārāparādhau cālokya daṇḍaṃ daṇḍyeṣu pātayet) || (MDh 8.126)
He [i.e., the king] should inflict punishment on those deserving punishment only after he has fully ascertained the proclivity, as also the time and place, accurately, and considered carefully the ability of the criminal and the severity of the crime. (Tr. Olivelle 2005a)
Kings stood at the apex of the political system and they were generally considered immune from accusations and punishment by worldly agents (e.g., NSm 15–16.21). So, how was royal wrongdoing recognized and addressed? There is provision in the literature for kings paying fines for their misdeeds. In Manu we read:
kārṣāpaṇaṃ bhaved daṇḍyo yatrānyaḥ prākṛto janaḥ |
tatra rājā bhaved daṇḍyaḥ sahasram iti dhāraṇā || (MDh 8.336)
In a case where an ordinary person is fined 1 Kārṣāpaṇa, the king should be fined 1,000—that is the fixed rule. (Tr. Olivelle 2005a)
This rule, however, raises a vexing issue. Who possessed the authority to determine that a king was guilty of an offense and, if so, to punish him? Certainly, our texts posit no such worldly authority standing over the king. On this verse, the commentator Medhāthiti states that the king should punish himself through an offering to Varuṇa (see below). But, there was clearly widespread discomfort with even this idea, as we find little support for Manu’s rule in other smṛtis. The medieval jurists Lakṣmīdhara and Devaṇṇabhaṭṭa go so far as to change the wording of the verse so that it applies not to the king, but to his underlings or associates (see Olivelle 2005a: 320).
Rather, it is another kingly god of Vedic origin, Varuṇa, who is said to punish kings. He is, to be specific, “the lord of punishment”:
īṣo daṇḍasya varuṇo rājñāṃ daṇḍadharo hi saḥ | (MDh 9.245ab)
Varuṇa is the lord of punishment, for he is the daṇḍadhara over kings (Tr. after Olivelle 2005a)
Varuṇa is conceived as a kind of cosmic sovereign over kings, who are his worldly cognates, and their relationship is mediated through daṇḍa . This passage is used by Manu to support a previous rule, in which he orders a king to offer to Varuṇa any fine taken from a mahāpātakin (one guilty of a grievous sin) by throwing it in the water (or giving it to a learned and virtuous Brāhmaṇa). Yājñavalkya applies this practice as the remedy to any unjust fine levied by a king (YDh 2.310). The de facto punishers of kings, however, are the agentless soteriological mechanisms by which unjust or greedy kings take on the sins of wrongdoers and find hell in the next life and by which good kings reach heaven (e.g., MDh 8.386; YDh 1.353; KātSm 960–1).
As to punishment itself, a few different typologies are to be found. Nārada states:
śārīraś cārthadaṇḍaś ca daṇḍas tu dvividhaḥ smṛtaḥ |
śārīrā daśadhā proktā arthadaṇḍās tv anekadhā || (NSm 19.60)
Punishment is known to be twofold: corporal punishment and monetary punishment.
The corporal is declared to be tenfold, while the monetary is manifold.
Indeed, a wide variety of amercements are levied in the legal codes (NSm 19.62–19.64, etc.), but prominent among them is a standard tripartite set of fines called the lowest, middle, and highest fines for violence (pūrva -/adhama -/prathama -, madhyama -, and uttamasāhasa ). These are not restricted to cases of sāhasa (“violence”), but borrow from them only the value of the fines assessed for various instances of such crimes. The classical jurists differ on their exact value (cf . MDh 8.138; YDh 1.361; NSm 19.37–19.38).
The ten-fold corporal punishment to which Nārada refers is given first by Manu:
daśa sthānāni daṇḍasya manuḥ svayaṃbhuvo ’bravīt |
triṣu varṇeṣu tāni syur akṣato brāhmaṇo vrajet || (MDh 8.124)
upastham udaraṃ jihvā hastau pādau ca pañcamam |
cakṣur nāsā ca karṇau ca dhanaṃ dehas tathaiva ca || (MDh 8.125)
Manu, the son of the Self-existent One, has proclaimed ten places upon which punishment may be inflicted. They are applicable to the three classes; a Brāhmaṇa shall depart unscathed.
They are: genitals, stomach, tongue, and hands; feet are the fifth; and then, eyes, nose, ears, wealth, and body. (Tr. Olivelle 2005a)
Manu’s inclusion of “wealth” (dhana ) in his list not only undermines Nārada’s later twofold division, but it also undermines the rule at MDh 8.124, for it is typically understood that Brāhmaṇas are exempt from corporal but not pecuniary punishment (although Medhāthiti, interpreting this couplet strictly, argues that it does). Nevertheless, this tenfold list gives us a good idea of the various corporal punishments ascribed throughout the legal codes. The commentators tell us that punishment of the “body,” the tenth place, is a reference to capital punishment.
We find another, fourfold typology of punishment in Manu (see also YDh 1.362; BṛSm 1.29.2ff., etc.):
vāgdaṇḍaṃ prathamaṃ kuryād dhigdaṇḍaṃ tadanantaram |
tṛtīyaṃ dhanadaṇḍaṃ tu vadhadaṇḍam ataḥ param || (MDh 8.129)
vadhenāpi yadā tv etān nigrahītuṃ na śaknuyāt |
tadaiṣu sarvam apy etat prayuñjīta catuṣṭayam || (MDh 8.130)
He should employ first the punishment of verbal reprimand; next a public denunciation; third, a fine, and finally, corporal punishment.
If he is unable to restrain them even with corporal punishment, then he should impose on them all these four. (Tr. Olivelle 2005a)
The notion of escalating degrees of punishment as a means of restraining offenders is somewhat at variance with the practice of prescribing specific punishments (or set of options) for an offense, but it conforms with more general instructions on punishment that emphasize its distinctiveness from the rendering of a verdict as well as the consideration of various externals by the judge (cf . GDh 12.51; MDh 8.126; YDh 1.363; NSm 19.45, etc.).
Aside from the well-known penance for theft (treated as a mahāpātaka rather than a dispute between parties), in which he was to strike the penitent with a pestle or club (ĀpDh 1.25.4, etc.), the king presumably delegated his power to carry out punishment to his appointed subordinates. They are little discussed by the classical jurists, if at all.
The king’s power to adjudicate disputes was routinely delegated to appointed judges (see below), but the dharma writers also recognize the authority of a variety of non-state legal forums. Nārada enumerates five venues of litigation comprising the greater legal system:
kulāni śreṇayaś caiva gaṇāś cādhikṛto nṛpaḥ |
pratiṣṭhā vyavahārāṇāṃ gurvebhyas tūttarottaram || (NSm Mā 1.7)
Families, guilds, assemblies, appointed judge, and the king: these are the venues of litigation, each latter one superior to each former.
While Nārada (following YDh 2.31 and further refined at BṛSm 1.1.58) presents these as a neat linear hierarchy, we should be cautious not to take this model as evidence of an integrated legal system. Even so, it does reveal how a jurist like Nārada thought about jurisdiction in the plural legal order of the classical period. The question we face here is how the right to punish was understood to be distributed among these various legal authorities.
The dharma texts address their instructions on vyavahāra (litigation in state courts) to the king, and they do so with reference to his unsurpassed power as chief judge and daṇḍadhara of the realm. At the same time, they recognize that this power is routinely delegated to appointees (MDh 8.9–8.11, etc.). In fact, the Arthaśāstra of Kauṭilya , our most important source from the classical tradition of statecraft, addresses its rules on vyavahāra entirely to appointed judges called dharmasthas rather than to the king (3.1.1ff.). In the dharma literature, such appointed judges are called by a few different names, such as sabhya or prāḍvivāka . All of these are the adhikṛtas in the formula above, professional judges appointed by the king. In the texts, they serve as his judicial substitutes in the fullest sense (with the exception that the king himself could overturn their rulings: YDh 2.32, etc.). So, the power to adjudicate disputes prescribed in discussions of vyavahāra is implied as much for appointed judges as for the king. Kane, following commentators such as Medhāthiti and Devaṇṇabhaṭṭa (see below), argues that the power to declare a verdict was distinct from the power to assign punishment, which remained with the king (1973: 391; cf . Jolly 1928 : 290–2). However, Bṛhaspati (1.1.91) holds that the appointed (Brāhmaṇa) judge could personally carry out two forms of punishments, verbal reprimands (vāgdaṇḍa ) and public censure (dhigdaṇḍa ), while only the king could inflict pecuniary or corporal punishment. At any rate, to the extent that an appointee could act as a full substitute, he must have had some power to rule as well as sentence. Most typically, the guilty parties are said to be handed over to the king after the verdict for punishment, whether for sentencing or simply the infliction of the punishment (NSm Mā 2.43; esp. BṛSm 1.1.88).
But, what of the right of the “lower courts,” the kula, śreṇi , and gaṇa , to punish? Bṛhaspati addresses the question directly:
kulaśreṇigaṇādhyakṣāḥ puradurganivāsinaḥ |
vāgdhigdamaṃ parityāgaṃ prakuryuḥ pāpakāriṇām || (BṛSm 1.17.17)
taiḥ kṛtaṃ ca svadharmeṇa nigrahānugrahaṃ nṛṇām |
tad rājño ’py anumantavyaṃ nisṛṣṭārthā hi te smṛtāḥ || (BṛSm 1.17.18)
The officials among families, guilds, and assemblies, dwelling in cities and fortresses, shall carry out the verbal reprimand, public denunciation, and abandonment of wrongdoers.
And, in following their individual duty, whatever favor or disfavor they confer upon men must be approved by the king himself, for they are declared to have been entrusted.
Recalling the four types of punishment listed by above, this verse restricts these lesser legal authorities from two kinds of punishment: monetary and corporal. We learn here also that they operate as extensions of the king’s judicial authority, with which these officials have been entrusted and to which they are responsible. Elsewhere, Bṛhaspati places further limits on them:
rājñā ye viditāḥ samyak kulaśreṇigaṇādayaḥ |
sāhasanyayavarjyāni kuryuḥ kāryāṇi te nṛṇām || (BṛSm 1.1.92)
Families, guilds, assemblies and the like that have been duly approved by the king may try cases among men, with the exception of sāhasa.
Here, the lesser courts are forbidden from adjudicating disputes that are sāhasanyaya , meaning here cases involving “rulings on violent crimes” (see NSm 14.2–5; Rocher 1954 –55). This would circumscribe many instances calling for corporal punishments according to the legal codes anyway, and it is likely that some convergence between these two rules should be understood.
Devaṇṇabhaṭṭa takes up the issue also in the Smṛticandrikā (III: 45), seeming to equate these lower courts with the “permanent” (pratiṣṭha ) and “impermanent” (apratiṣṭha ) courts mentioned at BṛSm 1.1.57, as opposed to the court overseen personally by the king (śāsita ) or by an appointed judge who possesses the king’s signet (mudrita ). He argues that judges in pratiṣṭha and apratiṣṭha courts could not adjudicate cases pertaining to sāhasa , and neither could they assign punishment (daṇḍadāpana ) or fines (arthadāpana ). Kane argues that they were essentially “arbitration courts” (III: 280).
A different perspective on the matter is offered by Medhāthiti in his bhāṣya on MDh 8.2, in which he comments extensively on the verse from Nārada cited above (Mā 1.7). There, he observes that the five different legal authorities enjoy different rights. He says:
The King’s right extends up to the infliction of punishments (daṇḍa ), while that of the Brāhmaṇa [i.e., the adhikṛta ] and the others [i.e., the kula, śreṇi , and gaṇa ] extends only up to the pronouncing of judgments (nirṇaya ), this latter right is distinct from the former. (Tr. Jha 1920–39: 8)
Medhāthiti, like Devaṇṇabhaṭṭa, distinguishes the right to pronounce a judgment from the right to carry out punishment. All of the courts render decisions, but only the king can punish. We have already seen, however, that Bṛhaspati recognized the right of the lesser courts to inflict certain forms of punishment (1.17.17), even if Medhāthiti and Devaṇṇabhaṭṭa disagree. And certainly, it appears that the lower courts enjoyed the right at least to enforce verdicts, for Medhāthiti’s earlier comments speak specifically about the steps taken by śreṇis to ensure their rulings were followed:
…the members of a guild fight shy of any matter relating to themselves going before the King, as that would lend the King’s officers an opportunity for interfering in the work of their guild; and hence they always take from the parties concerned sufficient security against their deviating from the decision arrived at, before they proceed to investigate a dispute; the understanding with the person standing security being that if the party deviate from the decision arrive at by the guild, he shall pay a stipulated fine, or he should not let him deviate from it. (Tr. Jha 1920–39).
Medhāthiti’s discussion of the gaṇa also refers to the enforcement of decisions on their part (Jha 1920: 7). All of this falls short, however, of ascribing to the lesser courts any right to punish, even the verbal forms endorsed by Bṛhaspati (1.17.17).
Somewhat different from these, however, are the rights assigned by Medhāthiti to the head of house:
Similarly the ordinary householder (gṛhin ) also would be an “authorised person” [i.e., appointed judge] so far as his own household-affairs are concerned,—this being in accordance with the declaration that “the householder is master in his own house” (svatantras tu gṛhe gṛhī ), which means that he is free to deal with all disputes (vyavahāra ) within his own household, up to the infliction of punishment (daṇḍaparyanta ),—specially with a view to proper discipline among his children and pupils; but he may deal with all cases, except the inflicting of bodily punishment (śārīra daṇḍa ), or the doing of acts conducive to depravity. What is meant is that in the case of minor offences (svalpa aparādha ) the householder (gṛhastha ) himself acts like the King, while in that of serious offences (mahat vyatikrama ), it is necessary to report to the king. (Tr. Jha 1920–39)
Here, the householder is analogized to the appointed judge and to the king himself, at least with respect to his own domain. This power is granted with respect to his “independence” (svatantra ) and it clearly applies to those who are dependent upon him, such as his children and pupils. He has the right to adjudicate disputes and punish minor offenses within his house, but he is forbidden from inflicting corporal punishment. Offenses that are more serious are referred to the king. Although we cannot strictly equate correlate the gṛha with the kula or Medhāthiti’s mahat vyatikrama with sāhasa in Bṛhaspati’s rule above, certainly this passage would seem to confirm the recognition of a right to punish. As much is recognized elsewhere in the tradition:
bhāryā putraś ca dāsaś ca śiṣyo bhrātā ca sodaraḥ |
prāptāparādhās tāḍyāḥ syū rajjvā veṇudalena vā || (MDh 8.299)
pṛṣṭhatas tu śarīrasya nottamāṅge kathaṃ cana |
ato ’nyathā tu praharan prāptaḥ syāc caurakilbiṣam || (MDh 8.300)
When they misbehave, a wife, son, slave, pupil, or uterine brother may be beaten with a rope or a bamboo strip on the back of their bodies and never on the head. If he beats them any other way, his liability is the same as for theft. (Tr. Olivelle 2005a)
In this passage, certain individuals are assigned the right to punish others physically for their offenses (aparādhas ), which punishment, however, was limited to striking the back with a bamboo strip or rope. The subject of this rule is not given in the text, but the identities of the various individuals that may be punished suggest that the subject may be the head of house (gṛhin ; gṛhastha ); although, Medhāthiti argues that the subject is whoever bears the relation, that is, a husband, father, master, teacher, and brother. Obviously, these are not exclusive statuses. Whoever is the subject of the rule, a limited right to punish appears linked to the notion of “independence” (svatantra ). Nārada identifies three individuals who are independent:
trayaḥ svatantrā loke ’smin rājācāryas tathaiva ca |
prati prati ca varṇānāṃ sarveṣāṃ svagṛhe gṛhī | (NSm 1.28)
Three in this world are independent: the king, the teacher, and the householder of each caste in his own home.
Later, he links this, in part, to the right to chastise or discipline:
svatantrāḥ sarva evaite paratantreṣu sarvadā |
anuśiṣṭau visarge ca vikraye ceśvarā matāḥ || (NSm 1.34)
To say that someone is independent means that he has persons dependent on him, and that he is empowered to discipline, expend, and sell them. (Tr. Lariviere 1989a)
It is important to note, as Medhāthiti emphasizes, that the householder’s right to punish is not merely for the purpose of beating, but is meant to have the salutary effect of helping to keep the offender on the right path (seemingly a common goal of legal authority: see Medhāthiti on MDh 8.2). Hence, milder forms of censure are to be tried first. As much is reflected in Āpastamba’s instructions to teachers, where he gives a sense of graduated punishment, without, however, endorsing physical violence:
aparādheṣu cainaṃ satatam upālabheta | abhitrāsa upavāsa udakopasparśanam adarśanam iti daṇḍā yathāmātram ā nivṛtteḥ | (ĀpDh 1.8.28–1.8.29)
When a pupil does something wrong, the teacher should always correct him. Instilling fear, making him fast or bathe, and banishing him from his presence are the punishments, and he should apply them according to the severity of the offense until the student has completed his studies. (Tr. Olivelle 2000 )
The right of such people to punish their dependents, however, seems to have been understood as exercised only under or within the sovereign authority of the king. So we read in Nārada’s instructions to teachers, where the same limit on physical violence as seen in MDh 8.299–8.300 is observed:
anuśāsyaś ca guruṇā na ced anuvidhīyate |
avadhenāthavā hanyāt rajjvā veṇudalena vā || (NSm 5.12)
bhṛśaṃ na tāḍayed enaṃ nottamāṅge na vakṣasi |
anuśāsyātha viśvāsyaḥ śāsyo rājñānyathā guruḥ || (NSm 5.13)
If [a student] does not obey, the teacher may punish him; he may beat him with a rope or a split bamboo cane as long as he does not hurt him. He must not beat him harshly, nor on the head or chest; but after chastising him, he must encourage him; otherwise the king must punish the teacher. (Tr. Lariviere 1989a)
Nārada’s rule makes clear that the teacher’s right to punish is limited and that the king is guarantor of that limit. For, if he exceeds it, the teacher is to be punished by the king. As to the precise extent of the limit, Nārada adds the important information that the student is not to be hurt (vadha ) by the beating. This underlines the impression that the king policed the right to punish and reserved for himself or his official delegates the right to inflict punishments considered to result in harm. What is less clear is the extent to which the householder’s gṛha can be correlated with the kula as one of the five legal domains. On the balance, it seems that Medthāthiti and others think of them separately, with the rules for the gṛhin expressed in Manu as exceptions for specific individuals carved out from the more general rules governing physical assault.
* * *
A systematic perspective on the right to punish in the dharma literature seems hardly possible given the nature of our sources, although we do see some agreement that all or certain forms of punishment were reserved for the king, with exceptions made for “independent” individuals such as heads of house or teachers. But, even if the passages above seem to converge on a rough consensus regarding the right to punish and its limits, we can find further passages that take a different perspective, such as the following from the Nāradasmṛti :
śvapākapaṇḍacaṇḍālavyaṅgeṣu vadhavŗttiṣu |
hastipavrātyadāreṣu gurvācāryāṅganāsu ca || (NSm 15–16.12)
maryādātikrame sadyo ghāta evānuśāsanam |
na ca taddaṇḍapāruṣye doṣam āhur manīṣiṇaḥ || (NSm 15–16.13)
yam eva hy ativarterann ete santaṃ janaṃ nṛṣu |
sa eva vinayaṃ kuryān na tadvinayabhāṅ nṛpaḥ || (NSm 15–16.14)
If a śvāpāka , a man who is impotent, a caṇḍāla , a cripple, a butcher, an elephant driver, a man who is uninitiated, their wives, or the wives of an elder or preceptor should violate customary rules, an immediate beating (ghāta ) is their punishment; the wise say that physical assault (daṇḍapāruṣya ) on these is not a crime. The virtuous man whom they offend is the very one who should punish them; the king has no role in their punishment. (Tr. Lariviere 1989a)
It appears that both in point of degree of harm as well as jurisdiction, this rule goes much farther than those previously discussed. For, daṇḍapāruṣya , the legal title covering physical assault, can include grave physical injury, and the commentator Bhavasvāmin argues that “beating or even killing these persons for violation of customary rules does not involve a crime and the ones who do the beating are not punishable” (Lariviere 1989a II: 188). Not only does this allow the infliction of harm by private parties, but it also allows it well beyond the narrow sphere of their particular dependents. This may have struck many, however, as too liberal of a policy, as Bhavasvāmin, despite endorsing the rule, advises the king not to encourage such behavior (188).
Insofar as the concept of “punishment” represents the legalization of domination, we can see the diversity of attitudes about it in the tradition as reflecting various claims about who has the right to enforce their will upon whom, some mediated through conceptualizations of the greater legal order and others ascribed to individual status, likely with some degree of overlap. And just as the royal power in premodern India never fully effaced the self-determination of all groups in the realm, so too the dictates of Dharmaśāstra never developed into a monolithic legal system effacing the plural legal order. What we have in this instance, then, is something of a heterogeneous record, partly descriptive and partly prescriptive, as to whose domination of whom was recognized as legitimate within legal imagination and what its limits were.