A Dharma Reader: Classical Indian Law

4. We have indications that political science and specifically Kautilya’s Treatise on Politics were significant within the intellectual world of the Guptas. An important Sanskrit drama, Mudrārākṣasa, whose plot is placed at the very beginning of the Maurya Empire in the late fourth century B.C.E., is closely connected to the Treatise on Politics. The hero of the drama is Canakya, who by this time was identified with Kautilya and considered the author of the Treatise. See Willis 2009; Balogh 2015.

5. See also the complex technical vocabulary employed by Kautilya within the context of court proceedings in ch. 10.1.

6. A book called Adhyakṣapracāra (Activities of Superintendents), which is the title of Book 2 of Kautilya’s treatise, probably existed as a separate document both prior to the composition of that treatise and after. It is referred to in Vatsyayana’s Kāmasūtra (1.2.10) and in Medhatithi’s commentary on MDh 7.61; 7.81. For further details, see Olivelle 2013: 11.

7. For a detailed study of these four areas of law both in the text of Kautilya and in later legal treatises, see Olivelle and McClish 2015.

8. The Sanskrit term jāti here may refer to some kind of social group based on birth and not simply to caste in its technical meaning.

9. Kautilya devotes the entire Book 11 to the topic of political saṅghas or confederacies. Commercial and professional bodies are often referred to as gaṇa and śreṇi.

10. These are translations of technical legal terms in Sanskrit: dharma, vyavahāra, caritra, and saṃsthāna, used frequently in other parts of the Treatise on Politics as well.

11. See, for example, AŚ 2.8.3; 2.22.15.

12. The Sanskrit term is pracāra. For this meaning of the term, especially within the context of the two other terms custom (caritra) and canon (saṃsthāna), see Scharfe 1993: 195–203.

13. See the articulation of this principle by Gautama (ch. 2.1: #1).

14. For greater details, see Olivelle 2004a, 2005c, 2006c. More comprehensive histories are found in Olivelle 2009b and Hiltebeitel 2011.

15. For detailed studies of the term in the Rig Veda, see Horsch 1967 (English version 2009) and Brereton 2009.

16. For a detailed study of dharma in this period, see Olivelle 2004c.

17. I have omitted the numerous citations of verses from the Rig Veda found in these texts and counted only once the numerous parallel passages that are repeated almost verbatim.

18. The Kauṣītaki, which parallels the Aitaraya, does not use dharma at all. In the Jaiminīya Brāhmaṇa of the Sāma Veda and the Gopatha Brāhmaṇa of the Atharva Veda the word dharma occurs in eleven passages. In all the reckonings of the term in the brāhmaṇas, I have omitted citations from the respective collections of hymns.

19. For a detailed study of this passage, see Olivelle 1996b. As I explain there, I take the compound dharmaskandha as a Bahuvrīhi. Traditionally, it has been taken as a Tatpuruṣa and translated “divisions of dharma.” The term skandha here refers to the torso or the main part of the body. Note the term ekaskandha for a coconut tree, which has only one trunk and no branches.

20. See below, ch. 6.

21. For an examination of dharma in this literature, see Lubin Forthcoming.

22. For studies and bibliography on Asoka and his ideology, see Falk 2006; Olivelle 2009c; Olivelle et al. 2012.

23. See Thapar 1997 for an overview. My own views on the Asokan imperial ideology based on dharma are found in Olivelle 2012b.

24. See Pollock 1985, 1989a, 1989b. See also the collection of studies on this topic in Dallapiccolla and Lallemant 1989, and my own reflections specifically with regard to “science” in relation to dharma, in Olivelle 2005a: 62–66.

25. I have used various strategies, because a single consistent translation would not do justice to every context. I have employed “treatise,” “authoritative treatise,” and the like, especially because scriptural texts, such as the Veda itself, are frequently called śāstra.

26. The authors of the earliest extant aphoristic texts on dharma refer to and cite seventeen of their predecessors. For these early authors and for an extended treatment of the literary history of the science of dharma, see Olivelle 2010. A much more voluminous history can be found in Kane 1962–1975 and Lingat 1973.

27. See Deshpande 2006.

28. See Katyayana’s vārttika 39 on Panini 1.2.64; Patanjali on Panini 1.1.47 (I: 115).

29. The Vaikhānasa Dharmasūtra was probably a text whose circulation was limited to the Vaikhanasa religious tradition. It did not have an impact on the broader tradition of the science of dharma and is never cited in later legal digests.

30. Aupajanghani (BDh 2.3.33); Bhallavins (BDh 1.2.11; VaDh 1.14); Eka (ĀpDh 1.19.7); Harita (ĀpDh 1.13.11; 1.18.2; 1.19.12; 1.28.1, 5, 16; 1.29.12, 16; BDh 2.2.21; VaDh 2.6); Kanva (Kaṇva, ĀpDh 1.19.3; 1.28.1; Kanva (Kāṇva, ĀpDh 1.19.2, 7); Kapila (BDh 2.11.28); Kasyapa (BDh 1.21.2); Katya (BDh 1.3.46); Kautsa (ĀpDh 1.19.4; 1.28.1); Kunika (ĀpDh 1.19.7; Kutsa (ĀpDh 1.19.7); Mahajajnu (BDh 3.9.21); Manu (ĀpDh 2.14.11; 2.16.1; GDh 21.7; 23.28; BDh 2.3.2; 4.1.13; 4.2.15; VaDh 1.17; 3.2; 11.23; 12.16; 13.16; 19.37; 23.43); Maudgalya (BDh 2.4.8); Puskarasadi (ĀpDh 1.19.7; 1.28.1); Varsyayani (ĀpDh 1.19.5, 8; 1.28.1).

31. The four early aphoristic texts on dharma, as we saw, cite the opinions of numerous other authorities on dharma. There is a refreshingly candid back and forth about these divergent opinions, something lacking in treatises on dharma produced in the first millennium C.E. See Olivelle 2010: 38.

32. These expressions are already found in documents prior to the Common Era: ĀpDh 2.15.1; GDh 11.20.

33. See Derrett 1976: 603. He is reviewing Dumont 1962. Dumont’s notions of hierarchy are further developed in his classic study Homo Hierarchicus (Dumont 1970).

34. The Sanskrit expression used here, sāmayācārika, is not peculiar to Apastamba; it is used by Gautama (8.11) and occurs as a topic heading in Kautilya’s Treatise on Politics (5.5). Apastamba himself (1.7.31) uses it a second time to say that even after a Vedic student has returned home the accepted practice (sāmayācārikam) is that he should behave toward his former teachers in the same way that he did while he was a student. In the Rāmāyaṇa (1.1.19) we have a significant statement: Rama is said to be adept and knowledgeable about laukike samayācāre, “accepted customs connected with the world.” The connection of samayācāra with laukika here is significant for the understanding of this compound.

35. Although this issue has not received much attention from modern scholars, it is a surprising fact that the term ācāra is absent in the Vedic vocabulary; it is not found in any Vedic text, even though the related term ācārya (teacher) is common. It appears to have been aneologism first employed a few times in the treatises on Vedic ritual and more frequently in the treatises on domestic ritual. The grammarian Panini (3.1.10) uses ācāra once to refer to behavior. It is likely that this somewhat new term was adopted as a central concept within the Apastamba school, irrespective of whether the texts of dharma and domestic ritual ascribed to him were written by the same individual. The beginning of his aphoristic text on domestic ritual echoes the beginning of his aphoristic text on dharma: “Next the rites that are gathered from normative practice” (ĀpGṛ 1.1.1). In both texts, the domestic rituals and the norms of dharma are derived from normative practice. Wezler (2004) summarizes his assessment of the two classes of texts: “Gṛhyasūtras as well as Dharmasūtras are verbalizations of certain regional or tribal-specific aspects of traditional social ‘practice’ of the Aryas. Although historically consecutive, they surely have factual points of contact. These verbalizations are textual ‘coagulations’ of the late Vedic period that were apparently regarded by the authors as fundamentally different from older parts of the tradition.” The term ācāra, as opposed to other similar terms such as caritra, appears to have a normative orientation and to refer especially to the customs of normative communities of Brahmans. For a recent study on the use of ācāra in the aphoristic texts on domestic ritual, see Lubin Forthcoming.

36. Patanjali’s use of the term also points to these conclusions. On Katyayana’s Vārttika on Panini 1.1.1 (I: 10–11), ācāra is opposed to jñāna (knowledge) and prayoga (application, usage). Here ācāra is the general behavior pattern, while prayoga is a particular act, both of which are opposed to jñāna (knowledge): one can know, for example, the various nonstandard words for a cow (gāvī, goṇī, etc.), but simply knowing these does not entail a fault or sin, only when one actually uses them (prayoga). An even stronger case for the meaning of ācāra as a habitual behavior pattern or practice is found in his comments on Panini 3.1.11 (II: 21), where the denominative word śyenāyate (“acting like a vulture”) is said to be used when a crow’s ācāra or behavior pattern resembles that of a vulture.

37. See AŚ 3.13.1, 4, 7, 13, 15, 21.

38. The expression used is agṛhyamānakāraṇa (with nonstandard internal Sandhi): see ĀpDh 1.12.8. See also ĀpDh 1.4.8–10. The opposite of this, that is, a practical motive for undertaking something, is called dṛṣṭārtha, something whose purpose is evident or detectable.

39. Deshpande (1993: 17–32) has shown that for Patanjali the terms laukika and vaidika refer to the two distinct subdomains of Sanskrit language. What is significant for our investigation is that laukika in the realms of both language and law refers to areas that are distinct from the Vedic and reflects the usages of living and historical communities. Patanjali is commenting on these terms that are used by Katyayana: Vārttika 2 on Panini 1.2.45 (I: 217); 15 on 6.1.1 (III: 3); 5 on 6.1.83 (III: 55); 2 on 6.2.36 (III: 125).

40. For these examples see Patanjali I: 2. The argument is taken up by Kumarila in his Ślokavārttika (Śabdanityatādhikaraṇa, 276). See Bronkhorst 2011: 143.

41. A similar strategy is employed in the Bhagavad Gītā, where the very first word is dharma, within the compound dharmakṣetre (“in the field of dharma”), and in the MDh, where the centrality of Manu is highlighted in the very first word: manum.

42. This theory departs from the general view in the ritual schools that a person should follow the ritual norms only as laid down in one’s own branch (śākhā) of the Veda. Each branch had its own Vedic and ritual texts; there did not exist a supra- śākhā Veda as the common property of all Brahmans or all Hindus. This notion was probably developed within the science of dharma and Vedic exegesis when dharma became normative for all without respect to ritual schools and Vedic branches. There emerged the concept of a general Veda expressed in the pithy hermeneutical maxim sarvaśākhāpratyaya, namely that rules in each Vedic śākhā are authoritative for all, even for those in other śākhās. See PMS 2.4.8–33; Kane III: 870; IV: 89–90, 453–55.

43. The term will be used again by Manu (2.6) in a verse totally dependent on Gautama, and by Yajnavalkya (1.7), who paraphrases Manu. Thereafter, this metaphor loses traction, even though the expression vedamūlatva becomes commonplace in later theological debates with respect to the epistemology of dharma.

44. On this point, see Wezler 2004; Brick 2006.

45. Klaus 1992: 86. One finds this meaning clearly in the Buddhist use of the term (Palisati) within the vocabulary of mental concentration (satipaṭṭhāna). The Buddhist usage places emphasis on the present, whereas within the discourse on dharma the emphasis is on recalling, that is, making present, what one knows from the past.

46. This is similar to what we will see in Medhatithi: chapters 7.2; 12.2.

47. For legal consultation, see Davis 2014.

48. The term śiṣṭa is absent in the ĀpDh, but Gautama uses it several times outside the context of the epistemology of dharma: GDh 9.70–71; 12.27; 19.2; 28.48, 50. The occurrences in chs. 9 and 28 have the usual meaning of cultured elite and are given within the context of constituting a legal assembly, while those at 12.27 and 19.2 appear to have the meaning of “what is enjoined” or perhaps “what one has been instructed to do.” A similar meaning is attached to the noun śiṣti at 2.42 (śiṣyaśiṣṭir avadhena), where the meaning is to discipline, in the context of chastising a pupil.

49. For the early history of this term, see Pollock 2005.

50. The term is used in other contexts in earlier literature: ĀpDh 1.1.9; 1.5.4, 8; 1.30.5, 9; 2.9.9; 2.23.9. The past participle śruta occurs in ĀpDh 1.8.27; 1.13.19, 20; 2.11.17; and elsewhere.

51. For this last epistemic source of dharma, see Davis 2007.

52. Not all ancient scholars accepted this kind of canon. The ninth-century commentator Medhatithi (on MDh 2.6) explicitly rejects the authority of such lists (see ch. 7.3: #4). Even those who accepted the list note that it is only partial, not comprehensive.

53. In Sanskrit such an opponent is called pūrvapakṣa, and he is a common literary figure in most scientific and scholastic writings.

54. A similar view is expressed in GDh 1.3.

55. See Kane III: 927.

56. See Bhattacharya 1943; Kane III: 926–68. A long list of such prohibitions is given by the thirteenth-century jurist Devanna Bhatta: SmṛC I: 30–32 (see pp. 176–77).

57. See ŚB 2.4.4.14; see also ŚB 4.5.1.9; 5.5.1.11.

58. The last verse is ascribed to the Saṃgrahakāra, the anonymous author of the Smṛtisaṃgraha, by Devanna Bhatta (SmṛC III: 620).

59. For a history of this term in the early literature, especially in Kautilya’s Treatise on Politics, see Olivelle and McClish 2015.

60. It is from this meaning that its philosophical use as “everyday truth” (vyāvahārikasatya), as opposed to the absolute truth (pāramārthikasatya), came about.

61. paṇyavyavahāro vyavahāraḥ: AŚ 2.8.8.

62. See, for example, Vijnanesvara (on YDh 2.6; ch. 13.1: #1), who says that a given lawsuit should deal with a plaint with respect to only a single title of law.

63. This same formal plaint is also called by numerous technical terms, some of which I have given above: pratijñā, sādhya, pakṣa, pūrvapakṣa, artha, abhiyoga.

64. See below, ch. 10.1I: #2; ch. 10.2: #3.

65. The history of this term is complex, and in its early usage it probably meant a judicial interrogator rather than a judge. See Olivelle Forthcoming.

66. See, for example, Davis 2014 and Derrett 1975a.

67. dharmaḥ śāstranyāyadeśaniyatā vyavasthā. Commentary on MDh 8.14.

68. śāśvato dharma anidaṃprathamato yā vyavasthā. Commentary on MDh 8.8.

1. For detailed arguments on the dates of this and other treatises on dharma, see Olivelle 2010, 2012a.

2. See the very similar instruction given to a student who has completed his Vedic study in TU 1.11.4.

3. This is the only occurrence of the term pariṣad (legal assembly) in Apastamba. For descriptions of such assemblies, see ch. 2.1: #3; 2: #2; 3: #5; ch. 4.6.

4. On this topic, see also GDh 1.3–4 (ch. 2.1: #1).

5. The same point is made in texts of Vedic exegesis; see PMS 6.1.15, along with Sabara’s commentary on it.

6. According to the commentator Haradatta, the reference is to black grains, such as beans, or to iron.

7. “Bath graduate” (snātaka) refers to a Vedic student who has completed his studies in his teacher’s house, taken the ritual bath signaling the end of studentship, and returned to his natal home. The argument here is based on a fundamental principle of Vedic exegesis that distinguishes injunctions, which alone are meaningful with regard to dharma, from “explanatory passages” (arthavāda) that do not have injunctive force (PMS 6.7.30). The latter type of passage is further subdivided into metaphor (guṇavāda), reiteration (anuvāda), and historical statement (bhūtārthavāda). Here Apastamba puts the texts cited by the opponent into the second subdivision. For a detailed discussion of these exegetical principles, see Kane V: 1225–56. The passages quoted cannot be identified in any known Vedic text.

8. The meaning is that customs of a particular region and family are authoritative vis-à-vis people belonging to that region or family unless those customs go against Vedic provisions, a principle stated in ĀpDh 2.14.10.

9. The reference is to the practice of levirate (niyoga) where a son is fathered by a brother or member of the family through the wife of a deceased brother.

10. Patanjali has traditionally been assigned to the middle of the second century B.C.E., and my own work on the theological vocabulary of the period confirms this broad dating (Olivelle 2012a).

11. See Patanjali 1.1.47 [I: 114–15]; 1.2.64 [I: 242–43].

12. See also Patanjali 1.1.1 [I: 2]. These and the subsequent rules are common in treatises on dharma.

13. Regarding these dietary prescriptions, see Olivelle 2002a, 2002b. In the case of fowls and pigs, their place of habitation defines their edibility; their wild counterparts may be eaten. On the five five-nailed animals, see also Jamison 1998. While animals with five nails cannot be eaten, five are exempt and may be eaten: hare, hedgehog, porcupine, tortoise, and monitor lizard, although there are variants of this list.

14. This verse is found verbatim in MDh 2.120.

15. For a discussion of the notion of śiṣṭa and the significance of this passage within the grammatical tradition, see Deshpande 2009.

16. The term śiṣṭi here probably refers to proper linguistic usage by śiṣṭas (cultured elite) in accordance with grammar (śāstra). Nevertheless, note the etymological connection among the terms śāstra, śiṣṭi, and śiṣṭa.

17. For the identification of these places and the extent of this sacred region, see Bronkhorst 2007: 1–3.

1. For detailed discussions on the dating of the aphoristic texts on dharma, see Olivelle2000 and 2010.

2. Here I depart from Rocher’s (1976) interpretation of Gautama’s expression: tadvidāṃ smṛtiśīle. He takes tadvidām as connected only with śīla, and takes smṛti (in the sense of a text)as standing alone. I have noted in the introduction the significance of placing the word “Veda” at the very beginning of his composition.

3. In this translation I follow the interpretation of Rocher 1976. The statement on transgressions is an example of the “conduct of those who know it,” and thus would appear to be authoritative. Yet, these transgressions do not constitute dharma, because “those listed later,” namely, recollection and conduct, have lesser power than the Veda, and when they are opposed to dictates of the Veda they become unauthoritative. For a different interpretation of this issue by Apastamba, see ch. 1.1: #8.

4. The Vedic supplements (vedāṅga) are six in number: phonetics (śikṣā), meter (chandas), grammar (vyākaraṇa), etymology (nirukta), astronomy (jyotis), and ritual texts (kalpa: these consisting of treatises on Vedic ritual, on domestic ritual, and on dharma). Subsidiary Vedas (upaveda) appears to refer to treatises on medicine, archery, and the like. The term Purana (in the singular) here must refer to some kind of text containing ancient tales. See the Bṛhadāraṇyaka Upaniṣad 2.4.10; 4.1.2. The term is also used several time in ĀpDh 1.19.13; 1.29.7; 2.23.3; 2.24.6.

5. For the meaning of ūha, see ch. 7, n. 102.

6. The list of the four orders is given at GDh 3.2 in the following sequence: Vedic student, householder, mendicant, and anchorite. So the anchorite is excluded from the assembly.

7. Note that Baudhayana, being the author of an Aphoristic Ritual Text, is more cognizant of the system of Vedic branches (śākhā). Gautama, on the other hand, appears to work at a level beyond the Vedic branches, where engagement with the science of dharma is a common intellectual enterprise, much like Panini with grammar, and less like engagement with a ritual text.

8. See also BDh 1.1.13, where Baudhayana uses the past participle smṛtaḥ, which became the traditional way of referring to statements in the smṛti texts.

9. The reference here, perhaps, is to the methodology of inferring the existence of Vedic texts on the basis of authoritative traditions (anumitaśruti). Note the issues we have already dealt with relating to the inferred Veda.

10. The reference is to the six Vedic supplements; see ch. 2, n. 4. Later texts take this as a reference to epics and Puranas: see ch. 8.2: #2; 3: #2.

11. That is, when the three epistemic sources mentioned in selection #1 are not available, a person may seek the advice of a legal assembly.

12. For these three, see ch. 2, n. 6.

13. The reference is to animals, such as horses, that have incisors in both jaws, while others, such as cows and goats, lack incisor teeth in the lower jaw. See Olivelle 2002b.

14. For the identification of these places, see Bronkhorst 2007: 1–3.

15. Punaḥstoma is a Soma sacrifice performed within a single day. It is prescribed for someone who has accepted too many gifts and feels as if he has swallowed poison. Sarvapṛṣṭhā is a sacrifice offered to Indra and said to be performed by a person seeking virility. See ŚrKo I: 635.

16. For the identification of these places, see Olivelle 2000, Appendix II.

17. The latter appears to be an alternative view.

18. The meaning and reading of pratilomakakṣadharmāṇaḥ are unclear. I follow the explanation given by Laksmidhara (Kṛtyakalpataru, Brahmacārikāṇḍa, 47). The term kakṣa, however, has also the meaning of an authoritative source in later exegetical discourse. If that is the meaning here, the translation would be “but not others, the dharmas given by degenerate authoritative sources” (the reference could be to Buddhist texts and the like).

19. The meaning of “Book of Causes” (nidāna) is unclear. The commentator Krishnapandita takes it to mean a treatise on various regions of India. It is more likely that the reference is to an ancient work of the Bhāllavins cited in the Bṛhaddevatā 5.23.

20. See ch. 2, n. 4.

21. We arrive at an assembly of ten by assuming that there are four individuals, each proficient in one of the four Vedas. For the three belonging to orders of life, see ch. 2, n. 6.

1. For detailed studies of this work and its compositional history, see McClish 2009; Olivelle 2013.

2. For further details, see Olivelle 2013.

3. Gain and loss: in the manufacturing process some substances, such as gold and silver, suffer loss (see AŚ 2.14.8), and others, such as yarn (see MDh 8.397), increase in weight.

4. Sample (prativarṇaka): the term refers to a sample, as it clearly does at AŚ 2.21.12 and probably also at AŚ 2.13.57, where we have prativarṇikā. The meaning appears to be that the value of a shipment is calculated on the basis of testing the quality of a sample.

5. The reference here is to the annual accounting done on the full-moon day of June–July. All state officials and superintendents are required to come with their revenue and account books to the capital. On this annual event, see Heesterman 1985.

6. The reference is to the census carried out by the collector (samāhartṛ) of the countryside within the kingdom.

7. The reference is to secret agents operating in the territory of an enemy king.

8. The term used is vyāvahārikī, referring to the interest rate on commercial transactions. We see here the close connection between convention (vyavahāra) and commerce and trade.

9. Seizure fine (sāhasa): the meaning and semantic history of this term as a particular kind of fine with lowest, middle, and highest amounts are not altogether clear. In the MDh (8.138)the three levels of fines for forcible seizure are: lowest 250 Paṇas, middle 500 Paṇas, and highest 1,000 Paṇas. The AŚ (3.17.8–10), on the other hand, gives them as 48 to 96, 200 to 500, and 500 to 1,000.

1. I have presented evidence for the thesis of a unitary authorship in my critical edition of this text: Olivelle 2005a.

2. For a longer discussion of the significance of verses, see Olivelle 2005a: 25–27.

3. For detailed studies of this issue, see McClish 2012, 2014.

4. On this source of dharma, see the detailed study by Davis 2007.

5. These are the mixed classes or castes born through the intermarriage between the four principal classes; see MDh 10.8–73.

6. That is, the Self-Existent One, who is both the creator and the father of Manu.

7. At MDh 1.35 is a list of ten great seers produced by Manu, who are viewed as his sons and pupils: Marici, Atri, Angirasa, Pulastya, Pulaha, Kratu, Pracetas, Vasistha, Bhrigu, and Narada.

8. The pronoun “him” here refers to the Brahman, who is the principal audience of the text. He is tasked with transmitting the teachings of Manu to the other social classes.

9. Such pure and exalted Brahmans purify those with whom they sit down to eat. See MDh 3.183–186.

10. See ch. 7, n. 99.

11. Later texts take “amplification” (paribṛṃhaṇa) to refer to such other supplementary texts as epics and Puranas: see ch. 8.2, #2; 3, #2.

12. See ch. 2.2: #1 (BDh 1.1.6).

13. See ch. 2, n. 6.

14. The last two verses are also found in BDh 1.1.11, 16 (ch. 2.2: #2).

15. All the commentators of Manu interpret the compound yugahrāsānupūrvaśaḥ to mean “in keeping with the shortening of the ages.” According to this reading, the different dharmas for the ages are caused by the decrease in their length. I think this is less likely. I prefer to connect hrāsa (shortening) with the human life span; indeed, the same term was used with regard to human life in verse 83. According to my reading, the meaning is: “in keeping with the progressive shortening (of the human life span) in each age.”

1. We see a parallel development in the medical tradition, where later texts ignore issues of epistemology and focus on practical issues such as recipes and therapy. I thank Dominik Wujastyk for this observation.

2. Medhatithi, in his commentary on MDh 2.6 (see ch. 7.2), rejects this enumeration of texts of recollection, saying: “For this very reason, the exhaustive list of authors of the texts of recollection: ‘Manu, Visnu, Yama, Angiras,…,’ is without any basis.” Yet these verses appear to have entered the text of Yajnavalkya quite early. They are found in all the extant manuscripts, and are commented on by the ninth-century commentator Visvarupa. They are also given in the extensive citation of Yajnavalkya found in the Garuḍa Purāṇa 93.4–6.

3. The meaning of “others” (itareṣām) is unclear. Commentators take it to mean the intermediate or mixed classes resulting from intermarriage among the four major social classes (varṇa). See MDh 2.18, where the term antarāla is used with the same meaning.

4. For the region where the black buck (kṛṣṇasāra) lives, see MDh 2.23 (ch. 4, #4).

5. There are six Vedic supplements (see ch. 2, n. 4) and four Vedas, bringing the total to fourteen.

6. For a detailed discussion of the composition of this text, see Olivelle 2009a.

7. For the significance of the Varāha incarnation, see Willis 2009.

8. The sixth of the seven nether regions underneath the earth.

9. The reference is to a sin or crime committed by a person. In Krita one abandons the region where such a person may live, while in Kali one has simply to abandon the individual who commits the sin.

1. For excellent analyses of such debates and the intellectual history of the period, see McCrea 2012, 2013, and Forthcoming. For an extended discussion with larger extracts from Vedic exegesis, see the forthcoming book by McCrea in this series: A Mīmāṃsā Reader: Classical Indian Hermeneutics.

2. See Benson 2010: 12–13; Keith 1921: 1–9. For the early literature of Vedic exegesis, see Parpola 1981 and 1994.

3. Sabara takes up the term artha in the root text for comment and gives it the meaning of something good or beneficial, while its opposite is anartha, something that is detrimental to a person’s well-being.

4. All these are rites included within the Vedic ritual repertoire but are intended to kill or harm an enemy. Consequently, they do not qualify as “beneficial” or “good” (artha) within the above definition of dharma. The Hawk sacrifice (śyena) is described in detail in the KātŚr 22.3.1–52; ṢaḍBr 3.8; 4.2; see also Kane V: 1114 n. 1818, and 1245. It is stated explicitly that “Hawk” here is simply the name of the sacrifice; it is not the substance that is sacrificed. It is, in fact, a variation of the Soma sacrifice. The image is that the rite will strike the enemy down swiftly like a hawk descending on its prey. The Thunderbolt sacrifice (vajra) is also a variety of Soma sacrifice, intended to harm the whole country or region (janapada): see KātŚr 22.11.28–36. The Arrow sacrifice is another variation, performed according to the rules of the Hawk sacrifice, except that it is done in a single day: KātŚr 22.5.30; ṢaḍBr 3.9.

5. A bedrock hermeneutical principle of Vedic exegesis is that one sentence or injunction can enjoin only one thing, not two. When a sentence does the latter, it has the flaw of vākyabheda, sentence split. Thus, the opponent argues, this aphorism cannot point out both these aspects of dharma.

6. The Sanskrit term for aphorism, sūtra, is derived from a verb meaning to sew or string. Thus, it is presented as stringing together many ideas. The argument is that the hermeneutical principle of split sentence does not pertain to the very peculiar linguistic field of aphoristic composition.

7. A similar rule is found at ĀśGṛ 2.4.1. The Eighth-Day rite (aṣṭakā) is an offering made on the eighth day after the full moon, generally for the benefit of deceased ancestors. The rite of aṣṭakā becomes the standard example of a ritual not given in the Vedas but still recognized as a Vedic rite, which is thus learned only from texts of recollection.

8. A prescription similar to this is found in VaDh 7.12. For an extensive discussion of this rule and its Vedic precedents, see Visvarupa, ch. 7.2: p. 112.

9. I have not found this rule verbatim anywhere, but this duty is given in MDh 8.264.

10. I have not found a source for this injunction.

11. A topknot is a tuft of hair, normally at the crown of the head, left uncut while the rest of the head is shaved. This tuft is normally left to grow long and is tied in a knot. I have found no verbatim source for this citation, but see VarGṛ 4.18. An interesting point in the order of these citations is that it follows the Sanskrit alphabet.

12. In the edition, the preceding section is not given as an introduction to the PMS aphorism (sūtra). But the last sentence (tad ucyate) indicates that the section was, indeed, such an introduction. The edition simply repeats the aphorism as part of the commentary.

13. The reference here is to the examples given above that are found only in texts of recollection.

14. For the distinction between what is found in the Veda and what is found in the world, see Patanjali’s discussion in ch. 1.1: 1.

15. The sense is that people belonging to the three upper classes (varṇa) claim that their recollection, i.e., the smṛti texts, are based on the Veda. And this claim is valid, because the same people perform the Vedic sacrifices prescribed in extant Vedic texts.

16. That is, the Vedic text that is presumed to be the basis of the text of recollection.

17. The argument here is that a text once known may have been forgotten, and hence it is now subject to inference: there must have been such a text that gave rise to this recollection.

18. This mantra, cited here with just the first words (pratīka), is given in the PārGṛ 3.2.2 and the Apastamba Mantrapāṭha 2.20.27: yāṃ janāḥ pratinandanti rātrīṃ dhenum ivāyatīm |saṃvatsarasya yā patnī sā no astu sumaṅgalī svāhā || A variant of this is found in AV 3.10.2. This and the following statements are connected to the examples given in the opponent’s statement in the introduction to PMS 1.3.1.

19. ṚV 10.4.1 cd: dhánvann iva prapā́ asi tvám agna iyakṣáve pūráve pratna rājan || Translation of Brereton and Jamison 2014.

20. This and the preceding passage indicate through inference the existence of injunctions regarding the construction of watering places and reservoirs.

21. The chant of particular melodies called Sāman is carried out during a Vedic sacrifice by a specialist priest called Udgātṛ. The sacrificial post is made out of an Udumbara fig tree (Ficus glomerata). The rule in texts of recollection goes something like this: audumbarī sarvā veṣṭitavyā, “The entire Udumbara post should be wrapped” (see MīKo III: 1344). This conflicts with a Vedic statement about touching the post, found only in the LāṭŚr 2.6.2.

22. The argument is that, as a direct perception cannot be annulled by a mere recollection or an inference, so a Vedic text that is directly perceivable (pratyakṣaśruti), here compared to direct perception, cannot be annulled by a text of recollection.

23. The rule regarding rice or barley is a classic example of an option arising from two equally authoritative texts that enjoin two different things. There are contradictory Vedicinjunctions: “He should offer a sacrifice with rice” and “He should offer a sacrifice with barley” (ŚB 11.3.1.3). The same is true with regard to the Sāman that should be recited, the Bṛhad or the Rathantara. For a detailed discussion, see Sabara on PMS 12.3.10–15; MīKo V: 2975; VII: 3803.

24. The meaning is that in the case of the prescription to touch the post, there is an explicit Vedic statement. In the case of the text of recollection prescribing the wrapping of the entire post, one has to depend on inference to establish the existence of a Vedic text that would provide its foundation. The latter, therefore, is weaker than the former.

25. This is the first interpretation, which takes this aphorism as part of the previous argument about texts of recollection that are in conflict with Vedic texts.

26. I have not been able to trace this rule. The Vaisarjana is an offering of ghee during a Soma sacrifice. See ĀpŚr 11.16.15.

27. This exact rule is not found in extant texts, but something very similar is found in KātŚr 14.5.35. See also KātŚr 14.1.20.

28. See the commentary on PMS 1.3.2, and n. 15 there.

29. The Holākā is referred to in KāṭhGṛ 73.1, and there the commentators say that it is done for the benefit of women. The deity is Rākā, and the rite is referred to in some regions by that name. I have not been able to find much information on the other two rites. The MīKo (2: 1022) says that the Āhnīnaibuka is the worship of trees such as Karañja (Pongamia pinnata or glabra; Pongam oil tree). The MīKo (2: 1134) says that the Udvṛṣabha is performed on the full-moon day of Jyeṣṭha (May–June) and consists of worshiping bulls and washing them. See Kane V: 1282–82.

30. The number of the tufts and the places on the head they were worn differed according to the lineage or Vedic tradition of each individual. Thus the BGṛ (2.4) states that one should keep one, three, or five tufts according to the dharma of his family. For a detailed discussion, see Kane II: 263f.

31. For a comparison of the notions of lost scripture in Buddhism and Vedic exegesis, see Kataoka 2013.

32. The two topics (adhikaraṇas) are contained in aphorisms 1.3.3 and 1.3.4: (i) authority of the Veda is greater than that of texts of recollection, and (ii) the unauthoritativeness of some texts of recollection is based on their having worldly motives.

33. “Different foundation” refers to such motives as greed. Here Kumarila goes against Sabara, who, as we saw in the preceding section, gives precisely this interpretation in the commentary on PMS 1.3.4 (ch. 6.1: #2).

34. I want to thank Kei Kataoka for helping with the interpretation of the expression viśiṣṭadhvanisthānīyena tenaiva. For the meaning of sthānīya in the sense of “equal to,” see Panini 5.4.10.

35. The intent here is to show that even though some aspects of a Vedic text, such as the various accents and articulation of sounds through different parts of the mouth, are humanly produced, yet they do not impinge on the authoritative nature of the Veda itself as authorless. In the same manner, when a human teacher or a humanly produced text of recollection points to a Vedic passage, that by itself does not make the Vedic passage humanly produced.

36. This appears to indicate that the previous verse presents the view of an opponent acknowledging the possibility that a particular text of recollection may not be based on the Veda. Calling him derogatively, “my dear fellow,” Kumarila rejects that view.

37. These are two branches (śākhā) of the black Yajur Veda.

38. The situation is this. A particular text of recollection is invalided by a contrary Vedic text. But thereafter one may well discover another Vedic text that would lend support to the same text of recollection, which would then be both validated and invalidated. Kumarila’s view appears to be that every text of recollection must be a priori assumed to have such Vedic validation, thus giving rise to an option.

39. On this example, see Sabara’s comments on PMS 1.3.4 (ch. 6.1: #2).

40. The identity of this text is uncertain; this appears to be the only reference to the work. Somesvara Bhatta, in his Nyāyasudhā, commenting on this passage, calls it by a slightly different title, “Chāndyogyānupapada” (the MīKo 3: 1733 corrects this to Chāndogyanupada), and identifies it as a “sūtragrantha,” a text of the aphoristic genre. I thank Larry McCrea for this information.

41. The cloth for wrapping the Udumbara fig post is referred to as Vaiṣṭuta, and the process of wrapping is given in LāṭŚr 2.2.6.1–14. Vaiṣṭuta (from vi + stuti) is a particular arrangement of verses in the Sāman chants. The word Śrī here probably refers to prosperity or to the goddess of prosperity. For the Śāṭyāyani Brāhmaṇa, see Gonda 1975: 349.

42. These “ears” are the two prongs at the top of the sacrificial post. At BhārŚr 12.10.4, for example, a piece of gold is placed between the two prongs of the post and an oblation of ghee is offered there, which would flow down to the bottom.

43. This explanation presupposes that the Sanskrit term used for wrapping is pariveṣṭana, the prefix pari denoting the full wrapping of the post.

44. The GDh (2.45–47) states: “To study a single Veda, he should live as a student for twelve years, and to study all the Vedas, twelve years each or until he has grasped them.” Thus, it is given not as a rule but as an option.

45. That is, this alternative is made for a person who may want to skip the order of a householder and go directly to another celibate order of life, such as renunciation.

46. Here Kumarila returns to two possible interpretations of the PMS 1.3.3, which states that a text of recollection that conflicts with a Vedic text should be disregarded. The first alternative is to say that it is preferable to follow the Vedic text when the two texts provide two different options. This is stated more explicitly in the very next verse. The second is that the entire aphorism pertains to texts of groups such as the Buddhists, who are outside the Vedic pale. This last point is taken up in the next section of Kumarila’s commentary.

47. The implication is that, if following the first Vedic text and not following the other is not an offence, then following the Vedic text and ignoring the text of recollection should also not create a problem.

48. This interpretation is based on two ways of dissolving the compound anapekṣa, either as a Karmadhāraya (descriptive determinative) compound: “nondependent,” or as a Bahuvrīhi (exocentric) compound: “in which there is no external dependence.” In either case, the meaning would be the same. The term pramāṇam (authoritative) is taken over from the previous aphorism, which is a common practice in this genre of literature. Thus, instead of a contradictory text of recollection being discarded, this interpretation makes it authoritative.

49. Here I follow Harikai’s (2008) emendation. Printed editions read śākyagrantha, which he emends following five manuscripts to read śākyanirgrantha.

50. The word uccāṭana is a technical term in the Tantric vocabulary. It is a magical act carried out by a Tantric practitioner that drives away evil forces and enemies. See the Tantrābhidānakośa I: 226.

51. The argument is that these texts of non-Vedic and anti-Vedic groups are popular, and therefore it is imperative for defenders of the Vedic worldview to undermine their credibility. The example given here is taken from common speech, where the term gavī is often used in place of the correct Sanskrit term gauḥ. The grammarian Patanjali (I: 8) also uses this example to show that the use of such common words does communicate the intention of the speaker, but it does not give him merit (dharma), because only the use of proper Sanskrit can accomplish that: see above, ch. 2.2: 3. This problem in the use of common terms, therefore, needs to be pointed out by grammarians.

52. The printed editions read śrutivihitaiḥ. I follow Harikai’s edition (2008) based on manuscript evidence: śrutismṛtivihitaiḥ.

53. This is one of the rare occasions when a Brahmanical author refers to the historical reality of other countries and their acceptance of non-Vedic doctrines, in this case, probably the Buddhist. See the comments of Halbfass (1983: 15–16) on this passage. This is also an interesting insight into how Buddhist theologians may have used their newfound popularity in distant lands, such as China and Tibet, as a tool for propaganda and for propagating their faith.

54. I follow here the reading of Harikai (2008), anubhavat tulyakakṣayā; the printed editions read: anubhavatulyakakṣayā.

55. The issue here is the relationship between word (śabda) and meaning. For many opponents of Vedic exegesis, including the Buddhists and Jains, this relationship is contingent and conventional. For the proponents of Vedic exegesis, however, the relationship is eternal and transcendent, especially as it relates to Sanskrit. This doctrine is the basis for their claim of the inerrancy of the Veda.

56. This statement is cryptic and not altogether clear. It may be that Buddhists and others cannot bear to acknowledge that their sacred texts are based on the same authoritative source, namely the Veda, as Brahmanical texts of recollections (in Kumarila’s vocabulary both kinds are called smṛti). Jha appears to interpret the phrase to mean that Buddhists cannot bear to acknowledge the Veda as the foundation because they think their own texts are equal in authority to the Vedas.

57. The fourteen repositories are: the four Vedas, Puranas, logic, hermeneutics, treatises on dharma, and the six Vedic supplements. See YDh 1.3, cited in ch. 5.1: #1.

58. This is the Shudra; and here Buddhists are considered Shudras by definition, confirming the suspicion I expressed (2004c: 40) that Manu and others use Shudra as an indirect way to refer to the Buddhists.

59. The expression alaṃkārabuddhau has caused problems of interpretation. Jha, in his translation, takes it to be the title of a Buddhist text. No such text exists, however. I want to thank George Cardona and Ashok Aklujkar for help in interpreting this difficult sentence. The meaning is that for the Buddha, what others may consider a transgression of his dharmais viewed as an adornment to be proud of. For a similar expression and citations of this passage, see Rajasekhara, Kāvyamīmāṃsā (C. D. Dalal; Gaekwad’s Oriental Series, 1), ch. 8, p. 38.

60. Gautama is the author of the foundational text on logic, the Nyāya-sūtras, and the reputed founder of the system of logic. These are the so-called orthodox systems of philosophy that acknowledge the supremacy of the Veda.

61. This is a paraphrase of MDh 4.30.

1. Asahaya’s commentary is of little use for issues concerning the epistemology of dharma, because his root text, that of Narada, deals only with legal procedure.

2. McCrea (2010: 123) states: “More notably, though, most writers on Dharmaśāstra seem to have had little or no interest in the epistemic justification of the authority of Dharmaśāstrain general, or of the specific texts they were concerned with. They seem prepared to take this authority for granted, and devote themselves instead to the specifics of interpretation: trying to determine exactly what actions the dharmaśāstras prescribe, rather than explain just why it is that we should believe them.” This is certainly an accurate description of how the epistemic issues are dealt with by jurists and commentators of the medieval period beginning around the twelfth century (see the next chapter). As we will see, the early commentators, represented here by Visvarupa and Medhatithi, are deeply engaged with these issues.

3. For a detailed discussion of the interaction between Buddhism and Brahmanism, see Bronkhorst 2011.

4. For the best and most detailed discussion of Bharuci, his works, and his date, see Derrett 1975, I: 4–27.

6. An injunction of this kind is not found anywhere in the Vedic texts. Professor Wezler (personal communication) has noted that this may be a case of an “invented” Vedic text; it is cited most frequently in discussions of noninjury (ahiṃsā).

7. All editions read: te śatruvadhalakṣaṇāṃ prītim anuṣṭhīyamānāṃ nirvartayanti. I think the original reading was anuṣṭḥīyamānā nirvartayanti, and the scribes/editors, taking the first word to be a feminine nominative, changed it to the accusative to agree with the preceding two feminine accusatives. However, before the operation of external sandhi, the original was anuṣṭhīyamānāḥ (nominative plural agreeing with the subject te). The verb anutiṣṭhati is used with regard not to internal feelings but to external actions.

8. This view is in direct contradiction to that of Sabara (ch. 6.1: #1).

9. This kind of inference (anumāna) is called sāmānyato dṛṣṭa. For a discussion of it in the commentary of Medhatithi, see Wezler 1999.

10. In my critical edition, the reading here is vicālayet (“he should not transgress”), but the reading in the manuscript accompanying Bharuci’s commentary is vicārayet.

11. The Sanskrit term I have translated as “general norm” is vyavasthā, which often means a legal decision. But the term appears to mean in these early commentators (also Visvarupa and Medhatithi) a general norm or legal principle, as well as specific rules enacted by groups or prevailing in various regions: see Medh on MDh 7.13. The exact meaning, however, in these contexts is not altogether certain (see Introduction, p. 45). See Bharuci’s use of the same term within the context of legal procedure in ch. 12.1: #1.

12. Bharuci here uses the word laukikadharma, clearly distinguishing this kind of dharma from the Vedic dharma (vaidikadharma). We have seen this distinction already in Patanjali.

13. The manuscript I used for this translation is Ms. T 555 of the University of Kerala Oriental Manuscripts Library, Trivandrum. Unfortunately, it is incomplete, breaking off in the middle of the section on the Vedic student.

14. Three colleagues—George Cardona, Albrecht Wezler, and Ashok Aklujkar—have helped me in various ways in understanding the prose of Visvarupa. But even they sometimes confessed to not fully comprehending his lines of thought. I thank them for their help.

15. The reference by the opponent is to the definition or hallmark (lakṣaṇa) of dharma given in the preceding verse: “When an article is given by individuals imbued with the spirit of generosity, at a proper place and time, to a worthy recipient, and following the proper procedure—that is the entire definition of the Law” (YDh 1.6).

16. The meaning is that the category of Brahman is broader in scope than renouncer (parivrājaka) and includes the latter; not all Brahmans are renouncers, but all renouncers are Brahmans, at least according to Brahmanical theology. The enumeration of the latter separately in a statement such as: “Brahmans should be fed, and also wandering ascetics,” is meant to emphasize the special position of the latter within the broader category. See Jacob 1907:37; see Samkara on VeS 1.4.16. So here the category of text of recollection (smṛti) would implicitly include practice (ācāra), but it is singled out as a special category.

17. The meaning here is not altogether clear. Perhaps what Visvarupa intends to say is that when a sundry text is cited as “smṛtyantara”—a very common practice in medieval texts—it is done by memory and not directly from a written text. When there is an actual text, then the name of the author is mentioned. So here we have a hint as to how the smṛtyantara category was used: when one remembered a floating verse and could not pin it down. Here, as in other places, artha is probably an abbreviated compound standing for padārtha, which, in Vedic exegesis, had acquired the meaning of an enjoined act. The subcommentary on Visvarupa, the Vacanamālā, confirms this meaning. It says that a smṛtyantara is not made into a composed text (upanibandhana) either because there are an innumerable number of them or because they contain optional material.

18. There is an explicit statement regarding this kind of renunciation in the context of a calamity befalling one’s family in the opening verses of Vidyaranya’s Jīvanmuktiviveka (Ānandāśrama Sanskrit Series, 20; Pune, 1901, p. 1): putradāragṛhādīnāṃ nāśe tātkālikī matiḥ |dhik saṃsāra itīdṛk syād virakter mandatā hi sā || “Fie upon this cycle of existence—should this be the thought lasting but for that time, a thought springing up at the loss of one’s wife, children, house, and the like, it is considered the faint level of detachment.”

19. The subcommentary Vacanamālā refers to Visvarupa’s introductory comments on YDh 3.45.

20. This sentence is corrupt: tanmūlatvaniyamoparasmṛteḥ. DhKo, V: 136, puts a question mark after it. My translation is tentative. The subcommentary does not comment on this phrase.

21. The meaning of the cryptic statement (samānam anyat) is unclear. Perhaps it simply says that the rest of the explanation of “desire arising out of right intention” is the same as the one given previously. See the similar expression iti samānam (“the rest is the same as above”) in the AŚ 5.1.13, 52; 7.6.32; etc.

22. Here we have a curious and, as far as I know, unique explanation of the different sources listed by Yajnavalkya. The well-known four aims of human existence (puruṣārtha), also called the quadruple set (caturvarga), are pleasure (kāma), success/wealth (artha), dharma itself, and ultimate liberation (mokṣa). The source of dharma is, in all likelihood, the first three: Veda, texts of recollection, and the practice of good people.

23. The subcommentary Vacanamālā also makes clear that the long passage from here until “To this we respond” (Author’s Response) presents various viewpoints of opponents in order to refute them seriatim. Some confusion is created by this method, because some opposing views are actually refuted not by Visvarupa but by another opponent of the first opponent. The subcommentary states: atraiva matantaraṃ darśayāmaḥ | iha smṛtimūlatvaṃ prati vivadante vādinaḥ | tatra svapakṣasthāpanārthaṃ pakṣāntarāṇi vikalpya nirākāroti kutaḥ punar ityādi dauhitrasmaraṇavad ityantena ||

24. The meaning is that having a few statements with a Vedic foundation makes texts of recollection not different from Buddhist and other false scriptures, where one can find some statements that agree with Vedic texts. This point is also made by Kumarila (ch. 6.2).

25. It is not altogether clear what “its opposite” (viparyaye) refers to—perhaps that the texts of recollection are not founded on the Veda, which is the understanding of the subcommentary Vacanamālā, which glosses: vedmūlatvābhāve. One may thus infer that these texts are not founded on the Veda, because the Veda cannot be understood as forming the foundation of everything. Indeed, the rest of the opponent’s argument is precisely aimed at undermining the claim that the Veda is the foundation of texts of recollection.

26. The reference is to the kalpasūtras, consisting of the aphoristic texts on solemn Vedic rites, on domestic rites, and on dharma. See ch. 1.1, p. 52.

27. The edition reads jyotiṣo ’pi hi, but I think this was an attempt by the editor to “correct” the difficult original reading, jyotiṣṭo ’pi hi, which is given clearly by the subcommentary Vacanamālā, with the explanation: jyotiṣṭomasyāpīty arthaḥ.

28. The following 41 verses composed by Visvarupa are terse and aphoristic, and cannot be translated without copious notes. I have omitted them. They are aimed at explaining and undermining Kumarila’s argument that texts of recollection are based on scattered Vedic texts.

29. As the subcommentary Vacanamālā makes clear, this first view rejected by Visvarupa is that of Kumarila Bhatta. The same view will also be rejected by Medhatithi.

30. In Vedic exegesis, only injunctions with a verbal form expressing, directly or indirectly, an obligation to perform specific acts are the meaningful elements within the Veda; they are independent. All other elements of Vedic texts, especially the mantras and explanatory statements (arthavāda), become meaningful only in relation to something other than themselves, that is, to the injunctions. See ch. 1, n. 7.

31. The technical term liṅga (indicative sign) refers to the second of the six modes (pramāṇa)that accompany an applicatory injunction (viniyogavidhi), which indicates the connection of a subsidiary element with the main action. See MNP 66f., 90f.

32. The meaning is that the mantras referred to in the injunction on the daily Vedic recitation have as their object the murmured recitation (japa). Therefore, they do not require any other rite, and are thus neutral or indifferent with respect to other rites.

33. ṚV 10.4.1 reads: “I begin the sacrifice to you and I propel my thought to you, so that you will become the one to be extolled at our invocations. You are like the first drink in a wasteland, o fire, for Pūru who seeks to attain you, you age-old king” (tr. Brereton and Jamison2014). The translators comment: “The usual high points of the birth and growth of Agni are covered in this hymn. What gives the hymn its character are the concentrated and somewhat unusual images that are studded through it. This is already evident in vs. 1, where Agni (who is fire, after all) is compared to ‘the first drink in a wasteland’; presumably the point of comparison is that both are eagerly desired and it is not certain they will appear.” The Author’s claim is that the content of the explanatory statement embedded within this mantra demands the existence of an injunction compelling people to construct water booths. For this and the following examples, see Sabara on PMS 1.3.2.

34. The TS passage reads in full: “Now confusion occurs in that they perform the same thing with a better and worse (instrument), for the ass is worse than the horse; they lead the horse in front to avoid confusion; therefore the worse follows after the better.” This passage deals with the ritual practice of placing the horse in front of the donkey as they are being led. This explanatory statement demands the existence of an injunction to validate the practice.

35. The term vyavahāra (translated as “expression”) is glossed by the subcommentary Vacanamālā as śabda (word or statement). I think this interpretation is correct, because the reference is to explanatory statements (arthavāda) that may form the basis for the inference of a corresponding injunction.

36. The edition reads viparyāya (inversion), but the subcommentary gives the reading viparyasa (error, delusion, misapprehension), which I have followed. It parallels the other term bhrānti (mistake) used in this passage.

37. This expression (vijñānaghana) is probably a reference to the Vijñānavāda school of Buddhism. This expression is also used by Samkara (on VeS 2.2.28) in his refutation of Buddhist doctrine, even though it does not appear to have been used by the Buddhists themselves. See also Bṛhadāraṇyaka Upaniṣad 4.5.13, which has prajñānaghana. The subcommentary Vacanamālā, however, takes this simply as a reference to this Upaniṣadic passage.

38. This is a very condensed argument. According to the subcommentary Vacanamālā, the statements of those with unimpaired eyesight can be a basis for arriving at reality, and so are the Vedic statements. That is why there is no flaw in the argumentation.

39. The argument is that such statements serve only a ritual purpose (kratvartha) and are not intended for the person performing the rite (puruṣārtha). The subcommentary Vacanamālā supports this interpretation with the maxim: phalasannidhāv aphalaṃ tadaṅgam, “when something is given in close proximity to an act that is fruitful, then it is a subsidiary element that does not have its own fruit” (see Sabara on PMS 4.4.19). Here, an explanatory statement cannot form the basis of a rule in a text of recollection, because that statement is subsidiary to the original Vedic injunction and not fruitful on its own.

40. A name, especially that of a deity or a sacrifice, does not have any injunctive power and thus cannot provide a foundation for dharma. The utility of names and their connection to the results of a ritual act are dealt with extensively in the literature of Vedic exegesis: see MNP 249f.

41. This same argument, as we have seen, is used by Kumarila, even though he would not support this opponent’s view that texts of recollection lack authority.

42. These are two branches (śākhā) of the black Yajur Veda.

43. A Vedic school called caraṇa is generally taken to be a synonym of Vedic branch or śākhā (see MīKo III: 1679; Panini 4.2.46). Visvarupa, however, appears to distinguish the two, using the former for the tradition of transmission and the latter for the texts so transmitted.

44. For the limbs (aṅga) or supplements of the Veda, see ch. 2, n. 4. The argument here is that mere retention (dhāraṇa) is not the same as actual recitation (pāṭha), where the exact words of the Veda are recited verbatim. Mere retention is the same as recollection, and thus these texts would fall into the latter category.

45. The vows (vrata) and recitation (adhyayana), as shown in the several texts cited by the subcommentary Vacanamālā, apply only to the Vedas and not to other texts such as the Vedic supplements and texts of recollection. This is taken as the reason these texts cannot be eternal, because they do not have a tradition of recitation in which their exact wording would be preserved.

46. The edition reads: smṛtayo nityā iti. My reading of the tenor of the arguments makes me inclined to see an avagraha. The reading would then be: smṛtayo ’nityā iti, in the negative.

47. The list headed by Manu referred to here is that of YDh 1.4–5. This list leaves out several authors of treatises on dharma; one such list is headed by Narada, and includes Pulaha, Gargya, Pulastya, Saunaka, Kratu, Baudhayana, Jatukarna, Visvamitra, and Pitamaha (cited in the subcommentary Vacanamālā).

48. The reference is to the views of the adherents of the schools of logic and Yoga, who claim that the Vedas are authoritative because they were authored by the Lord or Brahman.

49. The reference is to the treatise of Manu.

50. The argument is not altogether clear, and the subcommentary does not offer much help. If this rule referred to the study of the entire text of Manu rather than to individual injunctions, the argument appears to say, then the reference would have to be to the original composition (this appears to be the meaning of the term prakṛta used here) that Manu received from his father, the Creator. The verse immediately preceding this verse stated: “Manu, the wise son of the Self-Existent, composed this treatise” (MDh 1.102). Already in the introduction to the NSm (in Jolly’s tr.), we have the belief that Manu’s original treatise consisted of 100,000 verses arranged in 1,080 chapters; and this work was subsequently abridged by Narada and others. Its enormous extent makes it impossible for any being, let alone a human being, to be capable of studying it in its entirety.

51. The subcommentary Vacanamālā explains that texts of recollection clarifies and explicates the meaning of the Veda.

52. The subcommentary Vacanamālā explains that here the opponent, who also subscribes to the eternality of texts of recollection, rebuts the arguments made thus far by saying that what has been said does not apply to him. The difficult compound parāvṛtticaturaprajñāsahāyāḥ is explained thus: parāvṛttau paradūṣaṇabhaṅge caturā paṭīyasī prajñā buddhir yeṣāṃ sahāyo bandhuḥ. This may also a reference to Buddhist philosophers and logicians. See the term parāvṛtti (parāvṛtta) in Edgerton’s Buddhist Hybrid Sanskrit Dictionary (New Haven: Yale University Press, 1953), 320.

53. The text in the edition reads: sannikṛṣṭaphalāḥ kṛṣyādayaḥ | aniyataphalāś cāṣṭakādayaḥ ||The contrast between the fruits of the two in this reading makes little sense. I have followed the reasonable emendation proposed by the editors of the DhKo (V: 139): saṃnikṛṣṭaphalāḥ kṛṣyādayaḥ aniyataphalāś ca, asaṃnikṛṣṭaphalā niyataphalāś cāṣṭakādayaḥ. The reading of the edition was probably created by haplography. This passage is not commented on by the sub-commentary Vacanamālā.

54. The argument here is that the parallel between agriculture and acts such as the Eighth-Day rite given in texts of recollection does not imply that both are simply based on worldly practice. The author maintains that the latter both is well established in the tradition and can only be gathered from these texts.

55. This is a very concise and cryptic sentence. A lot is implied, and my translation is not literal but attempts to state what may be the meaning. I am indebted here to the explanation provided by Professor George Cardona. The subcommentary Vacanamālā does not comment on this passage.

56. The meaning appears to be that the authors of the texts of recollection gathered together those enjoined acts that were not found in the extant Vedic texts and presented them in textual form (upanibandhana). The subcommentary Vacanamālā gives the missing first half of this verse: kaścic chiṣṭo ‘tikāruṇyāt sarvadharmāgamārthavit | agranthakānām arthānāṃ cakāropanibandhanam || “A certain cultured individual who knew the acts of all the scriptures relating to dharma, because of his immense compassion, presented in textual form acts that were not contained in written works.” I take artha as an abbreviation of padārtha, which, as already noted, has the meaning of an enjoined act or rite in the tradition of Vedic exegesis.

57. The citation is somewhat elliptic but is derived from the Aitareya Brāhmaṇa 5.32 and 5.34. See also Kauśītaki Brāhmaṇa 6.12; Śāṅkhāyana Śrautasūtra 3.21.1–6. The passage is cited by Sabara on PMS 12.3.16. The reading of the edition, yady ukto, should be corrected to yady ṛkto. This passage deals with expiations (prāyaścitta) to be performed for ritual flaws, using bhūr for flaws in Ṛgvedic recitations, bhuvas for flaws in Yajurvedic recitations, and svar for flaws in Sāmavedic chants. The text ends with flaws where the precise Veda is unknown. The opponent’s point here is that the “unknown” (avijñāta) is equated to expiations contained in texts of recollection, and these are actually not derivative of Vedic injunctions. The text of the Aitareya in Keith’s translation reads: “To the gods said Prajāpati ‘If there is trouble in your sacrifice from the Ṛc, do ye offer on the Gārhapatya, with ‘bhūḥ’; if from the Yajus, with ‘bhuvaḥ’ on the Agnīdh’s altar, or on the Anvāhāryapacana at oblation sacrifices; if from the Sāman, with ‘svar’ on the Āhavanīya; if (the trouble) is unknown or a complete failure, running through all ‘bhūh, bhuvaḥ, svar’, do ye offer on the Āhavanīya only.’” For further treatment of this text, see Benson 2010: 317.

58. The meaning is that this way a text unknown to a particular individual would be granted the status of having a Vedic foundation, even if it does not.

59. The argument seems to be that if texts of recollection (smṛti s) containing these unknown expiations are themselves based on the Veda, then these expiations would not be unknown. The claim that they are unknown can only be sustained if the texts of recollection are independent of the Veda.

60. The reference and the argument are not altogether clear. The editors of the DhKo (5: 139) give the following explanation. When the king is vigilant, the transgressions of the bounds of good conduct are not tolerated. So, when we are vigilant your transgressions of the bounds of proper reasoning will not be tolerated! Likewise, as women married to men who are virtuous and capable of protecting them are not independent, so texts of recollection embraced (lit., taken in marriage) by the Vedas—which are authoritative on their own, thus imparting authority to them through their subservience—do not become independent.

61. The reference is to actions such as the Eighth-Day rite that are viewed as Vedic even though they are found only in texts of recollection.

62. The subcommentary Vacanamālā identifies this kind of worldly practice as that of the Buddhists and the like. The invariable concomitance is lacking here because the Buddhists do not base the authority of their scriptures on inference but on the immediate perception of the truth through yogic insight (yogipratyakṣa).

63. Here, Visvarupa makes a distinction between vidhyartha (an act enjoyed by a Vedic injunction) and kriyā (just any act that a person may perform).

64. The meaning appears to be that acts such as the Eighth-Day rite are supposed to produce results that come about long after the rite has been completed, and thus are not connected to the immediate outcome. This is different from normal worldly activities, where the result comes soon after the completion of the activity, for example, the harvest after the agricultural activity. Only the Vedic injunction can point to such a long-distance connection between an act and its result; but those who support the view of the independent authority of the texts of recollection, such as the Buddhists, do not accept the authority of the Vedas.

65. This refers to the thesis that texts of recollection are not independently authoritative. The opponent of Visvarupa has been presenting arguments in favor of such independence.

66. The subcommentary Vacanamālā gives here the reference given earlier: “Accordingly, he states: ‘He presented in textual form acts that were not contained in written works.’ Sacred scripture also, after introducing: ‘If from the Ṛc,’ and after saying: ‘from the Yajus; from the Sāman,’ states: ‘If unknown.’ And tradition states: ‘And what is unknown is what is in texts of recollection.’”

67. I have emended the reading of the edition, smṛtyarthe ‘bādhādarśanāt, to smṛtyarthe bādhadarśanāt, which is the clear understanding of the subcommentary Vacanamālā.

68. The edition here reads asāṃdṛṣṭikam (“without a visible motive”). I do not think this makes sense, because not having a visible motive is characteristic of Vedic injunctions. I have thus emended the text to read sāṃdṛṣṭikam.

69. The subcommentary Vacanamālā gives here the same reference as in n. 66.

70. The argument here is that a statement without an injunctive verb (e.g., the optative) may well be construed as an injunction if it deals with a subject unknown from elsewhere. In this context, the eating of the Soma is enjoined even though the sentence contains the verb in the indicative (bhakṣayati: TS 6.4.9.4).

71. For this example, see Sabara’s comments on PMS 1.3.1 (ch. 6.1: #2).

72. This passage is ascribed to Sankha-Likhita in DhKo V: 70. The entire text reads: āmnāyaprāmāṇyād ācāraḥ sarveṣām upadiśyate. By “practice” here the reference is to the “practice of good people” given as the third source of dharma in the root text of Yajnavalkya.

73. The reading in the edition is corrupt. The subcommentary Vacanamālā clearly identifies this as the text of Baudhayana; see ch. 2.2: #1.

74. I have been unable to identify this quotation. The subcommentary Vacanamālā also does not give a specific reference, but simply says that Vedic texts such as TS 2.2.10.2 and several texts of recollection he cites show that Manu was an expert in the Veda.

75. The subcommentary Vacanamālā gives several examples of the enumeration of authors of texts of recollection, including YDh 1.4.

76. This verse is probably Kumarila’s own composition presenting five alternative possibilities with respect to the composition of texts of recollection. It is given at the beginning of his commentary, Tantravārttika, on PMS 1.3.2. The verse is very concise and aphoristic. The meaning is that there are five assumptions with regard to the rules given in these texts: they may be (1) due to error, (2) based on personal experiences of the author (which have no authority), this being the position of Buddhists and other yogins, according to the Vacanamālā; (3) based on what others may have told the author; (4) the result of a willful desire to deceive on the part of the author; and (5) based on what the author had directly seen in Vedic texts. The last is the only reasonable assumption, according to Kumarila.

77. For the hermeneutical principle according to which a sentence cannot indicate two things, see Sabara, ch. 6.1: n. 5. Thus, the injunction on recitation makes one assume the existence of another injunction on teaching, because without teaching one would not be able to know, let alone to recite, the Veda.

78. Every injunction enjoining the performance of an act presupposes a person eligible to perform it. This is called by the technical term adhikāra, qualification or eligibility. So there is another type of injunction called adhikāravidhi dealing with who precisely is eligible to perform a particular enjoined act.

79. The subsidiary element (aṅga) here is the order in which the various acts in a rite are performed. This order is given in the six so-called pramāṇas, modes of evidence: direct statement (śruti), sense (artha), text (paṭhana), position (sthāna), chief matter (mukhya), and procedure (pravṛtti). For a detailed explanation of these, see MNP 199f.

80. This exposition is found in Visvarupa’s commentary on YDh 1.14, as pointed out by the subcommentary Vacanamālā.

81. This is a direct rebuttal of the opponent’s view expressed on p. 112: “Therefore, given that it would result in the denial of an enduring self, it would be difficult to sustain conventions based on Vedic injunctions.”

82. The meaning of upaplavamānaḥ here is unclear. I follow the subcommentary Vacanamālā, which takes it to mean upagacchan, having probably the meaning of conforming or corresponding; the instrument must correspond to the nature of the substance. You cannot take ghee with a knife.

83. This is a rather complex argument of the opponent, a complexity compounded by Visvarupa’s rather convoluted syntax. The gist of the argument is that, according to the proposal of Visvarupa, all rules found in texts of recollection would be derived through inference of one rule from another. The rather complex example of the Sruva spoon goes something like this. There is a rule to divide the ghee into four portions, which makes us infer an instrument such as a spoon. The injunction to divide the ghee using the Sruva spoon further calls for a rule to take the ghee up with this spoon, because ghee cannot be divided (cutup) like a cake, given that it is a liquid.

84. “Name” (samākhyā) is the sixth of the means of knowing a rule of ritual application (viniyogavdhi) (MNP 66f). Here, the opponent’s assumption is that the mantras associated with the daily Vedic recitation have names within them that refer to the Eighth-Day rite (aṣṭakā). This is denied by the author, because the ritual formulas are already associated with japayajña.

85. The Sanskrit words used in this passage, sāmārthya and śakti, are technical terms within the vocabulary of Vedic exegesis and refer to the capacity of words to signify, and are often used as synonyms of liṅga, the indicative force of words and one of the six means of knowing (pramāṇa) referred to in an earlier note. MNP 90, 100, 103.

86. See ch. 13, n. 107.

87. The Sanskrit term śreyas generally refers to anything that is truly beneficial to someone, and, as Rau (1957: 32–34) has shown, it generally refers in the Vedas to rich people as opposed to the poor (pāpīyas). It is frequently used in later literature with reference to the highest good, that is, final liberation from the rebirth process. It is often difficult to ascertain the exact meaning of this term within a specific context. Manu himself, however, provides a handy definition at MDh 2.224: “Some say that dharma and wealth (artha) are conducive to prosperity; others pleasure (kāma) and wealth; and still others, dharma alone or wealth alone. But the settled rule is this: the entire triple set [i.e., dharma, wealth, and pleasure] is conducive to prosperity (śreyas).” Medhatithi (on MDh 2.6; ch. 7.3: #4; p. 127), for example, claims that agriculture is also a cause of prosperity and gives a useful definition of the term.

88. This other work of Medhatithi is referred to elsewhere in his commentary as Smṛtiviveka (see Medh on MDh 2.6 under #4; p. 130). It was probably composed in verse. This work has not survived. See Kane I: 582.

89. For kalañja as red garlic, see Sabara on PMS 6.2.19; MNP, Edgerton’s note to 320. Both Sabara and ĀpDh 1.17.26 give kalañja along with garlic and onion.

90. See Medhatithi’s commentary on MDh 8.41 (below, ch. 12.2: #3), where he says that these dharmas were instituted by a distinguished ancestor within each lineage.

91. Here bāhya, “outside, external,” may be an abbreviated compound for vedabāhya found in MDh 12.95. But given the term bāhya is used in the very next paragraph with reference to Buddhists and other non-Vedic ascetics, it probably refers to these outside groups whose texts, although often bearing the name smṛti, are nevertheless without authority. So I take the compound bāhyasmṛti here and elsewhere in Medhatithi as a Tatpuruṣa compound.

92. It is the general view within the mainstream Brahmanical tradition represented by Vedic exegesis that the Veda has no author, whether human or divine; it is, therefore, self-existent. As such, the knowledge presented in the Veda is self-validating (svataḥ-prāmāṇya), just like perception. Even though perception is falsifiable through a later and more certain perception, this is not so in the case of the Veda, because there is no subsequent knowledge that can falsify or invalidate it.

93. For the meaning, see ch. 7, n. 91.

94. The reference is not altogether clear, but there is a class of Brahmans who are called Bhojakas. They were probably on the margins of orthodoxy, being connected in some way, possibly through intermarriage, with Iranian religious ideas and rituals. See H. von Stietencron, Indische Sonnenpriester: Sāmba und die Sākadvīpīya-Brāhmaṇa. Eine textkritische und religionsgeschichtliche Studie zum indischen Sonnenkult (Wiesbaden: Harrassowitz, 1966). Later in this same paragraph they are in a list along with Pancaratrikas and Pasupatas.

95. The reference is to the series of previous Buddhas according to Jha; but given the next sentence, the root or basis for the authority of Buddhist texts is the Buddha himself and his superhuman vision, and this basis has come down through an unbroken series of reciters of his words. The good and bad courses are the rebirth process. The editions of Jha and Gharpure read bhikṣūṇām in the genitive, and Jha translates: “good and bad conditions of Bhikṣus.” A variant in five manuscripts recorded by Gharpure reads: bhikṣāva. I think this is probably a wrong transcription of bhikṣave, written in the old Nāgari pṛṣṭhamātrā. Now, this is not standard Sanskrit, but the form is found in Pali and even more significantly, in Gandhari Buddhist manuscripts from northwestern India. Perhaps Medhatithi, living in Kashmir, had access to these kinds of Buddhist texts. We do find texts similar to the one he cites in several Buddhist texts: paśyāmy ahaṃ bhikṣavaḥ divyena cakṣuṣā viśuddhenātikrāntamānuṣyakena satvāṃ cyavantāṃ upapadyantāṃ suvarṇāṃ durvavarṇāṃ sugatāṃ durgatāṃ…(Mahāvastu, ed. Senart [Paris: Imprimerie Nationale, 1882–97], III: 448). See also Mahāvastu-Avadāna 3.448; Saṃghabhedavastu I.158.

96. The identity of anārthavāda (or ānārthavāda) is unclear. The explanation of the EDS (4: 2160, with the reading anarthavāda) is not helpful: “one of the heterodox sects (lit. having senseless deliberations).” The edition of DhKo (V: 90) gives the reading anāryavāda, but this is probably an emendation of the editors, and is also the lectio facilior. Perhaps, coming immediately after nirgrantha (Jain), the original reading may have been anekavāda, which is what Jain philosophy is sometimes called. James Hartzell has drawn my attention to a passage in the Buddhist text Kālacakratantra (edition of the Central Institute of Higher Tibetan Studies, Sarnath: 1986–1994, 87) where Buddhists are divided into four groups, one of which is the arthavādin. Another possibility, pointed out by Professor Wezler in a private communication, is that the reference is to the Ajivakas (Basham 1951). Their philosophy of niyati or fatalism/determinism would make all efforts such as rituals and asceticism ultimately meaningless (anartha). But the term remains obscure.

97. A sect or religious group called saṃsāramocaka is referred to also by Kumarila, Jayanta, and others. This group has been investigated by Wezler (1976: 329–48) and by Halbfass (1983:10–11; 1988: 329). They claim that killing is justified because it releases those killed from the cycle of births and deaths (saṃsāra).

98. It is unclear whether these are simply given in the ritual literature, or whether they are, in fact, Vedic rules. I have been unable to identify them. It could also be that these rules are given according to their meaning (arthataḥ) and not cited exactly as they may be found in a text.

99. There are conflicting rules regarding the holding of a vessel called Ṣoḍaśin, one saying that the sacrificer holds it and another saying that he does not. See TāṇB 16.1.8. Likewise, one rule says that the morning fire sacrifice should be offered before sunrise (TB 3.8.16.4), and another says that it should be offered after sunrise (TB 3.8.16; 5.1.33). The general hermeneutical rule in such cases is that the contradiction creates an option. The opponent’s argument, therefore, is that a similar option should be permitted between acts prescribed in available Vedic texts and acts prescribed by the scripture of these outside groups that contract the prescriptions of the Veda, because such acts may indeed be based on lost or unavailable Vedic texts. This argument sounds very much like that of Kumarila supporting his thesis of a scattered Veda.

100. For an exhaustive discussion of this source of dharma and the opinions of later commentators and scholars, see Davis 2007.

101. I have been unable to trace this citation.

102. The two terms used here, ūha and apoha, in general refer to the process of analytical reasoning. The former is the ability to deduce proper conclusions, while the latter is negative argumentation aimed at refuting false views. A verse given in several sources presents these as qualities of the mind: śuśrūṣā śravaṇaṃ caiva grahaṇaṃ dhāraṇā tathā | ūho ’poho ’rthavijñānaṃ tattvajñānaṃ ca dhīguṇāḥ || “Desire to learn, study, comprehension, remembering, reasoning, rejection, knowing the meaning, and knowing the truth are the eight qualities of the intellect.” Kāmandakīya Nītisāra 4.22. See also AŚ 1.5.5. These are probably the ones referred to as the “eight elements of intelligence” (buddhyā aṣṭāṅgayā) in a passage added after Rāmāyaṇa 6.101.22; see the translation by Goldman et al., 1415. The same verse is given in three manuscripts after MBh 3.2.16 or 17.

103. Here Medhatithi begins his actual gloss on the verse, commenting on each word. I have put the words taken from the verse for comment in bold type.

104. This is the broadest division of the Veda, mantra consisting of ritual formulas and brāhmaṇa consisting of everything else, but especially the injunctions.

105. Note that Jha’s first edition and Mandlik have the reading prāmāṇyahetuḥ, while Gharpure has prāmāṇyaparijñāne hetuḥ. The second edition of Jha has pramāṇaṃ parijñāne hetuḥ, which I have followed.

106. See ch. 7, n. 87.

107. The reference here is to an unseen and mysterious effect called apūrva that the ritual act creates in a person. This explains how the effect of the ritual act, for example, heaven, happens after a long time. Generally, this would go against the normal principle of causation. It is this apūrva that connects the cause and the effect over a period of time.

108. The Sanskrit expression is yāvajjīvam. We have the injunction yāvajjīvam agnihotraṃ juhoti (“He offers the daily fire sacrifice all his life”) in VarŚr 1.1.1.86; and yāvajjīvaṃ darśapūrṇamāsābhyāṃ yajeta (“He should offer the New- and Full-Moon sacrifices all his life”) in ĀpŚr 3.14.8, 13.

109. This principle, encapsulated in the maxim viśvajit-nyāya, is explained in PMS 4.3.10–16. It states that when no result is explicitly mentioned, one must assume that the result is heaven.

110. See ch. 1.2: #1.

111. Some glosses on Sanskrit terms are omitted here, because they make sense only within the context of the Sanskrit syntax and have little relevance to the English translation.

112. This example of a type of inference, known as pūrvavat-anumāna (inference of an effect from the perception of its cause), is given in Vatsyayana’s commentary on Nyāya-sūtra 2.1.38–39.

113. The author of the Vivaraṇa is Bharuci. See the headnote to the section on Bharuci (ch. 7.1).

114. For the Smṛtiviveka, see ch. 7, n. 88.

115. The reference here is to PMS 1.3.2; see Sabara’s commentary on this in Ch. 6.I.

116. Here Medhatithi appears to be saying that even though we know that there is a close connection between the rules contained in the texts of recollection and the Veda, it is not possible or wise to state exactly what this connection is. Later in this passage he will be more explicit: “Therefore, Manu and others are undoubtedly linked to the Veda with respect to this issue. It is, however, impossible to determine the specific nature of that link” (see ch. 12.2: p. 133).

117. In what follows, Medhatithi comments on the alternative explanations given above under Roman numerals.

118. ĀpDh 1.12.10: see ch. 1.1: #5.

119. Ritual injunctions are often divided into those that pertain to the ritual act itself (kratvartha) and those that pertain to the ritual actor (puruṣārtha). The former consist of rules of ritual procedure, such as the rule that the rice to be used should be husked by pounding. This serves a ritual purpose directly and benefits the ritual actor only indirectly. The latter, on the other hand, serve the purposes of the actor. An example would be the rule that a person who desires heaven should perform sacrifice. See Kane V: 1232–35.

120. The statement is from the TāṇB, 23.2.4: pratitiṣṭhanti ha vā ete ya etā upayanti | brahmavarcasvino ’nnadā bhavanti ya etā upayanti || “Those who have recourse to these sacrifices become well established. Those who have recourse to these become endowed with Vedic luster, they become eaters of food.” This is cited by Sabara on PMS 4.3.17 also to show that the present tense can carry injunctive force.

121. All three sentences of the above TāṇB passage have indicative verbs. There is a syntactic unity among them, because they follow each other sequentially.

122. The reference is to the theft of gold and drinking liquor in ChUp 5.10.9 mentioned earlier.

123. That is, the doctrine of the five fires.

124. This TB passage is cited in Sabara’s commentary on PMS 1.4.29. Here we have the injunction to put pebbles that have been wetted. We do not know the liquid with which they are wetted, but the explanatory statement immediately following this refers to ghee, from which we gather that the pebbles should be wetted with ghee. But the situation envisaged in the objection is dissimilar to this, because the explanatory statement does not immediately follow the injunction.

125. For the classification of injunctions (vidhi) into these four, utpatti (originative), adhikāra (eligibility or qualification), viniyoga (application), and prayoga (performance), see Kane V: 1228.

126. The special offering of ghee in the offering fire (āhavanīya) is called āghāra. Here the mantra contains the words “Indra” and “Vishnu,” which indicate that the offering is made to these deities. See TS 1.1.12; Keith’s tr. I: 14, and note 2 there, which gives the variant formulas in the different Vedic branches.

127. Having four feet indicates the completeness of dharma. Manu (1.81–82; see ch. 4.8)says that in the golden Krita Age dharma possessed all four feet, but in each subsequent age one of its feet was lost, so that in the current Kali Age only one foot remains.

128. The Sāmidhenī verses are ṚV 3.27.1; 6.16.10–12; 3.27.13–15; 1.12.1; 3.27.4; 5.28.5–6, which total eleven verses. The number fifteen is obtained by repeating four verses. See ĀśŚr 1.2.2–22. For the set of fifteen and seventeen and the hermeneutical issues involved, see Sabara’s commentary on PMS 10.5.27–33. The meaning appears to be that the original eleven verses are contained within the fifteen; it is unnecessary to look at the seventeen.

129. This fault consists of an infinite regress where each member in the line depends on the previous, which also needs validation. See the argument from a line of blind people given by Sabara: ch 6.1, #2.

130. The positive concomitance (anvaya) between X and Y is needed in order to deduce the presence of Y when X is perceived. Thus one deduces the presence of fire when one sees smoke. But if smoke has never been perceived by anyone at any time, then it is impossible to base a deduction on the basis of smoke. The opposite of this is the negative concomitance (vyatireka), where there is not even one case where smoke may be present without fire.

131. This and the following statement are very pithy and leave a lot to be understood. It is implied that there are some things we do infer without actual perception, for example, the motion of the sun. We really cannot perceive its motion, but we assume that it must move in order to be perceived in different places, in the same way as we know that motion is required for a human being to get from one place to another. The particular kind of inference is called sāmānyato dṛṣṭa, that is, through an inference based on a generally observed principle, such as that motion is required for an entity to be observed in one place at one time and at another place sometime later (see Wezler 1999). This is similar to but distinguished from circumstantial inference, dealt with in the next statement.

132. Circumstantial inference (arthāpatti) can be employed when a perceived reality cannot be accounted for in any other way than the one postulated. Thus, going from one place to another cannot be accounted for in any other way than by motion. The classic example is Devadatta, who is fat but does not eat during the day. So we must infer from the circumstances that he must eat at night. In the current case, however, the lack of other possibilities does not obtain, because it is possible for Manu and other writers to have simply erred or wanted to deceive others.

133. The issue is clearly dharma, which is the topic under discussion.

134. The new explanation still refers to the phrase in the original verse: “The recollection and conduct of those who know the Veda.”

135. See the Dhātupāṭha (ed. in Word Index to Panini-Sūtra-Pāṭha and Pariśiṣṭas, V. S. Pathak and V. S. Chtrao, Bhandarkar Oriental Research Institute, Poona: 1935), 523: śīla samādhau.

136. The Sanskrit here has a copulative (Dvandva) compound smṛtiśīle (re collection conduct), and Medhatithi’s labored explanation is focused on explaining how the compound can be understood in such a way as to make śīla qualify smṛti. The explanation is based on a particular kind of copulative compound called itaretara (see Patanjali on Panini 2.2.29; I: 434, 9–11). Here the two members of the compound are not independent but dependent on each other, the classical example being plakṣanyagrodhau (fig and banyan).

137. The Sanskrit reads smṛtiśīle ca (and), which is problematic for Medhatithi’s explanation. So he tries to move ca to a position after the last compound: tadvidām (of those who know it).

138. Perhaps “the science of logic” (nyāyaśāstra) here may refer to a treatise on logic, specifically to the Nyāya-sūtras of Gautama.

139. This enumeration is found in Visvarupa’s version of YDh 1.4–5.

140. These are significantly omitted in the list given by Yajnavalkya.

141. The identity of the Pṛśni hymn is unclear, but the ĀpŚr 21.13.8 identifies the ten formulas given at TS 3.3.5.1 as the ten mantras associated with Pṛśni (see MīKo 5: 2604). The reading of this sentence is corrupt. Mandlik and Jha read: pṛśnisūktaṃ tṛṇapāṇayo ’dhīyate aśvamedham aśvaṃ yathā samarpayantaḥ. Gharpure reads: pṛśnisūktaṃ tṛṇapāṇayo adhīyate āśvamedhamātraṃ ghāsaṃ samarpayantaḥ, without giving any variants but with a reference to Sabara’s commentary on PMS 2.4.2.8. Gharpure probably emended silently the received text following Sabara, who gives the practice among some: aśvamedham adhīyānāḥ kecid aśvasya ghāsam āharanti. I think the original reading of Medhatithi was something like the following: pṛśnisūktaṃ tṛṇapāṇayo ’dhīyate aśvamedham aśvasya ghāsam arpayantaḥ (or perhaps, aśvamedhāśvasya). It is likely that there is a confusion between the Aśvamedha section that is studied according to Sabara and the Pṛśni-sūkta given by Medhatithi. Perhaps both are the object of the verb adhīyate (“they recite”).

142. The same observation is made by Maskarin in his commentary on the GDh 1.2. The enormous number and variety of customary practices make it impractical to collect them all in one book.

143. This appears to be a salt mine located in Rajasthan. The origin of this maxim is unclear, but the meaning is that the salt in the mine overpowers and makes salty anything that may fall into it.

144. The example is derived from the Atirātra ritual where this prohibition is found. However, in other contexts, holding the vessel according to rule is proper, the prohibition being circumscribed to that ritual context. In the case of prohibitions that give satisfaction, the author argues that they are limited to contexts outside of that contentment. So the prohibition is inoperative when a learned person finds satisfaction in that act.

145. The example is poignant. People may say that whatever plant a mongoose chews is made into a destroyer of poison, but the truth is that a mongoose actually chews only those plants that already have the capacity to destroy poison. In like manner, whatever the cultured elite may find pleasing is dharma only because they find pleasure only in what is dharma.

146. This section is found in MDh 5.110–127. However, I do not find any statement in that discussion referring to the contentment of the self, although in some cases a person may clean a thing until he is satisfied that it is completely clean (MDh 5.126).

147. The Sanskrit term is para, which generally means “the highest.” This term, however, has another meaning, namely “other” or “different”; hence my translation, in this context, “farthest,” hoping to capture both meanings of the Sanskrit. The clear meaning of the term within the verse is highest, but Medhatithi is uncomfortable with the idea that the highest dharma could be obtained from an untouchable Candala. He is thus forced to resort to the second meaning of the term.

148. See Medhatithi’s commentary on MDh 1.2 given in section #2.

149. The sixfold strategy with the technical name ṣāḍguṇya is explained in Kautilya’s Arthaśāstra (7.1), and is also listed in MDh 7.160: peace pact, initiating hostilities, remaining stationary, marching into battle, seeking refuge, and double strategem.

150. The terms arthaśāstra and dharmaśāśtra in this context have been interpreted by commentators to refer not to Kautilya’s Arthaśāstra but to the sections of treatises on dharma that deal with matters of politics and law (e.g., chs. 7–9 of the MDh), and to other sections of those same texts that deal specifically with dharma.

151. The editors have the reading sainikaiḥ, “by soldiers,” and Jha gives this very translation. This makes little sense. I think the original was saunikaiḥ, “by butchers.”

1. This was already pointed out by McCrea (2010: 128).

2. See Kane I: 411 for references.

3. Olivelle 1995: 30.

4. Vedic exegesis accepts only injunctions as meaningful and as epistemic sources of dharma (see the long discussion on this in the section of Medhatithi in ch. 7.3). However, other sections of the Veda, especially the mantras and explanatory statements (arthavāda), are viewed as attached in some way to injunctions, and thus make syntactic wholes. Thus, they also can be an indirect source of dharma. See ch. 1, n. 7; ch. 7, n. 30.

5. See the long discussion on this in Medhatithi, ch. 7, pp. 126–27. It is unnecessary to mention this, because it is already known, but it is done so as to teach why texts of recollection also are authoritative.

6. Even Kane (1962–75) does not have a section devoted to him or even make any remarks about him or his works.

7. Here the singular is used with reference to a class of texts, and thus indicates all the Vedic texts, not just one.

8. That is, there is an assumption or inference that there must be a Vedic injunction serving as the foundation of the conduct.

9. The meaning of the expression ātmaguṇasaṃpad is unclear. The parallel term ātmasaṃpad is found in the AŚ 6.1.6 with a long list of qualites necessary in a good king.

10. This is part of the story of the god Dharma, who in the guise of a demon killed all four of Yudhisthira’s brothers. When Yudhisthira had answered the demon’s question, he got to choose which brother would live. He chose Nakula, even though he was not his uterine brother. MBh 3.297.65f. See Hiltebeitel 2011: 420–53.

11. Here Kulluka repeats the argument of Govindaraja and Medhatithi on MDh 2.6 (see ch. 7.2: #4, p. 124): “After reiterating the authoritativeness of the Veda that is established through reasoning, this verse communicates verbally the fact that the texts of recollection, such as that of Manu, have the Veda as their root.”

12. I give here the translation of Vijnavesvara’s reading of the verse, which differs somewhat from that of Visvarupa, that I have adopted in translating this verse in ch. 5.1: #2.

13. The reference is to the age when a boy is to undergo Vedic initiation.

14. See the edition and translation of this text, with a substantial introduction, by J.A.B. van Buitenen: Yāmuna’s Āgama Prāmāṇyam or Treatise on the Validity of Pañcarātra (Madras: Ramanuja Research Society, 1971).

15. Following DhKo V: 143, I emend nivārya of the edition to nirdhārya.

16. The Sanskrit samīcīna (being proper) is meant to explain the term samyak (right) in the compound samyaksaṃkalpaja (arising out of right intention).

17. See ch. 6.1: #3 (n. 29).

18. Verses 1.217–218 give the times for performing an ancestral offering (śrāddha), and the list ends with this statement, after which the author says: “these are declared to be the times for performing an ancestral offering.” Apararka limits the scope of this source of dharma to such instances.

19. That is, other than the sources given in this verse. The reference is to the dharma of outside groups such as the Buddhists, as well as Shaiva and Vaishnava sects.

20. Here begins a long digression concerning the authority of the sacred scriptures of various sectarian traditions, both Shaiva and Vaishnava.

21. The meaning appears to be that the inner kernel that contains the edible meat of the coconut is the Shaiva; the outer hard shell is the Kaula; and the dry fibrous husk parallels the Vedic rites. Thus far, I have been unable to identify these verses, although they must derive from a Shaiva Agama.

22. The lunar penance (cāndrāyaṇa) is a fast that follows the waning and the waxing of the moon. The MDh (11.217) describes it: “He should decrease his food by one rice ball a day during the dark fortnight and increase it likewise during the bright fortnight, bathing three times a day—tradition calls this the lunar penance.”

23. Found in Devī Purāṇa cited in DhKo V: 284, but not in Sharma’s edition of the text.

24. The maxim yāvad vacanaṃ vācanikam expresses a common interpretive strategy. A rule or a law cannot be made to state more or less that its words expressly say. That is, laws have to be understood literally. See Sabara on PMS 5.3.12; and Kane V: 1348.

25. Most of this description is found in Yoga Sūtra 1.24.

26. See n. 23 above.

27. The term siddhānta is a generic term referring to the Shaiva sacred treatises.

28. This is a syntactically convoluted verse. But the intent appears to be to present an “etymological” explanation of the term pāśupata that would undermine the authority of the texts bearing that name. The author uses the term paśu (a tame or domesticated farm animal) used commonly in the Pasupata vocabulary to refer to an individual soul, in contradistinction to god, Shiva, who is called pati (a term that can refer to the Lord as well as to a shepherd looking after the animal herd). Here, however, the word pata in pāśupata is derived not from pati, as is normally done, but from the fact that the paśu, by giving up his paśu nature, has become fallen (patita). It is this fallen nature of the paśu that is conveyed by the term pāśupata, thus robbing it of any authority. Such a fallen person parallels “those who have abandoned the path of the Veda” of the previous verse.

29. The Niḥśvāsa (perhaps more correctly Niśvāsa) Saṃhitā is one of the earliest Shaiva scriptural texts. It has been recently edited and translated by a team of scholars headed by Dominic Goodall (Forthcoming).

30. Vedic links refer to any type of ritual relationships, such as officiating at an initiation or a sacrifice, teaching, studying, and the like.

31. These seven traditions and their respective interpretations of the syllable OṂ are given in the BṛYogYSm in the verses following this citation: 2.69–103.

32. For an examination of this argument and of the interaction between Shaivism and Brahmanism in the medieval period, see Sanderson 2006b.

33. The reference is to Panini 4.3.101. Apararka gives here the manner in which the term śaiva is derived from the original śiva according to the metarules of Paninian grammar. Here there is the addition of aṇ, which creates the Vṛddhi strengthening of the initial vowel (Panini 4.1.83).

34. The technical term luk refers to an elision of a sound after some grammatical operations. Thus, one who studies a Shaiva treatise is also called “Shaiva.”

35. The Sanskrit term jāti has a double meaning. The first is birth or biological species; the second is the sociological division of human society into castes. Here the author uses this ambiguity to classify the entire biological/social universe into ten species/castes. I have not been able to identify this passage.

36. This is an implicit reference to the passage of Vyasa cited earlier: “Those who desire the purity of dharma do not want anything other than the Veda. It is the pure source of dharma; others are said to be mixed.”

37. The passage in the Bhaviṣya Purāṇa deals with the definition of a teacher called mahāguru, Great Teacher. The first line of the passage, which is omitted in the citation, reads: jayopajīvī yo vipraḥ sa mahāgurur ucyate—“The Brahman who makes his living from Jaya is called Great Teacher.” And it appears that the list of texts represents the category of texts called jaya, Victory. There is an alternate reading in the edition of this Purana: in place of saurāś ca mānavoktāḥ—“Sauras (relating to the sun) declared by Manu,” it has śrautā dharmāś ca nāradoktāḥ—“Vedic dharmas declared by Narada.”

38. This passage is ascribed by both Apararka and Madhava (PārM I.1: 133) to the Brahma Purāṇa, and by Devanna Bhatta in his Smṛticandrikā (I: 29) to the Ādi Purāṇa. See Brahma Purāṇa 213.164.

39. The meaning of the expression mūlakarma is quite unclear. The commentators of Manuare unanimous in taking it to mean some kind of witchcraft by which another person is brought under one’s power (vaśīkaraṇa): see MDh 9.290, where the meaning is clearly some form of witchcraft. In Vatsyayana’s Kāma Sūtra (4.1.9 and 6.2.56) also the reference appears to be to some form of magic potion made with roots to win the love of a woman. See also AV 4.28.6, where mūlakṛt (root cutter) is in apposition to kṛtyākṛt (witchcraft maker).

40. The meaning of “the mother and the Veda” is unclear. The compound mātṛveda may also be rendered “the mother Veda.” The sthāpaka is a particular official in the hierarchy of adepts. He is distinguished from the ācārya in the Mataṅgapārameśvara Āgama, Kriyāpada, ch. 10. Both undergo a special initiation rite. See, however, the different definitions of these two terms given in the Matsya Purāṇa cited below.

41. I have not been able to trace these verses in the published edition of the Matsya Purāṇa.

42. For the limbs or the supplements of the Vedas, see ch. 2, n. 4.

43. For the definition of this sacred region and its relationship to dharma, see MDh 2.23, cited above in ch. 4.4.

44. The two terms used here, ūha and apoha, refer to the two elements of rational analysis; see ch. 7, n. 102.

45. I am not sure whether the reading here is correct. If it is, the meaning is quite obscure. For the “repositories of knowledge” (vidyāsthāna), see YDh 1.3; ch. 5.1: #1.

46. The reference is to the above citation from the Devī Purāṇa.

47. The meaning of this simile is unclear. The eyes of a partridge (cakora) are believed to drink the nectar of moonbeams, and they turn red in the vicinity of poison (AŚ 1.20.8). Perhaps the eyes are viewed here as stealing the moonbeams, just as this person steals someone else’s dharma. The term ruci (paradharmaruciḥ) for pleasure is interesting, because it also means light and color. There may be a poetic interpretation here, where the man who performs another’s dharma is a thief in the same way as the eyes of a partridge that become red in the presence of moonlight. I thank Joel Brereton for this insight.

48. Regarding the Vedic supplements, see ch. 2, n. 4.

49. The reading of Apararka is arthair daśārdhaiḥ saṃyuktam. The meaning of artha here is uncertain: category, subject, etc. It is unclear what the fifteen are. The critical edition of the MBh has the reading abdair daśāhasaṃyuktam, whose meaning is equally opaque.

50. This is a variant of the text found in ch. 5 of this Upaniṣad. Note that the Sanskrit for soul is paśu (animal) and for snare is pāśa. See n. 28 above.

51. The kārukas are a Shaiva sect regarded as inferior to the Tantrik Shaivas. They are considered impure, and contact with them is avoided by other Shaivas. See Tāntrikābhidhānakośa II: 92.

52. Here Apararka returns to explaining the original verse of the YDh.

53. See ch. 2, n. 10. Apararka himself appears to take the term as referring not to the Vedic supplements but to other authoritative texts such as the epics and Puranas.

54. This half-verse is not found in the critical edition of the MDh.

55. The meaning here is that these sources of dharma permit us to infer a Vedic text as their basis, thus making that Vedic text perceptible.

56. See the section on Vasistha: ch. 2.3: #2.

57. That is, such a practice would not have either a worldly or a Vedic basis.

58. For this translation I have used the edition of Srinivasacharya (Mysore: 1914).

59. “Guru” refers to Brihaspati, who is the teacher of the gods. “Yogin” here probably refers to Yajnavalkya, who is often called Yogi-Yajnavalkya.

60. This verse is not found in the editions of the ViDh. For the Vedic supplements, see ch. 2, n. 4.

61. For the eighteen Puranas, see selection #1: Authoritative Texts.

62. I have here emended the text from dharmajñānānām, which is pleonastic because the knowledge of dharma is repeated later, to simply jñānānām.

63. The reference is to the tying of a bracelet on the bride at a wedding. This is already mentioned in Medhatithi’s comments on this verse in ch. 7.2.

64. The term akṛtrima here means that the Veda is not authored or composed by anyone. This refers to the mainstream view that the Veda is apauruṣeya, without a human or divine author.

65. This verse is not found in the extant text of Manu.

66. See ch. 1.1: #8. The reading here is somewhat different from the original.

67. This verse is not found in the extant text of Manu.

68. The reading here, anāmnāte (in a matter not scripturally laid down), is different from the edition of GDh, where the reading is anājñāte (in a matter that is unknown); see ch. 3.1: #3.

69. For the meaning of ūha (reasoning), see ch. 7, n. 102.

70. For this text of Baudhayana, see ch. 2.3: #2.

71. For this example, see ch. 7, n. 99.

72. For the Vedic supplements, see ch. 2, n. 4. The Pada text of a Veda presents the words separately, where all the euphonic combinations (sandhi) between words have been dissolved. The Krama text is an artificial version devised to assure the faithful preservation of the original text. In the Krama version each word is repeated with the word that follows. For example, if we represent the words with a, b, c, d, the Krama text would read: ab, bc, cd, etc.

73. This is a verse with a very difficult syntax; different readings are given in different texts that cite it (see DhKo V: 163). I follow the interpretation given by the editors of the DhKo, which fits the context of denying Shudras the eligibility to study the Vedas. The intention appears to be to say that the non-twice-born individuals, such as Shudras, are entitled to some rites but not to Vedic rites.

74. “Arhat” probably refers to Mahavira, the founder of Jainism. Carvaka was considered the founder of the materialist school of philosophy; see Bronkhorst 2007: 307–28, 363–66.

75. See n. 19 above.

76. For this citation, see ch. 2.2: #3 and n. 13.

77. For the interpretation of these verses, see ch. 4.8 and n. 15 there.

78. See ch. 8, n. 22; “painful fasts” are a set of penances called kṛcchra (see MDh 11.212–22).

79. In these two verses, the reference is to interaction with a sinner or a person who has fallen from his caste. In the former, what is to be abandoned becomes more restricted; in the present Kali Age, one has only to abandon the perpetrator. In the latter, the reference is to pollution resulting from interacting with a sinner. In the Kali Age, one is polluted “by action,” probably meaning by contracting ritual connections with that person. See ch. 5.3.

80. Here dharma refers to merit that one acquires through religious activities. See Medhatithi in ch. 7.3: #2.

81. A traditional method of religious suicide. One walks toward the north or northeast without food or water until one drops dead.

82. Given that there are two mealtimes in a single day, fasting for six mealtimes would mean fasting for three days.

83. The meaning is unclear. Gharpure, in a note to his translation, cites some texts that forbid the drinking of “new water,” that is, water from new rainfall.

1. The term liṅgataḥ can mean adjudication through any kind of sign, even forensic evidence, that may lead to the truth, witnesses and other kinds of legal evidence being only one aspect.

2. See Kane (III: 246), who interprets this use of the term by Gautama as “the means of deciding a matter.”

3. Two Sanskrit terms in this sentence are problematic. The first is the verb pratyāhṛtya. This verb is not found elsewhere in the literature and is not even given in dictionaries. It must mean something like bringing forward—that is, getting the facts from people in each group. These “facts” are referred to as artha, whose meaning in this context is unclear, given its vast semantic compass. I have translated it as “affairs” to include both the actual facts of the case and any rules of the association that may govern those facts.

4. For a study of the semantic history of this term, see Olivelle Forthcoming; Introduction, p. 44.

5. For the Vedic supplements and the unclear category of subsidiary Veda (upaveda), see ch. 2, n. 4.

6. The two commentators, Haradatta and Maskarin, differ in their interpretations. According to the former, the false witness should be held responsible for killing that number of animals or men and should be punished accordingly. Maskarin thinks that slaying means bringing ruin to that many of his ancestors. But see the passage of Baudhayana given in the next section.

7. The Sanskrit simply has the verb in the third person (“he should”) without a subject. The commentator Haradatta and Jolly, in his translation, think that the reference is to the defendant. But the entire discussion concerns the questioning of the witnesses, and I think here also the subject of the verb is the witness.

8. Clearly, here the term adharma refers to the sin or demerit acquired by a sinner or criminal, just as dharma often refers to merit rather than to laws or rules.

9. Baudhayana uses here the term sabhāsad (one sitting in the court), which is probably related to Gautama’s sabhya (assessor).

10. See MDh 8.18, cited in ch. 10.2: #2.

11. The person who should question the witnesses is referred to simply by a verb in the third person: “he should.” The person doing the questioning is left unidentified. It is likely that the reference is either to the king or adjudicator called prāḍvivāka.

12. The meaning of mānuṣyahīna is unclear. Jolly takes the expression to mean persons who lack human intellect, in other words, sociopaths, while the commentator Govinda takes it as a reference to persons who do not have relatives.

13. The Kūṣmāṇḍa are the four verses found in TĀ 2.3.

14. See the identical provision in the AŚ 3.8.1.

15. Veṇi as a technical term for the property of brothers who have formed a common household after dividing the ancestral property is unique and only given by the commentator Krishnapandita. Likewise, the term dhūmaśikhā for wages (bhṛti) is taken from the same commentator’s explanation.

16. For the meaning, see ch. 9, n. 6.

17. Verse 35 is so corrupt that it is not possible to translate it.

1. For a study on the use of Kautilya’s original composition by Manu, see McClish 2012, 2014.

2. I have opted to translate dharmastha as “justice” in preference to the normal “judge” because many civil matters, besides trying court cases, are part of his portfolio. For example, at AŚ 2.1.30 a man has to get his permission to become an ascetic; at AŚ 3.4.35 a widow needs his permission to remarry; at AŚ 3.12.14 a man holding a deposit needs his permission to sell it when the depositor cannot be found; and at AŚ 3.16.10 an owner discovering a lost or stolen property gets him to seize it. He was probably part of the second-tier bureaucracy (he is not listed in the salary list of AŚ 5.3), as opposed to his higher-level counterpart, the pradeṣṭṛ (magistrate), who is listed as one of the mahāmātras (high officials). We have in the justices of peace somewhat of a counterpart, and at least in the Anglo-American system, justices can also have judicial functions. There may be a historical connection between the three justices (dharmastha) of Kautilya and the three assessors (sabhya) noted in many treatises on dharma.

3. For the qualities that are required of a minister (amātya), see AŚ 1.9.1. It appears that a large number of officials carried this rank.

4. The term avastha falls into desuetude in later literature when it is replaced by pratibhū. We find its last use in MDh 8.60, which is dependent on Kautilya, where the term has been misunderstood by all commentators and translators, except for one brief gloss by Medhatithi.

5. See my discussion of this unusual term in both Kautilya and Manu in Olivelle 2004b.

6. For a detailed discussion of the four feet of legal procedure, see Olivelle and McClish 2015.

7. For a discussion of these criminal courts, see Olivelle 2012c.

8. The repetition tray as trayaḥ has been taken distributively by all translators, the first referring to dharmasthas and the second to amātyas. Thus, Kangle translates: “Three judges, (all) three (of the rank of) ministers.” We have, however, a nice parallel to this repetition in Panini 1.4.101 that speaks of three triads (tiṅas trīṇi trīṇi prathamamadhyamottamāḥ), glossed by the Kāśikā commentary as tray as trikāḥ. A similar usage is found in the Taittirīya Prātiśākhya (1.3; 1.10), the repetitions dve dve (in twos) and pañca pañca (in groups of five). The meaning in our passage also seems to be that benches consisting of three justices each should try cases. This, of course, contradicts the provisions of treatises on dharma, where a single learned Brahman judge substitutes for the king in a court of law, even though he is assisted by three assessors (sabhya).

9. I have emended the reading of the text (tirohitāntaragāranaktāraṇyopadhyupahvarakṛtān) here to read tirohitān, thus placing it outside the compound. Here I follow the suggestion of Rocher (1978: 20), with which Kangle also appears to agree (“kṛtān is to be construed with antaragāra onwards”) even though he maintains the above reading. Indeed, at AŚ 3.1.6 we have the adjective tirohitāḥ outside a compound and qualifying vyavahārāḥ (implied [anuvṛtti] from AŚ 3.1.2) while all the other words (AŚ 3.1.7–11) compound with kṛta: e.g., antaragārakṛtāḥ. In all likelihood, a haplography has occurred here in the dropping of “na” in tirohitānantaragāra -.

10. See ch. 3, note 9.

11. These two technical terms, consignments and deposits (nikṣepa, upanidhi), have a broad range of meanings. In the current context, the former probably refers to the raw material a client may give to an artisan for manufacturing an article. Given the reference to secluded women, who are forbidden to venture out of their homes, this appears to be the meaning; see the consignment of raw material for making yarn brought to the homes of such women at AŚ 2.23.11. These are dealt with in AŚ 3.12.33f. Deposits (upanidhi), on the other hand, refer to pledges or other items deposited in the care of someone. These pledges generally are used as collateral for loans: see AŚ 3.12.1f.

12. The reference is probably to people such as tavern keepers and prostitutes.

13. “Secret association” (mithaḥsamavāya): the term samavāya (AŚ 1.13.2; 3.1.25; 4.7.17; etc.) refers to any type of gathering, especially when there is an agreement to pursue a common goal. The most common secret association is consensual sex that constitutes the Gandharva kind of marriage (AŚ 3.2.6). But secret associations extended further than that (see AŚ 3.12.52).

14. The context for all these provisions is a joint family where the property remains undivided. Generally, the authority for property transactions rests with the head of the family, normally the father. If the father has retired, then the authority falls on his oldest son, and the father himself is regarded as a dependent. A brother excluded from the family probably refers to a brother who has either left or been cast out. The commentary Cāṇakyaṭīkā gives the alternate reading niṣkalena (for niṣkulena), which is interpreted as an impotent brother. All such individuals are incompetent to conclude a valid transaction.

15. “Person given as a pledge” (āhitaka) means someone given to a creditor as collateral. During the time the person is a pledge, he or she resembles a slave and cannot act independently: see AŚ 3.13.6, 16.

16. The age of majority in ancient India was sixteen years for a male and twelve years for a female (AŚ 3.3.1). At the other end, seventy years is viewed as the cut-off point; anyone older that that becomes incompetent to execute a transaction.

17. The reference is not to what immediately preceded but to valid transactions noted earlier (AŚ 3.1.6–11).

18. This appears to be a rider to the conditions of a valid transaction listed above. Within each familial, social, or economic group (varga), the prevailing customs regarding valid transactions should be honored, within the limits set forth in the rest of this provision.

19. It is a general principle, well articulated in YDh 2.23, that the proof of a later transaction is stronger than that of an earlier one. Thus, if someone proves that he lent money to someone, that proof is thrown out if the borrower can prove that he returned the amount at a later date. The YDh gives three exceptions: pledge, gift, and purchase. In these three cases, the earlier document is more powerful than a later one, because once these transactions are made, the other person does not have the legal power to alienate the property. The term ādeśa (directive) is unclear. I have followed Kangle in taking this to mean a charge given to a person to perform a certain function at a distant location (see AŚ 3.12.18).

20. The term vivādapada has the technical meaning of the various subjects within which a lawsuit may be filed (also called vyavahārapada in later sources). This meaning is clear in its use at AŚ 4.7.17. In other places, however (see AŚ 3.16.38; 11.1.4), a less technical meaning of legal disputes or filing lawsuits may be intended. Given that writing is involved here, I think the more technical meaning extending to the actual facts of the current case is intended.

21. The referent is unclear. The verb here is given in the singular, even though at the outset (AŚ 3.1.1) there are three justices in the court. Either the reference is to a court recorder or clerk, or we are confronted with material from a different source, such as a treatise on dharma, that did not recognize a bench of three judges. Nevertheless, note the presence of three assessors (sabhya) as part of the court even in treatises on dharma.

22. The term karaṇa, which I have translated as “time,” is a division of a day. There are eleven such divisions, and they correspond roughly to one half of a lunar day (tithi). Kangle and Meyer, in their translations of the AŚ, take the term to refer to some sort of government office, but why it should be repeated along with adhikaraṇa, which probably refers to the court hearing the case, is unclear. Coming after the day (divasa), I think it more likely that the term refers to a division of time. See the use of velā (time of day) at the same position in a citation given by Vijanesvara (on YDh 2.6; ch. 13.1: #1; KātSm 124, cited by Devanna Bhatta, ch. 13.2: #5).

23. This implies that the plaintiff and the defendant agree with regard to the accuracy of what has been recorded. If not, changes will be made to the official record. See YDh 2.6–7.

24. The meaning of deśa has been long misunderstood. Commentators take it as a reference to witnesses, and Shamasastry misses the point altogether, taking the term to mean a question. Kangle comes closest when he takes it to mean evidence. As I have shown elsewhere (Olivelle 2004b, 2005a; see also Samozvantsev 1980–81: 355; Vigasin and Samozvantsev 1985:137), in both Manu and Kauṭilya the term deśa refers to legally valid documentary evidence. This was the earliest designation for what would be termed later lekhya. The plaintiff, after first promising to produce a document, either does not do so or produces a faulty or defective document (hīnadeśa) or a non document (adeśa), which probably means a document that is unacceptable in a court of law. For this term, see also AŚ 3.16.29.

25. The meaning of puruṣa in this context is unclear. It can refer either to an official or, more commonly, to subordinates and servants of such officials. The context here is a court that travels to places where sessions are held. The losing party is liable for the costs of such travel and other court costs.

26. The reading and the meaning here are unclear. The manuscripts read aṣṭāṅga, whereas Kangle has adopted the reading aṣṭāṃśa, which, according to him, is the reading of the Malayalam commentary. Hartmut Scharfe, however, in a personal communication tells me that neither aṃśa nor aṅga is found in the Malayalam commentary, which gives this gloss: “The loser shall give the interrogating officer one-eighth of a Paṇa as expenditure.” In either case, the meaning is opaque. Kangle, following Shamasastry, translates: “one-eighth part (of a paṇa),” while Meyer, adopting aṣṭāṅga, takes it to mean one and one-eighth Paṇas. This sort of expression, however, is never used with regard to a Paṇa. The text is probably corrupt, but, if aṃśa is the correct reading, it must refer to some sort of share given by the defeated party to these servants (see dvyaṃśa at AŚ 3.5.8).

27. I take samavāya as a separate and fourth exception. Kangle takes it together with the preceding and translates: “association in caravans.” For samavāya as an association, often secret associations, see AŚ 3.1.11; 3.12.52; 4.7.17.

28. That is, the defendant has offered his response (the plea) to the plaint of the plaintiff. After receiving this response, the plaintiff has to offer his own reply that very day.

29. If the party being sued admits the claim made by the plaintiff, that would be the best outcome. This may be the course intended when the MDh 8.49 speaks of dharma as the first course in settling a lawsuit, because here the outcome is in accord with the truth. When an admission is not forthcoming, then a trial with witnesses ensues.

30. The reference appears to be to the witnesses who initially had the backing of both parties. Their testimony is superior to that of other witnesses. See also the term nibaddha (listed) used by Gautama (ch. 1.2: #2), which probably has a similar meaning.

31. Other sources, including the YDh, which depends heavily on the AŚ, permit a single witness in specific contexts: see MDh 8.77; YDh 2.72. The intent here may be to preclude a single witness specifically in litigations regarding debts.

32. These appear to be individuals who worked in some official capacity for an entire village and, given that they are mentioned in the salary lists (AŚ 5.3.23), may have had official standing. They may have had independent means, like the traders, and thus would have been able to have the work of cultivation carried out when land was seized from cultivators (AŚ 2.1.11). It is unclear whether the term is used here in the more common sense to mean a lowly and poor servant carrying out menial tasks for the village. Thus, for example, at AŚ 5.2.11, remnants are left from heaps of harvested crops for use by mendicants/beggars and village servants.

33. That is, a man who presents himself as a witness on his own (svayaṃvādin), without being formally nominated or appointed by either of the parties to the dispute. See NSm 1.143.

34. The meaning appears to be that the justices should simply divide the claim equally between the two litigants.

35. In the MDh (8.138) the three levels of fines are: lowest 250 Paṇas, middle 500 Paṇas, and highest 1,000 Paṇas. The AŚ (3.17.8–10), on the other hand, gives them as 48 to 96, 200 to 500, and 500 to 1,000.

36. For a detailed study of this statement, see Olivelle and McClish 2015.

37. See n. 35, and ch. 3, n. 9.

38. The story of the sage Māṇḍavya is narrated in the MBh 1.101. Once some thieves, with soldiers in hot pursuit, came to the sage’s hermitage while he was observing a vow of silence and hid there with their loot. When the sage would not reply to the soldiers’ questions, and finding the thieves in his hermitage, they impaled him. Learning the truth, the king removed him from the stake, but he could not pull the stake out. A part of the stake (aṇi) remained embedded in his anus, and thus he came to be called Aṇi-Māṇḍavya. The AŚ version of the story is some what different in that here the sage does not remain silent but confesses to a crime he did not commit, fearing torture. For a study of the legal significance of this story, see Wezler 1997.

39. This section is found in AŚ 3.12.38–51.

40. It is difficult to come up with this number with the text and the readings we have. There are eleven kinds in passage 22 and four in 21. One possible way to arrive at eighteen is to regard the two thigh bindings, two scorpion bindings, and two hangings as six kinds of torture, which is the solution offered by Ganapati Sastri. Prof. Wezler (personal communication)thinks that a line in 22 may have been lost due to haplography. This is possible, especially because one fails to see how drinking gruel would make a difference in burning a finger of a man.

41. Kharapaṭṭa literally means something like “hard/rough tablet,” and the reference is unclear. It may be the title of a treatise or of a chapter dealing with the mechanics of torture.

42. These are obviously various ways a justice may become partial to one party in a lawsuit by making it difficult or impossible for the other party to bring the matter to court. The first three acts all involve some kind of verbal threat. For the final item we have the verb abhigrasate. It is a hapax in the AŚ and is not recorded in any dictionary either. Its literal meaning should be something like “swallow up.” I think the meaning here is that the justice “swallows up,” that is, makes the lawsuit disappear by suppressing it in some way. We have the related word graseta in MDh 8.43, where it clearly refers to the suppression of a lawsuit brought by a private individual.

43. The person imposing penalties on the justices for malfeasance appears to be the collector (samāhartṛ; see AŚ 4.9.18, where magistrates also are the targets of punishment), while both the collector and the magistrates exercise control over superintendents (AŚ 4.9.1).

44. See n. 35 and ch. 3, n. 9.

45. Namely, eight times the amount under litigation. For this provision with regard to false witnesses, see AŚ 3.11.45.

46. The dependence of Manu on the AŚ has been amply demonstrated by McClish (2009, 2012, 2014). As already noted, the AŚ underwent a drastic redaction sometime after Manu.

47. See MDh 8.79, 181; 9.234; and ch. 13, n. 51. See my study of the term prāḍvivāka in Olivelle Forthcoming.

48. For more detailed comments on these selections, see my notes to the translation in Olivelle 2005a.

49. For the presence of counselors (mantrin) in a court, see VaDh 16.2 (ch. 9.4: #1).

50. The image here is of dharma (justice) that is pierced by a dart, which is injustice itself, and which the court is obliged to remove. The commentators Bharuci and Medhatithi, however, give a different explanation. Dharma is pierced by a dart when a judge decides wrongly and permits a miscarriage of justice to occur in his court. If the other officials of the court let it go unchallenged, then they are themselves wounded by this dart.

51. Here there is a play on the sounds of the Sanskrit terms, a kind of folk etymology. Bull is vṛṣa; to impede is alaṃ kurute; and low-born is vṛṣala, the final “la” coming from alam.

52. All the commentators take the term eka (one) to mean that one should not call a single witness. This is implausible, because the term occurs within a long list of individuals who are disqualified from being witnesses because of some disability. The term eka may, within this context, refer to what we would call a “single person,” that is, an individual who lives on his own and is not part of a larger household, either his own or of an extended family. Indeed, a single witness is permitted in verse 77, although this is admittedly a proverbial saying.

53. See Vasistha in ch. 9.4: #3.

54. The meaning is that the judge should not look into factors given above that would disqualify a witness in these kinds of cases. Anyone is permitted to testify.

55. Bühler, following the commentator Kulluka, translates: “(depositions) differing from that, which they make improperly, are worthless for (the purposes of) justice.” He connects dharmārtham syntactically with apārthakam. I follow the commentators Medhatithi, Raghavananda, and Ramacandra in connecting dharmārtham adverbially with the verb vibrūyuḥ; the meaning being that they tell a lie for a higher purpose. This point is spelled out in verses 103–104, where also we have perjury committed for the sake of dharma (dharmataḥ). Such false statements are made, for example, if the life of the defendant is at stake. For a long list of indicators that point to false testimony, see NSm 1.175–178.

56. Commentators and translators alike take deva in the compound devabrāhmaṇasāṃnidhye here to mean images of gods. This may well be the case. But deva can also refer to the king (see MDh 7.8; 11.83), and we have an exact parallel in MDh 8.60: nṛpabrāhmaṇasaṃnidhau, where the questioning of the witness is done in the presence of the king and Brahmans.

57. The commentator Nandana explains, correctly I think, that a Vaishya is made to touch these substances before testifying; and this is precisely what is stated in verse 113. Alternatively, the judge may have uttered an imprecation about the man’s cows, etc.

58. If the witness has no quarrel with Yama, the god of death and the judge of the dead, then there is no need for him to go on pilgrimate to the Ganges or the Kuru land (kurukṣetra) to expiate his sin.

59. The Kūṣmāṇḍa formulas are found at TĀ 2.3. The verse to Varuṇa is ṚV 1.24.15, and the three formulas addressed to water are found at ṚV 10.9.1–3.

60. For these fines, see n. 35.

61. The story in this verse probably refers to ṚV 7.104.15. Vasistha was accused of being a fiend who had eaten his own sons. He cleared himself by an oath that if it were true he should die that very day.

62. The procedures for various ordeals are described in NSm 20 (see ch. 11.2). In the fire ordeal, eight circles are drawn on the ground. Seven pipal leaves are tied to the palms of the person undergoing the ordeal and a hot iron ball is placed in his hands. He must carry it through the circles and drop the ball at the designated place. If his palms are not burned, he is declared innocent. In the water ordeal, an arrow is shot from a medium-sized bow. The person undergoing the ordeal must remain submerged in the water until a fast runner brings back the arrow. The third ordeal listed here is actually an oath; if his family does not suffer any mishap soon (during 14 days, according to commentators), he is innocent.

63. Vatsa was accused by his brother of being the son of a Shudra woman and thus not a pure Brahman. Vatsa went through fire to prove his pedigree. See Pañcaviṃśa Brāhmaṇa 14.6.6.

1. For a detailed discussion of the textual history and authorship of Kautilya’s treatise, see Olivelle 2013.

2. These fragments have been collected and edited: see Aiyangar 1941 and Kane 1933.

3. The commentator Vijnanesvara observes that “in a similar manner” refers to normative practice (ācāra). See the coupling of the two in the very next verse.

4. The Sanskrit term here, vyavahārapada, is ambiguous. It could simply refer to the subject or topic of litigation, such as the eighteen given in MDh 8.4–7. But given that Yajnavalkya refers to the four feet of legal procedure (catuṣpād vyavahāraḥ), I think it more likely that here also he is referring to the four feet of a judicial procedure.

5. Although Yajnavalkya does not identify the four feet, it is clear that he has listed them in the previous verses: plaint, plea, evidence, and proof. In later texts the last is identified as verdict.

6. The term sāhasa can also mean robbery or mugging, when property is taken by force in the presence of the owner. Thus, Manu (8.332) defines “violence” and distinguishes it from theft: “When an act is committed with force and in the presence of the victim, it is ‘violence’; when it is committed outside his presence, it is ‘theft.’”

7. The medieval commentators Vijnanesvara and Apararka take the term atyaya to mean some kind of destruction of life or property. I think the early commentator Visvarupa is correct in thinking that the term refers to an urgent matter, that is, a case that does not permit delay. See NSm Mā 1.39 (ch. 11.2: #1).

8. The meaning appears to be that the plaintiff attempts to settle his doubtful claim (e.g., a claim that the defendant has denied) not by proper means of proof, such as witnesses or documents, but by extrajudicial means such as arresting or threatening the defendant.

9. The issue here relates to who has the burden of proving his case, called by the technical term “prior litigant” (pūrvavādin), according to the explanation of Vijnanesvara. Most commonly this person is the plaintiff. Thus, his witnesses are called to testify first. There are conditions, however, under which the “prior litigant” status falls on the defendant (uttaravādin). Vijnanesvara gives this example of a suit concerning a piece of land that the plaintiff claims he received as a gift but is now occupied by the defendant. The defendant admits that the plaintiff received the land as a gift, but it was subsequently purchased by the king and given to the defendant. In this case, the claim of the plaintiff being annulled by the plea of the defendant, the onus of proof falls on the latter and he becomes the “prior litigant”; his witnesses are deposed first (see YDh II.23). Another explanation relates to the two kinds of plea: “special plea” (kāraṇokti or pratyavaskanda) and “prior judgment” (prāṅnyāya). In the first case, the defendant admits the charges but pleads innocent for a specific reason: e.g., admits that he borrowed the money as stated in the plaint, but says that he returned it. In the second case, the defendant claims that the same charges against him have been already dismissed by the judgment of another court. In each case, the burden of proof falls on the respondent or defendant, and he then becomes the “prior litigant.”

10. For the custom of waging a bet or stake at a trial, see Devanna Bhatta’s comments in ch. 13.2: #2. For a detailed discussion of this issue in the legal literature, see Lariviere 1981b.

11. That is, the judge must detect and dismiss all kinds of deceits and tricks that the litigants may use to win the case.

12. For opposing viewpoints on this issue, see Vijnanesvara’s commentary on this verse in ch. 13.1: #3.

13. Here the reading of Visvarupa (smṛter virodhe), which I follow, is markedly different from that of Vijnanesvara and Apararka, who read smṛtyor virodhe (“when there is a conflict between two texts of recollection”). Here the term nyāya appears to be a synonym of śāsana (royal edict), as in the AŚ 3.1.45.

14. Enjoyment (bhukti) is a technical term referring to legal possession and usufraction of the thing under litigation. If someone, for example, has tilled a particular field or milked a particular cow for a stipulated period of time, then the presumption is that he has ownership of it: YDh 2.24. See, however, YDh 2.27, where title to a property is given greater force than possession.

15. I follow here the reading of Visvarupa and Apararka: sarveṣv eva vivādeṣu. Vijnavesvara reads sarveṣv arthavivādeṣu. Note that property crimes (arthavivāda) are often distinguished from those involving injury caused by anger (manyukṛta). See Vijnanesvara on YDh 2.9 in ch. 13.1: #2.

16. For an explanation of the reasons for this, see Vijnanesvara’s commentary on this verse in ch. 13.1: #4.

17. The term dhana literally means money, but here, as Vijnanesvara points out (ch. 13.1:#4), it refers to movable property, such as horses and cows. The same term with this meaning occurs also in verses 25 and 26.

18. That is, the entire list beginning with “pledge,” given in the previous verse.

19. As I have already noted, the Sanskrit term vyavahāra refers both to a legal transaction (such as taking out a loan) and to a lawsuit. These two verses are dependent on the parallel in the AŚ given in ch. 10.1: #1, where the reference is clearly to legal transactions. Yajnavalkya uses the term within his discussion of lawsuits, but some of the disabilities referred to here apply instead to transactions. So, both meanings appear to be functioning here in a rather confused manner.

20. Probably outside the village or town. Valid transactions have to be both voluntary and out in the open in the presence of witnesses; hence the exclusion of those carried out inside a house or outside a village. See the parallel in the AŚ given in ch. 10.1: #1.

21. This topic is introduced here because the litigants in a lawsuit are expected to provide sureties capable of satisfying the verdict: YDh 2.10 (selection #1).

22. A surety for appearance is obliged to present the defendant in court at the day of the trial. A surety for trustworthiness simply vouches for the good character of the person. The third type of surety undertakes the obligation to pay either a debt incurred or the amount needed to satisfy the judgment of the court.

23. There are two ways several sureties could guarantee a loan. Each could guarantee a portion of the loan, or all could guarantee the entire loan individually. In the first case, each is liable only for that part of the loan guaranteed by him. In the second case, each is liable for the entire loan, and the creditor could press any one of them to pay all of it.

24. SmṛC II: 453 glosses nirdhūta with avadhūta, a particular kind of ascetic, which I think is the correct interpretation. People who have been excluded from their families, which is Vijnanesvara’s explanation, are comprehended under patita or outcaste.

25. The meaning of the unusual expression sabrahmacārika is unclear. Both Vijnanesvara and Apararka take it to mean the Vedic branch to which the person belongs.

26. The phrase tesamāḥ at the end of the verse is unclear with regard to both its reading and its meaning. Vijnanesvara assumes a reading without an avagraha: te samāḥ, and thinks it posits an even number of witnesses, possibly evenly divided between the two parties. Visvarupa, whom I follow, takes the reading to be te ’samāḥ (with “a” elided by Sandhi), requiring an uneven number of witnesses. Apararka and others give both readings. See DhKo I: 352.

27. That is, by the debtor and, if he is dead, by his sons and grandsons.

28. That is, so long as the underlying debt is not paid, the pledge that was the collateral for the loan can be made use of by the creditor and his descendants even beyond the third generation.

29. According to Vijnanesvara, “marks” refers to unique symbols that may be on the document. “Connection” refers to the previous dealings that the creditor and debtor may have had.

30. The meaning is that the amount under litigation should be more than 1,000 Paṇas for these kinds of ordeals to be used.

31. This is a variety of fire ordeal. An iron plowshare is heated and, after a ritual invocation, it is licked by the person undergoing the ordeal. If his tongue is not burned, he is judged to be innocent. See Lariviere 1981a: 215–16.

32. This is a very pithy aphorism. It is clear that the balance is used for Brahmans and others who are considered weak. The reading of pāda-c is unclear because of Sandhi: agnir jalaṃ vāśūdrasya, which can be either vā śūdrasya (so Vijanesvara) or vā+aśūdrasya—the latter is the reading of Visvarupa, which I follow. Traditionally they are viewed as meant for Kshatriyas and Vaishyas, respectively. Although left unstated, the last ordeal of poison is meant for Shudras.

33. This would be the place where the middle marker of the balance is when the two sides have equal weights.

34. That is, if there is a doubt whether he was burned or not.

35. This is an extremely pithy and syntactically muddled verse. According to Vijnanesvara, this is what happens. Three arrows are shot from the location where the man seeking exoneration is going to submerge himself in water. One quick runner runs to the place where the middle arrow (neither the longest nor the shortest) fell. Then at the third clap of the hand the man submerges himself in water, and another swift runner located at the middle arrow runs with the arrow to the original location. When he arrives there, if he does not see any part of the submerged man’s body, then he is judged innocent. For this and other ordeals, see Lariviere 1981a.

36. For various kinds of poisons, see AŚ 2.17.12.

37. Raghunandana Bhattacarya’s Divyatattva (Lariviere 1981a: 264) gives seven such symptoms according to the science of poison: 1. horripilation; 2. sweating and dryness of the mouth; 3. change of color; 4. trembling of body; 5. lack of eye control, sore throat, and hiccups; 6. heavy breathing and delusion; and 7. death.

38. According to Visvarupa, the instruction is given by the official administering the ordeal to the man undergoing it. The official tells him that he should not consider this as ordinary water but as divine water that will inflict punishment during future lifetimes if he tells an untruth. Vijnanesvara thinks that the reference is to a ritual formula addressed to the holy water and gives the following: “O Water, you are the life of living beings” (see Divyatattva 284in Lariviere 1981a). Apararka gives the simple formula: “By truth protect me, O Varuna.”

39. See Lariviere 1989, II: xix–xxii.

40. Cited in Lingat 1973: 102.

41. See Lariviere 1989, II: xxi; Lingat 1973: 102.

42. The term vyavahāra can mean both lawsuit and legal procedure. I think the use of draṣṭā in verse 2 (we see the use of the verbal forms of this term with the meaning of trying lawsuits: see MDh 8.1–2; YDh 2.1) makes it likely that we are dealing here with lawsuits, even though the meaning of legal procedure, which is the main topic of this chapter, is always lurking in the background. The meaning of the term shifts to legal procedure in verse 6.

43. The expression sottara is an abbreviated compound standing for sottarapaṇa (used in verse 5), which is the same as sapaṇa, used in the YDh 2.18. See n.10 above.

44. The term pada here has two meanings, referring literally to feet within the image of humans and animals, and in an extended meaning to the vyavahārapadas, the titles of law or subjects of litigation.

45. For a detailed study of these four feet and their literary and social history, see Olivelle and McClish 2015. Later authors interpret “four feet of legal procedure” to mean “four feet of a verdict.” The meaning then is that these are the four ways a judge would reach a decision.

46. This assumes that the customs of various regions and collectivities are written down in records kept at the royal chancery. See this practice in AŚ 2.7.2 (ch. 3: #1). These two verses of Narada are based on AŚ 3.1.39–40, given in ch. 10.1: #4. For an analysis of this expression, see Rocher 1979.

47. The four expedients (upāya) of political science are: conciliation, gifts, dissension, and military force (sāma, dāna, bheda, daṇḍa): see AŚ 2.10.47–56. These are somewhat out of place in a lawsuit, but see p. 313 for its application within the context of a lawsuit.

48. For how these four are affected by an unjust verdict, see MDh 8.18 in ch. 10.1: #2.

49. The reference is probably to a treatise on dharma. It is known from a variety of sources that an actual treatise was, at least sometimes, kept within the court, perhaps as an outward sign that the court is conducting its procedures in accordance with the principles laid down in it. See NSm Mā 1.29, given below.

50. The term kriyā here may also have the technical meaning of a plaint in a lawsuit. In the SmC (III: 88), Devanna Bhatta equates kriyā with sādhya, the charge that has to be proved. There I have translated this verse of Narada to reflect his understanding.

51. The Muṇḍaka Upaniṣad (1.2.4) identifies the seven tongues: “The Black, the Terrible, the Swift-as-the-Mind, the Blood-Red, the Smoke-Colored, the Sparkling, and the glittering Goddess—these are the seven flickering tongues of flame.” See also ṚV 1.146.1.

52. This is Yama, the god of death and the judge of people who die. He is impartial to everyone, looking only at the good and evil deeds they have performed on earth. This is spelled out in MDh 7.307: “As Yama, when the time has come, holds friend and foe alike in his grip, so the king should hold his subjects in his grip; for that is the Yama vow.”

53. The exact meaning of the term āgama here is unclear. I think it has a literal meaning of “arriving,” that is, the initial filing of charges in a lawsuit. This is the interpretation of Devanna Bhatta, who takes it to mean “hearing the words of the plaintiff” (arthivacaḥśravaṇam; DhKo I: 90).

54. For the story of Mandavya, see ch. 10, n. 38.

55. For a longer discussion of the conditions under which a transaction becomes legal, see ch. 10.1.

56. See YDh 2.12 and n. 7 above.

57. In the case of a person escaping detention, the term vineya (to be disciplined) is used, while in the case of a person wrongfully detaining someone, the term daṇḍabhāg (subject to punishment) is used. The distinction between the two terms, which often can have the same meaning, is unclear. Most commentators (see DhKo I: 118) take them to be synonyms, while Lariviere in his translation renders the former as “should be subpoenaed,” which is doubtful. See, however, the use of vinaya in verses 50 and 224 and vineya in 52 with the meaning of fine.

58. The legal age for males is sixteen and for females twelve. See ch. 10.1, n. 16.

59. The image is of an animal shot by a hunter with an arrow. This act creates an ownership of the wounded animal by the first hunter. A second hunter is not allowed to shoot the animal again. See the sexual metaphor of deflowering a virgin at MDh 9.43. Here, however, there may also be a moral argument against shooting someone who is already wounded.

60. The term āvedita used here has a technical meaning referring to the initial filing of charges: see YDh 2.6. It is at this time that the plaintiff has to inform the court about the evidence he will use to substantiate his case.

61. Here we have two technical terms: tīrita and anuśiṣṭa. The first appears to mean a lawsuit that has been adjudicated, while the second is where a final verdict has been rendered by the judge. Some think that the first is one where a verdict has been rendered, while the second is where fines and punishments have been imposed. See MDh 9.233; Kane III: 383.

62. Here the author is playing with the double meaning of sabhya: both refined/cultured and an assessor or court official. By telling a lie in court, he becomes unrefined and a false assessor.

63. The three kinds of proof are given above at NSm 1.65: document, witnesses, and enjoyment. The meaning here is that possession of an object, especially land, is needed to establish ownership, even when a title or witnesses are available—as stated in the very next verse.

64. For this and the following verses, see MDh 8.147–49.

65. I have read the text as strī dhanaṃ ca narendrāṇām. Others read strīdhanam as a compound; Lariviere translates: “The property of women and kings.” So far in the NSm only women (probably slaves) have been dealt with, not the property of women.

66. The verse is elliptic. I think the commentators (see DhKo I: 409) are correct in assuming that here also, enjoyment presented as evidence of ownership is not authoritative unless the person can show the title to the property.

67. I follow Devanna Bhatta in seeing the initial “Likewise” (tathā) as connecting this verse to the rule “When the very man” (see DhKo I: 410). The meaning is that, once a case has been brought against the man who first started using some property, when he dies while the court proceedings are in progress, then his son cannot claim possession as proof of ownership. This is an exception to the rule “Enjoyment alone is successful” and what is stated in the very next verse.

68. The exact meaning of this verse, especially of the last phrase, is not completely clear. Mine differs from Lariviere’s translation; I follow the reasonable explanation given by Vijnanesvara in his comments on YDh 2.89.

69. See above, YDh 2.92 and ch. 11, n. 29.

70. The terms “appointed” (kṛta) and “non appointed” (akṛta) refer to witnesses who are identified as part of the evidence that the litigants will present to court, and ones who are not so identified but may yet provide important information on a disputed matter. These are also called nirdiṣṭa and anirdiṣṭa.

71. This witness has been summoned specifically by the parties to a transaction and told repeatedly to remember what he has witnessed. A definition is given in KātSm 372: “When someone has witnessed the transaction and, in order to provide proof of that transaction, has been reminded of it repeatedly by the litigant, he is called a witness who is made to remember.”

72. An indirect witness is someone who has heard the testimony of a direct witness, but has not directly seen or heard the transaction himself.

73. Lariviere thinks that that man cannot be a witness with regard to those groups. I follow the clear explanation of the SmC (I: 177) that the members of the group who bear hatred toward that man are excluded from being witnesses.

74. The term vadhaka here may refer to butchers and those who slaughter animals. That is the meaning of vadhakṛt at NSm 1.67.

75. The Sarsvatīvilāsa (p. 141; DhKo I: 303) clarifies that this does not refer to simple disagreements among witnesses, which are only to be expected. When, however, a group of witnesses assert at the beginning that they are unanimous on a particular point, but later during the trial if even one of them disagrees with that assertion, then the entire group becomes disqualified.

76. With regard to the entire issue of the person who has the burden of proof, see Devanna Bhatta in ch. 13.2: #7.

77. This is one of the five types of appointed witnesses given above: NSm 1.130; n. 71 above.

78. Lariviere takes lubdhaka as a greedy man. The reading here is confused; many citations of this verse have the reading vyaṅgaika -(DhKo I: 306), explained as a person lacking a limb, and a single person.

79. The term vrātya here probably refers to people who have not undergone Vedic initiation at the proper time and are considered fallen from Brahmanical ritual practice.

80. For the meaning of the term kīnāśa, see MDh 9.150 and Medhatithi’s commentary on it.

81. The reference may be to people who engage in sorcery associated with roots called mūlakarma in MDh 9.290 and 11.64; see ch. 8, n. 39.

82. That is, physical and verbal assault.

83. The term artha here has been taken in a broad sense to refer to things or matters in general. Given the context of court proceedings, however, I think the term here may have the technical meaning of a court case or lawsuit.

84. For this story of Vasistha, see ch. 10, n. 61.

85. For this story, see MBh 13.94–96.

1. The words in the smaller font, here and in the section on Medhatithi, are not found in the verse but are implied and taken over from the previous verse. The commentaries on this verse would point this out.

2. See the previous note.

3. It is understood that it is to the court that he goes every day.

4. This sentence is probably corrupt. The DhKo I: 71 gives a conjectural reading, which I have not followed because it departs radically from the readings based on manuscripts.

5. Here also the DhKo I: 71 presents a very different reading.

6. Following the DhKo (I: 71), I have corrected the reading here from anudita to anūdita. See the similar use of this term in the commentary on MDh 2.6 (Jha, p. 56, second line from the bottom).

7. This is a very difficult sentence, possibly corrupt. The DhKo (I: 72) attempts to emend it as follows: atra kalpitavyavasthā yathā—lekhyaṃ, yathopabhogaḥ, sākṣiṇaḥ | anumānaṃ tu vastuniyatam. I think this passage has been corrupted. For lack of a better alternative, I have followed the text as emended by the DhKo. The meaning, however, appears to be that the kinds of evidence to be presented in court, such as documents, possession for a period of time, and witnesses, that are given in legal texts are really worldly norms (vyavasthā) and not Vedic dharma.

8. This is quite a difficult passage, and the reading is also uncertain. I follow the reading and punctuation provided in the DhKo (I: 72): yady api sarvaṃ laukikaṃ na śāstrakāravacanāt pramāṇaṃ bhavati | tathāpi laukikam eva, tasmin kvacic chāstram āśrayitavyam |

9. This verse comes at the end of Yajnavalkya’s description of the process of executing a legal document and states that this statement by the scribe should be written down at the very end, immediately after the signatures of the witnesses. Medhatithi, however, thinks that this sequence is not required for the validity of the document. He is a realist and does not think that the letter of the law is that important for arriving at the correct verdict in a case. Both Jha and Gharpure read lekhakas tattvato likhet, while DhKo (I: 71) has emended to this read tv antato likhet, following the reading of Vijnanesvara.

10. The editions of both Jha and Gharpure read here anupalakṣito ’pi. The DhKo (I: 72) gives the reading as anupalakṣaṇe ’pi. I think the correct reading should be anupalakṣite ’pi.

11. For the meaning of this phrase, see ch. 11, n. 26. The differences in the reading noted there do not affect the argument presented by Medhatithi.

12. The meaning of ta eva lekhyam is not altogether clear. The opponent’s argument seems to be that once a document has been witnessed, the witnesses become part of the legal framework of that document. So the document can be viewed as superior to mere live witnesses, who must relie on their memory.

13. Jha and Mandlik give the reading sākṣidvaidhānnyāyaḥ. I follow the reasonable emendation of the DhKo (I: 72): sākṣidvaidhanyāyaḥ.

14. The reading of Jha and Mandlik is nirupādhiḥ, but I think the emendation nirupadhiḥ given in the DhKo (I: 72) is superior. Or it could be nirupadhaḥ, which is derived from upadhā, a common term for fraud and deceit (see AŚ 2.5.9; 2.14.10).

15. The editions of Jha and Gharpure read: nāsty eva pramāṇāntarasya sākṣyāder avasaraḥ. This is clearly erroneous, because the intent is to allow for witnesses within these narrow circumstances. I have followed the emendation of the DhKo (I: 72): cātrāsty eva pramāṇāntarasya sākṣyāder avasaraḥ.

16. The plaintiff here is the man taking the loan. After he has executed the loan document, the creditor asks him to take only a portion on that day. He promises to give the rest of the loan on the following day, but he reneges on it. So, the creditor is holding a document for the entire amount of the loan, whereas he has given the debtor only a portion of it.

17. The text reads: vināpi sākṣibhiḥ sidhyet svahastaparicihnitam. This appears to be a variant reading of YDh 2.89, where the reading of the editions is: vināpi sākṣibhir lekhyaṃ svahastalikhitaṃ tu yat. The reading at YDh 2.93, pāda-d, however, is: svahastaparicihnitam.

18. Here I follow the reading of DhKo (I: 73): yadi hi tena na dhanaṃ dattam. The word na is omitted in Jha and Gharpure, perhaps through haplography.

19. I have been unable to trace the source of this citation.

20. Here I think after svahastalekhyam a phrase such as sidhyati is understood. The DhKo (I: 73), however, has the reading svahastalekhye.

21. The text in all the editions reads: prītyā tvayaitad uktam (This was stated by you out of affection). I am unable to make sense of this, because the debtor is trying to wiggle out of the contract. I have emended it slightly, because the orthography of e and ai are often confused: prītyā tvayy etad uktam.

22. Bhartryajna appears to have been a prolific commentator. He is cited not only by Medhatithi but also by other later authors as an authoritative writer. Kane places him in the eighth century C.E. For further details, see Kane I: 551–53.

23. The readings of the editions differ widely here. I follow the judicious reading offered by the DhKo (I: 73): vyākhyānāntarāṇi bhartṛyajñenaiva samyakkṛtānīti tata evāgavantavyāni sarvathā pramāṇamūlāni | smṛtikāraṇavyavasthānuvartitavyeti |. Jha (p. 76) gives a confused reading: -gantavyāni | sarvathā pramāṇamūlāni smṛtikāraṇam | vyavasthā tu kartavyeti |.

24. The Sanskrit harītakī refers to Terminalia chebula, the fruits of which are used as a laxative.

25. Medhatithi deals with the injunctions relating to the purification of articles made of metal, stone, wood, and the like in his commentary on MDh 5.110. What he wants to assert there is that even though these rules have injunctive verbs, such as the optative, they are not true injunction in the Vedic sense. They merely state the common practice in the external form of an injunction: “You should clean metal utensils using ash, earth, and water.” But such practices can be learned by observation, and the injunction simply reiterates the common practice. Thus, in the example given here, eating the myrobalan as a laxative is a common practice and recommended by medical practitioners. It is not something that has an unperceived purpose, which is required of all truly Vedic injunctions.

26. Success is artha and pleasure is kāma, two members of the triple set that includes dharma. Of the three, only dharma has a claim to be eternal.

27. The reading here is unclear and probably corrupt. I have followed the reading of the DhKo (I: 76): ayaṃ rājopānayanārtho daṇḍaḥ patati devatotsavārtho vā. Jha reads: rājño ’padeśenārtho, and Mandlik: rājñopadeśenārtho.

28. On the technical meaning of saṃpradāna as the recipient expressed in the dative of an act of giving (e.g., upādhyāyāya gāṃ dadāti), see Panini 1.4.32.

1. The term hetu here has a technical meaning. Rocher (1956) calls it probans, that is, the reason why the plaint is justified. Thus, if the plaint (probandum, or what has to be proven)says that the defendant owes the plaintiff 100 rupees, then the hetu is his claim that the defendant borrowed 100 rupees from him: “He owes me 100 rupees, because he borrowed it from me.”

2. These verses are ascribed to the KātSm 86–88, 97–98. See DhKo I: 125–26 for these verses.

3. For the meanings of vineya and daṇḍabhāg here, see ch. 11, n. 57.

4. The expression kṣamāliṅgāni has a technical meaning: the reasons the plaintiff permitted the defendant to keep and make use of his property without objection. This becomes relevant in cases where an article has been enjoyed by a defendant for a long time, creating a presumption of ownership. See Devanna Bhatta’s comments on KātSm 124–26 at ch. 13.2: #5.

5. See ch. 13.2: #5 where Devanna Bhatta ascribes this citation to the Smṛtisaṃgraha.

6. A nivartana is about 58 sq. meters.

7. These are the letters of the Sanskrit alphabet going from the guttural to the labial, with hard consonants first and then soft consonants.

8. This verse is not found in the edition of the NSm; see DhKo I: 141.

9. See DhKo I: 162 for this verse, variously ascribed to Narada, Brihaspati, Pitamaha, etc. See SmC III: 96 (ch. 13.2: #6).

10. See the following note for some examples of such difficult or ambiguous language. The compounds mentioned here probably refers to Tatpuruṣa compounds, where one has to guess the grammatical case of the first member in order to specify its syntactical connection to the following word. Different cases will provide different meanings, something that cannot be permitted in a legal document.

11. The example given here uses broken Sanskrit syntax and impossible compounds. I have omitted it, because it is impossible to show this in a translation. The example given by Vijnanesvara pertains to a debt of one hundred Suvarṇa gold coins inherited by a person from his father (suvarṇaśataviṣaye pitṛṛṇābhiyoge). The plea reads: gṛhītaśatavacanāt suvarṇānāṃ pitur na jānamīti. The only way a meaning can be derived from this is to rearrange the sentence with a proper syntax: gṛhītaśatasya pitur vacanāt suvarṇānāṃ śataṃ gṛhītam iti na jānāmi— “Because of the word of the father who took a hundred, I do not know whether (or do not acknowledge that) one hundred Suvarṇas were taken.” Here there is an ellipsis of śataṃ gṛhītam relating to suvarṇānām, and the compound gṛhītaśatavacanāt is impossible to dissolve.

12. Ascribed to Harita in DhKo I: 233.

13. This is one portion of a longer passage of two verses ascribed to Harita (DhKo I: 185). The entire passage reads: “When both a denial and a special plea are entered in a single case, or an admission coupled with another plea, what is the plea that should be taken up? Between a denial and a special plea, one should take up the special plea.”

14. That is, the defendant may choose to present evidence either first for the special plea (e.g., witnesses or a document proving that he returned the money) and then for the prior judgment (e.g., the court document), or vice versa.

15. The two feet in a case where there is a plea of admission are plaint and plea.

16. The term hīnavādin (or simply hīna) generally refers to the litigant who has lost the case (see NSm Mā 2.33, cited by Vijnanesvara on YDh 2.6). But here he appears to take the term as referring to a litigant who is guilty of an infraction pertaining to legal procedure. Later Vijnanesvara says that such a litigant is only to be fined; he does not lose the case.

17. The crime of not returning the money is tried under the first subject of litigation, the nonpayment of debts. The crime for robbery is tried under forcible seizure (sāhasa).

18. Editions of the Mit read parastrī (“another’s wife”) but paśustrī (farm animals and women, as a Dvandva compound) is the better reading (as noted by Gharpure), and it is also the reading in Lariviere’s critical edition of the NSm.

19. The logical reasoning is what was noted earlier. A man who lies with respect to one part of the allegation can be assumed to have lied about other parts as well. But the son here has not lied; he simply stated the fact that he does not know anything about his father’s debts.

20. The reference here is probably to YDh 2.20 that deals specifically with a plea of denial.

21. The meaning of the Sanskrit sthiraprāyeṣu is not altogether certain. Kane translates: “which are of a permanent character.” I follow the subcommentary Bālaṃbhaṭṭī, which says that disputes relating to debts have clear evidence in the form of documents and witnesses, while other disputes, such as adultery, would have to be decided through circumstantial evidence.

22. This maxim, ekadeśavibhāvitanyāya, is supported in YDh 2.20. This appears to be a variant of the maxim ekadeśavikṛtinyāya (see Jacob 1904: 16), which refers to the fact that a change in one part does not change the whole; or, in the present context, when one part of an animal is identified the whole animal is identified.

23. The reference is to cases involving transactions executed in the wilderness, etc., and to cases involving violence and denial of a deposit.

24. Note here that the edition has ca (“and”); I have taken it to be vā (“or”), following other citations of this verse (see DhKo I: 231) and the subcommentary Bālambhaṭṭī, which says that this provides an option (vikalpa).

25. Here we have quite a complicated compound: dvāramārgakriyābhogajalavāhādiṣu. Here my interpretation of kriyā and ābhoga is based on the subcommentaries Subodhinī and Bālambhaṭṭī.

26. I am not quite sure of the meaning of this half verse; it is not recorded in the DhKo, and Gharpure’s translation is wrong. It is not directly commented on by either Sulapani or Balambatta.

27. The argument is a bit complex. First, the relationship of enjoyment or possession to ownership is defined as pramāṇa, means of knowing something, in this case inference. So, when we see a man occupying a house, we infer that he owns the house, just as we infer the presence of fire on a mountain when we see smoke rising from it. But a means of knowledge does not produce what it reveals: enjoyment does not produce ownership, just as the smoke does not produce the fire. So, mere enjoyment cannot produce ownership. The causes of ownership are listed in texts such as Gautama’s cited by the opponent, but enjoyment is not one of them.

28. Gautama’s commentator, Haradatta, explains that here possession (parigraha) refers to a situation when someone takes fruits or grass that do not belong to anyone. The one who takes these first becomes their owner.

29. A restrictive rule (niyama) in Vedic exegesis is meant to restrict all possible or available options to a selected few. Thus, the opponent argues that Gautama’s statement is not an originative rule (utpatti) that would prescribe these as the causes of ownership. Rather, the various causes already known in the world are reduced to just eight by Gautama.

30. See ch. 3, n. 9.

31. The meaning appears to be that the enjoyment of a property for a specific period of time does not, ipso facto, create the loss of ownership in that property on the part of the owner and the ownership in that property by the person enjoying it. Such a transfer of ownership can be done only as a result of the judicial verdict after a legal proceeding.

32. At NSm 1.70 the time limit is set at ten years.

33. The meaning of the term sabhya in this context is unclear. The Bālambhaṭṭī glosses it with nirduṣṭo ’rthaḥ, a meaning that is not flawed.

34. The term for product, phala, has two meanings here: produce, like rice or mangoes, and long-lasting trees that may grow on the land. The former are destroyed when they are used or consumed, but the latter last in their original form even when their fruits are being enjoyed. One gets to own the latter even after twenty years, when one gets back the land.

35. We have noted already the multiple meanings of the term vyavahāra, which can mean both lawsuit and legal procedure. In Sanskrit both meanings are often present in its usage. I have been inconsistent in this chapter, using “lawsuit” when that meaning is primary, and “legal procedure” when the reference is to the court procedure rather than the lawsuit as such.

36. This verse of Katyayana has been subject to various interpretations: see Kane’s (1933) note to his translation of this verse. A central issue is the meaning of nyāyavistare. I have followed Devanna Bhatta’s interpretation, which takes nyāya in its technical meaning relating to legal reasoning.

37. Devanna Bhatta equates kriyā with sādhya, the charge that has to be proved. I have translated this verse of Narada to reflect his understanding.

38. For a detailed study of these four feet and their literary and social history, see Olivelle and McClish 2015.

39. The reference is probably to the court officials or to Brahmans who are present in court as advisors.

40. One can infer that a man is an arsonist when he is found near a fire with a firebrand in his hand. See NSm 1.155 in ch. 11.2: #2.

41. This is an interesting anecdote. In Kautilya’s Arthaśāstra (2.7.2; ch. 3: #1), it is explicitly stated that the king should have written records of the customs in various regions and communities of his kingdom.

42. This verse is ascribed to Pitamaha at SmC III: 58; see DhKo I: 105.

43. The author here picks up the argument that was interrupted by the objection after the citation of NSm Mā 1.10 (page 272).

44. See these four divisions at page 272, NSm Mā 1.8–9.

45. See ch. 11, n. 47.

46. For this rule, see MDh 8.18 in ch. 10.2: #2.

47. The meaning is that not all three are needed to have a dispute, just one would suffice.

48. This is both a substantive definition of a lawsuit or legal procedure (vyavahāra) and an explanation based on phonetic similarity. It is said to be vyavahāra because the legal disposition (vyavasthāna) is based on statements (vākya).

49. Here Srinivasacharya’s edition is incorrect: pramāṇābhāvaśaṅkhāpanodakam. Gharpure’s edition has the correct reading: pramāṇābhāsaśaṅkhāpanodakam.

50. The term used here for judge is prāḍvivāka: see ch. 10, n. 47 for a comment on this term.

51. Here we have a phonetic etymology and explanation of the term prāḍvivāka. The first derivation is from asking (pṛcchati) questions (praśna) and counterquestions (pratipraśna). The second is from speaking (vadati) at the outset (prāg) and affectionately (priyapūrvam).

52. Here Devanna Bhatta gives a bit more plausible etymology of prāḍvivāka, derived from two sources: prāḍ from the fact that he questions (pṛcchati), and vivāka because he is especially (viśeṣeṇa) charged with rendering a verdict.

53. In the long section omitted here, there is a discussion of trials adjudicated personally by the king or by a Brahman judge (prāḍvivāka) representing him. These are the higher venues for trials.

54. The reference is to disputes involving two villages. The adjudicators for such disputes are drawn from both villages.

55. There are various lists of low-caste and untouchable people, and many of them are headed by the washerman (rajaka): See Apa I: 279; Kane IV: 115. The most pertinent list is given by Devanna at SmṛC III: 65, where he lists eighteen “low” castes called by the technical term prakṛti. A passage ascribed to the Garuḍa Purāṇa in Hemadri’s Caturvargacintāmaṇi (IV: 38) gives a list of sixteen, headed by washerman, calling them Candalas who live in the village. In Maharashtra today there are lists of eighteen low-caste people, and this may have been true in other parts of southern India where Devanna Bhatta lived (personal communication by Ashok Aklujkar). A closer parallel is found in the Tamil country, where also there are eighteen such castes, again beginning with the washerman, listed in the Madras Tamil Lexicon: vaṇṇāṉ [washerman], nāvitaṉ [barber], kuyavaṉ [potter], taṭṭāṉ [goldsmith], kaṉṉ āṉ [brazier], kaṟṟaccaṉ [mason], kollaṉ [blacksmith], taccaṉ [carpenter], eṇṇeyvāṇikaṉ [oil merchant], uppuvāṇikaṉ [salt merchant], ilaivāṇikaṉ [betel merchant], paḷḷi [watchman], pūmālaikkāraṉ [garland maker], paṟaiyaṉ [drummer; Dalit, pariah], kōviṟkuṭiyāṉ [conch blower], ōccaṉ [priest at a goddess temple; another Dalit community], valaiyaṉ [fisherman], pāṇaṉ [tailor]. I thank Whitney Cox for this reference.

56. In the AŚ (1.2.1) the four sciences (vidyā) are listed as: critical thinking, triple Veda, economics, and government (ānvīkṣikī trayī vārttā daṇḍanītiś ceti vidyāḥ).

57. The reading of the editions is corrupt. I have emended the text to read: śāstraprayuktyā nṛpatiprayuktyā vā sabhāsiddhyabhāvāt.

58. The editions read sasādhanaḥ, but the text of the KātSm reads sahāsanaḥ, which is also the reading in the DhKo I: 133.

59. Apararka (p. 605) explains: vāmahastena kiṃcic cālayan, referring to shaking something with his left hand, or perhaps gesticulating.

60. The term uttaravāditva used here appears to be contrasted with prativāditva, the latter clearly referring to the defendant. Unless they are used here as synonyms, the former appears to refer to the role of entering the plea, which is the topic under discussion.

61. For this term, see Amarakośa 4.78 p. 67; this appears to be a local word.

62. The expression asabhyavādeṣu is uncommon; there are numerous alternate readingsgiven in citations of this verse. According to the explanation contained in the next two verses, these litigations concern especially heinous acts (later gurukārya is used) requiring the immediate attention of the court. The expression may also mean a suit that is not acceptable to the court. See the use of sabhya with the meaning of something that the court accepts in Devanna’s citation of MDh 8.164 at p. 300, and n. 82.

63. The editions and the DhKo (I: 133) read vādinā (or vādināṃ) tathā. The syntax becomes difficult with this reading, and Gharpure understands it to mean that it is the plaintiff who should place the defendant in front of the court, which is unlikely; in the preceding prose, both of them are so placed by the judge. I have emended the reading to vādinā saha.

64. Order and decree refer to oral and written commands of the king. Document (lekhya)refers to a private written contract. Record (paṭṭaka) probably refers to a title written on a tablet or cloth. Document (patra) refers to any legal record.

65. The meaning of “god-year” or year of god (devavarṣa) is unclear. If deva in this context means king, then it could refer to the regnal year.

66. Daṇḍa is a measurement of length, approximately 1.92 m. Tulā is a measurement of weight, approximately 37.76 kg. Prastha is a measurement of volume, approximately 0.31liters.

67. The reading here is probably corrupt. Srinivasacharya’s edition reads: pratibhuvā dhanikena kṛtā; Gharpure’s reads: pratibhuvi yā dhanikena kṛtā, also followed by the DhKo I: 151. I have followed the latter reading.

68. See DhKo I: 144–47; Rocher (1956) VyaC, #89. I have emended the text to read mukto for yukto.

69. These two statements are inconsistent, the first with regard to place (betel nuts do not grow in the middle region of northern India) and the second with regard to time (mangoes are not in season during the autumn).

70. It is a general principle of Indian law that the accumulated interest cannot become more than the principal owed; thus principal plus interest can never become double the principal borrowed. In gifts and the like, as we have seen, proof of an earlier transaction is stronger than that of a later one.

71. The meaning of saṃniveśa in this context is unclear. Kane (1933) thinks that the reference is to boundaries of the land.

72. 1,000 Palas = about 38 kg. One would never have a pan that is so heavy.

73. The meaning is that both the plaint of the plaintiff and the plea of the defendant have been written down by the plaintiff himself.

74. I follow the edition of Srinivascharya in reading atītakālo dviṣṭhaś ca; Gharpure and the DhKo (I: 143) read atītakāloddiṣṭaś ca. Clearly, the former is the lectio difficilior, and it also agrees with the explanation given in the following verses.

75. The term pada here may also be a reference to the vyavahārapadas, the subjects of litigation, as indicated by the statement of Katyayana given below.

76. The exact meaning of the term tyakta is unclear. It could well refer to grants or donations made by the king (here a tax exemption or a donation of a building), which cannot be subjected to litigation.

77. The meaning of the term prācīna here is unclear. It can mean “ancient,” which is how Gharpure translates it. If that is the case, Devanna Bhatta is exhibiting an interesting historical viewpoint, also evident in his comments above (p. 281) on KātSm 682, where he says that Brihaspati is a more ancient author than Katyayana.

78. This is also a citation from Katyayana, even though neither Kane (1933) nor Aiyangar (1941) includes it in their edition. See DhKo I: 155.

79. See the eight modes of acquiring ownership given in GDh 10.39–42 (above Vijnanesvara on YDh 2.24).

80. The draft of the plaint was written on a surface where corrections could be easily made. The earliest method appears to have been the ground or floor spread with sand. This is the method given in the court scene of the drama Mṛcchakaṭika (act 9).

81. Devanna Bhatta here makes a distinction between lokasiddha given in the citation and the parallel expression lokaprasiddha; the two often have the same meaning.

82. The reading of this verse here is somewhat different from that of the edition of the MDh. Here we have sabhyā na (not acceptable to the court) instead of the original satyā (untrue, invalid). Further in the original, bhāṣā probably referred to an agreement rather than a plaint, a meaning that term acquired at a later date.

83. The maxim of the time at which a conch is blown is explained by Sabara in his comments on PMS 6.4.42. When there is a rule that something should be done in a village at the sound of the conch, even if it is not blown on a particular day, the activity must be carried out. The meaning here is that even though the plea has not been entered, the time for emending the plaint ends when the prescribed time for entering a plea has passed. See Jacob 1904: 84.

84. I have followed the reading of Gharpure and DhKo I: 158: samyaksūkteṣu kāryibhiḥ, in preference to the reading offered by Srinivascharya: samyaksūtraṇakāribhiḥ.

85. The term for “plea” is uttara, which has the additional meaning of what comes after or what follows. In this case, the plaint is followed by the plea; so the plaintiff can still emend his plaint while the plea is being stated. What follows the plea is the presentation of evidence (kriyā), which, then, is technically uttara; and the defendant can emend his plea while the evidence is being presented.

86. The expression śrutārthasya in the text of Yajnavalkya is a bit ambiguous, because of the multiple meanings of the term artha. One of the meanings is lawsuit, which I have used in my translation of this verse in ch. 11. Devanna Bhatta apparently takes it to mean a point or a subject at issue, and thus takes the compound to be abbreviated and gives an expanded explanation.

87. The expression ajñātārthe ca vastuni is not altogether clear. Kane (1933) translates: “if the whereabouts of the subject-matter (of the dispute) are not known,” and Gharpure, “where the particulars of the thing are not known.” The meaning, I think, is that the exact location or identity of the subject under dispute is unknown or uncertain.

88. Here I emend the reading of the editions: kāryāṇām to kāryiṇām; see the previous commentary on KātSm 147 where, in the same context, Srinivasacharya’s edition reads kāryiṇām. Gharpure has misunderstood this section.

89. I think Kane is correct in interpreting mūla in the verse to refer to the original owner or seller of the property. See BṛSm 1.12.6 and MDh 8.202.

90. Here the term “thousand” probably refers to the number or quantity, with regard to items under dispute that are subject to counting: ten cows, a hundred coins, and the like.

91. Here we deal with the ambiguities of Indian manuscript writing systems, which do not leave a space after each word, and the issue of euphonic combination (Sandhi) where the last sound of a word often coalesces with the initial sound of the following word. Thus the statement mayādeyam could be rendered as mayā deyam (if we add a space between the two words: “By me it should be given”) or as mayādeyam (= mayā adeyam: “By me it should not be given”), with the coalescing of the final ā and the initial a into ā. The plea would thus become ambiguous.

92. This verse has been subject to multiple explanations: see Kane 1933, 151. The term tāmarasa is not a common Sanskrit word for a lotus. The statement also expresses its meaning in a roundabout or evasive way, because the response is presented as a question. This method is called vakrokti, a well-known figure of speech in Sanskrit aesthetic theory.

93. As pointed out in note 91, here also the confusion is created by the way Sanskrit is written in manuscripts and by Sandhi. Here the phrases sadādeyaṃ (always to be given or not given) and mayādeyaṃ (to be given or not given by me) could be interpreted positively or negatively.

94. For the four feet of a lawsuit, see above, p. 000. When there is a plea of admission, then the court proceedings come to an end after the second foot, that is, the plea.

95. This is an elliptical sentence. Vijnanesvara (above p. 000) completes it: “it becomes combined.”

96. For the use of these strategies with political science, see ch. 11, n. 47.

97. This verse has been subject to different interpretations. Vacaspati Misra in his VyaC cites it twice (158, 182; see Rocher’s note to 158) and gives two different interpretations. The second broadly agrees with the one given here. But in the first he gives quite a different interpretation, which, I think, is more plausible. According to this, the defendant first enters a special plea giving one reason, and later produces a stronger reason. It is the latter that he must prove.

98. The reference here is to an inquiry about a case of arson. See above, n. 40.

99. For the meaning, see ch. 13, n. 25.

100. The Sanskrit technical terms are kārya (case or charge) and kriyā (evidence).

101. Gharpure’s edition omits the term bhuktitaḥ (than enjoyment); the DhKo (I: 218) follows that reading. I think Srinivasacharya’s reading ato bhuktito bahuviṣayatvam is correct; the ending “to” of ato and bhiktito may have created a haplography.

102. Devanna Bhatta devotes a very long section of his work to documents, to which he refers the reader: SmC III: 125–52.

103. For the meaning of ancient texts, see above, n. 77.

104. For these three kinds of plea in KātSm 169, see the section on Plea (p. 301). The argument attributed to “some” is actually made by Vijnanesvara above in his comments on YDh 2.20 (p. 258), where he cites this very verse of Katyayana.

105. I follow here Gharpure’s reading, aviśeṣeṇa. Srinivasacharya’s reading, api śeṣeṇa, is probably a typographical mistake.

106. The intent is to show that this is not an actual injunction. For the classification of various statements, such as explanatory statements, see ch. 7, n. 30.

107. The commentary referred to here is that of Medhatithi: adharmajñagrahaṇāc ca liṅgān niścitachalaviṣayo ’yaṃ daṇḍa ity uktam—“What is stated is that, because of the use of ‘proficient in adharma,’ by inference, this fine pertains to a person whose subterfuge has been demonstrated.” While Medhatithi focuses on the evil disposition of the litigants, Devanna Bhatta, using the term udvṛtta, appears to allude to their pride and haughtiness, unless he is using this term as an equivalent of chala. It is also worth noting that the compound adharmajña is interpreted by both as “knowing adharma” rather than “not knowing dharma.” So the litigants are not simply ignorant; they actually know illegal ways of dong things. These two kinds of negations correspond to the two types found in Indian, especially Buddhist, philosophy: prasajyapratiṣedha and paryudāsa.