Chapter XIII. Witnesses (sākshinah)

Volume III pp. 330–360

Chapter XIII. Witnesses (sākshinah)

CHAPTER XIII WITNESSES (sākṣiṇaḥ).

The word sākṣin occurs in the Śvetāśvataropaniṣad⁴⁷⁶ (VI. 11) where it is applied to the one immanent Spirit of the uni- verse as All-seeing. Pāṇini⁴⁷⁷ V. 2. 91 explains the word sākṣin as meaning ‘one who has directly seen’ Gaut⁴⁷⁸ XIII. 1, Kauṭ III. 11, Nār IV. 147 state that when two persons litigate and there is a doubt or discrepancy between the two the determination of the truth in the dispute is due to witnesses Manu VIII. 74, Sabhāparva⁴⁷⁹ 68 84, Nār. IV. 148, Viṣṇu Dh. S. 8 13, Kāṭ 346 (q in V M p 317 and V. P p 106) provide that a (proper) witness is one who has himself either seen or heard or experienced the matter in dispute. This means that the evidence of a witness must be direct and should not be what is called hearsay. Medhātithi (on Manu VIII. 74) says⁴⁸⁰ that hearsay evidence occurs when a person who hears from another that has himself heard something from another comes forward as a witness and that hearsay is no legal evidence. Manu VIII. 76 prescribes that even though a man may not have been expressly asked or appointed by the parties (in the words ‘you are the witness to this transaction’), he is a proper witness when asked as to any matter in dispute that he has himself seen or heard The only exception to hearsay evidence allowed by the Viṣṇu Dh S. VIII. 12 is that⁶ when a person that is appoin- ted as a witness is dead or gone abroad, those who have heard what he said may give evidence It has already been stated (p 297) that a king was not to delay the examination of witnesses Kāṭ. (340-341 q by Aparārka pp 675, 677, Sm C. II. p 92, V. M p. 331) prescribes that the king (or the chief judge) should himself examine the witnesses that are present (in court) and should consider along with the sabhyas the statements made by witnesses and that, where it is doubtful who the witnesses on a disputed matter are, time should be given for producing the witnesses in order to ascertain the proper means of proof but that where it is clear who the witnesses may be he should make the hearing of the case proceed at once Kāṭ (352) further prescribes⁷ that when it is impossible to bring witnesses because they reside in a foreign country, evidence taken in writing before a man learned in the three Vedas and sent by him should be read in deciding the cause⁸ Gaut. 13. 2, Manu VIII. 60, Yaj. II 69, Nār. IV. 153 and others require that ordinarily there should be at least three witnesses in a cause. Br. (S. B E 33 p 301 v. 16) says that there may be nine, seven, five, four or three witnesses, or two only if they are learned brāhmaṇas⁹ Viṣṇu Dh S VIII 5 and Br. (S. B. E vol. 33 p 301 v 16) emphasize that a single witness alone cannot be examined for deciding a matter¹⁰. But Yaj II 72, Viṣṇu Dh S VIII. 9, Nār IV 192 state that one man alone may be a witness in a cause if he is endowed with the regular performance of his religious duties (is possessed of the qualities stated in Viṣṇu Dh S VIII 8) and is accepted as a witness by both sides Vide Manu VIII 77 also Br (S B E vol 33 p 301 v 18) allows a single witness to furnish valid proof if he is a dūtaka¹¹ (messenger), an accountant, one who has accidentally witnessed the transaction or if he is the king or the chief justice Vyāsa says¹² that even a single person may be enough in a cause, especially in heinous offences called sāhasa, if he be one whose actions are pure, who knows the dharma and whose truthfulness has been well attested before Kaut (III 11) says that a single witness (whether a man or a woman) may be enough in cases where the transaction in question was effected in secret (except the king and one performing austerities)¹³ Kat (353-355 q by V M pp 319-320, Sṃ C II 76, V. P. pp. 112-113) states that even a single person may be allowed to depose as a witness if he was taken in confidence at the time of making a deposit; so also a messenger sent by a litigant for borrowing an article (such as an ornament) or the manufacturer of an article or when a matter has been decided by the king or chief judge, the scribe, the chief justice or any one of the sabhyas may alone prove the plea of res judicata or the point as to what was said by the plaintiff or the defendant in a suit¹⁴ Vide Br (S B E 33 p 300 verses 13-14).

The qualifications of witnesses to be examined in a suit are stated in numerous places such as Gaut XIII. 2, Kaut III 11, Manu VIII 62-63, Vas 16-28, Śankha-Likhita (q in S. V. p 138), Yāj II 68, Nār IV 153-154, Viṣṇu Dh S VIII 8, Br (S. B E vol 33 p 302 v 28), Kat 347 (in Sm C II p 76 and V P p 111) The principal qualifications are;¹⁵ they should be born of a good family, should have lived hereditarily in the country, should be householders having a son or sons, should be well-to-do and men of character, should be trustworthy, should know dharma and act up to it, should not be covetous, and should have been cited by the parties as witnesseṣ Some Smṛti texts such as Kaut III 11, Manu VIII 68 (= Kāt. 351 and Vaṣ 16.30), Kāt 348 provide¹⁶ that ordinarily witnesses should be of the same varṇa or caste as the parties, that women should be witnesses in disputes between women, that men of the lowest castes (antyajās) should be witnesses for litigants of those castes and that a litigant of a lower caste should not try to prove his case by citing witnesses of a higher caste or a brāhmaṇa (unless the latter is an attesting witness on a deed). But most (even including Gaut and Manu) give an option and say that persons of all castes (even śūdras) may be witnesses for all. Vide Gaut. XIII 3, Manu VIII 69, Yāj. II. 69, Nār. IV. 154, Vaṣ 16.29 (sarveṣu sama eva rā) Nār. IV 155 and Kāt (349-50 q by Aparārka p 666 and V. P. pp 111-112) provide¹⁷ that in disputes between members of groups such as those who wear peculiar symbols (indicative of their sect), śreņis (guilds), pūgas (associations), companies of traders and all others who work in groups and who are therefore called (ārūyus and in the case of slaves, herds, wrestlers, elephant riders, horse-trainers and soldiers the heads (called tarṇṇ) of these groups are the proper witnesses Gaut. XI. 21 says that in disputes among members of classes of agriculturists, traders, cowherd-, money- lenders and craftsmen (such as carpenters and washermen), the help of other members of the same calling may be taken as witnesses and arbitratorṣ

Long lists of persons who are generally to be regarded as incompetent to be witnesses are given in Kaut. III. 11, Manu VIII 64-67, Udyogaparva 35. 44-47, Yāj. II 70-71, Nār. IV. 177-187, Viṣṇu Dh. S. VIII 1-4, Bṛ (S B E vol. 33 p. 303 vv. 29-30), Kāt. 360-364 (q by Aparārka pp. 669-670, Sm C II pp 77-78, V. P p. 119) Manu VIII 118 states the general grounds why oral evidence tends to be false viz. covetousness, infatuation, fear, desire of pleasures, anger, friendship, ignorance, minority. As the list of incompetent witnesses in Nār. is the longest,⁴⁹² that alone is set out here. One who has monetary ⁴⁹³ interest in the parties or the subject matter of the suit (such as a co-sharer, creditor and debtor of the parties), a friend (or relative such as an uncle); associate (in undertakings); enemy; one who has already been held to give false evidence; one charged with (or addicted to) sins; a slave; one who is in the habit of finding the weak points of others (or harming others); one who has no faith in religion; a very old man (above 80 years); a minor; a woman; an oil-presser; one intoxicated; a lunatic; one who is extremely inattentive; one distressed; a gambler; a village priest; one who undertakes long journeys (on the great roads); one who is a merchant engaged in sea voyages; an ascetic (who has renounced the world); one sick; one bereft of a limb; one who is the only witness; a brāhmaṇa learned in the Veda; one who does not perform the customary religious rites; an impotent person; an actor; an atheist; a vrātya; one who has abandoned his wife; one who has given up agnihotra (daily offerings into śrauta or smārta fire); one that officiates as priest for persons who are not entitled to perform vedic sacrifices; one who is an associate in eating food from the same vessel in which food is cooked (i.e. who is in commensality with a party); a former enemy (arīcara as one word); a spy; an agnate, a cognate (or born of the same womb); one whose evil doings in former lives are seen clearly; a public dancer (śailūṣa, or one who makes his women actresses); one who lives by (i.e. buys or sells) poison; a snake-catcher; one who is a poisoner, an incendiary, a mean person (kīnāśa, or parsimonious person); the son of a śūdra woman (from one of a higher caste); one guilty of a minor sin (upapātaka); one fatigued; a desperado; one who has relinquished all attachments; a person who is penniless (through gambling or other extravagance); a member of the lowest (untouchable) caste; one leading a bad life; a brahmacārin who has not yet returned from his teacher's house; an idiot; an oil-seller, a seller of roots; one possessed (by a ghost or demon); one hated by the king; a weather-prophet; an astrologer; one who proclaims to the public the sins of others; one who has sold himself (for money); one who has a limb too little (e.g. having four fingers to the hand); one who lives on the immorality of his wife; one who has bad nails; one who has black teeth; one who has betrayed his friend; a rogue; a seller of liquor; a juggler, an avaricious man; a ferocious man; an opponent of a śreṇi (guild) or gaṇa (association); a butcher; a hide-worker; a cripple; one excommunicated for a grave sin (like brāhmaṇa-murder); a forger (of documents or coins or weights &c); one who employs incantations and drugs for influencing others: one who is an apostate⁴⁹⁴ from the order of ascetics (pratyavasta); a robber; a servant of the king, a brāhmaṇa who sells human beings, cattle, meat, bones, honey, milk, water or clarified butter; a member of the three higher castes who engages in usury; one who has given up the peculiar duties of his caste (or station); a lūlikā; a bard; one who is the servant of a low person; one who has quarrelled with his father, one who causes dissensions Kant III 11, Manu VIII 65, Viṣṇu Dh S. VIII, 1 and several others say that the king cannot be cited as a witness⁴⁹⁵ (except possibly to prove the plea of res judicata or to prove what transpired before him when a litigation was going on )

The above is a formidable list of incompetent witnesses, therefore, most Smṛtis such as Gaut XIII 9, Kant III 11, Manu VIII 72, Yaj II 72, Nār IV 188-189, Viṣṇu Dh S III 6, Usanas (q. in Sm C II p 79), Kāt 365-366 (q in Sm C II p 79) expressly point out that strict examination of the character of witnesses should be entered upon in disputes about debts and the like which are of a fixed nature (of a civil nature), but that in complaints about all kinds of sāhasa (heinous crime), in robbery, adultery or the two kinds of pāruṣya (viz. defamation and assault) there should be no (strict) inquiry into the character of witnesses (that is, all can be witnesses in such cases) and that even those like slaves and fault-finders who have been enumerated among incompetent witnesses can become witnesses in grave matters²². The idea is that in such cases the witnesses need not be strictly examined as to their possessing the good qualities ordinarily required in witnesses But this does not authorize the examination of such persons as lunatics or idiots as witnesses, because they have not the power to understand things Manu VIII 77 emphatically states that even a single male, if he is free from greed, may be (sufficient as) a witness, but not women of good character, even though they be many, since the feminine intellect is apt to be unsteady But even Manu had to concede (VIII 70) that a woman or a minor or a very old person or a pupil, relative, slave or a hired servant may be a (competent) witness, if it is impossible to find another witness (in matters that occurred inside a house or in a forest or in crimes where life is lost). Kāt 367 (q by Aparārka p. 671) says the same thing. Uśanas²³ (quoted in the Sṃ C II p 79, V. P. p. 120) provides that even a slave, a blind man, a deaf person, a leper, a woman, a minor, a very old man may be (competent) witnesses in sāhasa, provided they are not interested in the matter of dispute (or are not related or partial to the parties). Nār. (IV. 190–191) gives it as his view that even when there is to be no strict inquiry into the character of witnesses in sāhasa, still a minor, a woman, a single person, one who forges deeds (or who is a cheat ), a relative and an enemy cannot be witnesses in sāhasa, since a child through ignorance, a. woman from her habit of untruthfulness, a forger ( or cheat ) because he is accustomed to doing evil deeds, a relative from affection and an enemy from the idea of taking revenge may speak falsely There was no total disqualification of women as witnesses, since Manu quoted above allows them to be witnesses in disputes between women and in cases where it is not possible to have other witnesseṣ Medhā- tithi on Manu VIII. 68 holds that women are disqualified as witnesses only where the plaintiff and defendant are both males but that where there is a litigation between a man and a woman or between women alone, a woman may be a competent witnesṣ Ancient Hindu Law is not singular in its partial rejection of the testimony of womeṇ Vide Best on Evidence ( 12th eḍ of 1922) pp 53-56 where it is shown how ancient Roman Law and the Medieval laws of several countries of Europe more or less excluded the testimony of women, how some States in the U S A. excluded the testimony of negroes and persons of coloured blooḍ By the Canon Law more credit was given to male witnesses ( Taylor on Evidence, eḍ of 1848 p. 655, note d ).

Nār. ( IV. 157-172 ) states that incompetent witnesses fall into five classes : (1) some²⁴ like learned brāhmaṇas, very old men, ascetics, persons practising austerities are incompetent because ancient texts ( vacana ) say so and there is no other ( special ) reason for their exclusioṇ The Vyavahāratattva (p. 214) points out that śrotriyas and others cannot be appointed as witnesses, but they may be what is called akṛta witnesses (i e. they may if they choose appear as witnesses in a cause). They are not competent like the king, not because they are untrustworthy but because it is not advisable to call theṃ They were privileged persons; (2) thieves, robbers, dangerous characters, gamblers, assassins are declared to be incompetent because there is the defect (doṣa) of untruthfulness in them; (3) witnesses are rendered incompetent on the ground of contradiction (bheda), when the statement of witnesses who have been accepted by the king for determination of the same matter do not agree; (4) he who without having been appointed (by the parties to a transaction) as a witness (to it) comes of his own accord to depose is termed a sūci (i. e. a spy) in the śāstras (and so was styled suyamukhi by Nār. IV. 157) and is unworthy to be a witness; (5) a person is incompetent as a witness by reason of intervening death (i e. he is mṛtāntara); when can any person bear testimony if the creditor (or claimant) is no longer alive, whose claim should have been heard by him specifically (but was not heard in that way)? Mṛtāntara is a witness whom one or other of the parties informed in a general way that there was some dealing between the parties and who was asked to be a witness (but was not told the details of the dealing) and then subsequently the party dieḍ In such a case the person told is not a good witness, as he cannot give the detailṣ But Nār. (IV. 94) mentions an exception to this last rule viz when a father while on the point of death tells his sons and the like 'these are the witnesses on such and such a matter,' there even after the father's death those witnesses would be competent though death interveneṣ

Witnesses are divided by Nār. IV. 149 into two sorts; (1) those appointed (kṛta) by the parties and (2) those not appointed (akṛta); the first being of five kinds and the last of six kindṣ The first five (acc to Nār IV 150) are a subscribing witness (likhita), one who has been reminded (smṛta) i e. one who is called as a witness without there being a document and who is present at a transaction and is again and again reminded of it by a party to it in order that that transaction may be effectively proved thereafter (Kāt. 371 and 372 q. in Mit on Yāj. II 68), a casual witness i e one, who, while a transaction is being entered into, comes there by mere chance and is then asked to be a witness (yadṛcchābhijñā or yādṛcchika, acc. to Br. ), a secret witness i. e. one who is made to listen the speech of the debtor while concealed behind a wall screen, an indirect ( uttara )²⁵ witness (i. e. one who learn from a witness who has seen or heard of a transaction when th to a distant country or is on the point of death, The six akṛta witnesses (acc. to Nār IV. 151–152) are co-villager (in disputes about boundaries), the chief judge, the king (before whom a suit was tried), one who is closely acquainted with the transaction of the parties (i. e kāryamadhyagata, acc. Br. ), one who is deputed by a party (for borrowing an ornamen or for settling a transaction, called dūtaka by Br. ), members of the family (as regards partition &c ) Br. mentions twelve kinds of witnesses that are practically the same as the eleven of Nār except one which he adds viz. lekhtā²⁶ (one whose name is caused to be written by a party in the presence of the witness when that party enters into some transaction in writing such as a loan with another) The only difference between ‘likhita’ and ‘lekhita’ seems to be that the first himself writes his own name as a witness on a document while the latter’s name is written as a witness by one party in the presence of the other and of the witness Br (S. B E. vol. 33 p. 299 verses 4-15 q. by Aparārka pp 666-667) defines at great length all the twelve and Kāt also defines most of them, but these definitions are passed over as not very important. Viśvarūpa remarks that these nice distinctions among eleven kinds of witnesses are made only for the sake of exposition and for the benefit of simple people²⁷. Before a witness for a party begins to depose, it is the duty of the opposite party to point out the ground or grounds, if any, of the incompetence of the witness put forward for examinatioṇ Kat. adds²⁸ that the latent defects of the witnesses of a party must be pointed out by the opposite party, but the patent defects will be considered by the members of the court at the time of giving their decisioṇ Vyāsa²⁹ states ‘the defects of a party should be pointed out by the opponent in open court by putting them down in writing and the witnesses should be called upon (by the judge) to refute them; if the witnesses admit (the faults pointed out) they do not deserve to be witnesses; if otherwise (if the faults are not admitted) they should be established by the opponent with evidence (other than those of other witnesses); for if the defects of a party were allowed to be established by the opponent, there would result the fault of a never-ending series (anavasthā), since the first party would then try to show defects in the latter set of witnesses and this may have to be carried on ad infinitum.’ Br. laid down that an opponent should not be allowed to point out the grounds of incompetency in witnesses after they have begun to depose and Kāt. adds³⁰ that he (that party), who points out, after the witnesses have deposed to a matter, faults in them in whom he found none at first and who can not set out a proper reason (for his not proclaiming the faults at first) should be fined in the first amercement. Br. further states³¹ that the defendant may point out the defects, if they exist, in the witnesses cited by the plaintiff; but if he finds faults which do not exist in the witnesses, he should be punished with a fine equal to the amount claimed (in monetary disputes) or equal to that which is levied from a false witnesṣ If the witness of a party is not able to refute the defect pointed out by the opponent, the party calling him should try to clear his witness of the fault; otherwise he should not succeed in his suit. When a witness is about³² to depose, he was, acc. to Br. (S. B. E 33 p 302 v. 23), to put off his shoes and turban, raise his right hand (in order to draw the attention of the Court) and after touching gold, cowdung or darbha grass was to depose to the truth Ap Dh S. II. 11 29. 7 says³³ that a witness should depose to the truth on an auspicious day in the morning in the presived over by the king (or chief justice) and after being exhorted as regards both (viz. the results of truthfulness and falsehood) and when accepted by all (the court and parties) as a competent witnesṣ Kaut also (in III. 11) says that witnesses should be examined in the presence of brāhmaṇas, waterjar and fire. Manu (VIII 79-80) provides that the judge should in a conciliatory manner exhort all witnesses assembled inside the court in the presence of the plaintiff and defendant as follows: ‘whatever acts of these two (litigants) between themselves you know as regards this matter (in dispute), narrate them all truthfully, since you are the witnesses in this case’. Yāj. II. 73 also states that witnesses should be made to depose in the presence of the parties and Gaut. XIII. 13 and Kāt (q by Mit on Yāj II. 73) prescribe that witnesses³⁴ should face the east or north, are to be examined in the presence of the images of gods and brāhmaṇas in the first half of the day and should be exhorted to tell the truth by being bound with oathṣ This examination in the public and after an oath went a long way in checking the tendency to depose falsely. The oath consisted of two parts, viz. (1) the requirement to tell the truth and (2) the exhortatory and imprecatory part. Both were administered by the presiding judge. Gaut. (XIII. 12-13) appears to prescribe that an oath (śapatha) was not to be administered to brāhmaṇa witness, but Manu VIII. 113 (= Nār. IV. 199) does not make this distinctioṇ Gaut. (XIII. 14-23), Manu VIII. 81-86, 89-101, Viṣṇu Dh. S. VIII. 24-37, Nār. IV. 201-228 contain very long exhortations addressed by the judge to the witnesses relating to the importance and high worth of truth, stating how the conscience of a man pricks him, what rewards await the truthful witness here and in the next world and what sin and terrible torments in Hell are the lot of an untruthful witness, what evil befalls even the deceased ancestors of an untruthful witness and how he is liable to be punished by the king. They are too long to be cited here³⁵. Shorter exhorta- tions are found in Yāj. II. 73–75, Vās 16. 32–34, Bauḍ Dh. S. I. 10. 19. 9–12 (S. B. E. vol. XIV p. 203), Br. (S. B. E. vol. 33 p. 301 verses 19–22), Kāt. 343. Nār. IV. 200 provides ‘let him (the judge) inspire them (the witnesses) with awe by (quoting) ancient texts, extolling the greatness of truth and denouncing falsehood’. Kaut. (III. 11), Manu VIII. 88 and 113 (=Nār. IV 199), Viṣṇu Dh. S. VIII. 20–23 add further modes of speech and exhortations according to the varṇa of the witness: A brāhmaṇa witness is to be addressed by the judge as ‘speak’ and to be exhorted to swear by his veracity, a kṣatriya one was to be addressed as ‘speak the truth’ and to be sworn by the animal he rides and his weapons (which he is to touch), a vaiśya is to be sworn by his kine, grain and gold and a śūdra is to be sworn by (imprecating on his head) all grave sinṣ Manu VIII. 88 provides that a brāhmaṇa should be asked to depose with the word ‘speak’, while VIII. 113 states that a brāhmaṇa is to be made to swear by his truth Some explain that the first rule applies to excellent brāhmaṇas and the second to inferior ones and rely on Gaut (XIII 12–13) who prescribes that an oath to speak the truth is not to be administered to brāhmaṇas but only to otherṣ Vide Haradatta on Gaut and Vivādacandra p 127 Manu VIII 102 (=Baud Dh. S I. 5 95 ) is careful to point out that the exhortation is not to depend purely on the caste but that brāhmaṇas ( and others ) who tend kine, who engage in trade, who are craftsmen or actors, who are menial servants and usurers should be treated (by the judge) as śūdras³⁷ (in the matter of taking oaths) Witnesses after they are assembled before the Court are to be ordinarily examined separately (vide Kāt quoted in ṇ 508 above ) But Gaut XIII 5 and Kāt 394 provide ³⁸ that if a certain matter is seen by the witnesses as a body then they may be examined together. Manu VIII 78 and Kāt 392 prescribe that what the witnesses declare quite naturally (without hesitation and fear) should be taken down and that when the witnesses have declared (what they know) they should not be questioned again and again by the king (or judge). From the Smṛtis and the 9th Act of the Mṛcchakaṭika it appears that it was the chief judge or the judges who put questions and that there was no elaborate cross-examination and re-examina- tion of witnesses as in England (which practice is followed in modern India). The only cross examination that appears to have been allowed was about the faults or grounds that made a person an incompetent witnesṣ In this respect ancient Indian practice resembled modern judicial practice in some European³⁹ countries other than Englanḍ Witnesses were compelled to attenḍ Kaut. III. 11 states that witnesses who are not far removed by time or space should be produced by the parties themselves, but those who are far away or who are unwilling to stir shall be got summoned by order of the king⁴⁰. Manu VIII. 107, Yāj. II. 77, Br., Kat and Viṣṇu Dh. S. VIII 37 say that if a witness knowing all facts and not prevented from coming by disease or similar causes refuses to come as a witness he incurs the sin of false witness, has to pay the amount in dispute and ten per cent thereof as fine to the king. This is in accor- dance with the principle enforced in modern times that it is the duty of a citizen to appear before a court of justice when summoneḍ Kaut (III 1) provides for subsistence allowance to witnesseṣ It is not quite clear whether parties could be regarded under ancient Hindu Law as competent witnesses on their own behalf. Yāj (IL 13-15) appears to suggest that parties could be regarded as witnesses in their own cause and held to be false from their demeanour as in the case of witnesseṣ Kaut. IV.⁴¹ 8 and the 9th Act of the Mṛcchakatika show that the accused was questioned by the Court as to his movements about the time of the alleged offence and till his arrest. Śukra ⁴² IV. 5. 184 defines sāksi in such a way as to exclude the litigant himself. Ordinarily witnesses were examined in the open court in the presence of the parties and never behind their back, but Kāṭ (387-389) prescribes that in the case of immovable property oral evidence may be taken on the property itself and in some cases even elsewhere than these two (viz the Court and immovable property), that is, in the case of the death of living beings witnesses may be made to depose before the corpse (of the animal killed) or, in the absence of the body, before ⁴³ some mark (such as the horn of an animal) Br. and Manu VIII 25 state that the truth of the statements of witnesses should be examined by marking their tone, their change of colour, their eyes, their gestures, and their demeanour ⁴⁴ Texts like Śaṅkha- Likhita (q by V. P. p 124 ), Nar IV 193-196, Viṣṇu Dh S VIII. 18, Yāj. II 13-15 and Kāṭ 386 point out the actions and appearances of a party or witness that deposes falsely, viz he appears restless, changes his place (i e moves from one spot to another), licks the corners of his lips, his forehead perspires, his face loses colour, he frequently coughs and heaves frequent sighs, he scratches the ground with his feet (toes), waves his hands and garment, his mouth becomes dry and his voice falters, he speaks incoherently, talks too much though not asked and does not reply straight to the questions asked and avoids meeting the eyes of the questioner Such a witness may be looked upon as untrue and the king or judge should bring him under discipline (so that he will be afraid to lie) Merely on seeing these signs a witness was not to be punished or entirely disbelieved, since these signs raise only a probability of untruthfulness (as remarked by Mit. on Yāj II 15 and V P p 124) When there are many witnesses but they differ in their statements certain rules are laid down by Manu VIII. 73 (=Viṣṇu Dh S. VIII. 39), Yāj. II. 78, Nār. IV. 229, Br. (S. B. E. vol. 33 p 303 v. 35 ), Kāt. ( 408 ) They are briefly these:⁴⁵ The statement of the majority of witnesses should be accepted, but if they are equally divided then the statement of those who are more pure (in character or more disinterested) should be accepted and if the meritorious are divided equally then those who are the best among them should be accepted as true. The Mit on Yāj. II. 78 adds that if there is a conflict of testimony between the statements of a few but eminently qualified witnesses on the one hand and those of a larger number of ordinary witnesses, then the former should be accepted as Yāj II. 72 requires i. e it prefers quality to mere numberṣ Kaut. III. 11 prescribes that when there is conflict in the oral evidence the decision should be according to the evidence of the majority of witnesses or of those who are pure in character or of those who are approved of by both sides as good witnesses or a mean should be drawn from their statementṣ Nār. IV 160 and Kāt. 359 state⁴⁶ that if one out of the (three) witnesses attesting a document or out of ( three ) witnesses that are cited by a party deposes in a way contradictory to what is deposed to by the other two, then all the three become incompetent witnesses on account of contradictioṇ This has been interpreted by the Vyavahāra mātrkā (p. 326) to mean that where all witnesses are of equal status and there is a majority of only one favouring one side, then all the witnesses are incompetent and oral evidence is in that case futile for decisioṇ

The important question is how much has to be proved by the party calling witnesses Yāj. II. 79 (=Viṣṇu Dh S VIII. 38), Nār II. 27 and Br ( q by Sm C. II. 91 ) state the general rule that that party whose averments are supported in their entirety as true by the witnesses becomes successful, while that litigant. whose claim is declared by them to be false certainly loses ⁴⁷. Nār. IV. 232 and 234, Kāt. 396, 398 lay down ⁴⁸ the rule that in civil disputes that are of a non-urgent nature (like recovery of debts), if witnesses depose to more or less (than what is averred in the plaint), then the claim is not established in its entirety, whereas if the witnesses depose to more or less (than the claim affirmed by the party) the deposition of the witnesses should be considered as not taken down or it should be omitted (from consideration) ⁴⁹, the witnesses in such a case are not liable to fine; they may deserve to be fined if they do not depose (to what they know). Similarly Nār. IV 233 and Kāt ⁵⁰ 399 say that where the depositions of witnesses are in conflict as to the place, the time, the property, the amount, the colour, the species (or caste), the form (or shape), the age (as stated in the pleading of a party), the depositions are to be regarded as good as not given at all. The rule amounts to this that any discrepancy between the essential details or particulars of a claim made in the plaint ( such as the time, place, number &c. ) and the case made out in the depositions as to any of these essential particulars will have the same consequence as tendering no oral evidence. This rule was not followed to the letter and besides other means of proof could be resorted to, as stated by the Mit., the Vivādacandra and others⁵¹. Kaut. III. 11⁵² states that when the witnesses depose to less than what is averred in the plaint, the plaintiff has to pay a fine ( of a fifth part ) on the excess and when they depose to more than the claim, the excess (held proved on the depositions) may be taken by the king.

In charges of adultery, heinous crimes ( sāhasa ) and theft the whole of the matter that is alleged may be held proved, when witnesses depose to only a part of the matter⁵³ (says Kat. 397).

Nārada IV. 165 prescribes that a litigant⁵⁴ should not secretly approach a witness cited by his opponent, nor should he try to win him over (by bribes or threats) through another; if he does so, he would suffer the consequences of a losing party ( hīna ).

There were rules about the late production of evidence. It has been already shown how if a litigant adduces weak evidence, though strong evidence is available, he would not be allowed to rely on the latter after judgment is pronounceḍ It is said by Nār. L 62 that when a law-suit has proceeded far enough (lit. has been cleansed or thrashed out), evidence such as a document or witnesses, would then become useless, if that evidence had not been announced before⁵⁵ After a reply is filed by the defendant, the plaintiff has to give a list of his evidence i e documents or witnesses (Yāj II 7) The meaning briefly is as follows : If he does not do so, proceeds with the trial and does not call all necessary witnesses or produce all documents and the case is almost finished, but judgment is not pronounced, even then he would not be allowed to produce fresh evidence at that late stage For if fresh evidence not previously disclosed were allowed to be given at that late stage, it would cause surprise to the defendant, who would then demand time for adducing evidence in rebuttal and then the plaintiff may again cite further evidence and this process is in danger of being carried on ad infinitum. If however witnesses were already cited but all were not examined (e g. if ten witnesses were mentioned and only three examined) and the plaintiff found that the evidence of the few witnesses examined for him is contradictory or hostile, then he may be allowed to produce even at a late stage the remaining witnesses (who may be seven i e. double the number of those already examined or who may be more respectable or purer in character). This proposition is contained in Yāj II 80 on which the mitākṣarā remarks that even if witnesses are not mentioned in a list, but before judgment is pronounced other more reliable or respectable witnesses or double in number (and of the same status as) of those examined are available, they should be examined by the court even at a late stage The rule is that as long as witnesses are available resort to ordeals should not be had Yāj. II 80 also has given rise to various interpretations Vide Mit and Aparārka on Yāj II 80, Sm C II p 94, V. P p. 130–134 The Mit thinks that the verse applies where the plaintiff or the party on whom lies the burden of proving a positive proposition finds that the witnesses produced by him are hostile or contradictory and that in such a case he may be allowed to cite other more respectable or more numerous witnesses to prove his proposition and to show that the first set of witnesses is false, while Aparārka holds that the verse says that when the party who has to prove his III ] Late production of witnesses

case cites witnesses which the opponent feels are false, the opponent may produce more respectable or more numerous witnesses to depose to the contrary and he would thus prove the first set to be false Aparārka relies on Kāṭ. (408) for this proposition⁵⁶. It may be said that Aparārka's interpretation appears more natural. V. P. (p 134) accepts Kātyāyana, but says that the interpretation of the Mit. (of Yāj. II. 80) is correct. Whatever view is accepted it leads to this that one set of witnesses not only may prove the truth of a party's case but also establish that another set of that party's witnesses are guilty of perjury.

Yāj. II. 82 provides that if a witness, having promised with other witnesses to depose to a state of things, denies at the time of his examination that he saw the matter in dispute, he should be fined eight times the amount that a defeated litigant would have to pay and a brāhmaṇa witness similarly guilty if unable to pay may be banished (or his house should be razed to the ground). Nār. IV. 197 says that such a person is worse than a false witness Manu VIII, 107, Yāj II. 76, and Kāṭ. (405) prescribe that if a person who has witnessed a transaction does not depose (i. e. remains silent) about it, though not suffering from disease or some other calamity, he should be made (after waiting for three fortnights) to pay the amount of the debt or other matter in dispute and also one-tenth thereof as fine to the king.

After the witnesses have deposed the chief justice and sabhyas have to consider the depositions of witnesses (as Kāṭ. 340 provides)⁵⁷ The Court has to find out which set of witnesses are to be believed and what witnesses are perjureḍ Perjury entailed, according to dharmaśāstra works, both secular and other-worldly consequenceṣ Āp. Dh. S. (II. II. 29. 8–9) states: ‘When a witness deposes falsely the king should fine him and, in addition, the witness incurs hell after death’ and Gaut. provides (XIII. 7 and 23), ‘a witness telling the truth goes to heaven, but falls into hell if he does the reverse and when he deposes falsely he is to be censured and fined’. Manu VIII. 118 states that witnesses depose falsely through covetousness, wrong ideas, fear, friendship, sexual passion, wrath, ignorance and minority and prescribes (VIII 120-122) varying fines for perjury arising from these several causeṣ Br. (S. B. E. vol 33 p. 301 v. 21) regards⁵⁸ a corrupt judge, a perjured witness and the murderer of a brāhmaṇa as equally sinful. Yāj. II. 81 and Kāt. (407) prescribe⁵⁹ that the litigant who procures perjured evidence and the witnesses guilty of perjury should each be separately fined double the amount of the fine prescribed for defeat in various disputes and a brāhmaṇa (guilty of doing the same) should be banished from the country⁶⁰ (but not fined). But when a person (of any of the three castes other than brāhmaṇa) is frequently guilty of perjury he should be fined and given physical punishment (as cutting the tongue or death) and a similarly guilty brāhmaṇa should be fined and banished (or degraded by being carried naked through the streets or deprived of his house) The Mit. on Yāj. II. 81 says that when Manu prescribes (VIII 380) that a brāhmaṇa guilty of a sin is not to be awarded death or corporal punishment and that he is to be banished with all his wealth, he refers to the brāhmaṇa being a first offender and not a confirmed one. Manu VIII 108 says 'where within seven days from deposing in a cause a (serious or painful) disease or fire or the death of a relative befalls a witness, he should be (being held as a false witness owing to these portentous happenings) made to pay the debt in dispute to the opposite side and a fine to the king' The Sṃ C. (II p 94) explains that this applies to a witness who is the only one on a disputed matter such as a messenger and the king or judge should wait for seven days to see whether any of these calamities befalls that witness and then hold him to be perjureḍ Kāt. (410) contains⁶¹ a similar rule, but he states that the king should wait one, two or three weeks or up to 45 days according to the value of the matter in dispute and the caste of the deponent Manu VIII 117 (=Viṣṇu Dh S VIII 40) prescribes that in III] whatever lawsuit it is established that witnesses perjured themselves, the judge should hear it again (from the beginning) or if the suit has ended he should set aside the decision and examine it agaiṇ

Gaut 13 24-25, Vas 16. 36, Manu VIII. 104, Yaj. II. 83, Viṣṇu Dh. S. VIII 15 prescribe that a witness may depose falsely in a cause when, if he were to speak the truth, a person of the four varnas is liable to suffer the sentence of death. Manu VIII. 105-106, Yaj II. 83, Viṣṇu Dh S. VIII 16 prescribe that as a penance for the falsehood a witness of the three higher castes may offer three oblations of boiled rice to the goddess of speech (Sarasvati) or he may offer into fire clarified butter with the mantras called Kūsmānda (Vaj. S. XX. 14-16 or Tai Ā. X. 3-5), or with the three verses beginning with ‘yaddevā devahedanam’ or with the verse sacred to Varuna (viz Rg. I. 24 15, ‘uduttamam varuna’) or with three verses of which ‘waters’ are the deity (Rg. X. 9. 1-3 ‘āpo hi sthā’). Viṣṇu Dh S. VIII. 17 adds that a śūdra guilty of perjury in the above case may by way of expiation feed ten cows for one day. It is strange that the dharmaśāstra works should permit perjury to save a culprit from being sentenced to death. Probably popular sentiment of old standing allowing the speak- ing of falsehood in certain cases was responsible for this In the Mahābhārata frequent reference is made to the permission to speak untruth in ordinary life when one’s own life or another’s life is in imminent danger Vide Śānti⁶² 45. 35, 109. 19. It is stated in Śānti 165. 30 that speaking falsely is not a sin in five cases viz in jest, to women, at the time of (i. e. in bringing about) a marriage, for the preservation of great wealth and for one’s life Vaṣ 16. 36 mentions those five occasions somewhat differently. Vide Manu VIII. 112 also for a similar rule. But so old a sage as Gaut (23 29) implies that he disapproves of this sentiment⁶³, when he states, 'according to some falsehood entails no fault (i. e. no expiation need be performed) in (bringing about or breaking promises made at) marriage, in sexual intercourse, in jest and in giving relief to one distressed'. Vide Gr. R. pp. 507-508 quoting Hārita Dharmasūtra on the same subject and mentioning four kinds of falsehood such as that by a witness, that in transactions of sale &c.

Nār. (IV. 235-236) states that when through⁶⁴ the care- lessness of the creditor (who dies) there is no document nor witness to prove a debt and the defendant denies it, then one of three methods (of proof) may be resorted to, viz. codana pratkālam (dunning the debtor each time to repay), yuktileśa (putting forth arguments) and śapatha (special oaths and ordeals)⁶⁵. Kāt. (233 q. by Sm C. II p 52 and Par M. III. 91) has a similar verse Yukti according to Nārada IV. 238 means⁶⁶ 'the creditor should follow (the debtor) by arguments and by himself remembering and by reminding the debtor of the time, the place, the relation (subsisting between the two)'. Others interpret yukti differently, that is, it means 'logical reasoning'. Kāt (214) defines⁶⁷ 'yukti' as 'the ascertainment or know- ledge of a linga' (i. e. a sign that leads on to an inference). It is stated by Bṛhaspati that anumāna (inference⁶⁸ or pre- sumption), as a means of deciding a law suit, is of three kinds, but it is inferior to witnesses (vide note 433 above). Vyāsa (q. by Sṃ C. II. p. 95) states that anumāna is the same as hetu and tarka V. P. (p 167) remarks that (long continued) possession and the frequent urging by the creditor both lead to the inference of title and the taking of a loan respectively and are therefore to be comprehended under yukti according to the definition of Kātyāyana Yet by popular usage yukti is res- tricted to presumptions arising from certain circumstances on the analogy of the maxim of 'gobalivarda' explained below under 'steya.' Yukti therefore means circumstantial evidence from which an inference may be drawn about a fact in issue in a judicial poceeding. Āp. Dh.⁶⁹ S. II. 11. 29 6 says that in case of doubt (judges) should decide from signs (i e by inference) or divine proof (ordeals). Vaṣ 19. 39 mentions the view of some sages that one who is found armed or wounded or in possession of the booty (stolen) may be declared to be (the thief or offender). Manu IX. 270 (= Matsya 227. 166) requires the just king not to condemn to death one accused of theft unless his guilt is proved as certain by proof that the accused had in his possession the articles stolen and tools for theft (for house-breaking). ‘An adulterer is proved⁷⁰ to be so by being caught in such acts as playing with the hair of another's wife, an incendiary by being caught with a firebrand in his hand (near the house set on fire), a murderer by being found (near the murdered man) armed and a thief when caught with some of the articles stolen in his possession’—say Śaṅkha-Likhita⁷¹. Kaut. ⁷² IV. 12 and Yāj. II. 283 similarly say that adultery may be inferred from ( the man and woman ) being caught in the act of dalliance with the hair or from the signs left (such as marks of nails or teeth on the lips, cheak &c ) after carnal enjoyment or from the admission of the woman concerned (or of both) Nār. (IV. 172-175) declares that there are six kinds of disputes in which the indications (or circumstances) themselves serve as witnesses and lead to a conclusion without the testimony of witnesses viz one found with a fire-brand in his hand (near the scene of arson) may be inferred to be the incendiary, one armed ⁷³ (found near the scene of murder) as the murderer, one caught dallying with the hair of another's wife as an adulterer, one with a spade in his hand near a breached embankment as the person who caused the breach in it and one armed with an axe as the feller of a tree, a person guilty of an assault may be inferred to be so from visible signs (such as his club or sword being smeared with blood) But Nār. IV 176 gives the warning that in such cases it is necessary to be careful in arriving at a conclusion, since a person in order to bring into trouble another whom he hates may create marks of injury on himself Kāt (337-338) further provides ⁷⁴ that where one litigant proves as against his opponent (such as a debtor) either the offer of a bribe, or the effacing of the means of recognition (such as effacing one's signature on a document) or holding out temptations (to the witnesses or sabhyas) or the concealment of his wealth (to avoid its being attached in execution of a decree), there the original claim (such as a debt) may be presumed to be proved even if he (the opponent) denies his liability. Judges often repeat from the Bench the words ‘witnesses may lie, but circumstances cannot’. But this so-called maxim is often dangerouṣ Circumstantial evidence leads to conclusions that are in several instances false. The ancient Hindu lawyers were quite aware of the dangers of drawing inferences from circumstantial evidence. Nārada’s warning has already been mentioned (p. 356). Kaut.⁷⁵ (in IV 8) remarks: ‘even one not a thief may by chance be on the way by which thieves pass and one meeting thieves is seen to be apprehended as a thief because he is mixed up in dress, arms and articles carried with the thieves or is found near the articles stolen, just as Māṇḍavya though not a thief declared himself to be a thief for fear of being subjected to torture; therefore (the king) should punish one only after thorough examination’. Māṇḍavya’s is a leading case on the danger of relying on purely circumstantial evidence. Bṛhaspati⁷⁶ also remarks that ‘a decision should not be arrived at merely by relying on the words of texts; for, if judges come to a conclusion without applying careful reasoning, loss of dharma results; a thief is held to be not a thief and a good man is held to be a wicked one in a judicial proceeding (not arrived at with proper reasoning). The sage Māṇḍavya was held to be a thief on account of the decision being arrived at without proper reasoning’⁷⁷.

The 9th Act of the Mṛcchakaṭika is a standing literary condemnation of conclusions about guilt drawn from circum- stantial evidence and of how judicial procedure is vitiated ( of vyavahāra-duṣṭatā, as the author says in Act I)

Nār IV. 239 provides that when even circumstantial evi- dence and presumptions therefrom do not enable the court to arrive at a decision, the judge should press the litigant with śapathas (oaths and ordeals) according to the place, time and strength of the litigant, such as fire, water, his spiritual merit and otherṣ Manu VIII. 109 also says that when there are no witnesses the truth would have to be found out by śapathaṣ Divine proof is called daivīkriyā or samayakriyā (Viṣṇu Dh. S. IX. 1). Some writers make a distinction between śapathas (special oaths) and ordeals (divyas), others (like Manu VIII 109-114, Nār. IV. 239) do not. For example, the Vivādacandra quotes Nār IV. 247 ‘when there is no witness in ‘a dispute between litigants, there the judge should decide by means of ordeals⁷⁸ and various śapathas’. Acc. to the Mit. (on Yāj. II. 96) and S V (p. 106) both special oaths and ordeals are divine proof (divya pramāņa), but in disputes of small value śapathas (oaths) were generally employed, while ordeals were resorted to only in serious disputes or crimeṣ The Mit (on Yāj. II 96), the V May p 46 and V P (p 170) state that ordeals are those that generally end in immediate decision of the dispute, while śapathas take time for decision (because after the oath the king has to wait for a week or more to see whether a calamity befalls the person taking the special oath). Vyāsa quoted by Sm C II. p. 96 employs the word śapathas for both special oaths (taken as to one’s ‘satya’) and ordeals like balance. Śankha-Likhita state that divine proofs are the balance, eating poison, entering fire, holding a piece of (hot) iron, offering one’s merit acquired by sacrifices and charity and the king should make him undergo other sapathas⁷⁹. Bṛhaspati says that where documents or witnesses do not lead to a certain decision or where inference from circumstantial evidence also is confusing, there the matter is to be decided by divine proof⁸⁰.

Special oaths were taken not only in judicial proceedings, but also in ordinary dealings to clinch one’s assertion or to clear one’s character or reputatioṇ Nār (IV. 243-244) refers to The special oaths taken for clearing themselves by Vasiṣṭha when he was suspected to be a yātudhāna (a demon or sorcerer) and by the seven sages together with Indra, who (sages) were suspected to have stolen lotus fibres (by each among them ⁸¹). Manu VIII. 110 also speaks of the śapatha taken before Sudās, son of Pijavana (when the sage was charged by Viśvāmitra with the offence of devouring his own hundred sons). Nārada (IV. 243) refers to Rg. VII. 104 15–16 in the first of which Vasiṣṭha made the imprecation,⁸² ‘May I die this very day if I am a yātudhāna or if I injured the life of a human being May he who falsely addressed me as yātudhāna be deprived of his ten valiant sons’⁸³ Manu VIII. 110 also refers to the same hymn⁸³. Manu VIII. 113 (= Nāṛ. IV. 199) referred to above (p 343) states the various oaths to be taken or administered to the parties according to their respective casteṣ Manu VIII. 114 further provides for special oaths for litigants by touching the heads of their sons and wiveṣ That sapatha taken by satya was common in very ancient times appears from Pāṇ V 4 66 (satyād-asapāthe)⁸⁴ Nāṛ. IV. 249 prescribes that in charges of grave offences ordeals may be resorted to and śapathas in matters of small moment Nāṛ. IV. 218 mentions ‘truth, horses and arms, kine, grains and gold, the feet of (the images of) gods and the ancestors, acts of charity and good deeds (sukṛta) are to be employed in śapathas as declared by Manu º Bṛhaspati⁸⁵ enumerates the special oaths mentioned by Manu and Nār. and adds that they are to be employed in small matters (civil and criminal). The Viṣṇu Dh. S. (IX. 5-10) prescribes that when the litigant is a śūdra and the dispute relates to a matter worth less than one, two, three, four or five krsnalas he should swear respectively by holding in his hand dūrvā grass, sesame, silver, gold or earth taken from ploughed land, that (IX. 11-12) beyond that value various ordeals may he offered and that oaths with similar objects in the hands may be offered to vaisyas, kṣatriyas and brāhmaṇas when the value of the subject matter is double, three times or four times of the value in the case of a śūdra. Manu VIII. 111 gives the admonition that a wise man should not make a false oath, because by so doing he incurs loss (of reputation) here and after death (by falling into Hell). Yaj. II. 236 prescribes a fine of 100 panas for one who takes an improper or profane oath⁸⁶.

The benefit of doubt is to be given, according to modern ideas on the administration of justice, to the accuseḍ This principle was put forward several centuries before Christ by Āp. Dh. S II 5 11 2 ‘the king should not punish when there is a doubt’ (about a man's guilt⁸⁷)