“SURPRISINGLY LITTLE WORK has been done on the comparative study of witnesses and witness testimony in different societies. The pervasive doctrine of modern Anglo-American legal theory, that witnesses are called ‘to establish the facts,’ has created an impression even among anthropologists and historians that the functions and activities of witnesses do not vary much from one culture to another. When lawyers have studied witness testimony in past societies their questions have been shaped by the Anglo-American law of evidence.” “If we start out from the assumption of modern courts that witnesses are called to ‘establish the facts of the case’ we shall misunderstand the Athenian data.”Footnote 1
With this critical note, Sally Humphreys opens her discussion of the role of witnesses in Athenian courts of law of the 5th to 4th centuries BCE. Humphreys recognized that modern conceptions about witnesses and their role in judicial proceedings often do not accord with the portrayals of witnesses in Athenian classical sources. She is not alone. Scholars who study witnesses and testimony as described in texts from other parts of the ancient world (be it the Great Code of Gortyn, legal documents from the Ancient Near East, or the Hebrew Bible) often find these descriptions hard to explain from a modern point of view.Footnote 2 Yet, while the challenges posed by such ancient depictions have been noted in scholarship, very little has been done to explain their overall meaning.Footnote 3 Are these challenges incidental and thus insignificant? Or are they indicative of a broader paradigm? If the latter, are the depictions of the role of witnesses in diverse ancient cultures related to one another, or should the role of witnesses in each of these cultures be studied on its own terms?
The current study compares a selection of sources pertaining to witnesses and testimony in a variety of ancient societies and their respective analyses by modern scholars. This comparison reveals that texts stemming from diverse ancient cultural contexts present modern researchers with similar problems, thus suggesting a shared ancient perception of witnesses and testimony that collides with the modern notion of the same. This, I argue, should lead us to identify two distinct paradigms of the role of witnesses and testimony, one implied in modern research and the other characteristic of ancient legal thought. The identification and articulation of the ancient paradigm and its conflict with the paradigm implied unwittingly by modern readers are essential if one is to understand the notion of testimony in ancient and in modern societies alike.
How are modern perceptions of witnesses and their role different from ancient ones? Humphreys, who admittedly touches on the subject only in passing, notes that the “Anglo-American” view sees witnesses as individuals who are “called upon to establish the facts of the case.” While this is, of course, correct, it is not the focus on establishing facts that makes the modern perception such a poor fit for many ancient depictions of witnesses and testimony. As I will argue, a recurring difficulty stems rather from a misunderstanding of the role of witnesses compared to that of judges. When addressing ancient materials, modern scholars expect witness testimony to be subservient to the judicial role. Witnesses are to provide information regarding the facts, while the judges are those who decide what to do with this information, and what should be the legal outcome resulting from it. In this worldview, judges hold the authority to decide the case and witnesses have no authority; their testimony is instrumental to the authority of judges. Admittedly, in the case of Athens, which was the focus of Humphreys’ work, the witnesses seem to lack much authority.Footnote 4 In many other ancient texts, however, an inverse paradigm appears, wherein witnesses — not judges — hold the authority of adjudication.Footnote 5 The witnesses’ statements — what we refer to as their testimony — are not brought to the judges for their consideration; in certain ancient texts the judges have no the authority to override the witnesses.Footnote 6 These two different paradigms of the role of witnesses represent two distinct models for the division of labor between witnesses and judges. I call these two paradigms “the instrumental paradigm” and “the authoritative paradigm,” respectively.
The line separating the two paradigms is sometimes a thin one. Even in adjudication that fits the instrumental paradigm, and particularly in many modern contexts, a ruling is dependent upon witness testimony, even if it is perceived only as a form of evidence. In antiquity too, witnesses’ reliability was sometimes assessed by judges before they could be trusted. Yet, despite these points of similarity, the two models differ in their configuration of the function of witnesses vis-à-vis that of judges. In the instrumental model, witness testimony is but a means of conveying information to someone else who is authorized to adjudicate, whereas, in the authoritative model, the witness is the adjudicating entity. I argue that, with the exception of Athens in the classical period, ancient depictions of witnesses are often challenging for modern scholars specifically because they do not adhere to the presuppositions that arise from the instrumental paradigm. If we read the same ancient sources through the lens of the authoritative paradigm, however, the difficulty in understanding the depiction of witnesses in such sources does not arise.
My argument builds on previous scholarship of ancient texts that has occasionally identified sources where “the witnesses play more than the role of simple observers […]. They are among those who decide the case;”Footnote 7 “La déposition des témoins étant décisive, la décision sur le litige revient aux témoins;”Footnote 8 “the ‘witnesses’ constitute a court.”Footnote 9 While correctly intuiting what I call “the authoritative paradigm” of the role of witnesses, these sporadic observations made in the context of specific texts and cultures have left little impression on the broader scholarly understanding of testimony in antiquity. They were apparently regarded as pointing out exceptions to a rule, and as such were never seriously considered as challenges to the general opinion that has continued to implicitly apply the instrumental paradigm wherever possible.Footnote 10 This article will demonstrate that the careful use of comparative evidence allows a broader picture to emerge, one that necessitates generalization if we are to accurately describe the legal ideas embedded in the ancient texts.
The article is comprised of three parts, in which I analyze three types of sources: (a) legal texts (laws as well as judicial private documents) where verdicts depend upon the testimony of witnesses; (b) literary as well as legal texts in which the terminology referring to witnesses and judges overlap; and (c) oath formulae that refer to deities as witnesses when they are invoked as guarantors of oaths.
As I will argue with regard to all three categories, just as texts from a variety of cultures of the ancient world exhibit similar features, so too the scholarly debates regarding these texts rely on similar (modern) assumptions, and thus invoke similar questions and arguments. The comparative perspective gained by this overview leads to the conclusion that there was a time in which the instrumental paradigm of the role of witnesses was not yet the rule. The authoritative paradigm was once dominant in several ancient societies. Recognizing this alternate paradigm for thinking about the role of witnesses not only sheds light on many ancient texts that have thus far remained obscure but also uncovers a fascinating — and until now untold — chapter in the history of legal thought.
A brief note on terminology is required before we begin. This article examines the conception of witnesses and testimony in antiquity in the regions that formed the beginnings of Western civilization: the Ancient Near East and its contemporary Mediterranean milieu. I refer to these societies when using the terms “antiquity” or “ancient” throughout the article. Other civilizations of past and present will not be discussed in this article. The period under consideration begins as early as the first written documents known to us, but its endpoint is less clear since ideas do not disappear overnight. Most of the texts discussed in the article are dated circa the first millennium BCE, although they often reflect much earlier ideas and concepts.
“If Witnesses Testify”: Conditioning Verdicts on the Words of Witnesses
Ancient texts regularly mention witnesses when describing standards of adjudication. A well-known example is found in Deuteronomy 17:6 which states: “On the word of two witnesses or of three witnesses the one who is to die shall be put to death; a person shall not be put to death on the word of one witness.”Footnote 11 Aristotle tells us of a law that similarly referred to the testimony of witnesses as mandatory for a conviction of a murderer: “at Cyme there is a law relating to trials for murder, that if the prosecutor on the charge of murder produces a certain number of his own relatives as witnesses, the defendant is guilty of the murder.”Footnote 12 How much discretion was afforded, according to these laws, to a judicial body that was presented with the testimony of the required number of witnesses, is unclear. Is the testimony of witnesses mentioned in these two sources meant to be decisive? Was a judicial entity, upon hearing those testimonies, bound to accept them?Footnote 13
Strikingly, the same dilemma awaits the modern reader in different ancient texts, even when they are clearly not dependent on one another. In what follows, I will demonstrate this recurring dilemma in some Neo-Babylonian court documents and in the Great Code of Gortyn. In both these far-removed contexts, we find conditional statements specifying that a legal resolution will enter into force “on the day that a witness comes” or “if a witness testifies.”
Neo-Babylonian Uruk
Certain surviving litigation records from a sixth-century BCE Eanna district in Uruk (near modern Samawah, Iraq) pose researchers with a serious challenge of interpretation. While some can be classified easily as “decision records” or “text-types including the resolution of disputes,”Footnote 14 others are harder to classify as their meaning and function are not entirely clear. Of special interest to us are documents that determine the outcome of the dispute conditionally, stating that it will apply “on the day that a witness comes and establishes (the case against PN).”Footnote 15 Let us consider two examples, both of which involve some accusation of theft from the local temple of the Lady-of-Uruk, for which the (maximum?) punishment is a fine of thirty times the value of the stolen goods. In the first case, Ina-ṣilli-Ištar is accused of stealing tithe-barley:Footnote 16
“On the day that a witness comes and establishes (the case) against mIna-ṣilli-Ištar, slave of mIddin-Marduk son of Eṭēru, that he opened the storehouses where mNabû-ušabši son of mNabû-zēraukīn placed the tithe-barley of the Lady-of-Uruk and took (it), […] he (mIna-ṣilli-Ištar) shall repay thirty-fold to (the temple of) the Lady-of-Uruk. If a witness does not establish (the case) against him, mIna-ṣilli-Ištar is clear.”
Nabû-ušabši accuses Ina-ṣilli-Ištar of theft of temple property. His punishment for this theft is set out in this document but this punishment is not yet enforceable; it will enter into force only “on the day that a witness comes and establishes (the case).” If a witness does not appear, then all charges against the accused would be dismissed. The second example is similar and demonstrates the formulaic nature of the documents:Footnote 17
“On the day that a witness or informer establishes (that) mNabû-ēṭir son ofmBēl-aḫašubši descendant of Eda-ēṭir had received silver or gold from mItti-Šamaš-balāṭu, the pilferer, or mKalbi-Bau, the goldsmith, the pilferer, son of mNādinu, […] whatever the witness establishes against him he shall pay 30-fold to (the temple of) the Lady-of-Uruk.”
This time Nabû-ēṭir is accused of stealing silver and gold from the temple. Here too, the future verdict against Nabû-ēṭir is stated in the document. Yet it seems that the Eanna officials who composed it did not have an opportunity to consider the testimony of witnesses pertaining to the accusation before drafting their resolution.Footnote 18 Rather, they have entrusted the future resolution of the lawsuit to a future witness, should one appear, stating that “whatever the witness establishes against (Nabû-ēṭir) he shall pay.”
These texts, sometimes referred to as “penalties pending evidence,” are exemplary of a larger group of text types where the decision that concludes the case appears to have been stated in the document before the judicial forum had been presented with the evidence, i.e., the witnesses’ testimony.Footnote 19 This feature is puzzling since it is unclear what is supposed to happen once witnesses eventually appeared. Do the parties return to the judicial forum for evaluating their testimonies? If this is the case, what was the purpose of stating the outcome of the case in advance? Does this outcome not depend upon the consideration of the testimony by judges? As noted by Shalom Holtz in his comprehensive study of Neo-Babylonian legal procedure:
“Phrased in somewhat modern terms, all of these texts raise the question of whether they were written during the evidentiary phase of the trial or during the sentencing. […] [T]here are those who interpret all three types of texts as “summonses,” which implies that the case remains open and awaits a final ruling. It may be, however, that all three text-types represent the end of the court’s involvement, […] as sentences which will take effect when evidence becomes available.”Footnote 20
Holtz further notes the various titles that were offered for these documents in previous scholarship which demonstrate a disparity between two positions.Footnote 21 Some scholars classify them as “preliminary decisions” (Felix Peiser) or “accusation pending trial” (Ellen Whitley Moore). These titles propose that the judicial proceeding has only just begun; the judges remain involved and “they would decide if the evidence had indeed ‘established’ the case.”Footnote 22 Conversely, others refer to these documents as “judgments in principle” (Denise Cocquerillat) or “conditional verdicts” (Bruce Wells).Footnote 23 These titles imply that the documents under consideration “reflect a concluding stage of the proceedings,” and that “there seems to be no further need for any adjudicating authority to render a decision.”Footnote 24 The disagreement between these two opinions is ultimately a question of understanding the legal procedure reflected in these documents and the way we understand the role of witnesses vis-à-vis judges as implied by them. If we assume the instrumental paradigm must apply to these documents, we will be inclined to think that they are but “accusations pending trials,” since the testimony of a witness must, surely, be revised by a judge who holds the authority to decide the case. If, however, these documents are in fact “conditional verdicts,” this means that the testimony is not subject to a judicial affirmation of any sort. Such a reading suggests that the authoritative paradigm is at play in the jurisprudential underpinnings of this ancient procedure.
Gortyn code
In the fifth-century inscription of laws from Gortyn (located today in southern Crete), a similar dilemma arises. Several passages in the Gortyn Code state the outcome of legal disputes to be applicable “if a witness testify.” Similarly, clauses addressing the authority of a dikastas Footnote 25 —a judgeFootnote 26 —grant him the power to decide the dispute “unless a witness testifies.” It is not always clear to what degree the dikastas was bound by the words of the witnesses or whether he had the discretion to decide against witness testimony. In some instances, the code is nearly explicit, as in the following example of a dispute over the ownership of a slave:Footnote 27
And if they contend about a slave, each affirming that he is his, if a witness testifies, [the dikastas] is to rule (dikazein) according to the witness, but if they testifyFootnote 28 either for both sides or for neither, the dikastas is to swear an oath and decide (krinein).
In this case, if a witness testifies the dikastas should rule according to the words of the witness. Only if the testimony is not decisive, because there are competing testimonies, or if no testimony is available, is the dikastas instructed to decide the case, apparently at his discretion, after taking an oath. This case demonstrates a difference in terminology applied consistently throughout the Gortyn Code, between two forms of resolution by a dikastas: a resolution based upon the testimony of witnesses (or an oath of denial taken by the defendant) is regularly referred to by the verb dikazein, rendering the meaning of “to give judgment,”Footnote 29 whereas a resolution made in the absence of witness testimony (or a defendant’s oath of denial), is referred to using the verb krinein, rendering a somewhat different meaning, which could be translated as “to decide.”Footnote 30 Scholars agree that the difference in terminology signifies two varying levels of discretion granted to the dikastas in his ruling: when instructed to dikazein, he has limited discretion, or rather, he is bound by the testimony (or the oath of denial). However, when instructed to krinein, he has a broader discretion.
While there is broad consensus among scholars regarding the aforementioned principle, opinions diverge substantially when it comes to other clauses of the Gortyn Code where explicit instructions for the dikastas on how to use witness testimony are not provided. What happens when a witness does testify in such a case?Footnote 31 Is the decision also automatic, based on the words of the witness, or is the dikastas permitted to consider witness testimony alongside other forms of evidence and arguments presented by the parties involved?
Alberto Maffi asserts that, whenever witnesses testify, the procedure will be decided automatically according to their testimony.Footnote 32 To him, there can be no krinein if witnesses come forward. Michael Gagarin, however, is hesitant to accept this interpretation. He holds that “there is no general rule for witnesses,”Footnote 33 and that it is perfectly possible for a dikastas, in certain circumstances, to consider the testimony of witnesses without being bound by it.Footnote 34 This difference of opinion comes to a head when we consider the interpretation of another clause of the Code, which sets forth a general rule applicable to a variety of cases. The opposing opinions are manifested in the different translations to this clause by different scholars. This is the translation offered by Kevin Robb:
Whatever is written, he [the dikastas] shall rule on; the dikastas shall give judgment (dikazei) as it is written, according to witnesses or oaths [of denial]; but in other matters he shall himself take an oath and decide (homnunta krinei) according to the pleas.Footnote 35
According to this reading, in any legal dispute, the dikastas must apply a two-stage test for decision: If the law covers the dispute and witnesses are available (or rather, the defendant is willing to take an oath of denial), he ought to make his ruling “bound strictly to the wording of the law and the testimony of witnesses”Footnote 36 (or the defendant’s oath of denial). Only in cases where this is not possible, i.e., in the absence of witnesses (or an oath of denial), must he decide the case (krinein) according to his discretion. A different translation of the same clause is offered by Gagarin:Footnote 37
Whenever it is written that the dikastas is to rule according to witnesses or an oath of denial, he is to rule as is written, but in the other cases he is to swear an oath and decide with reference to the pleadings.
According to Gagarin, the dikastas is bound by the testimony of witnesses, when such exist, only in cases where the law explicitly states that he is bound in this way. However, in other cases (according to Gagarin, the majority of cases mentioned in the code), the dikastas is not compelled to rule according to the witnesses, and even if witnesses come forward to testify, he can nevertheless decide the case at his discretion.Footnote 38
Gagarin’s reading is expected if one assumes the applicability of the instrumental paradigm. For him, the role of the dikastas was similar to that of a modern judge:Footnote 39 if not instructed otherwise, he “should consider the evidence on both sides, including witnesses, and any arguments that either side might present” and then reach a decision.Footnote 40 Others reject this assumption as anachronistic.Footnote 41 They argue that, if the code is silent at times concerning the guidelines that should direct the dikastas in his ruling, we ought to interpret the vague clauses in light of other clauses in the Code which clearly outline the limitations posed on the discretion of the dikastas by the testimony of witnesses.Footnote 42 While this terminology is not used explicitly, the reading preferred by these scholars construes the authoritative paradigm as more befitting the reading of the ambiguous clauses of the Code.Footnote 43 The choice between the instrumental paradigm and the authoritative paradigm is thus at the heart of the readings which different scholars prefer in both the case of the Gortyn Code and the case of the Uruk documents discussed earlier.
What’s in a Term: The Semantic Field of the Term “Witness” in Ancient Languages
The conceptual distinction between the role of a judge and that of a witness that we make in modern thought accords with our use of two different terms to denote each function. Distinct terms to denote witnesses and judges exist also in ancient languages. However, ancient languages occasionally present instances where the term reserved for a witness must be understood as referring to someone acting in a capacity that we would today describe as that of a judge. In what follows, I discuss this peculiar conflation attested in four ancient languages: Akkadian, Biblical Hebrew, Classical Greek, and Latin. Those occurrences of linguistic ambiguity were thus far treated in scholarship mostly as coincidences. Scholars who addressed them were unaware that a similar phenomenon existed in other languages and attempted to explain the problem away, usually on etymological grounds. However, I argue that if the same phenomenon of linguistic over-determination recurs in several languages, without etymological dependency between the technical terms used for witnesses and judges in those languages, it is no longer tenable to assume the dual meanings as mere coincidence. Rather, it should be regarded as a conceptual pattern, whose explanation must be sought in the common jurisprudential underpinnings of the ancient societies under examination.
Akkadian (in a Hittite text)
In Akkadian, one term for a witness is šibu (which during the Neo-Babylonian period continued to be used alongside the more popular term mukinnu).Footnote 44 Literally, it denotes “elder,” equivalent to Hebrew sav (שָֹℶ) and Arabic šāyeb (شايب). However, in the Ancient Near East (as in the Hebrew Bible) an “elder” is also a judicial role, as the elders (šībū) constitute a communal judicial institution.Footnote 45 Given this dual meaning, “[t]he term šibu may present some ambiguity for the interpretation of those texts wherein either ‘(city) elder’ or ‘witness’ could have been meant.”Footnote 46 In the absence of an alternative method of resolving this ambiguity, scholars choose between possible translations based on the respective roles ascribed to witnesses and judges, roles which are unwittingly structured along the lines of the instrumental paradigm.Footnote 47 However, the use of the instrumental paradigm in this way is sometimes misleading. One example is found in the Hittite law concerning lost property. This law is preserved in two versions, one being a revision of the other.Footnote 48 The Old Hittite (OH) version of the text reads as follows:
If anyone finds a stray ox, a horse, a mule [or] a donkey, he shall drive it to the king’s gate. If he finds it in the country, they shall present it to the elders (LU.MEŠ ŠU.GI-aš). [The finder] shall harness it (i.e., use it while it is in his custody). When its owner finds it, he shall take it in full value, but he shall not have him [the finder] arrested as a thief. But if the finder does not present it to the elders, he shall be considered a thief.Footnote 49
The New Hittite (NH) version of the same law reads as follows:
If anyone finds implements or an ox, a sheep, a horse, [or] an ass, he shall drive it back to its owner, and [the owner] will lead it away. But if he cannot find its owner, he shall secure witnesses (Hitt. kutruwaezzi) that he is only maintaining custody. Afterward [when] its owner finds it, he shall carry off in full what was lost. But if he does not secure witnesses, and afterwards its owner finds it [in his possession], he shall be considered a thief: he shall make threefold compensation.Footnote 50
In both these wordings of the law regarding the finding of lost objects, the same basic principle is stated. For a finder of a lost object to avoid an accusation of theft, he ought to declare his finding publicly. According to the earlier version, this declaration has to be made to local authorities: to the king’s gate, if the discovery takes place in the city, or to the elders if it is in the country. An elder here is designated through the Sumerogram ŠU-GI which signifies šibu. In the later version, the same result is achieved by introducing the lost object to witnesses. To designate the securing of witnesses, the NH version uses the Hittite verb kutruwaezzi; kutruwa(n) being the noun “witness.”
What should we make of the difference between the two versions of this law? Clearly, the dual solution is not preserved in the NH version of the law. There, the option of introducing the lost object to the king’s gate officials in a city is omitted entirely. Things are less clear, however, with regard to the option of presenting it to the elders, according to the OH version, as opposed to introducing it to witnesses, according to the NH version. Did the NH version attempt to change the norm that governed the finding of lost objects, by moving from the elders, who were city officials, to witnesses with no such capacity? Scholars have noted that the NH version has introduced several modifications to the earlier OH version of certain laws, some more substantial than others. It is the task of scholars to interpret and inquire about what is merely a stylistic modification, perhaps for the purpose of modernizing the language, and what is a more substantive alteration.Footnote 51
One should note that the Sumerogram ŠU-GI that is translated as “elder” in the earlier version could have equally been translated as “witness.” The fact that it was translated as "elder," indicating the capacity of an official, is an interpreter’s choice. This choice is informed by the fact that officials designated by the Sumerogram have a parallel function, according to the text, to that of the “king’s gate.” This parallel suggests a formal office-bearing authority. According to the instrumental paradigm, this sort of authority is not typical of witnesses, whose testimony is rather a means for the implementation of someone else’s authority—typically, a judge. This reasoning is, of course, not unique to this clause. One might suspect that it lies in the background of additional translations involving the Sumerogram ŠU-GI.
Biblical Hebrew
In Hebrew, the terminus technicus for a witness is ‘ed (עℸ).Footnote 52 It is regularly used when describing witnesses appearing before judges in a legal setting. However, some of its uses describe purely judicial roles. Two striking examples of this phenomenon refer to no other but Yahweh as an ‘ed. The first comes from the first chapter of the prophecy of Micah:
2 Hear, you peoples, all of you; pay attention, O earth, and all that is in it, and let Yahweh be a witness (‘ed) against you, Yahweh from his holy temple. 3 For behold, Yahweh is coming out of his place, and will come down and tread upon the high places of the earth. 4 And the mountains will melt under him, and the valleys will split open, like wax before the fire, like waters poured down a steep place. 5 All this is for the transgression of Jacob and for the sins of the house of Israel […] 6 Therefore I will make Samaria a heap in the open country, place for planting vineyards, and I will pour down her stones into the valley and uncover her foundations.Footnote 53
Prima facie, it seems that in verse 2 Yahweh is announcing his forthcoming judgment of the sins of Samaria, a judgment that is described in verse 6. The only factor which casts doubt on this reading is that in the same verse Yahweh is referred to as an ‘ed, presumably a witness and not a judge. This creates a problem because it is odd for Yahweh to be playing an instrumental role in divine judgment, such that one would expect a witness to play according to the instrumental paradigm. Rather, it seems clear that the text considers Yahweh as the ultimate judge.
Various solutions have been suggested to this problem. Some argue that ‘ed is used here in the sense of an accuser and not a judge.Footnote 54 Indeed, in other instances in the Hebrew Bible, the witnesses seem to be making accusations against a defendant in a trial. However, if Yahweh is the accuser, who is the judge?Footnote 55 In a standard legal arrangement, the accuser, like a witness, is subject to a higher authority, who presides over the case. Portraying Yahweh as fulfilling such a secondary role, either as a witness or as an accuser, seems equally problematic. Others suggest that we should think of Yahweh as possessing a dual function, both as an accuser/witness (in v. 2) and as a judge (in v. 3–6).Footnote 56 Against these readings, David Anderson and Noel Freedman suggest that ‘ed here should be read as simply meaning a judge.Footnote 57
Understanding ‘ed as fulfilling a judicial function seems to be the only plausible reading when we turn to a second example, from the book of Malachi:
Then I will draw near to you for judgment. I will be a swift witness (‘ed) against the sorcerers, against the adulterers, against those who swear falsely, against those who oppress the hired worker in his wages, the widow and the fatherless, against those who thrust aside the sojourner, and do not fear me, says the Yahweh of hosts.Footnote 58
Yahweh promises to judge and punish his people harshly for their extensive violations of the law and unjust behavior. This seems to be the role of a judge.Footnote 59 Yet in doing so, Yahweh is referred to as an ‘ed. Therefore, the ‘ed must be understood as possessing the adjudicative function associated with a judge.
Here again, some commentators, guided by the instrumental paradigm, have attempted to resist this conclusion and to maintain some conceptual distinction between the role of a witness and that of a judge. This effort yields difficult readings, such as the following: “Here the function of the witness within the court proceedings is not secondary but rather identical with that of the judge who can act swiftly because he was himself a witness.”Footnote 60 However, this reading appears incompatible with the text, since Yahweh is said to be “a swift witness” exactly when imposing judgment. This is very different from acting swiftly as a judge based on one’s previous knowledge as a witness. The phrasing thus suggests that in calling Yahweh an ‘ed he is actually said to be functioning as a judge.Footnote 61
Greek
For Greek sources, our focus will be on the term istōr (ἴστωρ) or histōr (ἵστωρ). Dictionaries translate the term as having two meanings: (1) witness and (2) judge.Footnote 62 The identification of istōr as a witness is based on inscriptions from several cities in Boeotia, where istores (Ϝίστορϵς) are mentioned in manumission procedures. Those function bearers could be termed either istores or martures, a standard word for witnesses.Footnote 63 While these inscriptions are dated as early as the third-century BCE, support for this reading is found in even earlier sources, where istores is used in archaic oath formulae. Again, these formulae could refer to divine entities either as istores or as martures.Footnote 64
In other early usages of the term, however, its meaning is less clear. Hesiod mentions the istōr as a person blessed with some kind of talent granted from birth.Footnote 65 This meaning is difficult to reconcile with the idea of an istōr as a witness according to the instrumental paradigm, which depends on a concrete event that may or may not be witnessed in the future. As noted already in scholarship, the term istōr appears in two much-debated passages in Homer’s Iliad as one who is required to put an end to a dispute between rival parties, a function which may be seen as demonstrating a talent of the kind referred to by Hesiod.Footnote 66 In one instance, Agamemnon is referred to by Idomeneus as the istōr who could be chosen to decide between his and Aias’ opinion concerning the outcome of a race.Footnote 67 The other mention, which is found in a more elaborated context, is in the ekphrastic trial scene on Achilles’s shield:
There was a crowd of citizens drawn to the meeting place: a dispute had arisen between two men, at loggerheads over the blood-price of a man who’d been killed: one claimed, in a public speech, to have paid it all, but the other swore he’d been given nothing, and both were determined to win the arbitrator’s (istōr) verdict. People were backing both sides, cheering one or the other, while heralds held them back, and the elders were sitting on polished seats of stone in the sacred circle, the loud-voiced heralds’ staffs in their hands: holding these they would rise to deliver judgment, each in turn; and there between them were set two talents of gold, to go to the one who delivered the fairest verdict.Footnote 68
This is one translation out of many offered for this passage, which poses significant interpretive problems.Footnote 69 One central ambiguity has to do with the role of the istōr in the scene apropos that of the elders who are mentioned shortly afterward. Is the istōr one of them or does he have a different role? This question arises because it is clear from the text that the istōr holds the authority to resolve the case at hand. The Greek reads: amphō d’ hiesthēn epi istori peirar helesthai.Footnote 70 To give just a handful of examples of translations of this line:Footnote 71 “Both parties insisted that the issue should be settled by a referee”;Footnote 72 “both were willing to appeal to an umpire for the decision”;Footnote 73 “Both then made for an arbitrator, to have a decision”;Footnote 74 “both were eager to take a decision from an arbitrator”;Footnote 75 “both men pressed for a judge to cut the knot.”Footnote 76 All these translations reflect the reading that the istōr puts peirar — end or limit — to the dispute. Given that the istōr is introduced in the text only through this single sentence, the decisive role of the istōr in any translation seems unavoidable.Footnote 77
The translation of the word istōr as “arbiter” or “judge” has raised objections. Some scholars have preferred a more neutral word, e.g., “umpire,” “referee,” or “one who knows.”Footnote 78 Most scholars assume that istōr is derived from the verb oida, “to see, to know,” which would fit with the identification of an istōr with a witness of some kind.Footnote 79 However, in light of the authority that is attributed to the istōr to resolve the dispute, commentators are hesitant to translate it as simply denoting a witness, even if they assume that one aspect of his role is also to bear witness and record something in his memory.Footnote 80 In other words, the instrumental paradigm holds back on the translation of istōr as witness. If we were to read the text according to the authoritative paradigm, on the other hand, there would be no tension between the title, a witness, and the role expected to be played by the figure carrying this title; namely, putting an end to a disputed case.
Latin
The Latin term of relevance for us is testis. A conflation of witnesses and judges in a Latin text that uses this term is found in the ancient procedure of the declaration of war. According to the description provided by the historian Titus Livy, in ancient times, a declaration of war was preceded by a visit of a fetial priest to the enemy’s territory, where he presented the Roman claims and demanded restitution. If those demands were not met, says Livy, then “at the expiration of thirty-three days, for that is the fixed period of grace, he declares war in the following terms”:Footnote 81
‘Hear, O Jupiter, and thou Janus Quirinus, and all ye heavenly gods, and ye, gods of earth and of the lower world, hear me! I call you as witnessesFootnote 82 that this people (ego vos testor populum ilium)’ — mentioning it by name — ‘is unjust (iniustum esse), and does not fulfill its sacred obligations. But about these matters we must consult the elders in our own land in what way we may obtain our rights.’
The declaration of war thus included a nomination of deities as witnesses. In what sense? What is the result of making the gods witnesses in this case? Scholars agree that the gods are not called upon as passive witnesses to observe the event alone: “the ‘witnessing’ that the gods should perform means that they, in the war about to begin with the agreement of the senate, [should] destroy the ‘unjust’ enemy people and provide a just Roman victory.”Footnote 83 This active part of adjudication and punishment ascribed to the gods is not the role typically associated with witnesses; rather, according to the instrumental paradigm, it is closer to the role of a judge.
A bolder expression of the same idea is found in a different passage where Livy quotes the words of Tullus Hostillius, the third king of Rome, on the verge of the war with Alba Longa. A few paragraphs before he discusses the declaration of war procedure, Livy mentions the whereabouts of that war, which began after the Romans have posed the Albans with an ultimatum of thirty days to fulfill Roman demands (a detail that echoes the fetial priests’ declaration procedure). Furthermore, when Alban envoys approach Tullus, he answered them as follows:Footnote 84
“Tell your king the Roman king make the gods witnesses (deos facere testes),Footnote 85 that whichever nation is the first to dismiss with ignominy the envoys who came to seek redress, upon that nation they will visit all the sufferings of this war.”
This text paraphrases the declaration of war formula, this time presenting it as part of the narrative. The narrated phrasing makes clear that the calling upon the gods as witnesses in the formula is intended so that they punish the unjust nation, "upon it they [the gods] will visit all the sufferings of this war” (in eum omnes expetant huiusce clades belli). The gods’ role is not a passive role of observers; according to Alan Watson, “it is self-evident in the words of Tullus that […] the gods are to act, to punish. They are judges in a criminal suit, not witnesses.”Footnote 86
In line with the instrumental paradigm, Watson assumes that the role of a witness and that of a judge are essentially different.Footnote 87 This assumption leads him to argue that in different periods the term testis had different meanings: originally it meant “judge,” as indicated by the fetial declaration of war; only later it has come to mean “witness.”Footnote 88 He further suggest an etymological explanation to support his theory.Footnote 89 Appling the authoritative paradigm of the role of witnesses would have led to a much easier reading of the ancient Roman declaration of war, rendering Watson’s historical reconstruction, as well as his etymological explanation, unnecessary.
Oaths Deities as Witnesses
The third category of texts to be discussed in this article is oath formulations. It is well established in scholarship that societies of the Ancient Near East and their Mediterranean contemporaries shared the same basic oath structure.Footnote 90 Witnesses play an important role in this structure and, arguably, this role hardly fits the perception of witnesses according to the instrumental paradigm. .
Essentially, an oath is a declaration accompanied by a conditional curse (self-curse, if the speaker is the one taking the oath).Footnote 91 The curse will apply if the declaration is false, or, in the case of promissory oaths, where a person undertakes to perform a certain act, if this promise is not kept. Furthermore, oaths regularly involve the invocation of divine entities, whether explicitly or through reference to a sacred object.Footnote 92 Of all these elements, our focus will be on the role played by the divine entities in the establishment of oaths.
The divine entities which function as oath deities are summoned to oversee the enforcement of the oath. “The role of these para-human beings in this context was to reward or punish those who either observed or contravened, respectively, the conditions imposed upon them by oath.”Footnote 93 If the oath is violated, the oath deities will impose the curses and punish the oath violators.Footnote 94 Similarly, as a positive means of enforcement of oaths, the deities may also be said to grant benefits and reward those who righteously fulfill their oaths.Footnote 95 Thus, some scholars refer to the role of oath deities as “guarantors” or “guardians” of the oath.Footnote 96 Clearly, the deities are also required to cast judgment and decide whether or not a violation took place before they impose appropriate sanctions or grant rewards—a fact that renders their role quasi-judicial. In some ancient oaths, this judicial function is expressed in explicit adjudicatory verbs.Footnote 97 However, oath deities are not termed “judges” of the oaths.Footnote 98 Instead, they are regularly referred to as “witnesses.”Footnote 99
The invocation of divine witnesses is a standard feature of oaths in the Ancient Near East and its surrounding Mediterranean cultures. This motif appears across a wide range of sources, where the presence of deities as witnesses serves to seal the oath and bind the parties involved. Thus, in many Hittite vassal treaties, we encounter the following oath formulation: “I have now summoned the thousand gods to assembly for this oath, and I have called them to witness—they shall be witnesses.” Summoning the gods as witnesses is what seals the treaty with an oath. A similar logic underlies the biblical account in which the people of Israel swear to the prophet Jeremiah to abide by Yahweh’s commandments, using the same fomula of calling their God as a witness: “May Yahweh be a true and faithful witness against us if we do not act according to all the word with which Yahweh your God sends you to us.”Footnote 100 The same structure recurs in Greek epic as well. In Homer’s Iliad, Hypnos’ adjuration of Hera is similarly phrased, as a call for divine entities to be witnesses: “Well then, swear to me now by the inviolable waters of Styx, with one hand on the fertile earth, one on the shimmering sea, so that all the gods with Cronos down below may bear witness, that you will grant me one of the young Graces, Pasithea, whom I’ve longed for all my days.”Footnote 101 The idea that an oath is taken by invoking a deity as a witness is explicitly articulated by Cicero in the following words: “For a sworn oath is a religious affirmation; and it is what you promised with this affirmation and, as it were, with a god as your witness, which must be kept.”Footnote 102
The association of oath deities with their title as witnesses to oaths is so strong that the very mention of a divine entity as a witness is, in itself, sufficient to indicate that an oath is being initiated, even in the absence of any additional oath terminology.Footnote 103 Moreover, the verb indicating the summoning of witnesses in both Hebrew (hē‘îd) and Greek (marturomai) functions as a synonym for the initiation of an oath.Footnote 104 But why were deities termed “witnesses” and not “judges,” “guarantors,” or something similar? Two lines of solution were suggested by scholars to this problem. As we shall see, both these solutions maintain a conceptual separation between the role of a witness and that of a judge, a separation that lies at the heart of the instrumental paradigm.
Either Witnesses or Judges
Some scholars, noting texts in which an oath is initiated without the divine entities explicitly summoned as witnesses, have suggested that a distinction should be made between two types of oaths: one in which the role of deities is (only) to witness the parties’ initial declarations, and another in which they function (only) as judges or guarantors who enforce the oath. An example of this attitude can be found in the following account by the late biblical scholar Moshe Segal: “[W]e have two kinds of oaths: one where Yahweh, or some other dear or holy entity, is called upon to be a witness […] the other, that includes a curse, where Yahweh is called upon to be a judge and to impose the curse to the oath-taker or the one whom the oath was imposed upon, if he violates it[…].”Footnote 105
However, the attempt to distinguish between two types of oaths, based on the wording of the formula is bound to fail. Many oaths clearly assign both functions of the deities, who are called upon both to witness and to punish violators. One biblical example of this is the oath exchanged between Jacob and Laban in Gen. 31:50. There, Laban says to Jacob: “God is a witness between you and me […]”; and a few verses later he adds (53): “The God of Abraham and the God of Nahor, the God of their father, let them judge between us.” In these words Laban calls the gods both to witness and to judge. Another example, among many, appears in the following verses from the Iliad (XIX 258–65): “Let Zeus, first, be my witness, highest and best of gods, and Earth, and Sun […] that never did I lay hands on the girl Briseïs […] If any of this is sworn falsely, may the gods give me all the many griefs they inflict on perjurers in their name.”Footnote 106 Here too, Hypnos invokes the gods as witnesses to his oath, while also requesting that they impose punishment in the event of perjury.
More importantly, all oaths arguably contain the element of a curse imposed by the divine entity who is expected to cast judgment —whether the curse is explicitly articulated or merely implied.Footnote 107 The curse is inherent to the conceptual structure of oaths and should thus be assumed to be lurking in the background whenever an oath is at stake. Conclusive evidence for this understanding is found in a mid-fourteenth-century BCE Hittite between Suppiluliuma I, king of Hatti, and Aziru, king of Amurru. Since this was an international treaty, two versions of the text were drafted — one in Akkadian, and one in Hittite, both of which have survived. This is how the section on witnesses is presented in the Hittite version of the treaty:
§16 I have now summoned the thousand Gods to assembly for this oath, and I have called them to witness, they shall be witnesses.Footnote 108
The Hittite text ends here. Nothing more is added, although there remains some free space on the tablet. When we turn to the Akkadian version, it presents us with two further paragraphs:
§17 All the words of the treaty and of the oath which are written on this tablet – if Aziru does not observe these words of the treaty and of the oath, but transgresses the oath, then] these oath gods shall destroy Aziru, together with his person, his wives, his sons, his grandsons, his household], his city, his land, and his possessions.
§18 But if Aziru observes these words of the treaty and o the oath which are written on tablet, then these oath gods shall protect [Aziru], together with [his person, his wives, his sons, his grandsons], his household, his city, his land, [and his possessions].Footnote 109
Since the Akkadian and Hittite texts are two versions of the same document, it is clear that the role of the deities is the same in both of these texts. Even if the function of the deities as those imposing curses and blessings is not explicitly stated in the Hittite version, it is nevertheless implied.Footnote 110
Both Witnesses and Judges
If we reject the idea that deities should be seen as either witnesses or judges of oaths, what alternative conceptual model could account for their complex function? A leading opinion among scholars is that the deities perform a dual role, being both witnesses and judges. This model is expressed in the following account by Małgorzata Sandowicz, in her examination of oaths in the Ancient Near East from the Neo- and Late-Babylonian periods:
The authorities appealed to are commonly referred to as witnesses of the oath. However, their power seems further-reaching: they are not only to witness a solemn statement, but also to mete out punishment in case the declaration is or turns out to be untrue. Thus, the function of the authorities is also that of guarantors.Footnote 111
Similarly, Frances Hickson notes, concerning oaths in a Roman context:
[T]he speaker either explicitly or implicitly makes two requests of the divinities by whom he or she swears; thus an oath is a type of petitionary prayer. The primary request is a self-curse that the speaker suffer some punishment if the words are intentionally false. The secondary request, that the divinities be present as witnesses to the speaker’s words […] (emphasis added).Footnote 112
Let us note that this dual model is also shaped by the instrumental paradigm. When the deities act — when they inflict punishment — they are judges (even if one labels them guarantors). By contrast, when they are passive receipients of information, they serve merely as witnesses. The two functions are conceptually distinct although they co-exist alongside each other.
This dual-function model, while accounting more fully for deities’ oath-related roles, is nevertheless unsatisfying. Semantically, this reading is problematic since it suggests that, although the dominant function of the deities is that of adjudication, they are referred to by their less significant function, that of witnesses. The question concerns more than just the use of one title when the deities supposedly perform two separate functions. What is truly at issue is the appropriateness of representing oath deities by their passive role as witnesses —recording the human undertakings— rather than by their more substantial and active role as enforcers of the oath.
Furthermore, several depictions of oaths in ancient sources appear to contradict the assumptions of the dual-function model. According to this model, deities are presumably termed “witnesses” due to their roles in the preliminary stages of the oath, in which they hear and observe the pronunciation of the conditional curse and the undertakings of the oath-taker. However, in some ancient texts, we find that oath deities are referred to as “witnesses” specifically in connection with their role in the later stages of the oath’s lifetime, long after it has been taken, when they are to oversee its fulfillment by imposing curses and punishments on violators. In other words, they are called to be witnesses percisely at the moment of enforcing judgment. A clear example for this is, yet again, the oath exchanged between Jacob and Laban in Gen. 31:46–50, mentioned earlier:
And they took stones and made a heap, and they ate there by the heap. Laban called it yəḡar-śāhăḏūṯā (Aramaic: a heap of testimony) but Jacob called it gal –‘ed (Hebrew, literally: a heap which is a witness). Laban said, “This heap is a witness between you and me today.” Therefore, he named it gal-‘ed, and miṣpāh (Hebrew, literally: a place of watch), for he said, “Yahweh will be watching (yiṣep̄) between you and me, when we are out of one another’s sight, if you oppress my daughters, or if you take wives besides my daughters, although no one is with us, see, God is witness between you and me.”Footnote 113
Here, Laban and Jacob name the pile of stones that they built “a heap of testimony” or “a heap which is a witness,” after their mutual call for the gods to witness their oaths. However, the text further suggests a striking analogy between these names and another Hebrew name for the same pile of stones, “a place of watch” (miṣpāh). Although not a synonym, this name is clearly meant to convey an analogous meaning. When Yahweh is a witness, he watches. Notably, this watching is to take place at a future time, when the parties are “out of one another’s sight,” in case Jacob might break his oath and violate his obligation; exactly when there is a need to “Judge”—to use the terminology of verse 53—between Jacob and Laban.
The second example appears in Thucydides’ descriptions of the events leading up to the Peloponnesian War between Sparta and Athens. Thucydides quotes the Athenian envoys to the Spartan representatives as they try to persuade them not to break the existing alliance between the two cities:Footnote 114
[S]o we urge you, while both sides still have the option of listening to good advice, not to break the treaty or transgress your oaths but let our differences be settled by arbitration according to our agreement. Otherwise, calling the gods of the oath as witnesses (θϵοὺς τοὺς ὁρκίους μάρτυρας ποιούμϵνοι), we shall try to defend ourselves against you, the authors of the war, following closely on the path you have set.
The Athenians are preparing for the possibility of war with the Spartans. The way they see it, such a war will be the direct result of a violation of an existing treaty that was sealed with mutual oaths. Like all oaths, these too involve oath deities. Yet their invocation as witnesses by the Athenians in this passage does not occur at the moment of oath-taking, but rather when it is violated. As witnesses the oath deities are made aware of the fault of the Spartans who broke the oath,Footnote 115 and are called upon to condemn the perjury of the Spartans and thus to authorise the Athenians to act on the violation and destroy their enemies.Footnote 116 Being witnesses, they are, in effect, judges.
To conclude, the attempt to separate between the gods’ role as witnesses to oaths and their function as judges of perjurers, which is informed by the instrumental paradigm, fails to provide a satisfying reading of the depictions of oaths in ancient sources. These roles cannot be ascribed to different oaths, nor can they be neatly allocated into two conceptually distinct aspects of one oath. Rather, the title “witnesses” seems to designate the authority with which judgment is cast and punishment is imposed upon oath violators.
Conclusion
In this article, I examined a variety of depictions of witnesses in texts from several ancient societies. Some are found in formal, legal texts; others appear in literary compositions, and still others in sacred religious pleas. Some involve human witnesses in actual legal procedures, while others feature deities casting divine judgment. Despite this diversity, all these depictions are challenging to the modern reader for the same reason: they resist a reading that views witnesses as performing an instrumental role. Rather, they portray witnesses as bearing the authority of adjudication, whether actual or metaphorical.
The first category I examined encompassed ancient rules of procedure. These rules seem to suggest that witness testimony binds judicial discretion. The second category included texts that demonstrtate the semantic field of the designated terms for witness in various languages, which sometimes appears to denote what we would better describe as judge. The third category of texts I examined was oath formulations that assume (whether explicitly or implicitly) adjudication of perjurers by divine entities invoked as witnesses. Texts from all three categories show a consistent pattern in a broad cross-cultural context, a pattern that hardly coheres with the instrumental paradigm, and accords much better with the authoritative paradigm. These findings thus suggest that witnesses in the societies that produced the texts under examination served as active authorities with the power to adjudicate, as opposed to the passive role they play according to the instrumental model.
By arguing that witnesses held an adjudicative authority, I do not suggest that they held this authority exclusively. Already in antiquity we find kings, priests, professional judges and lay juries possessing a substantial authority to adjudicate. My argument is simply that, more often than not, witnesses too demonstrate authority of adjudication, an authority that may also limit the discretion of the judicial body. In this articel I sought to draw attention to this authority of witnesses, demonstrating that it was neither accidental nor negligible, but rather stable and consistent across numerous texts from the ancient societies under review. The authoritative paradigm is thus well attested, and its extensive manifestations support the conclusion that it occupied a substantial place in the ancient jurisprudential imagination.
How should we explain the coexistence of the authority of witnesses and that of judges in the ancient world? Does it reflect an ancient competition between the two paradigms? Did they exist concurrently during some periods, before one became predominant? At this point, we can only speculate. More comparative research is required before we can articulate the process that generated these two competing paradigms of the role of witnesses vis-à-vis that of judges and the way their relations unfolded in the course of history.
The analysis of the notion of witnesses in ancient texts has clear implications for the study of ancient societies. At the same time, it also has great significance for the history of legal theory and legal thought. The conception of testimony as judgment rather than mere evidence implies an altogether different rationale for legal proceedings in ancient societies than the one with which we are familiar today. The fundamental three-part setup of a trial—consisting of the litigants, witnesses, and judges—is thus shown to be a flexible structure, susceptible to alteration according to varying ideologies and worldviews. Becoming acquainted with an alternative paradigm regarding the role of witnesess in adjudication allows us to question the ideological commitments of our own paradigme and to revisit it critically.
Supplementary material
To view supplementary material for this article, please visit https://doi.org/10.1017/S073824802500001X
Acknowledgments
I wish to thank Uri Gabbay, Shalom Holtz, Shai Lavi, Omer Michaelis, Yakir Paz, David Alan Sklansky, and Bruce Wells, for fruitful conversations and for their helpful comments that have greatly improved the article.
Open access
