WITNESSING

The Romans preferred documents in Greek that used dating formulae reflecting the Roman presence very accurately, as well as documents that were, when important, doubled with full interior texts, except in the specific cases in which attested copies were issued.

In the East, principals in, and witnesses to, a “simple” (undoubled) document wrote their names at the bottom of the document. In a double­document, when the writing of the document was completed, the top part of the recto side, with the inner text, was rolled up, flattened and then sewn shut, the sewing itself knotted multiple times (next to which knots the witnesses signed) on the verso.⁶⁶ Within these general patterns, however, Arabians and Judaeans made different choices when it came to how many witnesses they used.⁶⁷

Table 5.

5.1 Arabia

single or “simple”, no witnesses:

3 **p Hever 60 [receipt in Greek, 125]; **63 [renunciation of claims in Greek, 127]; **12 [receipt in Jewish Aramaic, 131].

single or “simple”, three witnesses:

2 *P Yadin 8 [purchase contract in Jewish Aramaic, 123]; *9 [waiver in Nabataean Aramaic, 122].

single or “simple”, seven witnesses:

1 *P Yadin 5 [written with the fibres; possibly a translation of a document of deposit into Greek; the seven witnesses are listed at the bottom of the recto, and then there are traces of witnesses on the verso as well; 110].

doubled, four witnesses:

64 Pliny, Historia naturalis 33.9-33.12, 17-21: East signs, 33.21.

65 See Cotton, “Subscriptions and signatures” (n 31); Cotton, Aramaic, Hebrew and Greek Documen­tary Texts (DJD 27) (n 32) 179-180 on the XEiQOXQyuTyc,.

66 As Yadin, “Expedition D” (n 5) 237 notes, “the tightness of the knots and the closeness of the signatures to them made it virtually impossible for the interior even to be opened without the signatures being damaged”.

67 This analysis of the number of witnesses to documents reports on what is currently known from publication, but is subject to revision: Cotton, “Diplomatics” (n 1) 51 and L H Schiffman, “Reflec­tions on the deeds of sale from the Judaean desert in light of rabbinic literature”, in Katzoff and Schaps, Law in the Documents (n 8) 185, 201 both note the need for a full-scale study of the number of signatures, but such a study requires a full re-examination of the documents themselves. L H Schiffman, “Witnesses and signatures in the Hebrew and Aramaic documents from the Bar Kokhba caves”, in Schiffman, Semitic Papyrology in Context (n 33) 164, 185 also argued that allowing others to sign for you was “a Greco-Roman practice” and indeed that (Schiffman, “Reflections on the deeds of sale”, 202) for contracts, the “entire procedure by which the witnesses' signatures are accompanied by those of the parties to the contract results from Greco-Roman legal practice...

5 *P Yadin 1 [debenture in Nabataean Aramaic, four witnesses plus scribe; 94]; *2-3 [sale contracts in Nabataean Aramaic, four witnesses plus scribe; 99]; *4 [agreement in Nabataean Aramaic, 99]; *17 [acknowledgment of deposit in Greek with possibly more than four witnesses - Lewis thought seven - but, more likely, the name of protagonist and scribe in the three missing lines, 128].

doubled, five witnesses:

8 *P Yadin 10 [marriage contract in Jewish Aramaic, 124-125; Schiffman [2003] 174 (as in n. 67) argues for three witnesses and two helpers]; *12 [extracts in Greek from acta, 124], *14 [summons in Greek, 125], *16 [copy of land declaration in census in Greek, 127], **P Hever 62 [copy of land declaration in census in Greek, 127], *P Yadin 18 [marriage contract in Greek, 128]; *P Yadin 23 [summons in Greek, 130], *26 [summons in Greek, 131].

doubled, six witnesses:^[159]

1 (Jewish Aramaic): *P Yadin 7 [deed of gift, 120; the last witness may be the scribe, Schiffman [2003] 173 (as in n. 67)].

doubled, seven witnesses:

4 (Greek): *P Yadin 15 [deposition, 125]; *19 [deed of gift, 128], **P Hever 64 [deed of gift in Greek, see Cotton [2003] 56-58; 129 (as in n. 1)]; *P Yadin 20 [concession of rights, 130].

5.2 Judaea:

single or “simple”, no witnesses:

2 (Jewish Aramaic): P Yadin 42 [lease agreement signed by two adminis­trators; 132], 43 [receipt signed by two administrators; 132].

single or “simple”, two witnesses:

2 (Jewish Aramaic): P Hever 13 [waiver of claims, 134 or 135], P.Sdeir 2 [acknowledgment of debt, 135].

single or “simple”, three witnesses:

5 P Hever 49 [acknowledgment of debt, 133, Hebrew], 8a [deed of sale, 134 or 135, Aramaic]; P Yadin 44-46 [leases of land in Hebrew, all 134].

single or “simple”, five witnesses:

1 P Mur 38 [restored; nd].

doubled, three witnesses:

5 P Mur 18 [acknowledgment of debt, 55-56, Aramaic; possibly only two witnesses, Schiffman [2003] 168 (as in n 67)], 29 [deed of sale, 66-70, Hebrew], 30 [deed of sale, 66-70, Hebrew], 19 [divorce, 111, Aramaic], 21 [marriage contract, first half second century, Aramaic; Schiffman [2003] 168-169 (as in n 67) suggests four].

doubled, five witnesses:

4 (Jewish Aramaic) P Hever 9 [deed of sale, late Herodian], ?21 [deed of sale, end Herodian-post H.], 22 [deed of sale, end Herodian-end Bar Kokhba; subscription and five witnesses?], P Hever 50 + P Mur 26 [deed of sale, Aramaic; undated; Schiffman [2003] 172-173 (as in n 67) argues for only two witnesses and several helpers]; 69 [marriage contract in Greek, 130].

doubled, seven witnesses:

1 P Yadin 11 [loan with mortgaged security, in Greek, 124], one party a Roman soldier.

The single documents have the fewest witnesses. Some use or require no witnesses at all, while others - purchase contract, waivers, acknowledgments of debt, leases, one deed of sale - use no more than three. At the other end of the spectrum are the double-documents with seven witnesses, of which there are five. This is a greater number of witnesses than any eastern, or indeed Greek tradition used: but seven witnesses were used for some of the most solemn of Roman documents, like wills.^[160] The four seven-witness documents from Arabia are a deposition (P Yadin 15); a concession of rights by Besas, son of Jesus, and the Roman lady Julia Crispina, guardians of the sons of Yehudah son of Elazar Khthousion, to Yehudah's older daughter (P Yadin 20); and two deeds of gift (P Hever 64 and P Yadin 19). In P Yadin 19, half of this gift is to pass only after the death of the giver. The one double-document with seven witnesses from outside Arabia was drawn up in Ein-Gedi (P Yadin 11), and in it the same Yehudah, some years earlier, borrows money from a Roman centurion named Magonius Valens.

It does not happen very often in what survives, but when it does it is very marked.

In the middle of the spectrum of witness numbers are documents with five. On the one hand, there is one simple or single document from Judaea with five witnesses - but only the signatures survive, so there is no way of deter­mining what sort of document it was. In the realm of the double-document, the options are considerably more interesting: there are three deeds of sale from Judaea, all mid-first century, and one marriage contract; two marriage contracts also survive from Arabia. And then there is one other set of documents, completely different. From Arabia there are three summonses to the court of the governor, two attested copies of land registration in the census, and one attested copy taken from the acta of the council of the city of Petra, all with five witnesses. These are, again, from the world of the official Roman document: the authorities approve or issue them. And again, there are none of them in, or none survive from, Judaea. So we have the Roman authori­ties associated with Roman protagonists, Roman courts, Roman practices, terms and dating formulae, the Greek language, and two levels of witness numbers in double-documents, seven and five. In Roman documents from elsewhere, seven witnesses are especially associated with depositions, with attested copies like military diplomata, and with the witnessing of legal acts “constructed” in the Augustan period and after, while the number five may be the standard “Republican” number of witnesses and five men also attest a copy of a senatus consultum in AD 138; so both numbers can have a Roman provenance and clear Roman parallels.^[163] It seems reasonable to conclude, from looking at these Arabian documents, that the Roman authorities in this province could make up whatever specifications they liked, including using five rather than seven witnesses for attested copies, but tended to keep their requirements both traditional and consistent, thus creating paradigms that their provincial subjects could learn.

Finally, to go back to the world of the lower number of witnesses in the doubled documents of Arabia and Judaea. So far, it would appear that a doubled document with four witnesses, an odd number indeed, is a practice confined to Arabia, and mostly to Nabataean Arabia at that.⁷³ On the other hand, this does not seem nearly as odd as having no doubled documents with three witnesses from Arabia at all, which is the case. From Judaea, on the other hand, there are deeds of sale in Hebrew, a marriage contract, an acknowledgment of debt, and a divorce, all with three witnesses. And here one final factor may be helping to determine the number of witnesses, and may explain something about the customs governing the documents. In the documents of both Judaean and Arabian provenance, all - or virtually all - of the protagonists are Jews.⁷⁴ The Jewish rabbis - writing, to be sure, some time after all these documents were written - had opinions about, in particular, how many witnesses a document should have. They also distinguished between simple and “folded” documents. Thus, as one says, “A simple document, its witnesses are two, and a folded one, its witnesses are three. A simple one in which a single witness's signature is written and a folded one in which two witnesses' signatures are written, both of them are invalid.”⁷⁵ Or another: “a simple document has witnesses' signatures within. And one which is folded has the signatures behind. A simple document on which its witnesses signed on the back, or a folded document on which its witnesses signed on the inside, both of them are invalid.”⁷⁶ There was also debate among the rabbis about gentile witnesses and the propriety of using gentile courts.⁷⁷ A pattern starts to emerge here: all the single or “simple” documents with two or three witnesses from Judaea are also either in Hebrew and/or dated by the year

73 Although, as Hannah Cotton points out to me, perhaps we should take the scribe as the fifth witness. Since, however, the other documents have been categorised by counting witnesses but not including scribes as potential witnesses in that count, it seems best to note the possibility that scribes could count as witnesses here but leave it an open question.

74 Parties to the transactions and scribes “for the much greater part Jews living under Roman rule”: Cotton, “The Guardian (ΕΠΙΤΡΟΠΟΣ) of a Woman” (n 31) 267 and (quotation) R Katzoff and D Schaps, “Introduction”, in Katzoff and Schaps, Law in the Documents (n 8), 1, 4.

75 Quoted (MBB 10:2) in C Hezser, Jewish Literacy in Roman Palestine (Texts and Studies in Ancient Judaism 81) (2001) 304.

76 Quoted (MBB 10:1) in Hezser, Jewish Literacy (n 75) 305.

77 Citations in Hezser, Jewish Literacy (n 75) 305 and Cotton, “Guardianship of Jesus son of Babatha (n 32) 102.

“of the freedom of Israel” or the year of “bar Kockba, the prince of Israel”, while the Judaean double-documents with two or three witnesses display similar signs of observant Judaism, like the use of Hebrew (P Mur 29-30), or references to the ketubbah (P Mur 21) and “the law of Moses and the Jews” (P Mur 19).⁷⁸ Although these Jews would thereby be conforming to a standard of Jewish behaviour that was enunciated as a standard only later, conformity even to what was a prevalent habit makes sense as an explana­tion, for these Jews could have drawn up their marriage contracts with five witnesses, for example, but chose to use only three.

F.