"Rotuli": Liturgy Rolls and Formal Documents Daly, Lloyd W Greek, Roman and Byzantine Studies; Fall 1973; 14, 3; ProQuest pg. 333 Rotuli: Liturgy Rolls and Formal Documents Lloyd W. Daly I N a recent number of this journal I raised a question to which I knew no answer.1 The question was how should one explain the format of Greek liturgy rolls. As rolls they may be thought of as survivals of the ancient book roll, but this easy explanation will not stand up to critical consideration, since liturgy rolls are written trans­ versa charta (i.e., with their lines of writing at right angles to the long axis of the roll) in direct contrast to the practice for book rolls, and they are frequently opisthographic. Meanwhile this question has also been raised and studied by G. Cavallo, whose concern is primarily focused on the Exultet rolls.2 He reviews the possible references to rotuli and rightly stresses (p.218f) the importance of a letter of Pope Zacharias (741-52) to Saint Boniface wherein, in response to a request for guidance on the question of the number of places in the mass at which the sign of the cross should be made, the pope says, Hin rotulo ... per signa sanctae crucis quantae fieri debeant, infiximus." He concludes, after examining the evidence for Greek usage, that the Italian rolls, including that of Ravenna, are modelled on Greek prototypes, that the illustrated Exultet rolls are of Beneventan origin, and that the tenth century is a hypothetically plausible time for the development of these latter. My approach to the problem is a somewhat different one. Regard­ less of the earliest date at which we can find satisfactory evidence of the use of the rotulus form for the liturgy, what could have suggested this apparent departure from the familiar form of the book roll or from that of the codex? The answer is, I now think, a simple one. The model taken for the format of liturgy rolls was not literary but docu­ mentary. If one considers the question in the abstract, the prayers 1 GRBS 13 (1972) 111f. I "La genesi dei rotoli liturgici Beneventani aHa luce del fenomeno storico-librario in occidente ed oriente," in Miscellanea in memoria di Giorgio Cencetti (Torino 1973) 213-29. The author has had the great kindness to send me a copy of the first proofs of this valuable article, which had not yet been published at the time of this writing. 333 334 ROTULI: LITURGY ROLLS AND FORMAL DOCUMENTS which form the content of the liturgy may be aptly paralleled by petitions in the civil sphere. The evidence on which such a solution may be offered is ample. There is evidence as early as Ptolemaic times of the use of narrow strips of papyrus written transversa charta for all kinds of private documents.3 The Ptolemaic petitions known as enteuxeis present a peculiarly interesting variation on this format, for although their breadth is greater than their height, their material was, as Gueraud interprets the evidence, cut from a roll (or at least a strip), and in relation to the roll the writing would have been transversa charta, to take advantage of the lay of the fibers on the recto of the papyrus.4 This line of argument infers, although it does not necessarily pre­ suppose, that the use of this format for documents was continuous from Ptolemaic times. I think that there are indications that it was. There is, e.g., a lease dated A.D. 169-70 (P.Mii. YogI. 63) which is com­ plete and measures 35x 12 cm. Incidentally, such leases appear to have been in the form of petitions from the lessee to the lessor. P.Dura 31, a parchment, is a bill of divorcement from A.D. 204 and qualifies for a place in this series by its dimensions, 27x 14 cm. Wilcken also illustrates complete documents from subsequent cen­ turies.s His number XIV is dated to A.D. 289 and is a contract for the sale of a camel. Again his number XVI is a contract for the sale of a slave and is dated to the year 359. This document is made up of four sheets glued together, and Wilcken believed that there had originally been a fifth. In any case the measurements of the papyrus as it survives are 69x 23-27 cm.6 It may be worth noting that although this contract was found in Egypt it originated in Askalon. So far as I know the use of this documentary format has not been carefully studied. Without presuming to speak with any authority, it seems to me from the examples cited, and they could, I am sure, be multiplied, that the format was probably used throughout antiquity as an alternative if, as and when circumstances (including the kind of papyrus available) suggested its use. If we come closer in time to the presumptive origin of liturgy rolls a Cf P.Berol. 13070 (letter), 13065 (pact), 7017 (testament), 11064 (testimony). , O. Gueraud, Enteuxeis (Cairo 1931) xix ff. I Tafeln ZJlr iilteren griechischen Palaeographie (Leipzig 1891). • The measurements and description are given by Wilcken in Hermes 19 (1884) 417-31. LLOYD W. DALY 335 as we know them, there are impressive parallels. From the sixth cen­ tury, e.g., there is a Fayum papyrus bearing the record of a lawsuit and the resultant settlement. The document formed a roll, the pre­ served length of which is over six feet, the width one foot. And the writing is, as Kenyon says, "across the width of the papyrus."7 A somewhat later example is the testament of Abraham, Bishop of Hermonthis, in the eighth century. This document is also "written in one large column across the breadth" of the papyrus and it measures "3 ft. 8 in. by 1 ft. 2 in."8 These are only examples. In fact Mitteis and Wilcken observe that this came to be the common style for large documents in the Byzantine period, when texts of over a meter in length were written in this format.9 Thus it seems clear that documents in this format would have been very familiar indeed in the east throughout antiquity and the middle ages. But another and, I believe, better parallel for the liturgy roll's format is to be found in Byzantine imperial documents on parchment, such as chrysobulls. The memory of the appearance of such docu­ ments as seen at Vatopedi on Mount Athos and in the Vatican Arch­ ives suggests to me the appropriateness of the comparison, but it can be checked against D6lger's fine set of facsimiles.10 These documents range in date from the typikon of John Tzimiskes of 971/2 (316.5 x 48.5 cm.) to the prostagma of John VIII Palaeologus of 1459 (70.5 x 33.5 em.) and in length up to well over five meters.ll These are, indeed, im­ pressive documents, and none more so than the letter of John II Comnenus to Pope Calixtus II of the year 1142.12 With this document one may compare the roughly contemporary liturgy roll Vat.gr. 2281.13 Not only the format but the layout of the text and the ruling 1 P.Land. 113= Greek Papyri in tlte British Museum, Catalogue I pp.199ff. 8 P.Land. I 77, pp.231ff. D Grundzuge und Cltrestomatltie der Papyruskunde 1.1 (Leipzig 1912) xlvii. Leo Santifaller has collected much material on this subject in his "Beitdige zur Geschichte der Beschreibstoffe im Mittelalter," MOIG Erganzungsbd. XVI.1 (1953), but my examples are not all to be found there. 10 F. D6lger, Facsimiles byzantinischer Kaiserurkunden (Munich 1931). 11 Cf the list in L. Santifaller, "Uber spate Papyrusrollen und fruhe Pergamentrollen," in Speculum Historiale (Freiburg/Munich 1965) 131-33. 12 Cf DOlger, op.cit. (supra n.10) Taf. III. 13 Franchi de' Cavallieri and Lietzmann, Specimina codicum graecorum VaticanoTlLm (Berlin 1929) pU5. Cf the description in A. Turyn, Codices graeci Vaticani saecltlis XIII-XIV scripti (Vatican City 1964) 23-25. 336 ROTULI: LITURGY ROLLS AND FORMAL DOCUMENTS are suggestive.14 What these comparisons suggest to me is that the format of liturgy rolls was deliberately copied after that of imperial documents. That the prayers of the corporate church should be given a vehicle similar to that long in use for mundane petitions and that one of the most visible and august instruments of the church should be given the outer form of the most august document of the empire seems wholly fitting. As to the date at which the roll form was first used for the liturgy, there is still no satisfactory evidence. Although there are very lengthy papyrus rolls of documentary content, as noted above, it may be doubted whether the liturgy, subject to constant use and wear, was ever put on papyrus in this form. Our earliest liturgical text in roll form on parchment is the so-called Ravenna roll, which is dated by its most recent student to the seventh century.lS The question of papyrus or parchment in this connection may be entirely without significance. While there is an abrupt shift from papyrus to parch­ ment for Merovingian documents after 673, which presumably repre­ sents a cutting off of the supply of papyrus resulting from the Arab occupation of Alexandria, papyrus continued to be available to the papal chancellery in Rome and to the imperial chancellery in Con­ stantinople.1s Cavallo, in reviewing the evidence for the use of the rotulus form in the Greek east, comes to the conclusion (p.223): "Nella liturgia greco­ orientale, quindi, il rotolo ha una sua costante tradizione, forse gia dal V-VI secolo, in ogni caso sicuramente dall' VIII-IX." I would agree to the IX since we have at least two examples of that century in Sinait.gr. 59P' and Vat.gr. 2282,18 but I would also agree that the evidence for earlier uses is not convincing. The fact would seem to remain that, no matter whether for east or west, the Ravenna roll is our earliest Ii In describing the script of the fragments of the Rotullis Beratinus I noted the high.loop­ ing ligatures between certain letters (GRBS 13 [1972] 115). These may also be observed as a striking feature of the chancellery hand as practiced in XI-century documents. Cf. Dtilger, op.cit. (supra n.lO) Taf. IX and x. 16 See the exhaustive study of P. Suitbert Benz, Der Rotulus von Ravenna (Liturgiewissen­ schaftliche Quellen und Forschungen 45 [Munster 1967]) 21. 11 Cf. P. Lauer and Ch. Samaran, us diplames originaux des Merovingiens (Paris 1908), and L. Santifaller op.cit. (supra n.9) 27-32. 17 Cf. V. Gardthausen, Griechische Palaeographie2 I (Leipzig 1913) 152. 18 Cf. B.-Ch. Mercier, La liturgie de saint Jacques (Patrologio. Orientalis XXVI.2 [Paris 1946]) 134. LLOYD W. DALY 337 extant specimen, and the statement of Pope Zacharias is our earliest certain reference to rotuli. Parchment had, by the seventh century, long since come to be the material of preference for codices, and writing on both sides of the sheet was normal. Writing on both sides of the liturgy roll might, then, be expected. It is impossible to tell from catalogue descriptions of liturgy rolls just how common this practice was, but I suspect that many copies are opisthographic in spite of the lack of any notation to that effect.19 In any case the use of both sides of the parchment may readily be explained as a result of the economy of material or, per­ haps more probably, as a means of keeping the size of the roll within manageable proportions. Further light is thrown on this question of the use of the verso by the magnificent XI-century illuminated roll in the Library of the Greek Patriarch of Jerusalem.20 The recto of this roll originally re­ ceived the text of the liturgy of St John Chrysostom and its accom­ panying illuminations. About a century later the verso was used for a copy of the liturgy of St Basil. It seems to me that the use of the verso of even such a sumptuous roll suggests again economy and perhaps convenience as a motive. It is worth observing also that the figure of St John Chrysostom (Grabar's fig.6) holds his liturgy in roll form, identified by its opening words, K";PL~ 0 {hoc, so as to form the initial kappa, with the length of the extended rotulus serving as the diagonal strokes of the letter. The following illumination (fig.7) shows the Saint again, forming part of an initial epsilon; he stands in the temple and holds the roll out in both hands, apparently offering it to the Christ on the opposite side of the column of text, and in front of him stands a lectern apparently ready to receive the roll. The similarity of the non-figurative border ornament of this roll to that of imperial letters did not escape Grabar (p.199), and he offers for comparison the letters of John II and Manuel I Comnenus.21 Finally, with reference to the Rotulus Beratinus, which was my point 19 The list of Santifaller, op.cit. (supra n.l1) 129-31, records only two out of 22 examples as opisthographic, but his information also apparently comes from catalogue descriptions. It should be noted that his list gives only examples. It does not, e.g., include any of the very numerous copies in the monasteries of Mount Athos or even all of the copies in libraries from which he does give examples. 20 A. Grabar, "Un rouleau liturgique Constantinopolitain et ses peintures," DOPapers 8 (1954) 161-99. 11 Cf. Dolger, op.cit. (supra n.lO) Taf. Ill. 338 ROTULI: LITURGY ROLLS AND FORMAL DOCUMENTS of departure in this speculation, it seems to me that the similarity of its script to that of this originally Constantinopolitan exemplar strongly supports the XI-century date proposed for that less elegant specimen. UNIVERSITY OF PENNSYLVANIA April, 1973