COLIN H. ROBERTS
LONDON. Published for
THE BRITISH ACADEMY
by THE OXFORD UNIVERSITY PRESS
[[From: Robert Kraft
Sent: 13 December 2006 04:12
To: Fang-Fang Lam
Cc: Robert Kraft
Subject: Permission Procedures?
Some weeks before T. C. Skeat died, I had written him to explore the possibility of producing an updated version of "The Birth of the Codex" (1987 edition by T.C.Skeat, based on the earlier work of C.H.Roberts). Unfortunately, I never received a response. It is a tremendously useful handbook, but difficult for modern students (and some scholars) with minimal skills in the relevant languages (Greek, Latin, German, French, Italian) to use efficiently. I propose to supply English translations as appropriate. It also needs to be updated with reference to more recent work on the subjects covered.
Since I have been working in this subject
area for several years now,
in the intersections of papyrology and the study of early Judaism and
early Christianity, I am very interested in producing such an updated
version for use in the current worlds of scholarship and general
interest. I would want to make it available electronically on the
internet (with appropriate links to images, and to enable regular
updating), but am not opposed to hardcopy byproducts as well. Since the
British Academy holds the rights to the book, I'm inquiring how to
pursue such a request.
Thank you for your help.
R. A. Kraft
Emeritus Professor of Early Judaism and Early Christianity
> Dear Professor Kraft
> Thank you for your enquiry regarding the possibility of producing an updated version of "The Birth of the Codex".
> The volume is still available and continues to sell as an important scholarly contribution in its own right.
> We would have no objection to a new independent publication which builds upon the work, but we do not wish to produce a revised, or updated edition of "The Birth of the Codex" itself.
> Yours sincerely
> Amrit Bangard
> Amrit Bangard (Ms)
> Publications Assistant
> British Academy=20
> 10 Carlton House Terrace
> London SW1Y 5AH
> Tel: 020 7969 5216
> Fax: 020 7969 5414
> www.britac.ac.uk =20
EVIDENCE OF LEGAL WRITERS
+ Other "Paraliterary Formats and Practices
LITERARY TEXTS OF THE FIRST FIVE CENTURIES
+ Ancient Bookselling and
+ Roll and Codex in early visual representations
CODEX IN EARLY CHRISTIAN LITERATURE
[[fairly simplistic and historically uncritical]]
CHRISTIAN ADOPTION OF THE
CODEX: TWO HYPOTHESES
[[expand with additional hypotheses]]
CODEX IN NON-CHRISTIAN LITERATURE
[[update, combine with chapter 8]]
predecessor of this monograph, The Codex,
was published in the Proceedings of
the British Academy 40 (1954) 169-204 and
was substantially based on two lectures delivered as the Special
Lectures in Palaeography at University College, London, in January
1953. When stock was exhausted, it was clear that
in view of subsequent discoveries and further work on the subject more
reprint was called for. Since at that
time I was not free to undertake the revision myself, Mr T. C. Skeat
agreed to do it on my behalf. The present book, a completely revised
some respects enlarged version of its predecessor, is the result of his
for the structure of the whole and the first seven Sections he is
responsible. We have, however, collaborated throughout and the work as
stands represents our joint views.
Two books have greatly lightened our task, Sir Eric Turner's The Typology of the early Codex (University of Pennsylvania Press, 1977) and the Abbé Joseph van Haelst's Catalogue des Papyrus Littéraires Juifs et Chrétiens (Paris, 1976) and to their authors we wish to express our indebtedness.
C. H. Roberts
wax tablet, mid third century B.C.E. [at least four tablets hinged
horizontally, inscribed on only one side each]: account of
expenses incurred on journey in
By courtesy of the
+Sarcophagus representation, boy with scroll (folded) and open horizontally hinged tablet codex, and group with similar objects including an open scroll held with one hand, a horizontally hinged tablet codex, and a bundle of scrolls on the floor [Rome, Vatican Museum, Gallery of the Candelabrum. Credits: Barbara McManus, 2003];
on thin leather, second
century A.D., with notes of labor employed and payments made. Actual
measurements 7.5cm x l 1.8cm.
By permission and with the kind assistance of Professor Dr. Wolfgang Müller, Direktor des Aegvptisches Museums und Leiter der Papyrus-Sammlung der Staallichen Museen zu Berlin (P. Berol. 7358/9)
III Papyrus Codex of the Pauline Epistles, third century A.D.: the conjoint leaves show Romans 11.24-33 on the left and on the right the end of Philippians and the beginning of Colossians. Actual measurements 19 cm x 30 cm.
By courtesy of the
IV Parchment Codex of Demosthenes, De Falsa Legatione, second century A.D.: the plate shows the two pages of a bifolium, slightly reduced, each with two columns to the page. Actual measurement of a page 19 cm x 16.5 cm.
By permission of the British Library (ref. Add. MS. 34473, art. I)
V Parchment Codex of the Bible, Codex Sinaiticus, fourth century A.D.: the bifolium shows in part Psalms, xix-8 -- xxiii 5. Actual measurement of a page 37.6 cm X C. 24. 7 cm.
By permission of the
British Library (ref. Add.
MS. 43725, ff
VI Painting of a young man holding an open Codex, from the catacomb of SS. Peter and Marcellinus, Rome, third century.
This, the earliest representation of the codex in art, is an exception to the practice whereby in the early centuries the roll is the symbol of the book in Christian as well as in secular art.
By permission of the
Pontifica Commissione di Archeologia Sacra
||δελτάριον, γραμματεῖον||codex, pugillares|
|parchment (leather)||διφθέραι||membranae volumina|
|cabinet for books
THE MOST momentous development in the history of the book until the invention of printing was the replacement of the roll by the codex; this we may define as a collection of sheets of any material, folded double and fastened together at the back or spine, and usually protected by covers. There has never been any doubt about the physical origin of the codex, namely that it was developed from the wooden writing tablet; there should have been little doubt about the time when this development took place, although it has needed the impact of successive discoveries, mainly but not entirely in Egypt, during the present [20th] century to induce scholars to take notice of what their literary authorities told them. But the questions why this change took place when it did, in what circles the codex was first used, and why it eventually supplanted the roll, are more complex and uncertain. The aim of the present work is to suggest at least provisional answers based upon a reappraisal of our literary sources coupled with an analysis of the evidence from papyri.
It is no part of the plan of this work to attempt to compile a bibliography of the vast literature (much of it now antiquated and inaccurate, or falsified by subsequent discoveries) concerning the codex, its origins and development. Any worker in this field must begin by expressing his obligations to Theodor Birt's Das antike Buchwesen in seinem Verhältnis zur Literatur, Berlin, 1882, supplemented many years later by his Kritik und Hermeneutik nebst Abriss des antiken Buchwesens (Iwan v. Müller, Handbuch der Altertumswissenschaft, I. Band, 3 Abt., München 1913). As a collection of the literary material Birt's work is indispensable and calls for few supplements, but the eccentricity of its interpretations makes it an unsafe guide even to these sources. Much can be learned from W. Schubart's Das Buch bei den Griechen und Römern (2nd edition, Berlin, 1921; the so-called 3rd edition, by E. Paul, Heidelberg and Leipzig, 1961, though embellished with additional illustrations, omits the notes which are so valuable a feature of the 2nd edition), [] which still remains not only the most readable but also the most reliable introduction to the whole subject. There are many valuable observations in K. Dziatzko's Untersuchungen über ausgewählte Kapitel des antiken Buchwesens, Leipzig, 1900, supplemented by his articles 'Buch' and 'Buchhandel' in Pauly-Wissowa, Real Encyclopädie. Considering the period when he wrote, Theodor Zahn's admirable treatment of the evidence for the Christian book in his Geschichte der neutestamentlicher Kanons, i, pp. 60 sq. (Berlin, 1888) is vitiated only by the then common assumption that papyrus implies the roll and parchment the codex. All these discussions, even to a large extent that of Schubart, were written before the full effect of the Egyptian discoveries had been appreciated, and these set the sources the authors quoted in a different light. A notable attempt to re-assess the question against the background of these discoveries is that of H. A. Sanders, The Beginnings of the Modern Book: the Codex,\1/ University of Michigan Quarterly Review, 44, no. 15, Winter 1938, pp. 95-111, while among studies which have appeared since the first edition of the present work, mention may be made of H. Hunger, O. Stegmüller, and others, Geschichte der Textüberlieferung der antiken und mittelalterlichen Literatur, Zürich, 1961, especially pp. 47-51 (Hunger), 346-50 (K. Büchner). F. Wieacker, Textstufen klassischer Juristen (Abhandlungen der Akademie der Wissenschaften in Göttingen, Phil.-hist. K1., 3. Folge, Nr. 45, 1960), especially in his § 4, 'Rolle und Codex, Papyrus und Pergament', discusses the transition from roll to codex in relation to his principal thesis, namely that the works of the classical jurists (Ulpian, Paulus, etc.) were originally published in rolls, and were transferred to codices circa 300 C.E., and that hand in hand with this transference went a re-edition of the works themselves. Tönnes Kleberg, Buchhandel und Verlagswesen in der Antike, Darmstadt, 1969, includes (pp. 69-86) an "Exkurs über die Buchherstellung und die Formen des Buches in der Antike" which provides an excellent summary of the question. Sir Eric Turner's The Typologv of the Early Codex, 1977, though a mine of information concerning all physical aspects of the codex, explicitly (cf. pp. 1-2) excludes any discussion of the origin of the codex form. The latest treatment, by Guglielmo Cavallo in his composite [] volume, Libri, Editori e pubblico nel Mondo antico, 1975, is considered in Section 12 below. It should be added that the task of assembling the data on texts other than Christian has been immeasurably lightened by the publication of Roger A. Pack, The Greek and Latin Literary Texts from Greco-Roman Egypt, University of Michigan Press, 1st edition 1952, 2nd edition 1965, here referred to as Pack-1 and Pack-2. In the predecessor of the present work the evidence was based on Pack-1; here it has been revised with the aid of Pack-2 and brought up to date with the aid of other bibliographies. For Christian texts the bibliographies of Kurt Aland and Joseph Van Haelst mentioned below (p. 38) have been of outstanding value.[+add references to new online resources]
\1/ See also the articles of C. C. McCown, 'Codex and Roll in the New Testament', Harvard Theological Review 34 (1941) 219-250, and ‘The earliest Christian books' in The Biblical Archaeologist 6 (1943) 21-31.
This introductory section
may suitably close with a warning. An overwhelming proportion of the
Two passages which sum up the difficulties and dangers in evaluating the material may be quoted here. The former is from T. Kleberg's Buchhandel und Verlagswesen mentioned above (p. 67): [] "This presentation could provide only some fragmentary witnesses from the history of ancient book trade. But we must recognize that actually everything that we know in general about this detail of ancient life consists of fragmentary episodes that, taken together, attest to different situations and must be filled out through inferences that are not always as well grounded. And so it is generally with most areas of ancient everyday life. The ancient authors very seldom provide us with complete coherent portrayals. At most we must content ourselves with individual sparse notices that are strewn about in the large portions of the surviving literature and in inscriptions."\4/ The same point had been made long before, and even more incisively, by Prof. F. Zucker in a review of K. Ohly's Stichometrische Untersuchungen; "I need to point out that in general, with respect to knowledge of books, we are dependent on the proposal of possibilities to a much greater extent than one often appears to recognize. The material is dangerously irregular, in some respects exceedingly rich, in others very poor. Above all one must be forewarned about filling out gaps in our knowledge on the basis of certain general assumptions that seem to us to be obvious."\5/
Darstellung konnte nur einige bruchstückartige Züge aus der
Geschichte des antiken
Buchhandels bieten. Aber wir müssen tatsächlich feststellen,
dass alles, was wir überhaupt von dieser Einzelheit des antiken
bruchstückartige Episoden sind, die zusammengestellt, von
aus beleuchtet und durch nicht immer gleich gut begründete
ergänzt werden müssen. So steht es übrigensmit den
meisten Gebieten des
antiken Alltagslebens. Die Schriftsteller der Antike bieten uns
äusserst selten vollstandige zusammenhangende Schilderungen. Meist
müssen wir uns mit einzelnen spärlichen Notizen
sich über grosse Teile der erhaltenen Literatur
und in Inschriften verstreut finden' [English
translation supplied by RAK]. A
very similar warning is given by Schubart, Das
Buch ...2, p. 36
\5/'Ich möchte überhaupt grundsätzlich bemerken, dass wir im Buchwesen in weit grösserem Ausmass als man vielfach anzunehmen scheint, auf die Erwägung von Möglichkeiten angewiesen sind. Das Material ist gefährlich ungleichmässig, in mancher Hinsicht überaus reich, in mancher überaus dürftig. Vor allem muss man davor warnen, Lücken unserer Kenntnis auf Grund gewisser allgemeiner Vorstellungen auszufüllen, die uns selbstverständlich erscheinen' (Gnomon 8  384) [English translation supplied by RAK].
PAPYRUS AND PARCHMENT
Nevertheless, since it has been
seriously claimed that
the increasing use of parchment in some way promoted the transition
to codex,\7/ it seems desirable to consider briefly both these
sources of information. The history of
papyrus from every aspect in the period which concerns us is amply
Naphtali Lewis, Papyrus in Classical
Antiquity (Oxford, 1974), a new and enlarged edition of his
well-known L'Industrie du Papyrus dans l’Egypte
(Paris, 1934). Until
recently no similar study has been
devoted to the history of parchment, but now a full-scale scientific
technical investigation is available in R. Reed, Ancient
Skins, Parchments and Leathers (Seminar Press, 1972).\8/ To
complement this there is a useful collection of the historical and
evidence in the University of California dissertation of Richard R.
Johnson, The Role of Parchment in Greco-Roman
1968 (published by University Microfilms in both microfilm and xerox
One of Johnson's principal services is
elucidate the confused and partly contradictory accounts of the
To explain the eventual
parchment a number of reasons have been put forward, and although most
have little bearing on the origin and development of the codex, they
briefly considered here. The
comparative qualities of
have often been compared, usually to the disadvantage of the
durability of both under normal conditions is not open to []
doubt. Many instances of the long life
of writings on papyrus could be quoted, but this is no longer
the myth that papyrus is not a durable material has at last been
authoritatively and, one would hope, finally refuted by Lewis (op.
cit., pp. 60-61). At
the same time Lewis finds no difficulty in
dispelling another popular delusion, namely that papyrus was
fragile and brittle material.\12/ He demonstrates that it was in fact
strong and flexible. Wieacker's claim
that parchment was preferred for the codex because papyrus was too
fold is totally without foundation.
question which has often been fruitlessly debated is whether papyrus or
parchment was the more costly material -- fruitlessly because
are almost wholly lacking. Richard R. Johnson (op. cit., pp. 113-117)
quotes a number of earlier opinions,\13/ but finally concludes that the
is both unanswerable and meaningless. The
great difficulty is that we have no comparative figures for the cost of
and parchment during the same period of time. Of
the few certain prices of papyrus rolls
collected by Lewis (op. cit., pp. 131-134)\14/
the latest (10 dr. 3 chalk.) is dated third century [[C.E.]], but as
must antedate the massive inflation which marked the latter part of the
century. Conversely, the only certain
price recorded for parchment is that given in Diocletian's Maximum
Edict of 301 C.E.;\15/ and there is no way in which the one can
be balanced against the
Despite all that has been said above, even the strongest supporters of papyrus [[perhaps??]] would not deny that parchment of good [] quality is the finest writing material ever devised by man. It is immensely strong, remains flexible indefinitely under normal conditions, does not deteriorate with age,\16/ and possesses a smooth even surface which is both pleasant to the eye and provides unlimited scope for the finest writing and illumination. Above all, it possesses one outstanding advantage over papyrus: whereas production of papyrus was limited to
Why, and when, parchment replaced papyrus
is a complex question detailed discussion of which is outside the scope
book. The manufacture of papyrus in
\17/ Lewis, op. cit., pp. 92, 94, n. 10.
\18/ Cf. Lewis, op. cit., pp. 90-4, and the article by E. Sabbe, 'Papyrus et parchemin au haut moyen age,' Miscellanea historica in honorem Leonis van der Essen 1 (1947) 95-103.
mentioned, parchment had the advantage over papyrus in that it could be
manufactured virtually anywhere. At
first sight this advantage would seem to be so overwhelming that one is
inclined to pose the question, not in the form 'Why did parchment
papyrus?', but rather 'Why did parchment take so long to replace
there is a technological factor which has not hitherto been
appreciated. Whereas the manufacture of
papyrus, like that of paper, is basically a simple and straightforward
and the technical skills necessary had in any case been elaborated by
Egyptians over thousands of years, the production of parchment poses
different problems, the nature of which can best be illustrated by the
following quotations from R. Reed, Ancient
Skins, Parchments and Leathers:
It is perhaps the extraordinarily high durability of the product, produced by so simple a method, which has prevented most people from [] suspecting that many subtle points are involved.... The essence of the parchment process, which subjects the system of pelt to the simultaneous action of stretching and drying, is to bring about peculiar changes quite different from those applying when making leather. These are: (1) reorganisation of the dermal fibre network by stretching, and (2) permanently setting this new and highly stretched form of fibre network by drying the pelt fluid to a hard, glue-like consistency. In other words, the pelt fibres are fixed in a stretched condition so that they cannot revert to their original relaxed state (pp. 119-120).\19/
Where the medieval parchment makers were greatly superior to their modern counterparts was in the control and modification of the ground substance in the pelt, before the latter was stretched and dried .... The major point, however, which modern parchment manufacturers have not appreciated is what might be termed the integral or collective nature of the parchment process. The bases of many different effects need to be provided for simultaneously, in one and the same operation. The properties required in the final parchment must be catered for at the wet pelt stage, for due to the peculiar nature of the parchment process, once the system has been dried, any after treatments to modify the material produced are greatly restricted. (p. 124).
This method, which follows those used in medieval times for making parchment of the highest quality, is preferable for it allows the grain surface of the drying pelt to be "slicked" and freed from residual fine hairs whilst stretched upon the frame. At the same time, any processes for cleaning and smoothing the flesh side, or for controlling the thickness of the final parchment may be undertaken by working the flesh side with sharp knives which are semi-lunar in form. . . . . To carry out such manual operations on wet stretched pelt demands great skill, speed of working, and concentrated physical effort. (pp. 138-9).
Enough has been said to suggest that behind the apparently simple instructions contained in the early medieval recipes there is a wealth of complex process detail which we are still far from understanding. Hence it remains true that parchment-making is perhaps more of an art than a science. (p. 172).
From these statements it will be clear that a parchment industry on a scale adequate to enable it to challenge the [] dominance of papyrus could not have been created overnight. Many years -- perhaps even centuries -- would have been required to work out the details of the process by trial and error, and to build up and train a sufficient labor force spread over the length and breadth of the
This brief survey will, it is hoped, be
sufficient to show that the transition from papyrus to parchment was of
entirely different character from, and quite unconnected with, the
from roll to codex, to which we will now turn.
[[add notes \20/
and \21/ or modify numbers]]
THE WRITING TABLET
THE writing tablet need not long detain us. It was commonly formed of two or more flat pieces of wood, held together either by a clasp or by cords passed through pierced holes; the central area of the tablet was usually hollowed slightly to receive a coating of wax, while a small raised surface was often left in the centre to prevent the writing on the wax being damaged when the tablet was closed. [[Such tablets are depicted in both a "vertical" form (opening away from the user) and a "horizontal" form (opening to the side), with multiple pieces sometimes attached accordian style (see n.23 below). These variations also mirror different ancient scroll formats (vertical and horizontal).]] Writing in ink or chalk was sometimes placed directly on the wood. It was one of the oldest, if not the oldest,\22/ recipient of writing known to the Greeks, who may have borrowed it from the Hittites.\23/ [[add information on its wide use in the Near East]] Homer knew of it, for it was on a folded tablet or diptych [["two piecer"]] that Proitos scratched the 'deadly marks' (Iliad 6.168 sq.) that were intended to send Bellerophon to his death. To the Greeks of the classical age the tablet had a tradition behind it and a dignity that the papyrus roll lacked;\24/ in Sophocles, Agamemnon orders the muster roll [list] of the Greek princes to be read from a tablet, and it is on a tablet that Zeus, in a fragment of Euripides, records the sins of men.\25/ In later Greece they [tablets] were the familiar recipient of anything of an impermanent nature -- letters, bills, accounts, school exercises [+add somewhere the details from Cribiore's research], memoranda, a writer's first draft. Already in the [] fifth century [[B.C.E]] tablets of several leaves were in use,\26/ but the nature of the material would set a limit to their number, and in fact no specimen surviving from antiquity has more than ten.\27/ [[add rabbinic reference to 12?]] The earliest surviving Greek tablets, seven in number, date from the middle of the third century B.C.E. All surfaces [[? both sides??]] were covered with wax, sometimes black, sometimes red; they contain rough accounts of expenses during a journey on the
\26/ Cf. Euripides, I.T. 727, δέλτου μὲν αἵδε πολύθυροι διαπτυχαί (of a tablet, then, many folding pages). Schubart's comment (op. cit., p. 175) that πτυχή (fold) is not strictly applicable to a hard material such as wood, and that therefore in this passage it implies a previous use of folded leather, papyrus, etc., is misconceived, since πτυχή can be used of the folds of doors. Cf. LSJ and Pollux, Onomast., ed. Bethe, i, p. 207 [= TLG 418, 2nd century CE]: καὶ Ἡρόδοτος (VII 239) μὲν λέγει `δελτίον δίπτυχον,’ οἱ δ’ Ἀττικοὶ ‘γραμματεῖον δίθυρον,’ καὶ θύρας τὰς πτύχας ἄχρι δύο, εἶτα πτύχας, καὶ τρίπτυχον καὶ πολύπτυχον (and Herodotus said "two-fold tablet," but the Attic commentators "two-paged notebook," and pages/doors the folds until two, then folds, even tri-fold and multiple-fold) [[check ET of Herodotous 7.239]].
the uses to which tablets were put see Schubart, Das Buch...2,
pp. 24 sqq., and notes, p. 175; the ninefold wax
tablet illustrated on p. 24 must originally have had ten leaves (see
article referred to by Schubart, p. 175). P. Fouad 74 of the fourth
century C.E. refers to and describes a
δελτάριον δεκάπτυχον (ten folded
little tablet). [[Jewish rabbinic
literature refers to
a 12 leafed version -- see Lieberman, etc.]]
\28/ Published by H.
I. Bell and Flinders
\29/ Ep. 3.5.15 sq. [[To Baebius Macer -- 10 Post cibum saepe - quem
interdiu levem et facilem veterum more sumebat - aestate si quid otii
iacebat in sole, liber legebatur, adnotabat excerpebatque. Nihil enim
legit quod non excerperet; dicere etiam solebat nullum esse librum tam
malum ut non aliqua parte prodesset. 11
Post solem plerumque frigida lavabatur, deinde gustabat dormiebatque
minimum; mox quasi alio die studebat in cenae tempus. Super hanc liber
legebatur adnotabatur, et quidem cursim. . . . 15 In itinere quasi solutus
ceteris curis, huic uni vacabat: ad latus notarius cum libro et
pugillaribus, cuius manus hieme manicis muniebantur, ut ne caeli quidem
asperitas ullum studii tempus eriperet; qua ex causa Romae quoque sella
vehebatur. 16 Repeto me correptum
ab eo, cur ambularem: 'poteras' inquit 'has horas non perdere'; nam
perire omne tempus arbitrabatur, quod studiis non impenderetur. 17
Hac intentione tot ista volumina peregit electorumque commentarios
centum sexaginta mihi reliquit, opisthographos quidem et minutissimis
scriptos; qua ratione multiplicatur hic numerus. Referebat ipse
potuisse se, cum procuraret in Hispania, vendere hos commentarios
Larcio Licino quadringentis milibus nummum; et tunc aliquanto pauciores
[ET Harvard Classics Letter #27] After a short and light refreshment at noon (agreeably to the good old custom of our ancestors) he would frequently in the summer, if he was disengaged from business, lie down and bask in the sun; during which time some author was read to him (liber legebatur), while he took notes and made extracts (adnotabat excerpebatque), for every book he read he made extracts (excerperet) out of, indeed it was a maxim of his, that “no book was so bad but some good might be got out of it.”  When this was over, he generally took a cold bath, then some light refreshment and a little nap. After this, as if it had been a new day, he studied till supper-time, when a book was again read (liber legebatur) to him, which he would take down running notes upon (adnotabatur). . . .  A shorthand writer [notarius] constantly attended him, with book and tablets [cum libro et pugillaribus], who, in the winter, wore a particular sort of warm gloves, that the sharpness of the weather might not occasion any interruption to my uncle’s studies: and for the same reason, when in Rome, he was always carried in a chair.  I recollect his once taking me to task for walking. “You need not,” he said, “lose these hours.” For he thought every hour gone [lost] that was not given to study.  Through this extraordinary application he found time to compose the several treatises [volumina] I have mentioned, besides one hundred and sixty volumes of extracts [electorumque commentarios] which he left me in his will, consisting of a kind of commonplace, written on both sides [opisthographos], in very small hand [minutissimis scriptos], so that one might fairly reckon the number considerably more. He used himself to tell us that when he was comptroller of the revenue in Spain, he could have sold these manuscripts to Largius Licinus for four hundred thousand sesterces, and then there were not so many of them.]
The correct designation in Latin for a plurality of tablets or for multi-leaved tablets was codex, whether the material used was wood, as was usual or, e.g., ivory. When Seneca [[the Younger]] enlarges\30/ on that inane studium supervacua discendi, [[vain passion for learning (digressing on ??) useless things ]], an infection the Romans had contracted from the Greeks, he cites as an example the enquiry [] whether Claudius Caudex, one of the consuls of 264 B.C.E., was so called 'quia plurium tabularurn contextus caudex apud antiquos vocabatur, unde publicae tabulae codices dicuntur.'[[because among the ancients a structure formed by joining together several tablets was called a caudex, whence also the Tables of the Law (?? public tablets??) are called codices]]. Already in the time of Cato the Censor\31/ the words tabulae and codex were interchangeable, and both are frequently found in Cicero for tablets used for business purposes.\32/ But neither now nor for a long time to come was there any question of the word codex denoting a book.
\32/ G.E.M. de Ste Croix, 'Greek and
Roman Accounting' in Studies in the
History of Accounting, ed. A.
C. Littleton and B. S. Yamey, 1956, pp.
41-3; P. Jouanique, 'Le codex accepti et expensi chez Cicéron', Revue
historique de Droit francais et étranger
46 (1968) 5-31.
The second passage, on which much ink has
been spilt to little profit, is the
statement by the elder Pliny that Cicero had reported a copy of the
parchment which could be enclosed in a [] nut
nuce inclusam Iliadem Homeri carmen in membrana tradit
The trouble is that none of the scholars
who have commented on this passage have investigated the subject of
writing, and therefore have no conception of what can be achieved by
and application. To take but a single
example, Harley MS. 530 in the
approximately, the Bible is six times as long as the Iliad. [+ Mani codex size, etc.]
FROM WRITING TABLET TO PARCHMENT
It would seem that it was
Romans, rather than the Greeks, who developed the
writing tablet to a size where it could accommodate lengthy accounts
distinguished, as the Greeks did not, between the large tablet and the pugillares
that could be held in a
closed hand). Certainly it was the
Romans who took the decisive next step, that of replacing the wooden
a lighter, thinner and more pliable material [[including very thin
wood, as at Vindolanda]]. We have seen
according to our literary evidence, the
have been made familiar with parchment as a writing material before the
of the second century B.C.E. But
also our sources suggest, it was intended as a substitute for papyrus,
probably have been used, like papyrus, in roll form.\41/ In any
case it is
probable that after the temporary interruption of supply [[in the early 2nd century B.C.E.]]
former predominance, though some knowledge of the usability of
Evidence from the last years of the Republic is scanty and of doubtful interpretation. We have already dismissed the suggestion that the codices librariorum which contributed to the funeral pyre of Clodius were of parchment [[above, at n. 34]]. A letter of Cicero to Atticus, written in 45 B.C.E., contains the sentence 'Quattuor διφθέραι sunt in tua potestate' [[four διφθέραι [leather copies of Varro's writings] are in your possession]], and it has been conjectured that these parchments were in the form of rolls; but there is so much uncertainty about the interpretation of the passage that it cannot safely be used as evidence.\43/ All we can perhaps infer is that Cicero's use of the Greek word διφθέραι (leather pieces) indicates that although the use of parchment as a writing material was known in cultured [] Roman society, it was not sufficiently familiar for it to have a recognised Latin equivalent. [[?? analogies show that language doesn't always develop that way!]]
Another passage which has often been quoted in this connection is Catullus 22.4-8:
Ad fam. 7.18.2.
\47/ Cf. R. R. Johnson, op. cit., pp. 60-61. [[G.O.Hutchinson, Cicero's Correspondence: A Literary Study (Oxford University Press 1998) 183-185 (he seems to assume papyrus). In the context, Cicero wonders if his own letters are being thus reused.]]
all the evidence points to the conclusion that a 'palimpsest'
[] was a papyrus from
which the writing had been removed to enable it to be re-used.
\48/ Kritik und Hermeneutik, p. 290; cf. R. R. Johnson, op. cit., p. 61.
However, by far the greatest source of
been the employment by modern palaeographers of the convenient term
'palimpsests' or its factitious Latin equivalent 'libri rescripti' to
re-written parchment manuscripts,
with the result that the word 'palimpsest' has become inextricably
the use of parchment, in defiance of all the ancient evidence. This
misconception has coloured all
discussion of the Catullus passage down to the present day.
\52/ Op. cit., p. 124
\53/ An Introduction
to Greek and Latin Palaeography, 1912,
p. 46, n 3.
\54/ For the Ptolemaic evidence see J. Vergote, Le Muséon 59 (1946) 253-8.
most recently E. G. Turner, The
Terms Recto and Verso: the Anatomy of
the Papyrus Roll (Papyrologica Bruxellensia 16), 1978, pp. 27-32.
[[or possibly wood?? see
nowVindolanda; Parsons review, reporting their argument: "...transversa charta means 'across
the fibres', so that Caesar will have sent his despatches from Gaul not
in the form of a rotulus, but
'in the forma of a note-book' (so they interpret memorialis libelli, plausibly,
against Turner) -- did he have this partly in mind, when he entitled
the collected version Commentarii?"]]
Positis capillis clearly refers to the depilation of the skin in the process of parchment making. Bicolor is more difficult to interpret, but probably refers to the difference in color between the flesh-side of parchment and the hair-side, which is often markedly [] yellower.\60/ This difference would not leap to the eye in the case of a parchment roll, in which all the membranes would be sewn together the same way round, but it would be very noticeable in the case of a parchment note-book in codex form, particularly if care had not been taken to arrange the leaves so that flesh- and hair-sides faced each other. The passage may thus be taken to indicate that the membranae, at this period if not earlier, were parchment note-books in codex form.
The final proof is provided by Quintilian (circ. C.E. 90), who gives the following advice: [[It is best to write on wax owing to the facility which it offers for erasure, though weak sight may make it desirable to employ parchment by preference. The latter, however, although of assistance to the eye, delays the hand and interrupts the stream of thought owing to the frequency with which the pen has to be supplied with ink (32) But whichever is employed, blank pages (tabellae) must be left in which one is free to make additions at will.]] Scribi optime ceris, in quibus facillima est ratio delendi, nisi forte visus infirmior membranarum potius usum exiget... relinquendae autem in utrolibet genere contra erunt vacuae tabellae, in quibus libera adiciendi sit excursio.\61/ Here the allusion to the wax tablet and the blank pages show that the codex form was in question. With Quintilian we have reached a stage in the history of the codex when it is more than a tablet but still less than a book.
\63/ So Schubart, op. cit., pp. 114 sq.
As was mentioned at the beginning
this section, all
the evidence points to the parchment note-book having been a Roman and
Greek invention. This is neatly
confirmed by the only Greek writer of the first century C.E. to mention the parchment note-book,
Apart from Paul [[quotes again?]], the only
two centuries C.E. to
parchment note-book is Galen. In his De Compositione
Medicamentorum he discusses a preparation alleged to be useful in
the spread of baldness and mentions that his friend Claudianus (himself
celebrated doctor) had come across it in a parchment note-book which he
acquired after the owner's death.\66/ Why Galen should have recorded
the form in
which the recipe was found is not clear, unless it was to indicate that
from a private compilation not intended for publication.
It is, however,
unnecessary to pursue the parchment note-book further, since already
end of the first century C.E. a surprising and, as it turned out,
had been taken in the evolution of the codex as a literary form; this
the subject of the next section.
MARTIAL AND THE FIRST APPEARANCE OF
A LITERARY FORM
itself the parchment note-book does not take us very far.
In the first two centuries of the Empire
polite society [[seems to have]]
acknowledged one form and one form only for the
[[literary]] book --
roll. Such was the force of convention
that even when the codex was in common use for books Augustine feels
apologize for writing a letter in codex form,\67/ and Jerome, who
that he is a gentleman as well as a scholar, writes his letters
rolls, even though he keeps his books in codices.\68/ The first hint
dominance of the roll is to be challenged comes towards the end of the
century. We have noticed (p. 18) that
Suetonius goes out of his way to mention Julius Caesar's idiosyncratic
writing his dispatches; and the reason why this impressed him may be
the works of his contemporary Martial, where we have the first
reference to literary publication in codex form.\69/ The evidence is
confined [[mainly]] to 1.2 -- a poem
introducing a revised edition of Books I and
II reissued together\70/ and to a number of verses in the Apophoreta;
all alike fall within the years C.E.
84-86.\71/ The former runs as follows:
|Qui tecum cupis esse meos ubicumque libellos||You who are keen to have my books with you everywhere|
|Et comites longae quaeris habere viae,||and want to have them as companions for a long journey,|
|Hos eme, quos artat brevibus membrane tabellis:||Buy these ones which parchment confines within small leaves [check uses of tabelli];|
|Scrinia da magnis, me manus una capit.||provide cylinders for the great [authors] -- one hand can hold me.|
|Ne tamen ignores ubi sim venalis et erres||So that you may not fail to know where I am for sale, and wander|
|Urbe vagus tota, me duce certus eris:||aimlessly throughout the whole city, with me as guide you will be certain:|
|Libertum docti Lucensis quaere Secundum||Look for Secundus, the freedman of learned Lucensis|
|Limina post Pacis Palladiumque forum.||behind the threshold of the Temple of Peace and the forum Palladium.\71+/|
[[to Maximus: To our honorable
brother Maximus we gave/sent letters, thinking that he would receive
them appreciatatively. Whether, however, we might be of some help, on
the present occasion, than you are able to ascertain, we deign to
rewrite. Know that without wordy letters to those most familiar to us,
not only laity but even bishops, is how this is written, to those we
only write so that they also might quickly write and using papyrus
which is appropriate for them to read, lest if this habit of ours is
misunderstood, it is deemed a wrong to you. Ad
honorabilem fratrem nostrum Maximum litteras dedimus,
credentes eum gratanter eas suscepturum: utrum
tamen aliquid profecerimus, proxima occasione, quam reperire
potueris, dignare rescribere. Sciat sane prolixas
epistolas ad familiarissimos nostros, non solum laicos, verum
etiam episcopos, sic quomodo ista scripta est, ad eos
scribere nos solere, ut et cito scribantur, et charta teneatur commodius cum leguntur,
ne forte istum morem nostrum
nesciens, factam sibi arbitretur injuriam. (ed note 0751 Sic
optimae notae Mss. At Lov.: Nos scribere non solere,
ut et cito scribantur, et certa teneantur.) see also n. 108a below and provide ET;
were literary letters expected to be on scrolls? The Augustine passage
seems only to indicate that papyrus is being used [not parchment?], not
that the format is a codex.]]
\69/ The word codex
is never applied by
Martial to the books in question. [[The
term is first used for a proper book by Commodian, Carmen
Apologeticum 11; Seneca used it earlier for the notebook (so
Resnick n. 5, above n.30)]]
possibly of Books I-VII. See also
Evan T. Sage, "The Publication
of Martial's Poems," Trans. Am. Phil.
168-176, who opts for only book 1.
||Tablets/notebooks of citrus wood
|Esse puta ceras, licet haed
||Suppose it to be wax, though it
is called parchment
|delebis, quotiens scripta novare
||You will erase whatever you want
to write anew.
||Vitellian tablets [for love
||The same [requesting money]
||Large sheets of papyrus
||Papyrus sheets for letters
||Homer's "Battle of Frogs and Mice"
||Homerus in pugillaribus membraneis.||Homer in hand-held parchments (notebooks?)|
|Ilias et Priami regnis inimicus Ulixes||The Iliad and Ulysses, enemy of Priam's kingdom,|
|Multiplici pariter condita pelle latent.||are there together, preserved in many folds of skin|
|14.186||Vergilius in membranis.||Vergil on parchment|
|Quam brevis inmensum cepit membrana Maronem!||How small a quantity of parchment has comprised vast Maro!|
|Ipsius vultus prima tabella gerit.||The first leaf bears his own countenance|
||Menandrou Qais [Greek]
|14.188||Cicero in membranis.||Cicero on parchment|
|Si comes ista tibi fuerit membrana, putato||If this parchment will be your companion, suppose|
|Carpere te longas cum Cicerone vias.||yourself to be making a long journey with Cicero|
||The "Monobiblos" of Propertius
|14.190||Titus Livius in membranis.||Titus Livy on parchment.|
|Pellibus exiguis artatur
||Compressed in tiny skins vast
|Quem mea non totum bibliotheca capit||for whom complete my library has
|14.192||Ovidi Metamorphosis in membranis.||Metamorphoses of Ovid on parchment|
|Haec tibi, multiplici quae structa est massa tabella,||This
mass that has been built up for you
with multifold tablets
|Carmina Nasonis quinque decemque gerit.||contains the fifteen lays of Naso|
||Calvi de aquae frigidae usu
||Calvus "On the Use of Cold Water"
\72/ Viz. 184 (Homer), 186 (Virgil) (see also 1.53.n2) ], 188 (Cicero) , 190 (Livy) , 192 (Ovid, Metamorphoses). [[ From searching ET for "page," and "book" -- See also 2.77 on the "double-sided page" for lengthy epigrams (saepe duplex unum pagina tractat opus -- often two pages of Marsus and of learned Pedro treat of a single theme) -- do a search for Martial's use of "pagina"? Also 11.16 "little book [libelli] ... shut my volume" (check word for "shut" -- posuitque meum ... librum)]] See also
praises, loves, and recites my little books, Rome
And every pocket, every hand has me.
(Laudat, amat, cantat nostros mea Roma libellos,
meque sinus omnes, me manus omnis habet.)11.3 but my book . . . is frequented by the hardy/rigid centurion
and Britain is said to recite our verses;
. . . But what immortal papyri-pages [verses] could I compose . . .
(sed meus . . .
a rigido teritur centurione liber,
dicitur et nostros cantare Brittania versus
. . . at quam victuras poteramus pangere chartas . . . )
\73/ On the question whether this should read membraneis or membranes see Birt, Buchwesen, p. 85; F. Bilabel, art. "Membrana" in Pauly-Wissowa, RE, points out that Martial's terminology finds a parallel in CIL 10.6.8, an undated inscription mentioniing pugillares membranacei operculis eboreis [[parchment notebooks ... ET]]; he suggests that it records a gift of books to the
. Templeof Apollo
\74/ Cf. the discussion in Section 9 of reasons, real or imagined, for the superiority of the codex over the roll.
\75/ E.g. by Birt, Die Buchrolle in der Kunst, p. 31, and Abriss, p. 353.It has also been questioned whether they were normal books in the sense of complete texts, or whether they were anthologies or extracts of some kind.\76/ This doubt is certainly misplaced in the case of the Homer and the Virgil (the epigram on the latter, with its immensum Maronem, would lose its point if an anthology were in question), while the Ovid is explicitly stated to contain the entire Metamorphoses. The
This has been doubted, both on internal and external grounds. The internal grounds are founded upon Martial's use of the word artatur. Originally Birt, in his Buchwesen (pp. 85 sq.), regarded these codices as containing the complete works in each case (except for the Cicero, where there is nothing in the text of Martial to oblige us to think that anything more than one or two works by Cicero were included), but by the time he wrote his Abriss he regarded them all as epitomes or anthologies, in defiance of the clear meaning of the Latin in at least two instances, namely Epigram 186 (Virgil) and 192 (Ovid). To justify his view Birt appealed (p. 349) to the meaning of the word artare, which he claimed implies an epitome or abridgment; but he begs the question by arguing that when Martial uses the same term for the collection of his own early epigrams in 1.2 (hos eme, quos artat brevibus membrana tabellis [[buy these ones which parchment confines within small leaves]]) this re-edition was merely a selection -- a theory for which there is no evidence whatever.
\77/ For coartare
= abridge cf. Cicero, De
Orat. 1.163, Seneca, Ep.
\78/ The Twelve (Minor) Prophets contain 3000
stichometry of Nicephorus,
and could thus have easily been accommodated in a very thin (cf. angustias)
\79/ Trans. Am. Phil. Ass. 82
(1951) 248-9. See also Evan T.
Sage, "The Publication
of Martial's Poems," Trans. Am. Phil.
Ass. ... (1919)
\80/ Cf. also L. Ascher, 'An Epitome of Livy
in Martial's day?' The Classical
Bulletin 45 (St Louis 1969) 53-54.
\81/ R. R. Johnson, op. cit., pp. 77-8 fully supports Oliver.
One other question which
these epigrams raise admits of no answer.
The gifts which are the subject of the Apophoreta
are divided into those intended for the rich and those
intended for the poor, and the objects are correspondingly expensive or
cheap. The epigrams are arranged in
pairs; in each pair, as Martial himself explains (14.1, Divitis
alternas et Pauperis
accipe sortes [[ET]]), one epigram describes an expensive
the other an
inexpensive present. But no theory that
papyrus books are necessarily dearer than parchment books, or that the
is the case,\82/ can be maintained without rearranging the order of the
epigrams. It is in any case highly
probable that the order in this book is disturbed; and it follows that
epigrams cannot be used in the profitless debate (see p. 7 above) on
question whether papyrus or parchment was the more expensive material.
Against this silence we can
the earliest extant fragment of a parchment codex in Latin – the
of a historical work, christened De
bellis Macedonicis, found at Oxyrhynchus (though not necessarily
Egyptian origin), which has been convincingly attributed, on the ground
its letter forms and its spelling to a date not far from C.E. 100. \84/ But for the moment this fragment
alone among the remains of Latin literature found in Egypt, the next
Latin parchment codex being perhaps the Leiden fragment of the Sententiae
of Paulus, assigned to the
third-fourth century.\85/ The most ancient Latin papyrus codices are no
older. In any case, the relative
scarcity of early Latin fragments from
\84/ P. Oxy. 1.30 = E. A. Lowe, C.L.A. 112.207 and Supplement, p. 47. E. G. Turner, The Typologv of the early Codex, p. 38, accepts a date early in the second century, but on p. 128, no. 497, it is dated first century.
\85/ E. A. Lowe, C.L.A., 10.1577, where the date is given as 'Saec. IV'. Turner, op. cit., p. 126, no. 473 says 'iii-iv.'
To sum up, it appears, so far
as we can see,
[[i.e. Secundus' ??]] experiment
was still-born. And if we ask why, an
obvious answer lies in the fact that at this time, and throughout the
century, Greek influence in Roman cultural life was perhaps more marked
any other period; and that in consequence an invention of the practical
genius in the field of literature (where convention, we may suspect,
the form in which a book appeared no less strictly than its
have been at a discount. An additional,
or alternative, reason may have been the technical difficulties,
Section 2, of manufacturing parchment on a scale large enough to enable
it to provide
a viable alternative to papyrus. [[revise??!! --
the impression from Martial is hardly that this is a tenuous
leave Martial, there is one final point which deserves consideration.
In the poems we have discussed the codex form
is so inseparably linked with the use of parchment that scholars have
regarded it as axiomatic that the parchment codex preceded, and indeed
the model, for the papyrus codex. Today
this is by no means so certain. The
whole matter has been debated at length in Sir Eric Turner's The
Typology of the early Codex, Chapter
3 (pp. 35-42), 'The
Priority of Parchment or Papyrus?' in which he asks the pertinent
the papyrus codex is confessedly modeled on the parchment codex, why
at an early date have developed idiosyncratic forms
a succeeding enquiry will show, may also have extended to its
To this question there is as yet no answer; and the possibility cannot
excluded that the papyrus codex, even if it did not antedate the
codex, may have developed in parallel with it.
At present the question is wide open.
\87/ Op. cit., p. 40.
THE EVIDENCE OF
As we have seen, for a century and more after the situation [R&S "experiment"!] described by Martial our literary sources are silent regarding the development of the codex. The evidence of the classical jurists is thus especially welcome, the more so since they would be expected to take account of general social attitudes rather than reflect the idiosyncracies of individual authors.\88/
point is well made by E. Schönbauer, IVRA
12 (1961) 137. [[On the Justinian Code in
general, and English translations of many of its provisions, see http://uwacadweb.uwyo.edu/blume&justinian/default.asp
appellatione continentur omnia volumina,
sive in charta sive in membrana sint sive in quavis alia materia:
sed et si in philyra aut in tilia
(ut nonnulli conficiunt) aut in quo alio
corio, idem erit dicendum. quod si in codicibus sint membraneis vel
chartaceis vel etiam eboreis vel alterius
materiae vel in ceratis codicillis, an debeantur,
videamus. et Gaius Cassius scribit deberi et membranas libris legatis:
consequenter igitur cetera quoque
debebuntur, si non adversetur voluntas
adapted from Samuel
P. Scott (1932 -- online
-- who struggles with the second sentence and takes it to refer to
however, the books are bound
in leather, or papyrus, or ivory, or any other substance, or are
composed of wax tablets, will they be considered to be due? Gaius
Cassius says that where books are bequeathed, the bindings are also
included"); for the process by which the material from Ulpian was
preserved, see http://www.law.berkeley.edu/robbins/RomanLegalTradition.html
-- in the 530s the emperor
Justinian caused to be created the Digest, which collected and
summarized decisions of earlier "classical" jurists (including Papinian
and Ulpian), the Code, which contained the actual laws at that
time, and the Institutes to summarize the Digest and
serve as a textbook for students; the Novellae [new laws of
Justinian] collection was
created about two decades later to update the Code and
summarize Justinian's legislation, although actual manuscripts of it
have not survived intact.]]
\91/ Emendati perhaps refers to the repair of minor blemishes in the material, cf. Lewis, op. cit., pp. 63-64.
\92/ Digest 32.5 -- 'Unde non male quaeritur, si libri legati sint, an contineantur nondum perscripti. et nonputo contineri, non magis quam vestis appellatione nondum detexta continetur. sed perscripti libri nondum malleati vel ornati continebuntur: proinde et nondum conglutinati\90/ vel emendati\91/ continebuntur: sed et membranae nondum consutae continebuntur.’ [[The ET is adapted from Scott]]
[] These passages, especially the former, deserve more detailed analysis than they have hitherto received. In the clause quod si in codicibus sint [if they are in codices] etc., the subject of sint [they are] cannot be, as one might expect, volumina [scrolls], since Ulpian is clearly at pains to draw a distinction between volumina [scrolls? or perhaps written entities] and codices [codices], and an expression volumina in codicibus [scrolls/entities in codex format] would thus be a contradiction in terms; it thus seems much more likely that the subject of sint is libri [books], understood from the opening words of the passage (librorum appellatione [The designation "books"]). In listing the various forms of writings of which the inclusion in the term libri is, for Ulpian, in doubt, he appears to divide them into two classes, the second introduced by the words vel etiam [or even], suggesting an even greater degree of dubiety than in the case of those previously mentioned. On this basis the two classes of disputed materials are as follows:
codices eborei [ivory codices]
codices chartacei [papyrus codices] codices alterius materiae [codices of any other material]
cerati codicilli [waxen codicils]
question now arises whether the term codices
membranei denotes, or includes, the parchment notebook which we
already investigated [Parsons
review says yes -- "Ulpian's first group could comprise, or at least
include, notebooks"] . The repeated
association of the parchment notebook with the waxed tablet from which
originated leaves no doubt [[!!]] that Ulpian would have placed it in
Group 2, covered,
we may presume, by the words alterius
materiae. It follows, therefore, that his Group 1 represents
papyrus and parchment codices in our sense of the term. [[also
possible is that the initial word volumina is not meant to specify
scrolls, but has already become synonymous with "book" (of any format),
as is also suggested by the quotation about Homer, below; such
vocabulary could also have been modified in the history of transmission
of these materials, when the terms did indeed interchange. -- get
be seen that in order to decide whether materials in both these classes
be accepted as libri, Ulpian
quotes an opinion of the first century
jurisconsult Gaius Cassius which has already been discussed on p. 21
Cassius writes that where books (libris) are bequeathed,
even parchments (membranas) are included.]].
It has been objected,\93/ that the quotation from Cassius does
answer the question posed by Ulpian, but this is to misunderstand
reasoning. It is true []
that Cassius only
specifically mentions membranae. Given
the date at which Cassius was
writing this must [[!! unless perhaps it
included Martial's parchment codices!]] mean parchment
Ulpian would have argued that since Cassius says et
membranae' the same must apply (consequenter)
to all the analogous forms (cetera) in
his Group 2, and, a fortiori, to his
It is clear that for Ulpian only
roll was fully
and unquestionably a 'book'; but it is equally clear that the
long be denied its place. Indeed his
contemporary and rival in the law, Paulus,
who succeeded him as
Prefect after his murder in 223, goes further and defines the book in
way that the codex is at last admitted on terms of equality with the
we may accept the attribution of the Sententiae,
or at any rate of this quotation, to him): When
books (libris) are bequeathed, volumes of papyrus (chartae
volumina) , or of parchment (vel membranae) and
wood-slabs (et philyrae) are included, and codices, as
By the designagtion
"books" (librorum) not merely volumes of papyrus (volumina
chartarum), but also any kind of writing
which is contained in anything is understood.\++/. The
book is now defined, and well defined, as
a self-contained unit, independent of material or format.
With this judgment the codex has arrived; but
it has still to become fashionable. This
confirms the verdict of the preceding section, namely that the
development attested by Martial
was not an immediate success, and that the codex emerged as an
a long period of gestation.
\++/ Sententiae 3.6.87: Libris legatis tam chartae volumina vel membranae et philyrae continentur: codices quoque debentur: librorum enim appellatione non volumina chartarum, sed scripturae modus qui certo fine concluditur aestimatur' [[ET adapted from Scott) -- get info on Sententiae]]
The passages from Ulpian and Paulus
are both discussed in detail by Wieacker,\94/ whose principal
hypothesis is that
the works of the classical jurists were originally published in roll
they were re-copied into codices about C.E.
300; and that hand in hand with this re-copying went an extensive
and alteration of the texts. Wieacker's
views, which in any case have been strongly challenged,\95/ do not
us except insofar as they involve his contention that the Ulpian and
quotations are not in their original form, but have been largely
re-edited. His precise motives in
wishing to discredit the evidence of these passages are obscure, but
he is concerned that any mention of the codex as a possible literary
this period\96/ might imperil his contention that the works of the [] classical
jurists were first issued in roll form, and were not
to codices until about C.E.300.
\94/ Op. cit., pp. 105-106.
Wieacker's objections to the text of the Ulpian quotation appear to be two: (1) some of the expressions used, such as 'in quavis alia materia' [in whatever other material], `ut nonnulli conficiunt' [as is sometimes done], 'aut in quo alio corio' [or upon any kind of prepared skins], ‘vel allerius materiae' [or other materials], are elaborations which have the odor of the classroom (‘Schulstubengeruch'), and (2) as already mentioned, the quotation from Cassius does not really answer the question posed by Ulpian; it is thus presumably, in Wieacker's view, an interpolation, and the consequences drawn from it (consequenter igitur cetera quoque debebuntur [it follows that everything relating to them will be included]) must likewise be rejected ('sicher unecht' [[surely inauthentic]]).
There is, of course, no doubt that the final clause
(si non adversetur
voluntas testatoris [if the
intention of the testator was not otherwise] is a later addition, since it runs counter to Ulpian's own
line of reasoning, which is a typical interpretation of terms without
regard to the probable wishes
of the testator [[but see
in this same section, regarding "libraries" and intentions]].
apart, however, Wieacker's arguments
are highly subjective, and
indeed the only conclusion that even he can draw is that there may have been some tampering with the text (`Wir halten eine (vermutlich
für möglich' -- [[we
consider a (probably pre-Justinian) textual modification to be
In the case of the Paulus quotation, there is general
the Sententiae as they have come
down to us were put together about C.E. 300, but this does not
imply that the individual opinions attributed
to Paulus [[who flourished two generations earlier]] are necessarily
unauthentic. Wieacker's sole argument
in this instance is that in accepting codices as libri the quotation
accurately reflects conditions obtaining about C.E. 300, and cannot therefore
go back to Paulus himself. It
would seem, therefore, that Wieacker rejects the entire passage as unauthentic.
This is, of course something of a petitio principii [minor point], since Wieacker has to prove not merely that the quotation suits conditions about C.E. 300, but that it does not suit conditions in the time of Paulus; and this he has failed to do.Both Schönbauer\97/ and Archi,\98/ in addition to their general criticisms of Wieacker's work, have specifically rejected his attacks [] on the authenticity of the Ulpian and Paulus quotations and it therefore seems justifiable to continue to use them here as evidence of the increasing prominence of the codex in the Roman world of the early third century.
\98/ Ibid., pp. 434-5.
Before we leave the jurists, there is one more passage from Ulpian which must be briefly mentioned. The text runs: If a hundred books (libri) are bequeathed, we must deliver to the legatee a hundred volumes (volumina), and not a hundred parts which anyone may select as he wishes, each of which constitutes a written book (ad libri scripturam); hence, when the works of Homer are all contained in a single volume (in uno volumine), we do not count them as forty-eight books (libros), but the entire volume of Homer (unum Homeri volumen) should be understood to constitute a "book" (pro libro).\99/ The meaning is perfectly clear [[??]], but what is remarkable, and puzzling, is Ulpian's choice of an actual example, namely a complete Homer in one volumen. In view of the clear distinction which Ulpian makes between volumen and codex, volumen here can only mean a (papyrus) roll. [[alternatively, it means a written entity, here defined by author! see above on the possible ambiguity of volumen]] But it is unnecessary to demonstrate that a complete Homer in a single roll is a physical impossibility; and since tens of thousands of Homeric manuscripts must have been in existence at the time, this impossibility would have been immediately obvious to Ulpian's readers. We must therefore conclude that the example was a purely hypothetical or imaginary one.
[[or that the meaning of volumen is already in flux and here it means codex-collection!]]
designation of "books"
all volumes are included, whether they are made of papyrus, parchment,
or any other material whatsoever; even if they are written on bark (as
is sometimes done), or upon any kind of prepared skins, they come under
the same appellation.
If, however, the
books are bound
in leather, or papyrus, or ivory, or any other substance, or are
composed of wax tablets, will they be considered to be due? Gaius
Cassius says that where books are bequeathed, the bindings are also
included. Hence, it follows that everything relating to them will be
due if the intention of the testator was not otherwise.
(1) Where a
hundred books are
bequeathed, we must deliver to the legatee a hundred volumes, and not
the hundred parts of volumes which anyone may select as he wishes, and
each of which will be sufficient to include the contents of a book;
hence, when the works of Homer are all contained in one volume, we do
not count them as forty-eight books, but the entire volume of Homer
should be understood to mean one book.
(2) Where the
works of Homer (Homeri
left, and they are not complete (non sit plenum), as many parts of the
same as can be
obtained at present will be due.
(3) Sabinus says
not included in legacies of books (Libris). Cassius adopts the
but he holds that parchment covers that are written upon (membranas quae scriptae
sint) are included.
He adds, afterwards, that neither book-cases, writing desks (armaria neque scrinia), nor other
furniture in which books are kept constitute part of the legacy.
(4) What Cassius
reference to blank parchments (de membranis puris) is true, for blank
sheets of papyrus (nec
chartae purae) are
not included in the term, "Books bequeathed" (libris legatis nec chartis legatis), and books are not due
under the term, "Sheets of papyrus bequeathed," unless, perhaps, in
this case the intention of the testator may influence us; as for
example, if one literary man should leave to another sheets of paper (chartas) as
follows, "I bequeath all my sheets of paper" ("chartas meas
universas"), and he had nothing else
but books (libros), no one will doubt
that his books (libros)
due; for ordinarily
many persons designate books as papers (plerique libros chartas
appellant). But what if anyone should
bequeath sheets of papyrus (chartas legaverit puras). In this case neither
parchments (Membranae), nor any
other materials used for writing, nor books which have been commenced
scribi libri) will be included.
4. Quod tamen Cassius de membranis puris scripsit, verum est: nam nec chartae purae debentur libris legatis nec chartis legatis libri debebuntur, nisi forte et hic nos urserit voluntas: ut puta si quis forte chartas sic reliquerit "chartas meas universas", qui nihil aliud quam libros habebat, studiosus studioso: nemo enim dubitabit libros deberi: nam et in usu plerique libros chartas appellant. Quid ergo, si quis chartas legaverit puras? Membranae non continebuntur neque ceterae ad scribendum materiae, sed nec coepti scribi libri.
(6) The legacy
of papyri does not include the material for making the leaves, nor such
leaves as are not yet finished.
6. Chartis legatis neque papyrum ad chartas paratum neque chartae nondum perfectae continebuntur.
(34) Where land with
equipment is devised, it is well established that the library, and any
books upon the premises, which the head of the household made use of
whenever he came, are included. If, however, a warehouse should be used
for the storage of the books, the contrary opinion must be held.
(43) He also gives it as his
opinion that where a house is devised with its equipment, ivory tables
and books are not included. This, however, is false, for everything in
the house by means of which the owner may be better provided and
rendered more comfortable will be included. No one doubts that the
furniture is something which contributes to the convenience of the head
of a household.
4. Ulpianus, On the Edict, Book XIX.
(1) Noxious drugs and poisons are embraced in this action; but the judge ought by no means to interfere in matters of this description, for it is his duty to perform the functions of a good and innocent man. He should act in the same manner with reference to books which it is improper to read (for instance, those treating of magic and similar subjects); all of these, however, should be immediately destroyed.18. Paulus, On Vitellius, Book II.
(13) Where a testator made a bequest as follows, "I leave my country-house in the same condition as I myself possessed it, together with the furniture, tables, and the urban and rustic slaves which shall be sent there, and the wines that may be in said house at the time of my death, and ten aurei in addition," as upon the day of the testator's death he had books, articles of glass, and a small clothes-press in the house, the question arose whether these articles should be included among those enumerated in the bequest. Scævola answered that only such articles as were specifically mentioned formed part of it.3. Paulus, On Sabinus, Book IV.
The following are embraced in
bequests of household goods, namely: cupboards, benches, bedsteads,
beds, even such as are inlaid with silver, mattresses, coverlets,
pillows, vases for water, basins, candelabra, lamps, and ladles.
In addition to these are
strong boxes and coffers. Some authorities very properly hold that
wardrobes and chests of drawers, if intended for the storage of clothes
or books, should not be classed as household goods, because the
articles for which they are designed are not included in that category.
The answer was that the law should not be employed to cause annoyance, and that terms ought not to be captiously construed, but that it was proper to consider with what intention the words were uttered; for, in accordance with this principle, if anyone was desirous of studying some branch of knowledge, he might state that he had an interest in such and such books being produced for his benefit, because if they were produced, after he had read them he would become a more learned and a better man.
Record books (for loans, etc.)
10.6. Alfenus, Epitomes of the Digest by Paulus, Book III.(1) Small writing tablets and memorandum books are not classed as household goods.
Where anyone bequeaths a
promissory note, it is understood that he had in mind not only the
tablets upon which it is written, but also the rights of action, the
proof of which is contained in the tablets. For it is clear that we use
the same "note" instead of the said rights of action; so when the note
is sold, we understand that the claim was also disposed of. Moreover,
where anyone bequeaths a claim, he is understood to have bequeathed
what can be recovered by an action at law.
Scævola, Opinions, Book III.
. . . The father had kept, in
the name of one of his sons, an account book of debts, and it was
afterwards decided and held that what remained in said book in the name
of his son was due to the latter; but not what had been already
collected and placed by
his father among the assets of his estate.
32.34. Scævola, Digest, Book XVI.
(1) A testator, having appointed
his son his heir to a portion of his estate, with other things left him
a preferred legacy in these words: "I request that twenty claims,
taken from my account-book (ex calendario), shall be given without fraudulent
my son Titius, after he has selected the same." The said testator,
during his lifetime, entrusted his son with the transaction of all his
business, and the son, after the will was made, and for ten years
before his father's death, during which time he acted as his agent,
contrary to the usual practice of his father as shown by his
account-book, lent new debtors large sums of money, and permitted the
old debtors who owed his father small amounts to increase their
obligations, in order that the aforesaid twenty claims might almost
fill the entire account-book of his father. The question arose whether
the son was entitled, as a preferred legacy, to the loans which he
himself had made. The answer was that he could only make a choice of
those which were in the account-book (in kalendario) of the testator at the time he
executed his will.
A father, having appointed
his son and his daughter his heirs, and left to each one of them
certain lands and book-accounts (kalendaria) by way of preferred legacies . . . .
41. Scævola, Digest, Book XXII.
A certain man left a legacy
in trust to Mævius as follows: "I bequeath whatever I possess in
city of Gades." The question arose whether, if he had any property in
the suburb adjoining the city, this also would be due to Mævius
the terms of the trust. The answer was that the meaning of the words
will also permit this extension. It was also asked, in the same case,
certain notes having been found in the account-book of the testator, he
being in the habit of loaning money in his native city of Gades, or in
the environs thereof, and having left the property which he had in said
Africanus, Questions, Book VI.
Where a testator appointed his
son and his grandson his heirs, and gave to his grandson under a trust
certain lands, and whatever might be on them at the time of his death
"with the exception of his account book," and, when he died, a sum of
money was found in his chest in which the notes and bonds of his
debtors were kept, it was held by several authorities to be hardly
probable that the testator had the said money in his mind when he
created the trust.
I go still further, and hold that
if the money had been collected and again invested in a similar manner,
the change of obligations would neither annul or diminish the effect of
the trust, so that if the same money was intended to be placed in the
account book, that is to say for the purpose of making new loans, it
would still be payable to the beneficiary under the terms of the trust.
91. Papinianus, Opinions,
(6) Under the term "house" is also understood a building joined to the same, if both were purchased for one price, and it is established that the rents of both were carried together on the books.
Scævola, Digest, Book XVI.
A testatrix left to her grandson the lands which she possessed in a certain district, as they were equipped, together with the wine, grain, and a book of accounts;
Scævola, Digest, Book VI.
The testator also left the claims in his account-book, and the money which was on said land.
Scævola, Opinions, Book III.
The question arose whether he would be compelled to surrender to the heirs the books in which the accounts were kept, as well as any sums remaining in his hands as shown by the entries of receipts and expenditures.
8) Moreover, where a person uses the article lent to him in some
other way than was intended, he is liable not only to an action on loan
but also to one on theft; as Julianus states in the Eleventh Book of
the Digest. He also says, "If I lend you a blank book and you cause
your debtor to write therein a note to secure you, and I then erase
this; if I lent the book to you in order that you might be secured, I
am liable to you in a counter action."
Scaevola libro 17 digestorum
pr. His verbis legavit: "Uxori meae lateralia mea viatoria et quidquid in his conditum erit, quae membranulis mea manu scriptis continebuntur nec ea sint exacta cum moriar, licet in rationes meas translata sint et cautiones ad actorem meum transtulerim". Hic chirographa debitorum et pecuniam, cum esset profecturus in urbem, in lateralibus condidit et chirographis exactis quam pecunia erogata reversus in patriam post biennium alia chirographa praediorum, quae postea comparaverat, et pecuniam in lateralia condidit. Quaesitum est, an ea tantum videatur nomina ei legasse, quae postea reversus in hisdem [isdem] lateralibus condidit. Respondit secundum ea quae proponerentur non deberi quae mortis tempore in his lateralibus essent et membranis manu eius scriptis continerentur. Idem quaesiit, an, cum emptiones praediorum in hisdem [isdem] lateralibus condiderat, praedia quoque legato cedant. Respondit non quidem manifeste apparere, quid de praediis sensisset, verum si ea mente emptiones ibi haberet, ut his legatariae datis proprietas praediorum praestaretur, posse defendi praedia quoque deberi.
Preamble to Justinian Code'
2) After having brought into perfect harmony the Imperial Constitutions hitherto involved in confusion, We have directed Our attention to the immense volumes of ancient jurisprudence, and have finally accomplished this most difficult task, proceeding, as it were, through the depths of the ocean, and aided by the favor of heaven.
(3) This having been concluded through the Grace of God, We summoned the illustrious Tribonian, Master and former Quæstor of Our Sacred Palace, along with Theophilus and Dorotheus, eminent men and professors, (whose skill, familiarity with the laws, and fidelity in obeying Our orders We have proved on many occasions) and especially directed them to draw up Institutes by Our authority, and with Our advice, that you may be able to learn the first principles of the law, not from ancient fables, but acquire them from the Imperial Splendor; so that your ears as well as your minds may absorb nothing that is useless or incorrect, but whatever is in accordance with reason in all things. And while, in former times, it was scarcely possible for those who preceded you to read the Imperial Constitutions in the course of four years, you may, now, from the very beginning, proceed to do so; being found worthy of such honor and happiness that both the beginning and the end of your instruction in the laws issue from the mouth of your Sovereign.
(4) Therefore, after the completion of the fifty books of the Digest or Pandects, in which all the ancient law has been collected, and which We have caused to be compiled by the said distinguished personage Tribonian and other eminent and most illustrious men, We have ordered these Institutes to be divided into the following four books, that they may constitute the first elements of the entire science of jurisprudence.
(11) There are also other persons who, for different reasons, are prohibited from contracting matrimony, and these We have permitted to be enumerated in the books of the Digest or Pandects compiled from the ancient law.
(1) The Divine Marcus published in
his semi-annual volumes of rescripts that an official belonging to the
Treasury could be excused from guardianship or curatorship as long as
he remained in office.
(33) Writing, also, even though it be of gold, belongs as much to papyrus and parchment as edifices or crops do to the soil; and, therefore, if Titius has written a poem, a history, or a speech, upon your papyrus or parchment, you, and not Titius, are considered to be its owner. But if you demand your books or parchments from Titius, and are not ready to pay the expense of the writing, Titius can defend himself by the exception on the ground of fraud; at all events, he can do so if he obtained possession of the said papyrus or parchments in good faith.(34) Where anyone has painted a picture upon the tablet of another, some persons think that the tablet should belong to the picture; and others are of the opinion that the picture, no matter what kind it may be, is a part of the tablet. It appears to Us preferable that the tablet should belong to the picture, for it is ridiculous that a painting by Apelles or Parrhasius should be considered an addition to a wretched tablet. Wherefore, if the owner of the tablet be in possession of the painting, and the artist who painted it demands it, but is unwilling to pay the value of the tablet he can be barred on the ground of fraud; but if he who painted the picture is in possession of the same, it follows that an action can be brought against him by the owner of the tablet; in which instance if he does not pay the expense of the painting, he can be barred by the exception on the ground of fraud, at all events if he who painted the picture obtained possession of it in good faith; for it is evident that if the artist or anyone else acquired the tablet surreptitiously, the owner of the same is entitled to an action of theft.
(12) It does not matter whether a will be written on tablets, papyrus, parchment, or any other substance.This rule We have introduced, not by way of innovation, but because it was more just, and also because Paulus in the books which he wrote on the works of Masurius Sabinus and Plautius, states that it was accepted by Atilicinus.
ROLL AND CODEX: EVIDENCE OF GREEK
LITERARY TEXTS OF THE FIRST FIVE
preceding sections the literary evidence for the emergence of the codex
has been examined. We now turn to the
evidence of the actual manuscripts which have survived from this
period. But before we do so, a few words of caution
must be given. An overwhelming
proportion of these manuscripts come from
The former of these points can be the more readily answered. Apart from the Delta and Alexandria, discoveries have been made in almost every region of Egypt, and serious though the absence of Alexandria is, it is probable that many of the literary papyri found at Oxyrhynchus, where wealthy Alexandrians possessed country estates, either were written in Alexandria, or, if local copies, would have reflected current fashions in the capital.\100/
question is much less easy to answer.
However, the ease of travel throughout the Roman world, the
movements of officials, merchants and others, and, above all, the
indeed other and more serious reservations to be made in the assessment
Egyptian evidence. The dating of
literary papyri is far from being an exact science, and estimates of
vary by a century or more. All we can
hope for is that the inevitable errors in dating will, at least to some
extent, [] cancel
out. A further difficulty is the
distribution over time. Relatively
abundant during the first three centuries, the output of literary
a dramatic falling-off after C.E. 300
which presumably reflects the general decay of Hellenism.\101/ However,
only take the evidence as we find it.
\101/ There may be other factors; for instance, the well-known scarcity of dated documents of the fifth century C.E. may be paralleled by a corresponding dearth of literary texts from the same period.
The statistics which follow are based on
the data in Pack-2 ([[1952\1]] 1965), supplemented by (a) F. Uebel, Literarische
Texte unter Ausschluss
der Christlichen in Archiv für
Papyrusforschung 21 (1971) 170-182, for publications up to
and (b) the section Testi recentemente pubblicati:
Testi lettereri greci in Aegyptus 51 (1971) 227-30; 52
53 (1973) 160-4; 54 (1974) 206-9; 55 (1975) 275-9; 57
202-47; 58 (1978) 225-87; and 60 (1980) 233-65. [[See
now Pack-Mertens\3 and the online
supplements by Mertens, dated January 2008.]] In the case of
parchment codices much use has been made of E. G. Turner, The
of the early Codex, 1977, which covers (see p. xxii) material
to November 1973. It should be made
clear that except in a few cases the figures are based on the estimates
given by the original editors. The
employment of dates spanning two centuries, e.g., second-third century,
certain disadvantage in that it gives the impression that there
diminution of literary activity during each of these bridging periods.
This was not, of course, the case but merely
reflects the predilection of editors for assigning a text to a
century. There is indeed a method
whereby this disadvantage can be eliminated.
William H. Willis in A Census of
the Literary Papyri from Egypt
(Greek, Roman and Byzantine Studies 9 (1968) 205-241) divides
spanning two centuries equally between the centuries concerned,
the grounds that of all the texts dated by their editors, e.g.,
century, there is a statistical probability that in fact 50% will have
written in the second century and 50% in the third.
This expedient has not been adopted here,
since it involves a re-interpretation of the judgments of the original
Fortunately the principal points which emerge from the survey remain
substantially the same whichever procedure is followed.
figures below cover all Greek (but not Latin) literary and scientific
Christian literature excepted; they omit items []
are, or appear
school exercises, single sheets, mathematical tables -- anything in short which
clearly not a [[literary]]
book. [[See the Catalog of
Paraliterary Papyri (CPP) for information on such materials.]]
waxed or wooden tablets, ostraca, and inscriptions on stone
metal are of course excluded. It should be
that since we are here concerned with format, i.e. whether a manuscript
is in roll or codex form, no distinction is made between texts
on papyrus and those on parchment. The unit in these
statistics is the book, i.e. where a single roll
or codex comprises two or more
different works it is counted as only one unit. However, where a
roll has been re-used by having a different literary work written on the
verso, the texts on recto and verso are each counted as
one unit, since the scribe of the verso text was presumably satisfied
with the roll format and only used a discarded roll
a new one for reasons of economy. [[problem:
reused rolls may only indicate necessity, not necessarily choice; is
the same criterion used in counting the "Christian" materials? The
following figures do not include "Christian" stuff. Categories
can be searched only by century (not by half century),
with set categories for papyrus (5615 -- apparently rolls), papyrus
codex (683), parchment (40), parchment codex (206), ostraca (249), wood
tablets (92), wax tablets (46), other (15 -- ivory, metal, stone); and
for languages Latin (190), Greek (6729), Other (27); total
for first five centuries CE, 5275 [6896!], of which 4247 [5487!]
are papyri rolls, 550 
papyri codices, 26 parchment, 172  parchment codices -- why don't the figures match?
ambiguous dates counted both ways?]]
pap + parch
||83.3 [81½]||11.2 [18½]|
||40.8 [26½]||44.6 [73½]|
is clear that the codex scarcely counted for Greek
before about C.E. 200. Nevertheless its
not, as has sometimes been suggested, entirely negligible. The
significance of these second-century codices for the origins and
the codex form in non-Christian literature will be
present, however, the fact remains that it
only in the course of the third century that the codex obtained a
significant share of book-production and it was not until about
C.E. 300 that it
parity with the roll. [[Interestingly, it seems
not to predominate in the preserved evidence from subsequent centuries
THE CODEX IN EARLY CHRISTIAN
As we have seen, in the pagan world of the second century the codex has barely a foothold [[see further chapter 12 below]]. In the contemporary Christian world the position is very different, and it is to this that we must look for the origins of the modern book. The assembling of statistics in the field of Christian papyri has been immeasurably lightened by two recent publications, viz. Kurt Aland, Repertorium der griechischen Christlichen Papyri: 1, Biblische Papyri, 1976, and Joseph van Haelst, Catalogue des Papyrus littiraires juifs et chrétiens, 1976, the second being of especial value for the present investigation since it includes for the first time a survey of all Christian papyri, both biblical and non-biblical. The data from these two publications have been supplemented by the bibliographies of Kurt Treu, Christliche Papyri VI and VII, in Archiv für Papyrusforschung 26 (1978) 149-159, and 27 (1980) 251-258 [[update needed]] respectively. It should be noted that whereas Aland's work is strictly limited to texts on papyrus, the publication of van Haelst, despite its title, includes texts on all kinds of material, as also do the bibliographies of Treu.
\102/ E.g., in the case of a small fragment written on both sides, the difficulty of distinguishing between a codex and an opisthograph roll, cf. E. G. Turner, The Typology of the Early Codex, pp. 9-10.Of these 14, five (nos. 5, 6, 9, 10, 14) are opisthograph, i.e. the biblical text is written on the back [[against the fibers]] of a re-used roll [[including literary (##5, 14[Latin]) and documentary (##6, 9, 10)]], which thus imposed the roll format. This reduces the number of 'genuine' rolls to nine. Of these nine, three (nos. 7, 8, 12) are probably of Jewish origin,\103/ and two more (nos. 1 and 3) possibly so.\104/ This leaves only nos. 2, 4, 11 and 13 to be considered. The last-named is an eccentric production, being written on the verso of a roll the recto of which is left blank. Various complicated explanations of this phenomenon have been proposed,\105/ but for the present purpose we can reasonably leave it out of account. No. 2 is opisthograph, but has biblical texts on both sides. The Exodus is presumably Christian, since κύριος is abbreviated (although υἱός and Ἰσραήλ are not). Nevertheless, from any point of view the item is clearly an oddity, and we are thus left with only two normal rolls of Christian origin, viz. nos. 4 and 11. As regards no. 4, the Psalms were used for such a variety of purposes, devotional, [] liturgical, magical, etc., that this exception has less significance. The Isaiah has the nomen sacrum for κύριος, and is therefore presumably Christian and a genuine exception.\106/ It may be added that no text of any part of the canonical New Testament is known written on the recto of a roll. [[For some other early Christian writings in roll format, see LDAB 2459 (Irenaeus, 2/3rd c), 3071 (Gospel Harmony, parchment, 3rd c; nom sac IH, QU); 6616 (parchment liturgical rotulus, 8th c.); see also the 5720 papyrus Christian Book list (CPP 388) from the 4th c., blank back]]
K. Aland, Studien
zur Uberlieferung des Neuen Testaments und seines Textes, p. 114.
\106/ Cf C. H. Roberts, op. cit., p. 31, n. 1: 'What is true is that the contracted form of κύριος is in the first three centuries the mark of a Christian manuscript'. Van Haelst's verdict of 'probablement Juif' is presumably based on the fact that it is a roll.
If we examine these 172 biblical
manuscripts from a
different standpoint, we find that there are eleven which in our
opinion may be
assigned to the second century and are thus the earliest Christian
in existence. All are on papyrus and in
codex form. The following list\107/
includes references to the work of van Haelst, and we have also,
because of the
particular importance of dating, appended references to E. G. Turner, The
Typology of the Early Codex,\108/ in
which he records both his own judgments and those of other scholars. [[The following items are rearranged
(while still retaining the R&S numbers) with the probably earliest
\108/ The references are to the identification numbers of the
manuscripts in the 'Consolidated
List of codices consulted' at the end of the book.
On the date of this manuscript, which occupies a key position
Christian texts, see Roberts, op. cit., Appendix
11, pp. 78-81, where the conclusion is reached (p. 81) that 'on present
evidence a second century date, though possible or even probable, is
necessary and a provisional verdict should be second/third century.'
We have excluded from the list a second-century codex of Genesis,\110/ since in spite of the codex form we consider it to be of Jewish origin.\111/ The above eleven are without exception Christian. To these may be added four other Christian non-Biblical texts which in our opinion are to be assigned to the second century:
Oxy. 4.656. Van Haelst 13. Typology
OT 9. ii/iii (E.G.T.; iii ed.; ii
Bell/Skeat). [LDAB 3094]
possible early Jewish
uses of the codex for what became scriptural literature, see Treu
(especially the Excursus)
Augustine provides two direct references to Jewish scriptural
codices in the 4th century, as Bill Adler pointed out to me. Augustine,
De fide rerum quae non
videntur 6.9 (PL 40.178): The Jews provide support to Christians
“through their codices (in codicibus), even if they are enemies in
their hearts.”[Sed cum legunt, non mirentur quod ista illi
codices sunt, propter inimicitiarum tenebras non intelligunt.] Ep.
71.5 (to Jerome; PL
33), chapter 3: "the words in the Hebrew codices [to
the book of Jonah] were correctly rendered in the Greek
version, and in the Latin one taken from it." [ quoted by Jerome in
response, Ep 75.6-7(21-22) PL 33 «Quidam frater
noster episcopus, cum lectitari instituisset in Ecclesia
cui praeest, interpretationem tuam, movit quiddam longe
[Col.0263] aliter a te positam apud Jonam
prophetam, quam erat
omnium sensibus memoriaeque inveteratum, et tot aetatum
successionibus decantatum. Factus est tantus tumultus
in plebe, maxime Graecis arguentibus, et inclamantibus
calumniam falsitatis, ut cogeretur episcopus, (Oëa
quippe civitas erat), Judaeorum testimonium flugitare.
Utrum autem illi imperitia an malitia, hoc esse in hebraeis
codicibus responderunt, quod et graeci et latini
habebant atque dicebant. Quid plura? Coactus est homo
velut mendositatem corrigere, volens post magnum periculum
non remanere sine plebe. Unde etiam nobis videtur
aliquando te quoque in nonnullis falli potuisse.» to which
responds with more careful terminology: Sin
autem Judaei vestri, ut ipse asseris, malitia, vel imperitia
hoc dixerunt esse in voluminibus Hebraeorum, quod in
graecis et latinis codicibus continetur; manifestum est eos
aut hebraeas ignorare litteras, aut ad irridendos cucurbitarios
voluisse mentiri. -- Ep 75.6-7.(21-22)] A close
study of Augustine's use of book terminology (liber, codex, volumen,
libellum, carta, cartula, membran*) would be useful here: see Louis
Holtz, "Les mots latins de/signant le livre au temps d'Augustin," Les de/buts du codex, ed Alain
Blanchard (Actes de la journe/e d'e/tude organise/e a\ Paris les 3 et 4
juillet 1985 par l'Institut de Papyrologie de la Sorbonne et l'Institut
de Recherche et d'Historie des Textes, Bibliologia elementa ad librorum
studia pertentia 9; Turnout 1989) [pp.??]. Resnick thinks that
"Augustine is likely to be only projecting Christian usage of the codex
upon the Jews" (3 n.8). Resnick also implies that while "it seems that
the Christian community self-consciously decided upon the codex in
contradistinction to both Jewish and pagan practice" (4), and
discusses at length the rabbinic requirements of scroll usage for
liturgical purposes as possibly formulated with distinction from
Christian codices also in view, he then notes [citing Saul Lieberman,
Hellenism in Jewish Palestine (NY 1962) appendix 3] the use of
non-scroll formats for other purposes (including scriptures for private
study): "Lieberman even suggests that the earliest Christians borrowed
the codex form for their literature from this Jewish practice. [n. 52:
"Lieberman remarks: 'the employment of the note-book was the most
suitable way of indicating that they [the rabbis] were writing the Oral
Law for private, or unofficial use, and not for publication' ibid 205]
This possibility is viewed favourably by C. H. Roberts [n.53 Birth of
Codex 59], and directly contradicts the established view: namely, that
the Christians borrowed the codex form from the Romans. . . . But if
Lieberman is correct, the history of the codex must be dramatically
rewritten ..." (11). Resnick goes on to repeat, with apparent favor,
the idea that "since the earliest disciples apprehended the person of
Jesus especially through his unwritten words, which formed an oral
tradition [n.56 to Gamble, "Christianity: Scripture and Canon" in The
Holy Book in Comparative Perspective, ed. F. M. Denny and R. L. Taylor
(Columbia SC 1985) 36ff], [] it is possible that Jewish-Christians
would have used t he same form to record this oral tradition as the
rabbinic community used to record its Oral Law without violating the
ban against its publication. In Lieberman's view, then,
Jewish-Christians, 'did this because otherwise they would have
transgressed the law . ... We would naturally expect the logia of Jesus to be originally
copied in codices'" (11-12). Ultimately, Resnick focuses on the
"theological" motivations that solidified the codex as the accepted
form of Christian scriptures, without expanding further on the question
of its original introduction into Christian usage.
be emphasized that, as will have been evident from the
Typology references, not all scholars agree about the dates to be
assigned to these fourteen manuscripts.
Some would find our list
too inclusive,\112/ others too restricted;\113/ but about some in the
(nos. 1, 2, 8, 10, 12) there is unanimity.\114/
\113/ Conspicuous among manuscripts which some scholars have placed in the second century are: the Bodmer St John (P. Bodmer II; van Haelst 426; Typology P66. iii [LDAB 2777]); the Chester Beatty Papyrus IX, Ezekiel, Daniel, Esther (van Haelst 315; Typology OT 183-iii E.G.T.; Wilcken, Galiano ii [LDAB 3090]); and Chester Beatty Papyrus 8, Jeremiah (van Haelst 304; Typology OT 202 iv E.G.T.; ii or ii/iii Kenyon [LDAB 3084]).
\114/ On factors influencing the dating of early Christian papyri see Roberts, op. cit., p. 12, n. 2.
As already stated, all the above biblical manuscripts (nos. 1-11) are codices, while of the other four (nos. 12-15) the only true exception is the roll of Irenaeus (no. 15), since the Hermas (no.13) is written on the verso of a roll carrying a documentary text and the scribe thus had no option in his choice of format. The distinction between biblical and non-biblical texts would not have been so obvious to the users of these as it is to us, and both the Egerton Gospel and the Shepherd of Hermas might have been regarded as indistinguishable from the canonical books of the New Testament. [[add note on Hermas in MSS?]] Even if we give this extended meaning to the term ‘biblical,’ the conclusion remains the same, namely that when the Christian Bible first emerges into history the books of which it was composed are always written on papyrus and are always in codex form. [[add note on canonization?]] There could not be a greater contrast in format with the non-Christian book of the second century, a contrast all the more remarkable when we recall that Egypt, where all these early Christian texts were found, was the country where the papyrus roll originated and where the status of the roll as the only acceptable format for literature was guaranteed by Alexandria with its dominating position in the world of books.\115/
\115/ Of course, the fact that all known
specimens of early Christian papyri come from
Within these limits are found 118 texts, 14 of which are written on parchment and 104 on papyrus. For 83 of these the codex form was chosen. The remaining 35 are rolls, 3 of them opisthograph. By themselves these figures are not particularly instructive; we have to consider not so much the date of writing as the category of the work. Given that there is no example of any of the four canonical Gospels being written on the recto of a roll (i.e. in roll form by choice) we might expect that any other Gospel which resembled them in narrating the life or recording the teaching of Jesus (e.g., the Egerton Gospel) would circulate in the same format; this would not necessarily apply to such works as those of the Gnostic Gospels which are in fact theological treatises, or the Infancy Gospels. Such a differentiation would be all the more likely if the Christian adoption of the codex originated, as suggested below (p. 59), in the use of tablets for recording the [[Jewish]] Oral Law. Similarly, whereas we have seen that the codex form was closely associated in the second and third centuries with the books essential for the Christian mission, viz. the books of the Jewish scriptures ("Old Testament") together with such Christian works as were deemed authoritative, we might not expect to find the codex so widely used for works of general Christian literature: a theological treatise such as that of Irenaeus already noticed (p. 41) might well have been expected to adopt the form conventional for academic works. The picture, however, is not quite so clear-cut as this.
The first of the categories into
which the 118 texts
divide is that of Apocrypha. Among the
very varied works ranged under this heading there are 10 possible
examples of Gospels as we have defined them. Both
the second century examples are codices,
as we have noted above (nos. 12 and 14).
Of the remaining 8, both manuscripts of the Gospel of Thomas are
so too is the
Fayum Gospel and an Oxyrhynchus
fragment dated to about C.E.
200 and now
plausibly assigned to the [[subsequently??]] banned Gospel of
Peter,\116/ while a roll is
format of the only surviving manuscript of the Greek Diatessaron, if
may be fairly classified here. The
remaining 3 manuscripts are codices. The
other apocryphal texts, 23 in number, and including Infancy Gospels and
various Apostles, are exclusively codices. [[give updated list, as
\116/ See D. Lührmann in ZNTW, 72, 1981, pp. 216-26.
In the Patristic section of 39
texts, two works, the
Epistle of Barnabas and the Shepherd of
Hermas, both of which had some claim to canonical status, are
former by 2 codices, the latter by 2 rolls (one of which is
opisthograph) and 9
codices.\117/ Of the remaining 26 texts 6, including 3 manuscripts of
are rolls and 20 are codices. [[give list!]]
\117/ Two further papyri of this work, both of the third century and both codices, will be published in P. Oxy. 50.
of the persistent use of the roll in the liturgy of the Eastern Church
below, p. 51, n. 6[[??]]) it is not surprising that 6 of the 11 texts
section are on rolls. [[list]]
now to consider possible reasons for this remarkable predilection of
Christians for the codex form, and endeavour to formulate hypotheses
would at the same time explain the divergence of treatment accorded to
and non-biblical texts.
WHY DID CHRISTIANS ADOPT THE CODEX?
INADEQUACY OF PRACTICAL
reasons adduced by Martial\118/ in favour
of the parchment codex, even if Secundus's experiment had not been the
it pretty certainly was, are quite inadequate to explain what was not
preference by the Christian communities for the codex form, but an
devotion to it, and that too [[not only]] for the books of the New
Testament but for
of the "Old Testament" as well, and from the earliest period for which
has survived. Indeed so universal is the
Christian use of the codex in the second century that its introduction
date well before C.E.
is, moreover, significant for the history
of the early Church that Christian book-production methods should have
themselves from Jewish so completely and at so early a date: that the
Christians transcribed the books of the "Septuagint" onto codices
complete the severance was. [[?? very problematic]]
See Section 5 above.
been widely assumed that the codex must have possessed some significant
practical advantages over the roll. A variety of such explanations have
put forward, and it will be convenient to discuss them in detail at
point. Before doing so, however, one
thing must be made clear, namely that there are two quite separate
firstly why Christians adopted the codex for their writings from the
and secondly why (though only over a period of several centuries) the
eventually displaced the roll in the field of non-Christian literature
well. The practical reasons which we
shall consider here will not necessarily have the same force for both
processes, and this is a factor which must be borne in mind.
1. Economy. This is the reason most commonly put forward, and one of the most obvious and apparently convincing. Since the codex makes use of both sides of the writing material instead of [] only one, the cost of producing a manuscript would be reduced. There was, of course, no reason why the verso of a roll should not be used for a continuation of the work on the recto, but the very small number of such rolls (the roll in Apocalypse 5.1, βιβλίον γεγραμμένον ἔσωθεν καὶ ὄπισθεν [a book written inside and outside], is perhaps the best known example) shows that this was not regarded as a satisfactory method. But how much would the saving by using the codex format amount to? It is unlikely that the cost would be halved, since the expense of writing would be the same in either case, at least if the manuscripts were professionally written. The Christian manuscripts of the second century, although not reaching a high standard of calligraphy, generally exhibit a competent style of writing which has been called 'reformed documentary’\119/ and which is likely to be the work of experienced scribes, whether Christian or not; certainly there is nothing in the nature of privately made copies such as the celebrated manuscript of Aristotle's Constitution of Athens.\120/ And it is therefore a reasonable assumption that the scribes of the Christian texts received pay for their work. Nothing is known of the general level of book-prices in antiquity, but some very rough calculations suggest that by employing the codex format the cost of producing a book might be reduced by about one quarter.\121/ In a specific instance, the early third century manuscript of the Pauline Epistles in the Chester Beatty collection might have cost about 28 drachmae if written in roll form, and about 20 1/2 drachmae if written in a codex -- a saving of 7 1/2 drachmae. The Pauline Epistles in the Beatty codex are about twice as long as an average Gospel, the saving on which would thus be only half this amount.
Roberts, op. cit., pp. 14-15
\120/ One possible exception is P. Baden 4.56, of which Aland says 'es handelt sich sehr Wahrscheinlich um, eine private Abschrift' [[it most probably represents a private copy]].
\121/ T.C. Skeat, ‘The length of the
papyrus roll and the cost-advantage of the codex,’ Zeitschrift
und Epigraphik 45 (1982) 169-175 [[reprinted in ...]].
It seems very unlikely that this reduction in cost would have been sufficient to account for the fundamental change from the roll to the codex. In fact, if economy was such a decisive factor, one would expect to find some traces of other attempts to make the most economical use of the writing material; but such traces are conspicuous by their absence. Scripts are of a normal size, and are not noticeably small or compressed, although had they wished [] to do so there would have been nothing to stop Christian scribes from adopting a script as small as that of the second century codex of the
\122/ The calculation is as follows: the Beatty codex originally consisted of 216 pages, each of 2 columns = 432 columns, the average width of which is 5 cm. The intercolumniations are about 1.5 cm wide, and 432 x 6.5 cm = 2808 cm = 28 m.
it seems to have taken several
for the full potential of the codex to be recognised.
Up to the third century no surviving codex is
known to have had more than 150 leaves = 300 pages,\123/ and many are
smaller. But thereafter they grow to
much greater proportions. One of the
Coptic Manichaean codices (the Psalm-Book, fourth-fifth cent.) had at
pages,\124/ while the great parchment codices of the entire Bible, the
Vaticanus and the Codex Sinaiticus (fourth cent.) contained at least
1460 pages respectively, while the Codex Alexandrinus (fifth cent.) had
least 1640. By now the advantage of the
codex was evident to all, and not merely for Christian literature;
first 35 rolls of Ulpian's Ad Edictum
were republished in three codices containing the text of 14, 11 and 7
respectively,\125/ while Gregory the Great remarks that within the
compass of 6
codices he has compressed a work which had occupied 35 rolls.\126/
The largest so far known is the Philo codex from Oxyrhynchus, which had
least 289 pages, cf. Turner, Typology, p.
82. [[how much Philo did it contain]]
\125/ F. Wieacker, Textstufen klassischer Juristen, pp. 127-8.
the question whether the quality of comprehensiveness may have
early Christians in their choice of the codex, the answer must be much
as in the case of compactness, namely that it is doubtful whether it
been an important factor as early as C.E.
reference. It has been suggested that it would have been much
easier to locate
a particular passage in a biblical text written in codex form than it
in a roll, and that this would have been a decided advantage in the cut
thrust of theological debate. Since a
codex could be opened at a particular place much more quickly than a
be unrolled to find the same passage, this certainly appears to be a
argument: one thinks of Augustine in the famous 'Tolle, lege' episode,
kept a finger in the codex of the Pauline Epistles to mark the place of
providential passage he had found. But
it must be remembered that in the ancient world there was no such thing
exact quotation in the sense of giving the precise location of a
particular passage. The only available
means of so doing was by
means of few examples, both Greek and Latin\128/ of the position of a
indicated either by stating by how many stichoi
it came from the beginning of the work, or, more rarely, from
end. This would, of course, give only an
approximate idea of where to look for the passage, unless the reader
prepared to count the text in stichoi himself. For more
immediate and accurate location the
text would have to be equipped with marginal stichometry, e.g. for
hundred stichoi to be noted in the
margin, and this is far from common. And
of course if the manuscript was so equipped the passage could be
irrespective of whether the manuscript was a roll or a codex.\129/ How
was made of stichometry as a means of reference is illustrated by the
in manuscripts of poetry or drama, where line-numeration could very
been introduced, this seemingly obvious step was never taken.\130/
[[add information on passage titles used in some references]]
\128/ K. Ohly, Stichometrische Untersuchungen, 1928, pp 109-18.
\129/ H. Ibscher, in Jahrbuch der Einbandkunst 4 (1937) 4, actually claimed that for the purpose of reference the roll was just as convenient as the codex ('Selbst als Nachschlagewerk eignete sich die Buchrolle genau so gut wie der Codex' [[as a work of reference, the roll was just as convient as the codex]]).
there are some examples of verse and drama texts with
every hundred lines, these clearly would have been of little use for
reference. [[cite some examples??]] To be of practical use, it
would have been necessary to mark the text at much more frequent
every five lines, as in modern editions. The absence of any such system
proves that the stichometric
just mentioned could not provide, and were not intended to provide, a
has also been suggested that page-numeration, which is a
feature of many early Christian
codices\131/ (the earliest are perhaps the Chester Beatty Numbers and
and the Egerton Gospel), was devised to facilitate reference. But in
the whole of ancient literature there
is no example of a page-reference being given, and the reason is
namely that no two manuscripts are identical and pagination will thus
different in every case.\132/ Moreover, had this been the intention the
pagination would have been inserted at the outset, whereas in fact it
been added by a later hand or hands. It
is much more likely that pagination, which in any case is not
merely a device for keeping the pages in the right order during the
binding and -- perhaps even more important -- to ensure that none were
missing. All this is confirmed by the
fact that in later centuries\133/ pagination is replaced by quire
which fulfils the same function. [[give examples of use of both? GMatt
\131/ On the pagination of early codices see E. G. Turner, Typology, pp. 75-77.
\132/ Compilers of medieval library lists
capitalized on this by often
‘secundo folio,' i.e. the first word
on the second leaf of a
manuscript. This would necessarily be
different in every case, and provided a ready means of identifying
containing identical texts, such as Bibles and service-books.
\133/ Cf. Santifaller, Beiträge zur Geschichte der Beschreibstoffe im Mittelalter mit besonderer Berucksictigung der papstlichen Kanzlei, 1953, pp. 164-5. Pagination does not begin to reappear until the latter part of the twelfth century.
\135/ The general principle, so far as there is one, seems to be that anything intended for continuous or repeated reading or reference was invariably in codex form. This includes literary and scientific works of all kinds, monastic chartularies, collections of statutes, law-books, etc. On the other hand there are certain specialised categories of rolls, such as Rolls of Arms or Mortuary Rolls which are invariably in roll form although they could just as well have been written in codices. Outside the great series of rolls produced by central government, the most prolific source of rolls in England are Court Rolls, the records of manorial courts, which have survived in vast numbers and which continued to be engrossed on rolls down to the middle of the seventeenth century. According to L. Santifaller, op. cit., p. 183, papal records were still being kept on rolls in the third century C.E., but were probably transferred to codices in the fourth century.
\137/ On these problems see E. G. Turner, Typology, pp. 73-4.
[] 8. Conclusions. We have now to consider the extent to which the foregoing arguments might have influenced the early Christians in their choice of the codex. In contrast to the slow and piecemeal process by which the codex ousted the roll in secular literature, the Christian adoption of the codex seems to have been instant and universal. [[i.e. for preserved materials, assuming it is representative]] This is all the more striking because we would have expected the earliest Christians, whether Jew or Gentile, to be strongly prejudiced in favour of the roll by upbringing, education and environment. The motivation for their adoption of the codex must therefore have been something overwhelmingly powerful, and certainly none of the reasons considered above appears capable of producing such an effect. We must therefore seek alternative explanations. Two different hypotheses will be here discussed, although neither can be claimed as more than tentative.
THE CHRISTIAN ADOPTION OF THE CODEX:
It is remarkable that the
never developed a specific word to designate the codex form. It is true
that the Latin codex was
transliterated as κώδιξ, but
certain official, governmental or legal connotation (cf. Atsalos, op.
cit., pp. 143-144): e.g.
in the proceedings of the
Council of Chalcedon (C.E. 451),
where ἀπὸ κώδικος ἀνέγνω =
recitavit [read from a codex] in
the Latin version, of reading from a
register of Imperial letters. By this
time βιβλίον itself had
already come to imply a codex, and
the Latin version of the Acta: cf. L.
Santifaller, op. cit., p. 172, 'in
der lateinischer Ubersetzung wird für βιβλίον
der Regel das Wort codex ...
the Latin translation as a rule the word codex is used to render βιβλίον [biblion]]]
The nearest approach to
Greek term for codex seems to have
been the word σωμάτιον [animal skin?], cf.
Ep. 395, where it
is used for a
parchment codex in contradistinction to ἐν
χάρτῃ i.e. in
a (papyrus) roll; and Ep. 231, where Basil writes ἐν
[in papyrus] while his correspondent
Amphilochius prefers to write ἐν σωματίῳ [on skin?].
should be noted that σωμάτιον
could designate either a parchment
or a papyrus codex. This is clear from
late tradition, preserved by Eusebius
and Jerome\142/ associates Mark with the foundation of the Church of
and the connections of this
Church, when it emerges into the light of history, are with the West
than the East.\143/ If the Gospel of Mark, in the form of the parchment
postulated above, had reached Egypt, it is likely that it would have
copied on papyrus, so much more readily available than parchment, and
codex might thus have been created.
The foregoing is the hypothesis put forward in the predecessor of the present work,\144/ but it must be admitted that the arguments [] against it are formidable. In the first place it is hard to see why the notebook format should have been retained in conjunction with a writing material, namely papyrus, not at that time commonly used for such a purpose. The assumption would have to be made that Mark's original manuscript, or copies of it in the same notebook format, already enjoyed a measure of authority when they first reached Egypt, and that the codex format itself thus acquired a symbolic value, not least because it stood out in sharp contrast both to the Jewish Roll of the Law and to the pagan book; and that for these reasons when it came, inevitably in Egypt, to be copied on papyrus, the codex format was preserved.\143/ Cf. Roberts, op. cit., p. 59. The statement there that letters on the date of Easter were exchanged between the churches of Rome and Alexandria in the later second century is incorrect, there being no evidence for this earlier than the third century.
\144/ Pp. 187-9.A second objection is that the obscurity of the early history of the
the obscurity of the early Alexandrian church, and possible reasons for
see Roberts, op. cit., pp. 49-51, 71. [[note also
that Mark's alleged connection with Rome may be more important than
\146/ J. A. T. Robinson, op. cit., p.
see also van [[sic! von??]] Campenhausen, op. cit., p.
171, n. 112. [[but if the hypothesis of Mark as the source of Mt &
Lk is taken seriously, Mark's early influence is central!!]]
\147/ Roberts, op. cit., p.
59, n. 5 and p.
The figures are those of Th. Lefort in Muséon
66 (1953) 16 sq., quoted in Roberts, op. cit., p.
On the date of the Epistle of Barnabas see now J. A. T. Robinson,
who claims that there is nothing in the Epistle which could not have
written circ. 75 C.E., and would himself place it not long after the
case for a common origin of the two
innovations is prima facie strong,
and if it is accepted, the
beginnings of the Christian codex cannot be associated with Rome and
(the hypothesis which has been discussed above), since the earliest
manuscripts either do not employ nomina
sacra at all, or do so in an uncertain or irregular fashion.\152/
would likewise seem to be ruled out in view of the obscurity of the
Church referred to above. If these two
areas are excluded, there remains only two early Christian churches
sufficient authority to devise such innovations and impose them on
of Antioch\155/ for at least some part in the origin of both nomina
sacra and the codex are strong. It
was one of the principal places where Jewish
Christians, dispersed from
\156/ Acts of the Apostles 11.19.
\157/ Acts of the Apostles 11.20. [[uncritical]]
Roberts, op cit., pp. 34-35.
\159/ For this account of Jewish writing
habits we are greatly indebted
Lieberman, Hellenism in Jewish Palestine,
1950, Appendix III, 'Jewish and Christian Codices' (pp. 203 sq.).
further be noted that in writing the Oral Law on a tablet the form
indicate that no real publication was intended, whereas to publish in
of a roll would be regarded as a transgression of the Law.
On publication and the Oral Law see
Lieberman, op. cit., pp. 84 sq. [[unknown by unknown!! tablet
in NE hist?? see also Resnick for more careful presentation (and
interpretation) of the rabbinic details; Parsons review comments:
"Lieberman ... collected Mishnaic references to the practice of writing
down the oral law (rabbinic sayings and decisions) in pinakes (as
opposed to the scrolls of the Written Law); he suggested that Jesus'
disciples too had recorded their master's sayings on pinakes, and from
such pinakes of papyrus developed the use of the codex. The authors
elaborate this suggestion, and put it forward as the more plausible
hypothesis. But of course it too is open to attack. (i)
Methodologically. it leaves the Western codex as a separate
development. (ii) Evidentially, Lieberman's material is itself not very
convincing. Passages in which Rabbis are said to have written in
pinakes simply use that word, and do not specify the material (as Dr.
S. P. Brock informs me); and the one passage which defines three types
of pinax -- papyrus, was and smooth wood -- regards the first two as
potentially unclean. Clearly there were pinakes of papyrus, and no
doubt these were papyrus notebooks; but the specific connection with
rabbinical practice is not made."]]
\160/ Kelim 24.7 (The Mishnah, trans. H. Danby, 1933, p. 639). For the date see Lieberman, op. cit., p. 203.
\161/ Various theories have been propounded,
suggesting that some, or
of the canonical gospels, and Acts, were written at
this, it could be argued that the
used tablets for recording the Oral Law, but in no case did this usage
the codex. [[no evidence]] On the other hand, the use of
the roll in Judaism was so rooted in tradition and prescribed by the
[[???]] that such
a development would have been impossible.[[!!! but see the Resnick article for
more accurate detail]]
The Christians, however, would have had no such
the adoption of a form of book which like the nomina sacra
would have differentiated them from both Jews and
pagans, as already noted, might have constituted an additional
If the foregoing hypothesis is correct, it follows that the parchment notebook (membranae) can have played very little part in the invention of the Christian papyrus codex. It is true that
\162/ One of the earliest Christian parchment codices would appear to be a page of Acts, P. Berlin. Inv. 11765 = van Haelst 479, which is assigned by Roberts to the second-third century. E. G. Turner, however, in Typology, under number NT Parch 76, dates it fourth century on pp. 29, 159. In the 'Detailed List of Early Parchment Codices' in Typology, p. 39 (cf. also p. 94) there are no Christian items ascribed to the second-third century. Five are ascribed to the third century, viz:
But concerning these five items some reservations must be made. No. 1 is dated ?iii, i.e. third century with a query, on p. 160. No. 2 is dated 'iii ed.; E.G.T. iv?' on p. 163. No. 3 was dated early fourth century by the original editors (and by H. J. M. Milne). No. 4 is possibly Jewish, cf. Roberts, op. cit., pp. 33-34, 77. No 5 is dated third-fourth century by Cavallo..
To sum up, although
neither of the two
hypotheses discussed above is capable of proof, the second is decidedly
more plausible. One final point to be
considered is the date at which the Christians may be presumed to have
the codex, or rather the date by which general agreement was reached in
Church that the codex was the only acceptable format for the
scriptures. [[!!!]] Neither hypothesis provides any chronological
background. So far as the first is
concerned, if the Gospel of Mark provided the ultimate model, we do not
when the Gospel was written, when copies of it could have reached
or how long it would have taken for papyrus to replace parchment as the
material. The second hypothesis is
equally unproductive. If
only hard evidence thus remains that of
manuscripts themselves. We have seen
that there are a number of Christian papyrus codices dating from the
century, including at least one which is agreed to be not later than 150
manuscripts are all, so far as we can judge, provincial productions,
and it is
thus in the highest degree unlikely that they are the earliest codices
produced. All in all, it is impossible
to believe that the Christian adoption of the codex can have taken
later than circ. 100
may, of course have been earlier);
date will be assumed in the following Section. [[entirely simplistic
and relatively uncritical]]
[[Add a section on letters:, and the Pauline corpus hypothesis:
THE CHRISTIAN CODEX AND THE CANON OF
has sometimes been suggested that the adoption of the codex by the
Christians in some way influenced the development of the canon of
scripture. No ancient writer alludes to
this, and there is no direct evidence, so whatever can be said on the
must necessarily be conjectural.
the Christian Bible as a whole, any possible influence of the codex on
contents can be immediately dismissed.
Manuscripts of the entire Greek Bible are excessively rare at
period, and in any case the history of the "Old Testament" canon,
predominantly [[??]] upon the Jewish canon, is quite different from
in the case of the New Testament as an
entity it may be doubted whether the existence of the codex has ever
effect upon the canon. Manuscripts of
the complete New Testament in Greek are by no means common. Of the
2,646 minuscule manuscripts listed by
Kurt Aland in his Kurzgefasste Liste der
griechischen Handschriften des Neuen
Testaments, 1963, pp. 61-202, only 56 contain the New Testament
while a further 136 contain the New Testament minus the
Apocalypse, which the Eastern Churches long regarded with
suspicion. The same picture is broadly
true of the Latin\163/ and other early versions.
Any influence of the codex on the contents of the New Testament
therefore have been on smaller groups, particularly the Gospels and the
To take the Gospels, if the establishment of the four-Gospel canon is linked in some way with the adoption of the codex, there are three possibilities. Either the canon came first, and favored the adoption of the codex, which made it possible to include all four Gospels in a single volume; or the adoption of the codex came first, and realisation of its possibilities favoured the establishment of the four-Gospel canon; or, whether by chance or design, both
[] developments took place simultaneously, without either necessarily influencing the other.
have already seen that the
the codex cannot be dated later than circ.
C.E., and much therefore
upon the date to be assigned to the establishment of the
at any rate whether this took place before or after100 C.E.Unfortunately
there is at present no general agreement on the date or circumstances
the canon emerged. A recent writer on
the subject, Hans von Campenhausen, in The
Formation of the Christian Bible, 1972, would place the emergence
of the canon between the time of Justin and that of Irenaeus (pp.
between circ. 160 and circ. 185,
adding 'to define the date
more precisely than this is not possible'.
His reason for this is that 'the four-Gospel canon was not a
creation, "constructed" at one blow, nor was it disseminated from a
single centre. Its formation was gradual
and the result of earlier presuppositions, and it was in the end
accepted.' As for the circumstances, he takes the view that the canon
being either by direct reaction to the activities of Marcion or, which
thinks more likely, because Marcion created a situation in which the
obliged to define what was authoritative and authentic.
Elsewhere, however, he seems to allow a
rather wider span of years for the emergence of the canon, as when he
'when the beginnings of the four-Gospel canon are placed, as they must
in the second half of the second century' (op.
cit., p. 238, n. 156).
to Irenaeus the four-Gospel canon is something divinely established and
consonant with the forces of nature, and the date of circ.
185 can thus safely be taken as the terminus ante quem for
the creation of the canon [[for Irenaeus and his influence at least]].
But it is by
no means so certain that the age of Justin provides us with a terminus
post quem. Professor
Moule, for example, leaves it
open whether the four Gospel canon is earlier or later than
(circ. 140): 'Did that interesting
heretic find four Gospels already recognized together by about 140
and did he deliberately drop off
and John (as well as the unacceptable parts of Luke)?
Or was it rather that the catholic Church,
after seeing what havoc Marcion wrought by his one-sided use of
brought the four Gospels together to restore the balance and make a
fourfold harmony? This is the same problem
as confronts us for
the whole [] New
Testament canon: was Marcion's the first canon,
and is the
orthodox canon the Catholic Church's subsequent reply?
Or did Marcion play fast and loose with an
already existing canon? There is at
present no conclusive evidence for the existence of a pre-Marcionite
canon. Marcion may have been the
catalyst we have already hinted at. We
cannot be certain' (The Birth of the New
Testament, 3rd ed., 1981,
climate of uncertainty it is very difficult to trace any possible link
the four-Gospel canon and the adoption of the codex.\164/ All that can
that so far at least no critic has suggested a date for the creation of
the canon as early as 100
and we may thus reach the tentative
the adoption of the codex pre-dated the four-Gospel canon.
If this is so, we have now to consider
whether the canon was influenced by the existence of the codex.
\164/ The most detailed proposal to link the adoption of the codex form with the four-Gospel canon is that of G. Rudberg, Neutestamentlicher Text und Nomina Sacra, Uppsala, 1915, pp. 36-46. Accepting the thesis of Hermann von Soden that his three families of Gospel manuscripts, I, H and K, all derived from a single I-H-K archetype, Rudberg concluded that this archetype implied a codex, since a single roll could not contain all four Gospels: cf. op. cit., p. 36, 'Diese technische Einheit der Evangelien, mit dem I-H-K Text, kann nicht eine Rolle gewesen sein und auch nicht mehrere ... Wir müssen ein Buch, ein Codex annehmen'. [[ET This technical unity of the Gospels, as found in the I-H-K text, can not have been a roll and also not ?? . . . We must assume that it was a book -- a codex.]] The adoption of the codex and the establishment of the four-Gospel canon were thus intimately connected and each presupposes the other. However, von Soden's theories have not found acceptance, and in any case the position has been radically altered by subsequent discoveries.Pre-existence of the codex was certainly not essential for the creation of the canon. The Jews, after all, [[presumably]] created their own canon of the "Old Testament" without any benefit of the codex, and no doubt the Christian Church [[leaders]] could have decided upon the four-Gospel canon irrespective of whether at the time the four Gospels were circulating as four rolls, four codices, [[some combination of rolls and codices,]] or one codex. One area in which it has been claimed the codex exercised a decisive effect is the canonical order of the Gospels. As Campenhausen puts it: 'Any publication which established a fixed sequence of gospels is conceivable only as from the start in the form of a codex' (op. cit., p. 173). But against this is the evidence of variations in sequence, notably the so-called 'Western Order' of Matthew, John, Luke, Mark which is also found in the fourth century Freer Gospels, a manuscript almost certainly of Egyptian origin, while Campenhausen himself points out that although Irenaeus in discussing the origins of the Gospels treats them in [[what became]] the
[] canonical sequence (apparently because he believed this was their chronological order), he elsewhere always uses the order Matthew, Luke, Mark, John (op. cit., p. 195, n. 243).
But perhaps the strongest argument against any definite link between the four-Gospel canon and the codex is the extent to which, both during and after the second century, Gospels continued to circulate individually or in smaller groups or in conjuction with other books of the Bible, and that too not only in Greek but in the Versions also. Examples from the fourth century onwards are given by Zahn,\165/ but there are plenty of earlier date. The earliest extant Gospel manuscript, the Rylands John (P52) probably never contained more than that Gospel (see below), as certainly was the case with the somewhat later Bodmer John (P66) and the third-century P5, a bifolium consisting of conjoint leaves from the beginning and end of a single-quire codex of the same Gospel, while another notable example is the fourth-century codex of John in Sub-Achmimic from Qau. For groups of less than four Gospels we have the Bodmer Luke and John (P75), the codex of Matthew and Luke divided between
In the next
century the Chester Beatty Gospels and Acts (P45) originally
leaves (= 220 pages), and this was achieved by the use of a larger page
25.5 x 20 cm), giving a larger written area (about 19 x 16 cm),
together with a
smaller script than in most of the second-century codices.
For the present, therefore, a second-century
codex of all four Gospels seems unlikely; and there is much to be said
Campenhausen's conclusion: 'the fact that Irenaeus and the Muratorian
regard the fourfold gospel as a spiritual unity is a theological
nothing to do with book production' (op.
cit., p. 174).
have spoken only of the Gospels. Of the
remainder of the New Testament, the most obvious group which might have
influenced by the codex is the Pauline Epistles. These
form a body considerably shorter than
the Gospels (5,095 στίχοι according
to Rendel Harris),\167/ so a codex containing all of them would have
proportionately more feasible in the second century, as the Chester
codex shows that it was in the third. But there is no evidence
whatsoever to indicate that the codex
any part in their selection or circulation.\168/ [[add Trobisch stuff]]
\169/ The title itself is not original, being taken from Birt's Abriss, p. 351.
\170/ This is
a widely-held belief. Cf, e.g., Cavallo, op.
p. xx.: 'a dar vita alle prime communità crisitane
fu una plebs senza ruolo economico, politico e
intellettuale' [[ET to
give life to the
first Christian communities was a constituency (plebs) without
political and intellectual status]]. But
E. A. Judge, The Social Pattern of
the Christian Groups in the First Century, 1960 , p. 60: 'Far from
socially depressed group, then if the Corinthians are at all typical,
Christians were dominated by a socially pretentious section of the
of the big cities.' Also p. 61: 'Pliny
accepted the fact that Christians represented a broad cross-section of
from Roman citizens downwards, but reserved his surprise, apart from
numbers, in which he is an alarmist, for the ominous fact that the new
was infecting not merely the cities but the countryside.
Until then, however, we may safety regard
Christianity as a socially well-backed movement of the great
suggestion that there was a kind
distinct sub-culture whose favorite reading was popular romances in
is an attractive one, but hard evidence for this is slender. If we
analyze the surviving fragments of
||P.Colon. inv. 3328
||P.Mil.Vogl. 3.124 (AchillesTatianus)|
|| P.Schub. 30 (P.Berol. inv. 16971) (AchillesTatianus)
Achilles Tatius (Pack-2 no. 1).\171/
the end of the section 'Romance,’ Pack-2
lists other possible romance texts; of the eight antedating the fifth
not one is a codex or opisthograph; one is an ostrakon. [[Pack-Mertens
3 (online) displays 683 items in response to the request for papyri
codices in the June 2006 online edition, and 206 for parchment codices.
For the period up to the year
300, M-P\3 claims 214 papyri codices and 30 parchment; for the
period up to 200 CE, only 51 papyri and 5 parchment. Many of
these are dated ca. 200.]]
we take other types of literature which
reasonably be regarded as popular, one obvious genre is the so-called
the Pagan Martyrs or Acta Alexandrinorum.
For these texts Montevecchi gives the following figures:
Every one of
is a roll. [Not too surprising,
since most of them are dated relatively early. M-P\3 lists one
2226.01 "Acta Alexandrinorum" ? (Acta Maximi ?) BKT 9.177 (P.Berol. inv. 21273) Fayoum ? III CP]