An Introduction to the New Testament

By Adolf Jülicher


Chapter 2 - Section 50


§ 50. The Outward Form of the Texts down to about 1500 A.D.

[Cf. for this and the following sections C. R. Gregoryʼs ʽProlegomenaʼ in C. Tischendorfʼs ʽNoyum Testamentum Graece,ʼ ed. 8, vol. iii. (1884, 1890 and 1894.]

1. The exact time at which papyrus gave way to parchment as writing-material for the sacred books cannot now be determined. It probably happened at different times in different places—in Egypt naturally later than elsewhere: but soon after the Mohammedan invasion in the seventh century, papyrus seems even there to have disappeared entirely, even from domestic use. At any rate, all that Theodorus of Mopsuestia, who died in 428, knows of it, is that many years ago, in the time of Paul, men possessed the Divine Scriptures in the form of rolls. Jerome tells us that between 340 and 880 the bishops of Caesarea saved the library formed in that place by Origen and Pamphilus from decay by laboriously transcribing everything it contained on to parchment. Thus the greater part of this library must originally have consisted of papyrus rolls, and we may probably consider the period about 3800 as that of the general transition to the use of parchment. In the persecution of Diocletian it is in almost every case the codices of the Divine Law which are sought for by the authorities and given up by cowardly Christians; if in later times the ʽvoluminaʼ are still spoken of, it only means that the old name had been retained for the new thing.

It was, in fact, very difficult to convert stiff parchment, ill adapted as it was to the process of gluing, into rolls; the usual practice was to fold the leaves over in the middle, and then to lay several of them one inside the other, or one on the top of the other; booklets thus produced could be fastened together by the binder in any desired number, making a volume resembling the form of our present books.1 As a rule, such a folio consisted of four double leaves (quaternio) and more rarely of five; one, two, or three were scarcely ever used except at the end of a book, when a complete folio was not needed. Both sides were written on, and thus it consisted on an average of sixteen pages, like a printerʼs sheet of to-day. Some particularly strong material, such as wood, sometimes covered with leather or silk, was chosen for the binding of the folios, which, when put together, were often very thick; for the finely dressed parchment of ancient times now disappears for the sake of greater durability. The fact that economical owners were often tempted to make more than one use of their parchment is in a sense a misfortune, but often turns out the reverse. If a library already contained several copies of the New Testament, but not the works of some revered father of the Church, the addition was made without expense by scratching out the original writing (in case it was not already faded enough) in one of the New Testament parchments, and writing the desired text over the old, or between the lines, or occasionally, but not often, crosswise. Such manuscripts are called Palimpsests (codices rescripti, and sometimes even bis rescriptt). The original writing, which can often only be made legible by means of chemical reagents, is generally the most interesting to us; whatever fragments we possess of the Gothic translation of the New Testament and of the oldest Syriac version of the Gospels have come down to us for the most part from ʽCodices rescripti.ʼ¯ On the other hand, from the fourth century onwards the Bible manuscripts were often prepared with extravagant splendour; parchment of marble whiteness and of the greatest delicacy was procured, gold and silver letters were painted on a ground of purple—as in the ʽCodex Argenteusʼ of the Gothic translation of the Gospels at Upsala —and the cover richly adorned with jewels and fitted with costly clasps; while the decorations which were inserted in the margins of the manuscripts, especially at the beginning of the book, belong to the most valuable material for the history of Christian Art.

Paper, a cheaper writing-material than parchment, at last took its place in the cultivated world about the shores of the Mediterranean; it was apparently invented by the Chinese, and made out of linen rags. It was known to the Greeks as early as the eighth century, and from this time onwards leaves of linen, as formerly of papyrus, are to be met with between the parchment pages. The traditional material, however, was long preferred for New Testament manuscripts. It was not till after the fourteenth century that the parchment manuscripts disappeared entirely, and the ʽCodices bombycini,ʼ and ʽchartaceiʼ replaced the ʽCodices membranacei,ʼ though retaining in all other respects the appearance of the older books.

9. When the Emperor Constantine commissioned Bishop Eusebius of Cesarea to provide him with fifty copies of the Holy Scriptures for the newly built churches of his capital on the Bosphorus, he expressly desired that they should be very legible and of a convenient size for general use. In the latter respect tastes and necessities varied with the times, but in general the tendency to a decrease in size is unmistakable in the history of the codices. When Jerome bewailed the unwieldy bulk of the codices then in vogue he was probably not thinking only of their thickness. Among the parchment manuscripts still extant we may find examples of the large folio, the quarto, and the small and dainty octavo—the last a sign of a comparatively modern age. The manuscripts we possess of the Greek New Testament never, so far as we know, exceed a size of 18 inches in height by 16 in breadth; a very general medium size is 12 by 8. The parchment pages were originally considerably higher than the average of those made of papyrus, and also of correspondingly greater breadth; thus if the copyist still wished to keep to the usual length of the lines on a papyrus page, and was yet unwilling to leave such enormous margins unused, he simply divided each page of the parchment into several columns, clearly separated from one another by a small space: the Sinaiticus has four such columns, the Vaticanus three, but it is more usual to find only two. Even some of the quite ancient manuscripts, however, have their lines running across from margin to margin, and when it became the custom to cover the text with all manner of auxiliary apparatus, equally wide margins were needed for every portion, so that this also contributed to the abandonment of the older fashion.

3. A change in the nature of the writing-materials, however, need not necessarily have brought about a change in the characters used. Not until the ninth century are the uncial letters, which had been retained until then, supplanted by the cursive hand, but even then in such a manner that the conservatism of the Church long clung to the older custom—in fact, until late in the eleventh century—as is proved by a great number of lectionaries. The cursive hand is also called the Minuscule, and the uncial the Majuscule. But it is not principally the height or even the general size of the single letters which makes the distinction between the two methods of writing; large and coarsely written minuscules on the one hand, and very fine and delicate uncials on the other, are not uncommon. Naturally, moreover, the change did not come about without some preparation. The uncial writing had gradually dropped more and more of its old beautiful features, the letters had become narrower and more pointed, and had begun to slant to one side; the practice of joining several letters together was growing commoner; the differences in length—as for instance between Iota and Rho—increased; we find in fact that a semi-uncial hand was developing. In the case of the cursive hand still more, almost a new alphabet had at last been produced; we can still perceive its relationship to the original form of the letter, but everything has become smoother, partly through abbreviation and partly through the separation of words, though always with the tendency to make the fewest possible strokes, and to lift the pen as seldom as possible. This form of writing, too, underwent many developments; it borrowed again and again from the old uncial letters, and it is the foundation of our modern Greek hand. It is not the fault of the cursive hand, but of its innumerable abbreviations, that the manuscripts of the later Middle Ages are in general so difficult to read; whole words are often represented in them by a single hieroglyph, while in the old manuscripts such abbreviations are put rarely found, and then only in the case of constantly recurring words (e.g. ΚΝ for κύριον, ΑΝΩ, for ἀνθρώπῳ,, ΠΝΑ for πνεῦμα). With the minuscule, again, it now becomes the rule to separate the words by dots or by a space, and to insert punctuation and signs; but after the eighth century these are also found in uncial codices, and are apparently not merely the insertions of a later hand. Individual scribes well versed in the rules of grammar made accented copies (κατὰ προσῳδίαν) of the Books of the Bible as early as the fourth and fifth centuries, but this attempt had no permanent success.

Most of the good codices are carefully and evenly written; the scribe drew lines to help himself (and in the case of fine parchment it was only necessary to do this on one side), sometimes single ones, in which case the letters were merely written upon them, and sometimes double, when they were inserted between the two; the space between two such rows would then be about equal to the height of the row itself. The number of rows on each page depended on the shape of the codex and the copyistʼs manner of writing: in the Sinaiticus there are 48 on a page; in the Vaticanus, although it is much less in height, 42; in the Codex H of the Pauline Epistles, 16, although it is about an inch higher than the Vaticanus; in the codex Δ of the Gospels, the number varies between 17 and 25. A single column of the Sinaiticus takes on an average about 12 letters, of the Vaticanus about 17, of the Alexandrinus about 21, of the Codex Ephraemi S about 40; thus, counting the columns together, there are respectively 48, 51, 42, and 40 letters on each line. As the line represents a mere unit of space (and not of sense), words are sometimes broken off at the end of them without a hyphen, e.g. πει ρασμον, but this hardly ever occurs in the middle of a syllable.

4. But searcely a single writer of the New Testament manuscripts known to us was content to reproduce his original without any regard to the meaning of the text—that is, without giving his readers some assistance towards the understanding of it. At the beginning of a new paragraph the Sinaiticus makes the first letter project into the left-hand margin, and from the fifth century onwards it became usual to distinguish these initial letters by their special size as well— later even to paint them with some colour, mostly red. Then from the single letter several came to be treated in this way, till at last the whole first word was coloured. But the needs of the reader (and of the church reader in particular) were met comparatively early by a much more comprehensive plan. The New Testament text was split up into a series of sense-units, written in such a manner that the beginning and end of each unit must be clearly perceptible, whether it filled the space of one or more actual lines. This was, however, a costly undertaking, as by this method half lines and more had constantly to be left blank; and indeed it was probably on this account that the system disappeared, even before a better substitute was found for it in a rational system of punctuation. This Colometric2 method of writing appears to have been introduced into the sacred literature of the Greeks by Origen—at first only for the Psalms, in which the nature of Hebrew poetry determined the limits of the sentences automatically. Thus in his ʽHexaplaʼ he could give a comprehensive view of all the seven different Greek translations beside the original text. When Jerome had his Latin translation of Isaiah written out in separate versicles of this sort (ʽper cola et commataʼ—and even Cassiodorius made the mistake of applying the words to the punctuation-marks so named in modern times! ) he warned his readers against the error of supposing that they were dealing with poetical verses, and excused himself by saying that this method was employed in the works of Demosthenes and Cicero, although indeed he was quite conscious of the novelty of applying it to the prose books of the Bible.3 For this reason alone, then, those New Testament manuscripts in which this practical method is adopted could not well have been written before the fifth century; the most famous of this kind are: for the Gospels, Codex D; for the Acts, Codex E; and for the Pauline Epistles, Codex H.4 The average length of one of these sense-units differs very much according to the different ideas of the writers as to what might be called the ʽsmallest complete sentenceʼ; the Laudianus (Codex E) has particularly short units, but those of most of the others are also rather shorter than our present verses. Where a ʽcolonʼ required several lines, the auxiliary lines were designated as such by inserting them between the usual ruled lines; but it is clear that all kinds of confusion must have arisen in this respect on recopying.

Unfortunately, this method of writing in units of sense has often been designated the stichometric method; but stichometry is in reality not a manner of writing at all, but a system of measuring off the texts when written. Hven as early as the Codex Sinaiticus, notes are inserted in the margin beside most of the Pauline Epistles—though it is true they are in a somewhat later hand—giving the number of stichiz in these Epistles.Στίχος,, Latin versus, is a mechanical division, and it is not till the time of the Byzantines that we find it used to denote a sentence. The intermediate stage between the two meanings is furnished by the poetical Books of the Old Testament (βίβλοι στιχηραί) because there every ʽverse,ʼ i.e., the smallest complete phrase, filled exactly one line. In the case of prose works this attention paid to the lines is at first sight somewhat surprising, and in reality we find that all the Pauline Epistles in the Simaiticus take up many more lines than the number of στίχοι given. But the stichus had long become a technical term in the bookselling trade, a unit of measurement for written work familiar to every expert; thus Josephus reckons the contents of his ʽArcheologyʼ at 60,000 stichi, and Origen, without making any definite calculation, can say of the second and third Epistles of John, that they were less than 100 stichi long. The hexameter was the foundation of this unit of measurement; 12 to 19 syllables, or 32 to 44 letters, were probably the usual amount for a stichus. Prices could only be settled accurately with the calligrapher or the bookseller by the help of the stichic measurement, and it is no wonder, then, that the number of stichi was calculated in the New Testament Books too, and the result noted down in the post-scriptum. But this does not necessarily mean that the books were written out so as to correspond with this number (i.e. in lines of exactly the length of a hexameter); the conditions of space often prevented this, and the end was attained by inserting the number of stichi in the margin at intervals of 50, and also at the ends of longer paragraphs, while the numbers for each individual Book of the New Testament were added up in a separate note.

From the sixth century onwards we scarcely ever find a Greek manuscript in which the numbers of the stichi are not given in this way, and those for the Acts and the Epistles usually agree with those of the so-called ʽText of EKuthalius,ʼ though even in their case attempts at a different mode of reckoning are by no means unknown. When we remember the endless copying and re-copying which these very unstable figures must have undergone, we must, of course, expect to find many mistakes among them, for they were probably never corrected by the process of re-counting. When, as sometimes occurs, the numbers of the ῥήσειε or ῥήματα are given instead of, or as well as, those of the stichi, it means that a different authority from that for the stichi has been followed, though with the same intention; the totals of the ʽsentencesʼ are too nearly identical with those of the stichi to admit of the supposition that a different principle of reckoning was adopted in their case.

[For the following ef. the ʽCollectanea Monumentorum veterum Eeclesiae Graecaeʼ of L. A. Zaccagni, published in Rome in 1698 (vol. i. pp. liv-xci and 401-708). See also ibid. p. 724: ʽHuthalii Episcopi Sulcensis Actuum Apostolorum et xiv Sancti Pauli aliarumque Catholicarum Epistolarum editio ad Athanasium juniorem Episcop. Alex . . . . graece et latine edita.ʼ Also J. A. Robinsonʼs article on ʽEuthalianaʼ in ʽTexts and Studies,ʼ iii. 3, 1895; and E. von Dobschiitz on ʽButhaliusstudienʼ in the ʽZeitschrift für die Kirchengeschichte,ʼ part xix. 1898, pp. 107 fol., and on ʽButhaliusʼ in the ʽProtestantische Real-Encyclopädieʼ (edited by Hauck), part v. 1898.]

5. But the New Testament text was not only copied out in more or less practical form; as in the Masoretic version of the Old, it underwent a peculiar form of elaboration, and was in fact surrounded by a mass of auxiliary notes of all kinds. I am not referring here to the ʽCatenaeʼ (see pp. 599 fol.), although in the later Middle Ages scarcely a single Greek text of the New Testament was allowed to appear without them; nor to the increasingly copious postscripts giving information as to the original language and the author of each document, and the place and time of its composition; nor to the tables of contents5 at the beginnings, and all the later amplifications of the older and shorter superscriptions.6 In addition to these a learned apparatus of the most diverse character and value was added to the text, and vestiges of this are still to be found even in the latest printed editions.

In this sphere of activity the master and pioneer appeared until recently to be a certain Euthalius of Alexandria, whose work was dated by its first editor, Zaccagni, at 458. Thé mystery in which this remarkable book used to be wrapped is not yet quite cleared away, but, owing to the searching investigations of Robinson and von Dobschiitz in particular, we now possess the certain knowledge that the Euthalius of Zaccagni did not constitute a literary entity at all, but was a compilation put together by different hands from materials belonging to different periods, practically complete as early as the fourth century, though enlarged even after the year 600 by additions from other sources. Whether a Euthalius was at least one of the revisers—possibly the editor of the year 396—remains an open question until these manuscripts have been more accurately and fully examined, for then only will it be possible to determine his share in the work of compilation. The very diverse elements that go to make up the Corpus Euthalianum are held together by one interest only—that of presenting the Apostolic writings to the Church conveniently arranged and adapted for study, according to the approved models of Greek scholastic learning. We do not yet know whether the text which the so-called Euthalius used as the foundation for his work was a particularly good one; but in any case he wrote it in ʽsense-unitsʼ from beginning to end, furnished it with stichometry, carefully identified all the quotations to be found in it, both sacred and profane, prepared indices for these quotations, and made the consultation of them easy by a complicated system of inserting figures in the margin opposite the place containing the quotation. In addition to all this he contributed short prefaces to the Epistles, chronological sketches of the life and death of Paul, and other embryonic attempts at an ʽIntroduction to the New Testament.ʼ

But probably the most useful part of all this work was his division of the Acts and the Epistles into longer and shorter sections. The Acts, for instance, we find divided into forty chapters (κεφάλαια), of which the first and second together form what is now our first, and the third our present second. In most of these, again, subdivisions (ὑποδιαιρέσεις) are added, always beginning lower down than the beginning of the chapter proper; e.g., in the Euthalian chapter iil. they begin at what are now verses ii. 14, ii. 17, ii. 22, ii. 37, and ii. 42. The numbers of these sections are again noted on the margin of the text, by means of red pigment. But the indices to the chapters and sections do not consist in simple numbering, or in the mere giving of the initial words, but an attempt is made in them—and by no means unskilfully—to summarise the contents. The seventh chapter of Romans, for instance (verses vi. 1-23 in our version), is thus described: ʽConcerning the good conduct which ought to accompany faithʼ; chapter xvii. (=vv. xii. 1-3) thus, ʽInjunctions concerning virtue towards God and menʼ; section α (vv. xii. 3 fol.) thus, ʽOn concordʼ; section δ (vv. xiii 1 fol.) thus, ʽOn obedience to the higher powers.ʼ Lastly, considerably larger sections are formed by putting together several chapters to make one lesson (ἀνάγνωσις).7 These, too, are of very varying length, but the author of these old pericopae evidently had the object in view of dividing the whole body of the Apostolic writings into Lessons embracing a complete ecclesiastical year of fifty-seven services.

This ideal could never be maintained in the public worship of the communities, and thus the Lessons of Euthalius never attained any very wide acceptance. But his chapters and all the rest of his arrangements played all the more important a part in the Greek and Syriac Bibles. He never won complete and sole recognition, however—still less in the case of the Acts and Catholic Epistles than in that of Paul—and the Gospels, which he never touched, had already been satisfactorily arranged in chapters before his day.

The Latins gave the name of breves to what the Greeks called κεφάλαια (and also τιτλοί, περιοχαί, and περικοπαί) a word which had at first undoubtedly signified the summary of contents at the beginning of the chapters, and was not applied until later to the chapters themselves. The now universally adopted system of division was introduced in the beginning of the thirteenth century by Stephen Langton, Chancellor of the University of Paris, principally for the sake of convenience in quotation and reference, and with this object he aimed at as close a similarity as possible between the lengths of the chapters. This innovation soon made its way into all Latin Bible-manuscripts; and as it was in the West too that the first printed versions of the Greek Bible appeared, it naturally followed that the approved arrangement should also have been introduced into those versions. The fact that in an arrangement so indispensable in our eyes to the scientific and edificatory use of the Scriptures, unity was not attained until after a thousand years of diversity, can only be explained by the circumstances of the times; we can, in fact, barely understand that up to the end of the fourth century such divisions were dispensed with altogether; for when earlier writers, such as Clement of Alexandria, Origen, or Dionysius of Alexandria, speak of Pericopae and chapters in connection with New Testament Books, they only mean divisions according to the sense, which the observant reader perceived to be wholes complete in themselves, but which need not for that reason have been marked upon the text. And in fact that they were not so marked can be proved from the language of Jerome.

Eusebius, who was the first to undertake the subdivision of the four Gospels (he made 1162 chapters out of them), did so with the sole object of giving the reader a synoptic survey of the parallel passages within them. To accomplish this, therefore, he seeks and carefully marks out the passages in each Gospel for which parallel passages can be found in the three others, in two of them, or in one, or for which there are no parallels at all; then counts up the sections thus obtained in each case (e.g., 355 for Matthew, 232 for John), some of which are infinitesimal, and others (especially in John) of considerable length, and prepares a table of ten rubrics (κανόνες), in the first of which he sets down the passages common to all four Gospels, in the second those common to Matthew, Mark, and Luke, and so on. The tenth gives those passages peculiar to Matthew, Mark, Luke, or John, in four separate columns. When the numbers of these chapters as well as those of the rubric to which each belonged were correctly noted in the margins, it would not be difficult to obtain a synoptic view of any given portion of the Gospels with tolerable rapidity and with sufficient accuracy to satisfy the demands of that age. The plan of this work, of which Eusebius speaks in his dedication to Carpianus, had occurred to him while making use of a ʽDiatessaronʼ by Ammonius of Alexandria8; this man had wished to attain the same end—though at the expense of Mark, Luke, and John—by adding to the complete text of Matthew the corresponding sections from the other Evangelists. Unfortunately, this Husebian apparatus was too complicated to be handed on without corruption, and a few mistakes would have vitiated it all; but it is characteristic of the conservatism of the Church that almost all the Gospel manuscripts from the sixth to the sixteenth century possess it, although all interest in these comparative studies had long died out. Far more useful to the clerical owner were the marginal notes, α (ἀρχή) and (τέλος), which marked the beginning and the end of the Lessons for Sundays, Saintsʼ days and festivals, and are regularly found in all New Testament manuscripts after the ninth century, while accurate indices of these pericopae may also be found attached to them.

All this supplementary matter, which bears witness to the labour of the Church on the sacred text, does not deserve special attention on account of its possible value in the history of the Church, of literature, or of culture—for no very excessive intelligence, after all, went to the production of it— but it is often full of significance for the history of the New Testament Text, as giving useful indications concerning the origin, antiquity, birthplace and mutual relationship of the different manuscripts. As a rule, it is the mistakes it contains which render the best services in this respect.



1) τεῦχος, πυκτίον, σωμάτιον in Latin = codex.

2) A κῶλον, according to Augustine, De Doctr. Chr. iv. 7 = Lat. membrum, phrase.

3) ʽNovo scribendi genere distinximus.ʼ

4) See § 52, 2.

5) ὑποθέσεις.

6) E.g., instead of πράξεις ἀποστόλων, πράξεις τῶν ἁγίων ἀποστόλων,, and later still ʽwritten by the holy Apostle and Evangelist Luke.ʼ

7) Acts has 16 Cath. Ep. 10; Pauline Ep. 13.

8) Probably about 250.