Biblical Majuscule

Biblical majuscule is so called because it is illustrated particularly in the great Bible manuscripts of the fourth and fifth century such as Codex Vaticanus (discussed below), Codex Sinaiticus (currently dismembered and housed in various libraries but viewable on line) and Codex Alexandrinus (British Library, Royal 1.D.VIII). However, it was in fact used for writing all kinds of sacred and secular texts, not only Bibles. It makes its first appearance in papyri of the 2nd century C.E. and continued in use alongside minuscule well into the minuscule era.

According to the description provided in Cavallo’s book (mentioned in the preceding page), this script, which does not present any particular difficulties for decipherment, is characterized by:

The last two characteristics are the ones which Cavallo particularly identified as distinguishing Biblical majuscule; regarding the last characteristic, Cavallo argued (pp. 4-5) that the chiaroscuro effect was due to the use of a pen with a wide nib held at an angle of about 15°. The result of tracing certain letters according to Cavallo’s description with a nib held at a 15° angle may be seen below (the width of a nib at 15° is indicated in blue):

Many if not most people who have taken a course in Greek paleography in recent decades have been introduced to Biblical majuscule by studying a black and white reproduction of page 144 of Codex Vaticanus 1209 (shown below in color), because this page was featured as Plate 1 in Enrica Follieri’s widely-used collection of plates, published in 1969:

If you zoom in on this page and compare its script with the sample letters above, you will find that the latter are a reasonable approximation of the former; but you will also notice some discrepancies, especially with the last two characteristics in our list above. It is true that the letters in the manuscript generally lack proper "serifs," but many letters do have oblique or circular flourishes at the end of the strokes. As regards the angle at which the pen was held, examination of circular letters such as epsilon, omega and sigma shows that the narrowest parts of the "circle" do indeed correspond to positions compatible with an angle of about 15° (i.e. approximately 11 o’clock and 5 o’clock); and the pattern of thick and thin strokes often corresponds to the chiaroscuro produced by such an angle in letters like alpha, beta and mu, as shown in the sample letters above. However, the effect is very weak, to the point that many letters seem in fact to be composed of strokes of more or less uniform thickness; and it is also clear that many strokes do not follow the expected pattern: in particular, diagonal strokes rising from left to right (e.g. the upper right-hand strokes in ypsilon or in kappa), which should be very thin, are not; and while all three strokes of nu should in principle be of approximately equal thickness (as in our sample letters), in fact the middle, diagonal stroke is generally noticeably thinner than the other two.

There is also a more fundamental problem with the notion that the Vat. gr. 1209 can serve as an example of the canon described by Cavallo. Looking at the page above, you will notice that certain letters appear faded, including seven letters in a row (ς δερρεις) in the second line of the third column. The reason for this was explained by Cavallo himself (p. 53) and in somewhat greater detail in P. Versace’s recent book I Marginalia del Codex Vaticanus, pp. 43-50; 67-68. The original script of the manuscript, having started to fade rather badly, was retraced, letter by letter, probably twice (as demonstrated by Versace), namely once in the tenth or eleventh century, with ink only slightly darker than the original ink; and once again, probably in the sixteenth century, with the black ink which is seen today in most of the manuscript. The retracing was done very carefully, but you can see the outlines of the original letters peeking out from many of the black letters in the manuscript. The scribes who retraced the text occasionally skipped single letters or entire words, sometimes by accident but more often because they considered the skipped letters (or sometimes entire phrases or verses) to be mistakes; these are the ones which appear faded, and they are also the only fully visible bits of the original script.

As a result, it is actually only the letter nu which here represents the fourth-century "perfection" of the Biblical majuscule. In p. 144 of the manuscript, shown above, there are also several letters which were rewritten by neither of the medieval retracers, including several instances of "ephelcystic nu;" two mu’s (first column, lines 3 and 28) whose omission transforms the Hellenistic verb form λήμψονται into the Attic λήψονται; and some other individual letters which you may want to find and interpret yourself. In the second line of third column, at Nb 4:25, the first retracer, who was clearly comparing this manuscript to another one with a slightly different text, decided that the other manuscript’s reading τὰ σκεύη ("the instruments") was preferable to the reading τας δερρεις ("the curtains") which the original scribe had written in the Vat. gr. 1209, so he added a reference sign (an s-shaped squiggle) above the alpha in τας, drew the same reference sign in the margin, adding the word σκεύη underneath it, and indicated that the letters ς δερρεις were to be disregarded by not retracing them. The second retracer followed suit, also retracing the marginal reading (but not the reference signs). The result of all this, for our purposes, is that we have eight more clearly visible letters of the original hand.

There are similar cases in a number of other pages (the longest stretch of original script is about four lines on p. 1479, at 2Co 3,15-16), but it has always been rather impractical to study the original script of the manuscript, since this would have involved looking up bits of faded text on hundreds of different pages, without ever gaining an impression of what a page written in this script actually looked like as a whole. Today, however, it is relatively simple to make a collage of the remains of the original script, which is what you see below. (This collage includes only passages where more than a few letters are preserved together, and has been "cleaned up" by eliminating traces of the script on the other side of each leaf, which often show through, as well as other blemishes. It may be that some of the letters seen here were rewritten by the first retracer: it is impossible to tell for sure, outside of those very rare cases where all three "levels" of script are clearly distinguishable in close proximity, as in the example of ειπεν above; however, this seems rather unlikely, as you will see if you compare the ε in ειπεν above with any of the ε’s in the image below. It seems clear, in fact, that the second retracer [unlike the first] was not collating the manuscript against another copy, looking for more passages to leave out, but rather merely repeating the work of his predecessor, respecting his choices in the process.)

Transcribing this collage (and checking your transcription against the one which can be found here) will familiarize you with this script. For further practice in reading Biblical majuscule, you may use the transcriptions from other parts of the manuscript which are provided here (skipping for now the first transcribed page [p. 23], which is a later addition in minuscule), keeping in mind that what you are reading is not really a fourth-century script but a sixteenth-century retracing of it.

Another exercise which you can perform with this page is to attempt to verify the hypothesis that two different hands may be distinguished this manuscript. According to Milne & Skeat, Scribes and Correctors of the Codex Sinaiticus, pp. 87-90, Scribe A copied pp. 41-334 (Gn 46:28-1Rg 19:11); Scribe B copied pp. 335-624 (1Rg 19:11-2Esd); Scribe A copied pp. 625-944 (Ps-Tob); and Scribe B copied pp. 945-1518 (Hos-Dn, and the entire New Testament). However, these results were arrived at not by observing the scripts of the two scribes (this was thought impossible due to the retracing, since Milne & Skeat did not have the benefit of the collage above) but by observing other habits of theirs, e.g. the way they decorated their colophons, the way they used the sign > to fill out short lines so as to justify their right margins, etc. The boundaries between these sections are indicated in the collage with turquoise lines; can you identify differences between the script of Scribe A and that of Scribe B? You may find it easier to perform this exercise using the image provided here, where the samples are sorted by scribe and where you will also find some observations which you can compare with your own.

Since this is your first transcription exercise, it will be well to note certain general features which you will need to notice and understand while transcribing Greek manuscripts, before returning to consideration of Biblical Majuscule specifically. One feature which you cannot fail to notice is the lack of word division. It is worth noting that "word" (and therefore "word division") is at best a very fuzzy concept and that the Greeks were, in a sense, quite right not to incorporate such a concept into their writing system. Still, the Greek grammatical tradition going back to ancient times makes it very clear that the Greeks shared our intuitive notion of what a "word" is, so it is perhaps surprising that their script, with few, mostly late exceptions, normally appears as an uninterrupted flow of signs (known as scriptio continua). Despite this, you should follow the customary practice of introducing word division into your transcriptions, and you will find word division in the transcriptions provided here.

On the other hand, Greek scribes are not at all bothered by division of a single "word" between two lines, and no sign comparable to our hyphen is used to warn the reader of this. However, scribes usually do endeavor to make line breaks coincide with syllable breaks, following the usual rules of Greek syllabification: namely, single consonants are assigned to the following syllable, as are those consonant clusters which can occur at the beginning of a word; in consonant clusters which cannot begin a word, the break is places so as to form a cluster on the right (which may consist of a single consonant) which can begin a word. Thus we have, in Vat. gr. 1209, νυ|κτος (p. 62, c. 1, ll. 2-3 from the bottom); ε|στιν (ibid., c. 3, ll. 11-12); αι|σχυνθησονται (p. 715, c. 1, ll. 24-25); αν|δρα (p. 1207, c. 2, ll. 22-23); δεν|δρον (p. 1252, c. 2, ll. 9-10 from the bottom). These rules are generally followed even in compounds like ε|ξηλθαν (ibid., c. 1, ll. 11-12); προ|σηλθαν (ibid., c. 3, ll. 15-16), though exceptions do occur (e.g. ε|χθρος in p. 1252, c. 2, ll. 2-3; in addition, some scribes are much less careful in this respect than the scribe[s] of Vat. gr. 1209). Syllabification rules at line break may even apply across word boundaries, as in ου|κ ην (p. 62, c. 1, ll. 5-6); ε|π᾽ αυτο (p. 144, c. 1, ll. 9-10); με|θ᾽ ημων (p. 1207, c. 1, ll. 27-28). These last three examples suggest that the scribe was thinking in terms of syllables rather than words; and this may indeed explain why scriptio continua continued to be the normal practice of Greek scribes for so long.

You will also notice, in studying the collage above, that the text is almost entirely lacking accents and breathings. It has often been asserted that the accents and breathings which you can see in most pages of this manuscript were all added by the medieval retracers; but we can see here that the original scribe does appear to have used them occasionally: see the rough breathing on the relative pronoun ἁς in l. 23 (left column; the breathing is shaped like the left half of a majuscule eta); the circumflex on the genitive article τοῦ in l. 25 (left column); and the grave accent on the neuter article τὸ in l. 9 (right column). It is true, however, that the use of these diacritics remained "optional" until well into the age of minuscule. Another feature which may puzzle you is the use of the sign which we call "diaeresis" (two dots over a vowel) but whose usage here shows clearly that the scribe did not think of it as marking a "separation" between two vowels (a diaeresis appears on the following words: left column l. 2 ϋιων; l. 15 ϊδου; l. 25 ϊωας; ll. 44-5 ϊουδα; right column l. 5 ϋμων; l. 6 ϋποκατω). What exactly he thought it denoted is rather mysterious, and this will remain so for many if not most scribes down to the time of the latest Greek manuscripts in minuscule. This sign is occasionally used in manuscripts as it is in modern Greek typography (i.e. to indicate non-diphthongization of two contiguous vowels), but more usually it is placed (seemingly at random) only on ypsilon or on iota, preferably (but not only) when they begin a word. Diplomatic transcriptions often reproduce these diaereses as they appear in Greek manuscripts, but since they do not have any apparent significance they may also be left out.

This collage also gives you a very small sample of another category of difficulties which will become much more prevalent in later scripts, namely that of abbreviations. Here these are limited to a just a few, namely the horizontal line above a vowel (which is often the last letter in a line), which signifies nu, here lines 8-9 πα̅|τα = πα(ν)τα; l. 16 ηλθε̅ = ηλθε(ν), etc.; the abbreviation for και in the form of a kappa with a tail attached to the lower oblique stroke, here l. 19 εξεχομενων κ(αι); and the so-called nomina sacra, a series of words which are conventionally abbreviated by contraction and indicated by a sign similar to the nu abbreviation we have already seen, here c. 1 l. 6 and c. 2 l. 16 κ̅ς̅ = κ(υριο)ς; c. 1 l. 26 κ̅υ̅ = κ(υριο)υ; c. 2 l. 11 ι̅ν̅ = ι(ησου)ν; c. 2 l. 25 κ̅ν̅ = κ(υριο)ν.

Returning now to the specific case of Biblical majuscule, we may note that, since it was in use for a very long time, exemplars which use it are not really datable paleographically. However, Cavallo’s hypothesis was that the seven characteristics listed at the top of this page will be found primarily in manuscripts copied during the script’s mature phase (or "perfection," to use his vocabulary), which he dated to a period running from the second quarter of the third century to the third quarter of the fourth century (i.e. about 225-375 C.E.). Scripts which seem intended to follow this canon but which depart significantly from one or more of its characteristics may be assigned a date before or after this range, depending on whether the departures from the canon are interpreted as arising from its still-incomplete state or, on the contrary, as arising from its progressive decadence. Since we are disregarding papyrus manuscripts here, and since parchment manuscripts are extremely unlikely to antedate the maturity of the Biblical majuscule, Cavallo’s chronological indications may in practice be taken to mean that departures from the canon (in parchment manuscripts) indicate a late date (i.e. later than 375 C.E.). Examples of this are the already-mentioned Codex Alexandrinus, which is dated to the fifth century C.E.; as well as the two manuscripts below, Vat. gr. 1288 (Cassius Dio, also dated to the fifth century C.E.) and Vat. gr. 1666 (a manuscript of the Dialogues of Pope Gregory the Great [590-604 C.E.], translated into Greek by another Pope, Zachary [741-752 C.E.]). The latter manuscript bears a subscription on f. 185v which allows it to be dated precisely to the year 800 C.E.; it is in fact the earliest known Greek manuscript which is explicitly ​dated.

Note how, in these two manuscripts, all but the first three characteristics of Biblical majuscule listed above are disregarded to a greater or lesser degree. Angles tend to become curves (note especially the merging of the two median strokes of mu, and of the two small strokes of alpha, which become a loop); the lower stroke of ypsilon descends below the bottom line only slightly or not at all; the contrast between thick and thin strokes tends to be binary (i.e., the intermediate thickness is avoided); and many letters have very pronounced serifs. These tendencies are more marked in the second manuscript, whose script is also more irregular (in particular, letter heights are more variable). (Note that the first manuscript is entirely lacking accents and breathings, while the second has them on most words, written by the original scribe.)

(On to Alexandrian Majuscule)