2. INTRODUCTION TO MINUSCULE BOOKHANDS

The vast majority of medieval Greek manuscripts are written in a quadrilinear "minuscule" script which is very different from the "majuscule" script we have studied so far. Unlike the majuscule script, the minuscule one is not entirely intuitive for those who have learned to read printed Greek books, so it will be worth while to study the minuscule alphabet before going on to decipher manuscripts written in this script. Obviously the letter-shapes and the description below refer to an idealized abstraction of the script, which, when realized in an actual instance of writing, always incorporated specificities belonging to the individual scribe and/or to styles which came and went over the centuries (in other words, no scribe ever wrote exactly as described here). It is of course the history of such developments which we will be studying in the following pages; here we are seeking to establish the basic scheme that these developments are based upon. (For comments on individual letters, see here.)

The letters are shown here as they appear in isolation (note that the "optional" upturned tails of alpha, beta, eta, kappa, lambda and mu are often included when these letters are written in isolation). However, a defining feature of Greek minuscule, especially as compared to the majuscule scripts we have already seen, is that the letters are not normally written in isolation but rather strung together into a continuous stream, even across word breaks (scriptio continua).

How this "stringing together" works in detail is best learnt by experience in transcribing, but a brief description is offered here, referring to the table above with the dots in the letters. Normally, letters are joined by placing them so that a dot on the right side of a letter coincides with a dot on the left side of the following letter (please note that the dot on iota should be thought of as being on its left side). A few letters (namely zeta, iota, xi, omikron, rho, phi and omega) have no dots on their right sides: these letters are not normally joined to the right. Two letters (namely nu and upsilon) have red dots on their right side: these may only connect to a red dot in the following letter; if the following letter has no red dot, the two letters are simply not joined. There are also a few pairs whose dots cannot be aligned (the only one which is at all common is epsilon-sigma) and which are normally written separately (though the sequence epsilon-sigma is more commonly written with a ligature, as explained below). Scribes may also freely leave a small space between any pair of adjoining letters, especially (but not only) in situations where their strokes would "get in each other's way" (e.g. in the pair chi-lambda). You may see how all of this works in practice by studying the sentence below:

There is an obvious analogy between Greek majuscule and minuscule on the one hand, and modern Western "block capitals" and longhand "cursive" handwriting on the other. The analogy is accurate in terms of the opposition between bilinear and quadrilinear scripts; but It is important to realize that early Greek minuscule is not necessarily "cursive". Cursive scripts (including our own "cursive" longhand) are those which allow the writer to keep the tip of her pen on the page (almost) constantly, lifting it only very occasionally. For example, in longhand, you can, if you so desire, write arslongavitabrevis without lifting the pen at all — though you will need to dot the i's and cross the t. There are styles of Greek minuscule (especially in later periods) which are truly cursive in this sense; but minuscule bookhands in fact commonly trace each letter separately, indeed often in several distinct strokes, even though the letters are arranged so that they produce an impression of continuous flow (the same is true of many modern, Western calligraphic scripts).

In order to read a minuscule manuscript, in addition to the individual letters you will also need to be able to recognize ligatures. A ligature is defined here as a sign representing two letters, which cannot be neatly analyzed into its constituent parts, i.e. in which at least one of the strokes which normally compose the letters is modified, omitted, or belongs to both letters at once (N.B. some paleographers reserve the word "ligature" for this last case and use the word "nexus" as we are using "ligature" here). It would be quite unusual (though not unheard of), for instance, to find the sequence epsilon-tau written in a Byzantine manuscript as it appears above (in the words δε τεχνη); the normal way of writing ετ is the one you see in the table below, which also includes the other ligatures which are commonly used in early Greek minuscule:

Regarding the first ligature, besides gamma, also chi may join with a preceding alpha in a similar way. — The next two are hardly ligatures at all (by the definition given above), but are included here because they may be confusing; note that the upper, optional extention of alpha is used here to avoid giving the impression that a sigma is involved (as in the sigma-pi ligature). Besides alpha, also eta and kappa may be joined to pi or tau in a similar way. — The three ligatures involving a final nu (εν, ην, υν) can be rather confusing: the principle appears to be that the descender of the nu curls up to meet the preceding letter (which, in the case of ην, is the full eta, including the "optional" extension on the right); the ligature υν is in principle identical to the (less common) letter sequence βυ, which is why scribes will usually take care to mark βυ, when it occurs, in some way (by including the optional, upper extension of beta, by extending the stroke which joins the two letters, or simply by leaving a space between them). — Beginners often mistake the ευ ligature for ει. — The double sigma ligature is easily mistaken for pi (note that the ligature σσ, like a single sigma, is not joined to the left on the headline [nor fused with other ligatures, as described below], precisely to avoid ambiguity with pi). — The double tau ligature is in principle identical to the sequence tau-gamma, which however never occurs in Greek. — Beginners often mistake the υσ ligature for iota-sigma; remember that iota is not connected to the right. — Finally, note that ligatures may occasionally be fused together, e.g. the sigma in εσ may itself be in ligature with pi to produce εσπ (the last figure in the table above); similarly επτ, εστ, etc.

The origin of the Greek minuscule script has been the subject of a great deal of speculation. It seems to appear out of nowhere, fully developed, in the first decades of the ninth century, when it begins to supplant the old majuscule canons as the normal Greek bookhand. In reality there is little doubt that this bookhand developed out of informal, cursive scripts which had been in use (though not as bookhands) for a long time before the ninth century; it seems probable, moreover, that the Studite monastery in Constantinople (about which we shall have more to say in the next page) played a central role in the development of the particular script which was to become the standard Greek minuscule bookhand (see C. Mango, "L'origine de la minuscule," in the proceedings of the 1974 paleography conference in Paris, pp. 175-180). However, the triumph of this particular script was not a foregone conclusion. There appear to have been several contemporary unsuccessful candidates, as it were, for the position of standard Greek minuscule script, and a few remains of some of these scripts are still extant (the ones seen below are both dated to about 800 CE, though this is really no more than an educated guess). These are rather difficult scripts, and you are unlikely to encounter them again; but they are of interest to understand the early development of the Greek minuscule bookhand. If you enjoy a challenge, you may want to try your hand at transcribing them, or come back to them once you have gained greater expertise. One of these is the so-called "Damascene" or "Hagiopolitan" script, of which a fairly extensive sample may be found in the oldest Greek manuscript on paper, Vat. gr. 2200, shown below (for further details about this manuscript, and pages with transcriptions, see here).

Another "runner-up"is the "Sinaitic minuscule" which you may admire below.

Transcription of ff. 218r-217v: (ὑποστά)σει μονογενῆ ἒν δυσιν οὐ|σίαις γνωριζόμενον υ(ιο)ν οτ(ι) | ὩΔΗ Γ ΠΡ(ωι) ΤΩ ΠΡΟ ΤΩ(Ν) ΑΙΩΝ(ΩΝ) | Τῶι ἐπὶ τῶν ὤμων τὸ πλα|νώμενον πρόβατον ἄραντι | και καθελόντι ἒν τῶι ξύλωι τὴν | αὐτοῦ ἁμαρτίαν χ(ριστ)ωι τῶι θ(ε)ωι | βοήσωμεν· ὁ ἀνυψώσας | τὸ κέρας ἡμῶν ἁγιος εἶ κ(υρι)ε | τῶι ἀναγαγόντι τὸν ποιμένα | τὸν μέγαν ἐξ αίδου χ(ριστο)ν· καὶ τῆι | αὐτοῦ ἱεραρχίαι δια τῶν ἀ|ποστόλων τὰ ἔθνη ποιμά|ναντι ἐν αληθείαι καὶ πόθωι | πιστοὶ πν(ευματ)ι λατρευσωμεν | τῶι ἐκ τῆς παρθενου σαρκω|θέντι βουλήσει ἀσπόρως | υ(ι)ῶι και τὴν τεκοῦσαν μετὰ τό||κον θεϊκῆι δυναστείαι παρ|θένον φυλάξαντι ὧ επι πάν|των βοῶμεν θ(ε)ωι ἅγιος εἶ κ(υρι)ε· | ΩΔΗ Δ· ΠΡ(ωι) ΡΑΒΔΟΣ ΕΚ ΤΗΣ ΡΙΖ(ΗΣ) | Τίς οὗτος ὡραῖος εξ εδὼμ | καὶ τούτου τὸ ἐρύθημα τῶν ϊ|ματίων ἐξ αμπέλου βοσωρ | ὡραῖος ὅτι θ(εο)ς ὡς βροτὸς | δε αἵματι σαρκὸς τὴν | στολὴν πεφοινιγμένος | ὧι μελωδοῦμεν λαοὶ δοξα | χ(ριστο)ς τῶν μελλόντων ἀγαθῶν φα|νεὶς αρχιερεὺς ἡμῶν τὴν | ἁμαρτίαν διεσκέδασεν | καινίσας ξένην ὁδὸν τωι | ἰδίωι αἵματι εἰς κρείττο|να και τελειοτέραν εἰσέδραμεν | δε σκηνὴν πρόδρομος ἡμῶν | εις τὰ ἅγια.

Even if you put off working through pages from these manuscripts until later, it will be useful to take a close look at them now. First of all, be aware that these scripts include a number of ligatures which are not featured in what became the standard Greek minuscule (e.g. ligatures with pi on the left, such as pi-omikron in l. 13 of the left-hand page of the Sinai manuscript, which looks like the sigma-pi ligature in standard Greek minuscule). On the other hand, we also find here some ligatures which are not common in early standard minuscule but which will reappear in much later styles, three or four centuries later; perhaps the most remarkable is the omikron-iota ligature found in the "Hagiopolitan" manuscript, e.g. p. 262, ll. 19 and 20. The Sinai manuscript also displays a very unusual notation of iota adscript (e.g. f. 218r l. 4 Τωι; l. 6 τῶι ξύλωι; etc.), namely a sort of half-way solution between adscript and subscript, where the letter is reduced in size (sometimes looking rather like a full stop), but is still written beside the preceding vowel and not under it (cf. N. G. Wilson, "Miscellanea Palaeographica", GRBS 22 [1981], pp. 395-404, who cites a similar finding in another Sinai manuscript, gr. 735).

Also of interest are the accents and breathings. In the "Hagiopolitan" manuscript, these are written only sporadically; in the Sinai manuscript, they appear on most words. This reflects the situation which generally obtains in early Greek minuscule, where accents and breathings are optional up until about the middle of the 10th century (after which they are always written). Note that the breathings in the Sinai manuscript are mostly of the archaic variety, i.e. shaped like the left half of a majuscule eta (rough breathing) or like the right half of an eta (smooth breathing), but that the modern, more cursive variety (similar to our inverted commas) also appears (compare the breathings in ὡραῖος ὁτι at the beginning of line 8 on the right-hand page). Both types occur in early minuscule manuscripts, but later ones (very approximately from the eleventh c. onwards) usually prefer the "cursive" breathings. Accents are not always placed exactly where we would expect them; in particular, they usually appear over the first letter of a diphthong, not the second, as is the modern convention; and the circumflex normally follows a breathing rather than being placed above it (see ὁῦτος in l. 5 of the right-hand page of the Sinai manuscript). Note also that the usage of the accents does not always follow modern conventions (e.g. the preposition ἒν with a grave accent in the first line of the Sinai manuscript).

It is worth taking a moment to consider the purpose of the system of accents and breathings. As guides to pronunciation, the breathings were almost entirely useless in the Byzantine period, since initial aspiration of words had long since been lost (they were of course still useful for distinguishing between homographs like ἥν and ἤν). The accents did indicate the syllable of a word which carried the tonic accent, but this was presumably not very useful to native speakers of Greek (except for distinguishing pairs like βίος and βιός); and the distinction between the three types of accent was no longer reflected in pronunciation. Consider, however, the same phrase which we looked at earlier, now supplied with diacritics:

If you know the rules of Greek accentuation, you will find this much easier to read than the unaccented version we saw earlier, because the diacritics only rarely leave you in doubt as to word division: the grave must be on a final syllable; the acute must be on a penultimate (or, under certain conditions, antepenultimate) syllable; and of course the breathings always indicate the beginning of a word. This is not a perfect guide (for instance, the accents alone would not prevent you, on your first attempt, from dividing τέχνημα κρή… instead of τέχνη μακρή), but in terms of legibility it is a great improvement upon unaccented scriptio continua, and this is presumably the main reason why you will find diacritics systematically provided in all but the earliest Greek minuscule manuscripts.

Finally, we may note the development of abbreviations. The few lines of majuscule in the Sinai manuscript are heavily abbreviated, mostly by suspension, meaning that the reader is simply invited to supply the missing letters, with no real guidance as to what they might be. Fortunately, this type of abbreviation, which is quite common in Latin manuscripts, is rather unusual in Greek ones and especially in Greek minuscule. In the minuscule parts of the Sinai manuscript, abbreviations are used sparingly, apart from the nomina sacra, which we have already met in majuscule scripts, and an abbreviation for και which is shaped like a Latin "S" and which will remain common in standard Greek minuscule (there is also an abbreviation by suspension of ὅτι, written as an omicron with a tau added above). The "Hagiopolitan" manuscript makes a rather more liberal use of abbreviations, including nomina sacra and the S-shaped και, as well as the horizontal stroke above the headline representing final nu (p. 261 l. 10 τουτων), which we have already met in majuscule but which is rather unusual in most minuscule scripts. Besides και, we also find here some further word-abbreviations which will be common in standard minuscule (a list of these may be found in the lower part of this page), namely γαρ (a barred gamma), p. 262 l. 11; and a curved oblique stroke with dots on either side signifying ειναι (p. 261 l. 7 from the bottom; also ἐξειναι written with the same abbreviation, p. 261 l. 13). There are also some examples of a new category of abbreviations, those which represent certain letters or letter combinations, primarily (but not only) at word-end, and which are added to the preceding letter rather like diacritics. We will see much more of these in standard Greek minuscule, and a list of the most common such abbreviations may be found in the upper part of the page already referred to. The ones which appear here are:

The remainder of this pathway will be devoted to various categories and instances of the standard Greek minuscule script whose basic scheme is described at the top of this page. The problem of categorizing Greek minuscule bookhands is much more difficult than in the case of majuscule bookhands, where the extant manuscripts fall with relative ease into a small number of groups or "canons" with well-defined characteristics; and indeed no one has followed Cavallo's lead and proposed a system of "canons" to describe the bewildering variety of minuscule scripts, although this ought to be possible, in principle, if the theory of script development which he invoked were correct. (The Perlschrift is a possible and partial exception to this lack of identifiable "canons" in minuscule: see the remarks in section C of the page on Thirteenth-Century Hands.) Doubtless the difference is due in part to sample size: the hundreds of extant majuscule manuscripts are naturally (indeed necessarily) less diverse, and thus give a greater (though perhaps false) impression of relative simplicity, than the tens of thousands of extant minuscule ones. In addition, minuscule scripts naturally have more morphological variables: besides the criteria which we have used to distinguish between majuscule scripts (axis angle; module or size of the body of the letters; "aspect ratio", length of ascenders and descenders, angularity, serifs and flourishes, stroke contrast; specific letter shapes), minuscule scripts may also be distinguished by the "connectedness" of the letters (in terms of their relative positions), their cursiveness (in terms of ductus), and the frequency of ligatures and/or abbreviations. Letter-shapes also display greater diversity in minuscule, partly because the ligatures add another 20-odd signs, which can be written in peculiar ways, to the 24 letters of the alphabet; and partly because scribes had the option (and made increasing use of over time, as we shall see) of substituting individual majuscule letter-shapes for minuscule ones, so that many letters could be represented by two (or more) completely different shapes. In theory, more variables, i.e. more criteria, might allow a more precise differentiation between scripts. In practice, however, since (as we already observed in considering our initial collage of majuscule hands) no two criteria produce the same groupings, we are in fact left with an enormous, inchoate mass of observations which in most cases defies classification.

There is one factor which somewhat mitigates this difficulty, namely the fact that many more minuscule manuscripts than majuscule ones are subscribed and/or dated, or display other signs of their origins. By connecting such "external" information with "internal" observations about the scripts themselves, it is possible to place most minuscule manuscripts, if not into a neat classificatory framework, at least somewhere in a finite set of "nebulae" (the term was used by L. Perria in her contribution to the proceedings of the 1998 Paleography Conference in Cremona, vol. 1 p. 159, with specific reference to 9th- and 10th-century scripts) which can be thought of as standing in chronological, geographical or institutional relationships to each other. The upshot is that when you are studying an unidentified Greek minuscule manuscript and wish to describe its script or assign a tentative date to it, you will often not be able to go through a checklist allowing you to determine definitively that it belongs to this or that group or century. You will need to carefully compare it with samples from each of the "nebulae" and decide where it fits best, and base your description and/or dating on that.

You may think of each subsequent page in this pathway as presenting one (or in some cases several) of these "nebulae". They are presented in rough chronological order; but of course many of them are mutually contemporaneous, and you should not suppose that they arose one out of the other in any genealogical sense. In fact, there is no consensus among paleographers about the precise contours of these "nebulae" (with a few clear-cut exceptions), but we will generally follow the groupings found in L. Perria's book Graphis: Per una storia della scrittura greca libraria.

Moreover, there is one broad chronological development which is not precisely reflected in the succession of "nebulae" presented here and which we need to consider globally at the outset, namely the gradual reintroduction of majuscule letters. You will have noticed that certain "lower-case" letters in modern Greek typography, which derives its letter-shapes ultimately from earlier bookhands, are nonetheless completely different from the minuscule letter-shapes presented at the top of this page. In the case of β, ε, κ, λ, π and ω, it is not a coincidence that the "modern" lower-case Greek letters more closely resemble the shapes of majuscule Β, Ε, Κ, Λ and Π and omega than they do the minuscule letters at the top of this page. Only the very earliest minuscule manuscripts are in "pure" minuscule; from the end of the ninth century onwards, scribes began to insert individual majuscule letters into their handwriting (not as "upper-case" letters with a differentiating function, but rather as optional variants, with no regard to their position within a word). The reintroduction of majuscule letters occurred gradually, with majuscule lambda, alpha and lunate sigma appearing in minuscule scripts of the late ninth century; majuscule gamma, eta, kappa and pi becoming common in minuscule scripts of the early tenth century, and most other majuscule letters (except ypsilon, likely because of the risk of confusion with minuscule gamma; of course the shapes of letters like iota, omicron and rho are substantially identical in the two alphabets) joining them, at least occasionally, by the mid-tenth century onwards. By the time typographers created their first Greek letters in the late fifteenth century, imitating the handwriting of contemporary Greek scribes, the shapes derived from the majuscule versions of these six letters had become far more common than their actual minuscule equivalents, which is why we write these letters as we do. Since they were introduced gradually, the proportion (or percentage) of majuscule letters in a ninth- or tenth-century script may sometimes be used as a criterion for dating it, with low proportions militating for a date earlier in that period, and high proportions for a later one. Mostly, however, the use of optional majuscule letter-shapes from the mid-tenth century onwards simply adds another variant (actually about fifteen new variants, since scribes could now choose between two different shapes for most letters of the alphabet) to the many which we have already noted and which scribes could use to differentiate their scripts from those of their colleagues and which we may use to categorize them.

(On to Old Round Minuscule)