2 sebastian kempgen

8/12/2019 2 Sebastian Kempgen

1/16

, . 10.

24

WHY OT (Q)? WHY nOT?

A NOTE ON THE DEVELOPMENT OF A CYRILLIC LETTER

Sebasti an Kempgen (Bamberg, Germany)

1. IntroductionSometimes it can be fruitful to ask yourself very basic questions, and once such aquestion has been put forth, it is quite astonishing to find that nobody seems tohave touched upon it before. In the present paper, I will ask one such simplequestionand I will try to give an equally simple answer.

In grammars or manuals on Old Church Slavonic (OCS) one willsometimes find assertions such as the preposition ot is always being written asq. In addition, in the world-wide character encoding standard, Unicode, thisletter (in its uppercase and lowercase forms, Q q) is defined as a separate

character in the Cyrillic section but not in the Glagolitic section, which wasadded to the Unicode standard later than the Cyrillic part. Because all charactersthat are being added to Unicode have to go through a long process of applicationand approval, this status quo supposedly reflects current philological knowledge.1Fig. 1 shows the Cyrillic Old Church Slavonic portion of Unicode.

F ig. 1: Old Chur ch Slavonic portion of the Cyril li c Uni code character table

So the simple question to ask is why? Why otand not another preposition? Isthere a convincing explanation that could be given to a student asking such a question?Of course, one cannot be sure from the outset that this question canbe fully answered,

but one can neverless try to reconstruct some of the reasons that could have lead to thedevelopment of this letter, and maybe offer new insights on how it developed.

2. Classical GrammarsLet us first take a look at some classical grammars, manuals, text books and primersof Old Church Slavonic (OCS), how they introduce the Cyrillic alphabet and whatthey are saying with regard to these letters, the Cyrillic omega, , and the ot, q.

1This does not mean, however, that there cannot be any mistakes in Unicode and the decisions

leading to the inclusion of individual characters. The Cyrillic omega is the best example of this: in a

rare exception to its policies, Unicode. Inc. changed the shape of characters number 047C and

047D in a recent revision from omega with titloto the correct form,omega with velikij apostrof.


2/16

S. Kempgen.Why ot (Q)? Why nOT?...

25

2.1. In his classic Manual of Old Bulgarian,Leskien2uses a table to introducethe OCS Cyrillic and Glagolitic alphabets along with Greek correspondences.Leskien does not mention q at all. However, he wisely choose to present the two

shapes the Greek Omega had at the time when the Slavic alphabets weredeveloped (see Fig. 2).

Fig. 2: Omega and ot in Leskiens alphabet chart3

In the glossary of his book, Leskien writes the preposition as well as the prefix asozn, while the texts themselves give ozn and zn, when they are transliteratedfrom Glagolitic, like Zogr., but the Cyrillic sources are typeset using only ozn(Supr., Sava). Leskien has a sample of q in a heading fromAssem., but the same page

actually a photomechanical reproduction of a page from the edition of Kurz4

hasquite a few instances of the preposition otbeing written as a:

Fig. 3: Superscript OT fr om Ass. (Kurz 1955, 2)

2.2. Kulbakin has a chart of the Glagolitic letters with their Cyrillic counterparts5which means q is missing. He tries to demonstrate how w might have been composedfrom s and z both in Glagolica and Kirillica (), and he offers an interestingobservation regarding : covering Cyrillic OCS texts, he comments with regard to Und.(end of 11th century) that appears more frequently than in earlier texts. Indeed it islogical that the frequency of q cannot be higher than the frequency of itself, whiletheoretically every occurrence of could at the same time be an occurrence of q.

2 Leskien A. Handbuch der altbulgarischen Sprache. Grammatik Texte Glossar. NeunteAuflage. Studienbuch. Heidelberg 1969.3Leskien A. Handbuch der altbulgarischen Sprache ..., S. 4.4Kurz J.(ed.): Evangeli Assemanv Kodex Vatiknsk 3. slovansk. Dl II.Vydal Josef Kurz. Praha,

1955,. 2.5Kul' bakin S. M. Drevnecerkovnoslovjanskij jazyk. Izd. 2-e, izm. i ispravl. Xar'kov, 1913, p. 16.


3/16

, . 10.

26

2.3. Trubetzkoy6 includes a long chapter on the writing systems, their

development, the derivation of Cyrillic letters from Glagolitic ones etc.; heintroduces both the Glagolitic and the Cyrillic alphabet on their own, but in his

work, too, r is completely missing! Interestingly, however, he says Im migenGebrauche von Ligaturen stimmen die kyrill. Denkmler mit den glagol. berein.[...] ist auch in den kyrill. u [] die einzige oft gebrauchte Ligatur7.

2.4. In his Old Church Slavonic Grammar, Luntalso has a chapter on the writingsystems8; he introduces the Glagolitic alphabet in the form of a table, and lists theCyrillic correspondence for each Glagolitic letter. The Cyrillic OCS alphabet is notintroduced as such, and the consequence of his approach is that r is never mentioned.Lunt mentions the (Glagolitic and Cyrillic) letters i/w and the hypothesis that theycould be interpreted as a ligature of s and z9, but nothing more10.

2.5. The Old Slavonic Grammar byBielfeldt11introduces the sounds and lettersof the language, but is completely missing from the corresponding table12andthe following pages which, of course, also means that q is missing anastonishing fact

13. In the grammar section and in the texts, the preposition is

simply spelt as ozn.

2.6. So, reviewing these classical grammars shows that none of them evenmentions glagolitic L/ Mor Cyrillic A/ Qeither as a special writing conventionnor as a letter-like character. In other words: these manuals, which are still in use

6Trubetzkoy [Tr ubeckoj] N. Altkirchenslavische Grammatik. Schrift-, Laut- und Formensystem.

Herausgegeben von Rudolf Jagoditsch. Zweite Auflage. Wien, 1968 (11954).7Trubetzkoy [Tr ubeckoj] N.Altkirchenslavische Grammati. .., S. 42. When he notes the origin of u

from two separate letters, one could rightly ask whether it is correct to write these two separate

letters using OCS Cyrillic. It could be argued that this ligature has not been composed from these

two Cyrillicletters but that it has been borrowed as a complete entity from Greek writing, and that,

in Greek, it developed as a vertical ligature of und .8 Lunt H . G. Old Church Slavonic Grammar (Slavic Printings and Reprintings, III). Second,

revised edition. The Hague: Mouton, 1959 (11955),.1423.9Lunt H . G.Old Church Slavonic Grammar...,.21.10 By the way, typographically it does not make real sense to use the modern form in this

context, as Lunt doesin this context one should use only the old form (w).11Bielfeldt H . H.Altslavische Grammatik(Slawistische Bibliothek, Nr. 7). Halle (Saale) 1961.12Bielfeldt H . H.Altslavische Grammatik..., S. 24.13 A strange solution is found in this book to typeset the w: a Greepsi() is used instead of the Cyrillicletter, and a Greek is used for the izhica (y)! Actually, once you think of it, there is an optical relationship

between the w and thegraphically, the Cyrillic letter could be derived from thepsisimply by making itsbottom straight, not curved. So far, the Cyrillic letter has rather been perceived to be a s with an added stroke

going down, which has lead to the hypothesis that is a ligature of a s with a z; with the psiin mind, it could

indeed also be described as consisting of a long central stem with two arms. However, it is unlikely that this

is the origin of the Cyrillic w, because the alphabet also has its own true version of the psi, i.e. .


4/16


27

today and by no means are only of historical importance, do not answer ourquestion at all.

2.7. Some information can be gleaned from Schaeken/Birnbaum14, who mention theGlagolitic and Cyrillic Omega and illustrate the letters with a sample from Assem.They state: In den altkirchenslawischen Handschriften kommt die graphischeVariante k/ nur in beschrnktem Mae vor, und zwar hauptschlich bei derInterjektion ! (gewhnlich im Unterschied zur gleichlautenden Prposition o), ingriechischen Wrtern und Namen []), bei der Prposition oder bei dem Prfixoz(n) meist in Form der Ligatur q bezeugt und auerdem als Zierbuchstabe(Initiale) am Satzanfang. In addition, they comment that apart fom u and q, CyrillicOCS rarely uses ligatures

15.

These observations are in stark contrast to the early grammatical tradition. Aseven a cursory look at the material presented by Jagi16proves, in the works from thecenturies following the period of Old Church Slavonic, Qhad often simply replacedin alphabet charts, i.e. it was a common character in these works on OCS.

3. Glagolitic Texts3.1. Let us now turn to the text themselves.Diels17states that Glagolitic K and Cyrillicsometimes reproduce a Greek /, but otherwise serve only aesthetic purposes. Hedescribes very precisely the distribution and occurrence of the two o-letters inGlagolitic and Cyrillic manuscripts18. According to him, we find K in Glagolitic textsmainly at the beginning of sentences and/or in headings, and also in Greek o! andosann! He then goes on to say that Ril. sometimes has K in word-initial position,similarly Grig. and Ass., while in Kij. and Ochr. the letter is missing completely

19.

From his observations he infers that K does not correspond specifically to GreekOmega, although this characters shape could have been its origin, but rather fulfills anaesthetic purpose. For Cyrillic texts, Diels notes that the findings are not identical in

every respect. According to him, Savva regularly uses in the preposition oz inheadings (see 4.2, below). He notes the generally wider use of in Savaand Supr., theabsence of this letter in Chil.,Hilf.andLavr., and concludes his paragraph by notingthat the use of this letter is restricted to ozn in Und.and in Slu.

14 Schaeken J., H. Birnbaum. Die altkirchenslavische Schriftkultur: Geschichte, Laute und

Schriftzeichen, Sprachdenkmler (mit Textproben, Glossar und Flexionsmustern).Altkirchenslawische Studien II (Slavistische Beitrge, Bd. 382). Mnchen, 1999, S. 80.15Schaeken J., H. Bir nbaum. Die altkirchenslavische Schriftkultur: .., S. 82.16 JagiV. Codex slovenicus rerum grammaticarum (Rassudenija junoslavjanskoj i russkojstariny o cerkovno-slavjanskom jazyke). Berlin, 1896 (Reprint Mnchen 1968).17Diels P. Altkirchenslavische Grammatik. Mit einer Auswahl von Texten und einem Wrterbuch.Bd. 1-2 in einem Band. Heidelberg 1932, . 24.18Di els P. Altkirchenslavische Grammatik..., 40f.19Di els P. Altkirchenslavische Grammatik...,. 41.


5/16

, . 10.

28

3.2. Let us illustrate Dielss observations with a few samples of actualmanuscripts; first, an example from Zogr.of the preposition otbeing written aslinear JTin a heading (see Fig. 4)20:

Fi g. 4: Linear OT in Zographensis (Jagi 1911, tabl. III, 5c)21

Another header from Zogr., however, uses the Omega-O for the preposition, see Fig. 5.

F ig. 5: Omega-OT in a header i n Zogr. (Ev. Lukas.)

Trubetzkoy22

asserts that in the oldest texts, the Omega-K was only used for theinjection o!, while the preposition owas written using the normal graph for o,i.e. Glagolitic J23. The restricted use mentioned by Trubetzkoy is, however, nottrue any longer for later texts: there are cases where the preposition o is written

20 The Cyrillic lines above the graphic are interesting from a paleographical perspective, as they

contain a stroked (!) Cyrillic I (see the first letter of iskoniin the beginning), a character so far notencoded in Unicode as a separate Cyrillic letter.21JagiV.Glagolieskoe pismo. In: Jagi, I.V. (ed.), nciklopedija slavjanskoj filologii, Vyp. 3,Grafika u slavjan, SPb. 1911,. 51230.22Trubetzkoy [Trubeckoj] N .Altkirchenslavische Grammatik. Schrift-, Laut- und Formensystem.

Herausgegeben von Rudolf Jagoditsch. Zweite Auflage. Wien, 1968 (11954), S. 27.23 The question, by the way, why OCS needed the Omega at all, if it had such a restricted use, can

and should be answered not only with regard to aesthetics. As Vaillant (Vaillant A. LAlphabetVieux-Slave. Revue des tudes Slaves 32, 1955, p. 21) correctly points out, the character isessential theologically to encompass everything from Alpha to Omega.


6/16


29

with Omega-K, see the preceding figures as well as the beginning of the secondline in the following example (Fig. 6):

Fig. 6: Superscript-OT i n Zographensis (Jagi 1911, table III, 5c)24

Fig. 6 also shows, thatZogr.uses the linear variant as well as the superscript-OT inheadings. The same superscript-OT is also to be found in Ochr

25; next, an example

fromMar.showing the linear form twice (see Fig. 7):

Fig. 7: Marianus: Linear OT (Jagi 1911, tabl. III, 6a)26

Now lets have a look atAssem., the text mentioned by Diels among those usingthe Omega-O. In Fig. 8, there is an example of a header, and in Fig. 9 of asentence-initial use of a superscript-OT using the Omega-O.

F ig. 8: Assem. Omega-O superscr ipt (f ol. 41)27

24JagiV.Glagolieskoe pismo ...,. 51230.25See JagiV.Glagolieskoe pismo ..., table VII, 13.26JagiV.Glagolieskoe pismo ...,. 51-230.27Asemanievo evangelie. Faksimilno izdanie. Sofija, 1981.


7/16

, . 10.

30

Fig. 9: Assem. Omega-O with superscri pt, sentence-in it ial posit ion

Assem.has a certain tendency to having Q in headlines, but often we can find A instead:

Fig. 10: O wi th superscript T (Assem. fol. 24)

Thus, in headings we find both Land M, and we also find Mat the beginning ofsentences; however, in running text we mainly find Lbut sometimes M, toothere is no complete consistency.

3.3. Let us sum up the review of Glagolitic texts: Here, we find Las well as Mfor thepreposition ot. If the texts make any distinction at all, Kis used in headlines and atthe beginning of sentences, while Jappears in running text. This also explains why Lshould be more frequent than M, but Mhas the more prominent place because of itsoccurence in headlines. Neither combination, however, is considered to be a fixedletter-like combination: they occur alongside other superscripted characters,especially in headlines. Figs. 8 and 10 are very instructive in this respect: they showthat the superscripted OT occurs in headlines where every word has a superscript; sothese headings are no proof with respect to the preposition OT. The use of thesuperscript at the beginning of sentences, however, is distinctive.

In other words: in Glagolica the superscript-OT is not a fixed entity but still anO with a superscriptT. However, there had already developed a certain tendency towrite the preposition OT as Min headlines, and the distribution of these characters areclearly better described for each text and not for the alphabet system as such. It islogical, then, that as we have seen from the short overview of some classicalgrammars nobody claims that the Glagolitic alphabet had a separate letter M-superscript, although the combination itself, just as L, does frequently occur. In thisrespect, the Glagolitic differs from the Cyrillic alphabet. The Glagolitic T, by the way,

lends itself very easily to be used as used as a superscript, because of the very lowalternate character shape it has (t).On a side note, we might draw attention to the different hypotheses on the

origin of the Glagolitic K: Trubetzkoys view is that it can be derived from GreekOmega, i.e. , (obviously, by connecting the upper halves to form the large circle


8/16


9/16

, . 10.

32

make the characters better visible) shows the single instance of H and a sample ofGon the same page, in the same phrase.

Fig. 13: Two OT-l igatures in Savvina kn iga (fol . 134b)

epkin29, by the way, does not seem to be aware of the one instance of b in theheadings, but he mentions one occurrence of qin the running text where the authorsof Savvina kniga 1999claim it wasnt used at all. Fig. 14 shows this occurrence:

F ig. 14:rin runni ng evang. text in Savvina kni ga (fol . 87)

Diels states (see section 3.1) that Suprasl. uses more often (for example, in thepreposition o). This may be true in general, but does not include a specific use of thepreposition ot superscripted or not which other texts exhibit. The reason issimple: Suprasl. consists mainly of itija and slova whose headlines have alinguistic structure which differs from those characteristic of the gospels: they dontuse the preposition ot at all! They use either none or v, na, or o. Therefore, inSuprasl., the does not stand out optically at all, and it surely does not play animportant aesthetical role. In other words: the use of q in headings depended heavilyon the liturgical character of the text itself, at least in the beginning, and on the layoutthat was chosen for the text in question: if anything, in Suprasl. the headers are

smallerin type than the rest of the text, see Fig. 15, lines 13 below the divider.

F ig. 15: H eader with oznin Suprasl30. (chapter 32)

29epkin V.N. Rassudenie o jazyk Savvinoj knigi . S priloeniem dvux fototipieskix snimkov.

Sb. Otd. Russkago Jazyka i Slovesnosti Imp. Ak. Nauk, tom LXVII, No. 9 SPb. 1901, p. 44.


10/16


33

In Fig. 15, we have a rare example of a header where ozn is used twice, and thus hada chance to use the headline form qbut uses the linear form instead. The broader

use of itself can, at the same time, be seen in the name W(first line). Incontrast to this, all Gospels with their prominent and fixed header structure ofEvang'elie ot[nameGen.] offered an ideal position for the systematical use of thedigraph Qand making it popularfrom there it could later spread to other uses.

4.3. In other words, already in some of the oldest Cyrillic texts, like Savva, we find astrong tendency to use the digraph qin headers, and the normal linear form ozn inrunning text. We have to speak of a tendency, because various deviations from thisgeneral rule occur. In headers, Cyrillic Qis often used where Glagolitic texts still hada mixture of Mand L. Not all texts, however, are uniform in this respect, and sometexts completely do without , while others, like Suprasl., simply have no prominentand no suitable headlines where the digraph could be put to good use.

5. Why OT? Why nOT?This leads us back to our initial question, why it is the preposition OT thatdeveloped its characteristic superscript shape, and the picture is beginnng to

become somewhat clearer. First, as we have seen, a certain tradition of writingexactly the preposition OT using the Omega-O is already present in Glagolitictexts, in some texts more prominent than in others. Second, the preposition OToccurs indeed much more frequent than other prepositions in the most prominent

place, the headings of the Gospels. This should answer why it is the prepositionOT and not another preposition, such as do to. Third, as we have seen, inGlagolitic texts the superscripted occurrences of OT cannot be interpreted as aligature or as a special character on its own. This clearly is a development of thenewly introduced Cyrillic alphabet. Fourth, there is one reason that certainly has

affected the history of the superscripted OT becoming a fixed entity with acharacter-like status, and it is astonishing that grammars do not mention this atall: O and T are not only the elementsof this complex character, OT is also thenameof the character r itselfbut also the name of . It is very probable that thename ot which was at the same the most characteristic occurrence of the letteritself, i.e. as the superscripted preposition ot, influenced the basic shape of theletter so much that in the later tradition, it simply replaced it instead of wefind r in alphabet charts. The case of r is the only such case in the Cyrillicalphabet where the name of the character names all parts that make up thecharacter itself. The Glagolitic alphabet did not have the character Mas such, and sothe name ot for the Glagolitic character was chosen foranother reason (if it is theoriginal name at all). Moreover, one cannot be completely sure what came first: thename or the shape. We have reason to assume though that the shape came first: the

word ot is no good example for the hard words that are the names of theneighbouring characters in the alphabet, and so doesnt serve the purpose of the

30Zaimov J., M. Kapaldo.Supraslski ili Retkov sbornik. T. 1-2. Sofia, 198283.


11/16


12/16


35

6.2. Let us take now take a look at some Greek sources. Fig. 17 shows a headline fromthe so-called Theodore Psalter35, which was written in 1066 in Constantinople. Herewe see a very prominent ligature in a heading. There are other identical samples.36

F ig. 17: Greek li gature (Theodore Psalter, 11th centur y, fol 27v)

The next illustration (Fig. 18), although from the 13th or 14th century (seeTrobisch37), is also remarkable. The writing is Greek minuscle, but the ligature a

and the digraph are very carefully written und are easy to identify as such. Thesample also shows the general tendency of the T to be written with a high stem,even if it stands alonewe find the same in Cyrillic manuscripts.

F ig. 18: Greek l igatureaand (Dr esden Gospel, 14th centur y)

Fig. 18 serves as an ideal illustration for a central argument of the present paper.And once one has seen the point, it seems so obvious that it is astonishingnobody has come up with the observation before. The clue is: one simply has tochange the reading direction of the Greek words to get the Slavic word, from top-down to bottom up:

Greek reading direction: Slavic reading direction:

In fact, the top-down direction of TA, TO, TOY, TC seems be an exception in Greekbecause Gartdthausen38 stresses that Greek abbreviations always have to be readbottom-up and that only these ligatures differ in that respect. In other words, when

35Theodore Psalter: http://www.imagesonline.bl.uk/britishlibrary/, see record number 18999.36 The illustration is available online at: http://www.imagesonline.bl.uk/britishlibrary/, record

number 18999; an identical ligature occurs in the same Psalter on f.28, see record number 11311.37Tr obisch D. J. A Dresden Septuagint Manuscript of the Psalms. 2003. Online available at:

http://www.bts.edu/trobisch/CodexG/a_dresden_septuagint_manuscript_.htm38Gardthausen V. Griechische Palographie. Leipzig, 1879, 114f.


13/16

, . 10.

36

switching from Glagolica to Kirillica, Slavic scribes continued using superscripts, butat the same time they narrowed their use down to a well-known character shape thatwas present in Greek and thus was a familiar feature for any Greek looking at an OCS

manuscriptit certainly made them look more Greek-like! And because E and D(H)were the most prominent vertical ligatures in Greek, and both were carried over to thenewly developed Cyrillic alphabet, this means that the basic set of Greek uncial verticalforms in use in the 10th century was also used in Cyrillic. The selection between GandHhad already been predetermined by Glagolitic writing traditions (Mand Kmainly inheadings, in senence initial position and/or in the preposition OT, with L rarelyoccurring in running texts) and was further supported by the fact that Cyrillic wasdeveloped under the influence of Greek uncial forms, not minuscle forms.

6.3. The idea proposed here can be supported by more observations. Thesuperscripted Greek forms do not occur in manuscripts only, but also in frescoes andon icons. The inscription in Fig. 19 is actually a very good illustration because it is

part of a famous 11th century mosaic from the Hagia Sophia, i.e. from the heart ofByzantium, depicting the emperor Constantine and his wife Zoe. Here, the is asuperscript, and it is an article. Fig. 20, also from the Hagia Sophia, is interesting,

because it shows the use ot a as well as in ligature form in the word avtokrator onthe leftbut this time, in the first line we have the linear form Z. In other words:the digraphs or ligatures a and were used in Greek not only when writing inflectionalforms of the article the, but also when these character sequences occurred as such inother words. Another illustration (see Fig. 21) is the famous Pantokrator mosaic fromthe 12th century Cathedral of Monreale (Sicily) where we also find the ligature . Ofcourse, it was never obligatory to use these ligatures in Greek.

F ig. 19: Greek arti clein a mosaic fr om Hagia Sophia (f ir st half of 11th cent.)


14/16


37

F ig. 20: Greek l igatures aandin another mosaic fr om Hagia Sophia (1118)

F ig. 21: Greek li gature in a mosaic (Monreale, Sicily, 12th cent.)

6.4. The similarity between Greek and Slavic goes deeper, actually, than justwriting. In Greek, the forms of the article, Hand G, are actually very frequentlyused word-forms and occur in prominent positions in headlines or titles, as wehave seenthe Slavic preposition is well-known to occur in every heading of thefour Gospels39. And, to stress this basic fact once again, both in Greek and inSlavic, these shapes are complete words, not just characters which means theyare equal on this level, too.

6.5. There is an interesting small difference between Greek and Slavic, though: inGreek, the superscript T seems to play the more important in the shape of thecharacter while in Slavic its just the opposite: the Tis usually (but not always)smallbut at least it is not larger. First, this is a clear consequence of the reading

39I am not sure if this has been noticed before, but I am under the impression that the use of the preposition

itself could have been influenced by Greek, where the headings of the gospels usually start with TOEYAGGELION KATA [name] or TO KATA [name] [HAGION] EYAGGELION. Isnt it one of thereasons to use the Slavic preposition otin the first place to to mimic the Greek preposition kata?


15/16

, . 10.

38

direction: In Greek, the Tis the root of the word and thus important, while theOmegaor Ois the desinence and of lesser importance. But what about the Slavicwriting? Here, we have two forms: in running text, both are either of the same

size, being of equal importance for the word, or the T is smaller. However, inheadings we typically have a small superscript. To me, this is clearly an influencefrom Glagolitic writing where the superscript T was always used in its lowvariant, i.e. t. Superficially, the Glagolitic superscript T and the Cyrillicsuperscript Tcould even be thought of to be the same character, with Glagolica

just using the outline, not the solid form:

Glagolitic shape: L MCyrillic shape: A Q

Thus, the OCS Cyrillic header forms are a continuation of how the superscript T wasactually written in Glagolitic texts (i.e. low), but it was the Greek article and its usethat served as the blueprint for positively selecting the Slavic preposition otto stand

out among the otherwise mainly linear writing of the early manuscripts.

6.6. Stressing the reading direction of the Cyrillic superscripts (bottom up) couldpossibly also be used as a case in point against interpreting the w as a ligature of s and zbecause it would have to be read top-down, which seems less usual for Cyrillic40.

Preposition r: ligature w:

6.7. Another argument supporting our claims regarding the Greek influence on theZligature can be derived from Greek. Around the time the Cyrillic alphabet wasdevised, Greek went through a period of important changes and began to develop itsmany minuscle ligatures. The ligature Eis the oldest among all Greek ligatures, being

already attested in the 6th century; in the 10th century, ligatures for the inflectionalforms of the article began to spread (H F G). In other words: at the time we aretalking about, when the Greek helped to serve as a blueprint for Cyrillic r, there wasnot yet have an abundance of uncial ligatures in the Greek language, which wouldneed an explanation why certain were chosen and others were not. That obviouslystill was not the case; Greek had few firmly established and common uncial ligaturesat the time, and from them all thosethat made sense as Slavic words (prepositions)were taken over into Cyrillic, namely u and r. If we take the history of Greek itselfinto account, this will, by the way, also explain, why Kirill could use a ligature forOU in Glagolica: because the same ligature already existed in Greek while others did

40 Regardless of whether this is the true origin of the character or not, it is a fact that Slavic scribes

did not see these two components in the letter any more. As epkin (epkin V.N. Rassudenieo jazyk Savvinoj knigi. S priloeniem dvux fototipieskix snimkov. Sb. Otd. Russkago Jazyka iSlovesnosti Imp. Ak. Nauk, tom LXVII, No. 9 SPb. 1901 , .36) points out, there is one case inSavv., where the scribe corrected as, which should have been aw,by using a superscriptzinstead,

i.e. he wroteR., in other words: he followed the same reading direction as inq.


16/16


39

not. Obviously, the superscripted was not yet considered a ligature in Greek at thetime, but when the Cyrillic alphabet was developed, this view seemed to change.

If we take a look at tables of Greek minuscle ligatures, we can discern a clearfact: of all the many forms listed in such tables, practically only h and g caneasily be reinterpreted and reused as South Slavic words simply by changing thereading order, and these (along with older e) are indeed exactly the ligatureswhich were taken over into the Cyrillic alphabet not a result that seems to havehappened by chance, it seems, but a fact that can be systematically explained.

OT (Q)? ?

()

r M( L) , r ( , ). , , , , , , : , ;

ot Q . , , , , - , , r Q . , . : Q r ( U u) , , .

2 sebastian kempgen

Documents