Home » In English, Избрано, история

G. Sotiroff, Glagolitic Script and Linear B (Research from 1970)

2014.05.11 5 коментара

CANADIAN SLAVONIC PAPERS *, Vol. XII, No. 3, Fall 1970, 303-331


Glagolitic Script and Linear B **



The existence of two distinct Slavonic scripts, and still more so the fact that they have been used side by side for well over a thousand years, has aroused the curiosity of many scholars, out of which has sprung a vast literature aiming to establish when, where, by whom and for what purpose the second of these two scripts may have been invented. The latest attempt to unravel this mys­tery seems to date from 1963. In a richly documented study, V.A. Istrin discusses in detail the two Slavonic alphabets and the question as to which of them comes first.1 After a warning to the reader that ‘at the present stage of our knowledge it is not possible to settle this question in a definitive way’,2 Istrin proceeds to give his own opinion, without however hinting at the possibility of a connection between Glagolitic writing and Linear B. The purpose of what follows is to explore this possibility and some of its implications. Before doing this, it is necessary to review briefly the hypothesis put forward by Istrin himself. This will be followed by an analysis of the Glagolitic alphabet and by some comments relating to the antiquity of Slavonic writing.



One of the two extant Slavonic alphabets, according to Istrin, was certainly invented in 863 A.D. The other one – whichever it may have been – ‘could have been created in Moravia only after the death of Methodius (885 A.D.) and the expulsion of his disciples (886 A.D.)’.3 The fact that the Teutonic priesthood in Moravia at that time forbade the use of Slavonic liturgical books is sufficient reason for Istrin to accept Moravia as the birthplace of the second Slavonic script. The purpose of this invention, he believes, must have been to get away from Cyrillic – suspicious in the eyes of the Teutonic priesthood because of its affinity with the Greek alphabet, and therefore with Byzantium, – to copy the Slavonic books with the new (and different) characters ‘and thus to pull them out from under the papal curse, to save them, at least in part, from persecution and annihilation’.4 Thus, Istrin thinks, ‘if the second alphabet was created in Moravia, this could only have been the Glagolitic one’. The possibility that both alphabets originated in Bulgaria is discounted because of the absence of any reference to such a major event in the old Bulgarian manuscripts.5 As to the inventor of the Glagolitsa, Istrin believes this to have been Gorazd, one of the disciples of Methodius.

Istrin thinks that the above hypothesis also explains the complicated design of the Glagolitic characters as compared to the Cyrillic ones. He agrees with E.F. Karskiy, and actually produces a table in support of the latter’s view, that ‘all that was required was to turn around some of the Cyrillic letters, while supplying others with curlicues, instead of dashes or dots; this done, one gets the corresponding Glagolitic designs’.6

Furthermore, in the Glagolitic script Istrin sees an artificial creation and ‘apparently the product of the creative effort of a single man’.7 Finally, he observes that the Glagolitic numerals differ from the Cyrillic ones, but limits himself to saying that the Glagolitic system of numerals is ‘artificial’,8 without subjecting these numerals to scrutiny.



We might start by settling the preliminary question which has exercised, off and on, all students of early Slavonic literature. This question is, whether or not Saint Cyril was the author of the alphabet which bears his name. Istrin himself says that we cannot be certain of the answer, although he clearly believes Saint Cyril to have been the author. In this Istrin is right, if the following considerations have any validity:

a)     Saint Cyril undertook to compose a Slavonic alphabet upon the request of the emperor in Tsarigrad (Constantinople), as this is attested in the anonymous ‘Life’ of Cyril (chapter XIV). It is unthinkable that the Emperor would have given such an order if he did not expect a definite political advantage from this move. The immediate political purpose of this initiative was to pull neighbouring Bulgaria away from the influence of Rome, and to tie it, so far as possible, to the Eastern Empire. We must remember that, at the beginning of the 9th century, Byzantium had suffered severe setbacks owing to the rising power of Charlemagne in the west, and of Tsar Kroum in Bulgaria. If the new alphabet were to influence the Bulgarians (and other Slavonic peoples) in favour of Byzantium, it had to be similar to the Greek one, not dissimilar. The Cyrillic alphabet satisfies this condition, the Glagolitic does not.

b)     The Cyrillic numerals are practically identical with the Greek ones. Thus, the users of the Slavonic liturgical books – if they wanted to be able to refer back to Greek originals – did not have to memorize a separate numerical system. The Glagolitic numerals differ considerably from the Greek ones. If it was Saint Cyril who invented the Glagolitic numerals, he would have done the opposite of what he had set out to do.

c)     The Greek script which was in use in the 9th century A.D. is simpler, more functional, and more elegant than the Glagolitic script. Saint Cyril would not have been nearly as brilliant a man as he certainly was if he had neglected the opportunity of developing an equally simple, functional and elegant Slavonic script, for the sake of the curlicues for which the Glagolitsa is known.

For all these reasons, we conclude that the only alphabet Saint Cyril could have invented is the one which bears his name.

Was, then, the Glagolitic script invented later on, or was it already in existence when Saint Cyril undertook his task? A direct answer to this question is not available, of course. However, being reduced to conjecture, which we are, does not mean that we are reduced to making mere stabs in the dark. There are some basic facts which, when taken into account, help considerably in orienting this sort of inquiry towards positive results.

Why is it, we may ask, that the name of Gorazd, if it was he who invented the Glagolitic script, did not become attached to his invention, as the name Cyril became attached to the Cyrillic alphabet? We know that the Gothic alphabet was invented by Ulfila. We know that it was Cadmus who brought the Phoenician alphabet to Europe, where it was adopted first by the Macedonians (Pelasgians), and subsequently by the Greeks. Authors like Mauro Orbini and Anselmo Banduri tell us that the Glagolitic script was invented not by Gorazd, or some other disciple of Cyril or Methodius, but by Saint Jerome, who died in A.D. 420.9 Neither Istrin, nor anyone else, had disproved this statement. The claim, therefore, that it was one of the disciples of Cyril and Methodius who composed the Glagolitsa, is gratuitous.

Yet another point must be considered. Composing a new alphabet, and getting it to be accepted for general use, is anything but a simple matter. In the 9th century A.D., when writing materials were much more scarce and expensive than they are today, only wealthy and influential people could afford experi­ments with books. To all intents and purposes, kings alone could afford the outlay involved in the design and promotion of a new alphabet. Then, as today, nobody could have succeeded in such a venture without the backing of a strong government. It is also quite certain that no ruler would embark on such a major project without some weighty reason. Neither is it likely that such an event would have remained unrecorded, particularly if it came on the heels of another alphabet fulfilling, more conveniently, the same purpose. Istrin is aware that the Glagolitsa could not have been invented in the 9th century without a solid reason, and he looks for this reason in the need to save the Slavonic books in Moravia from persecution and annihilation. Unfortunately, this simply cannot be accepted as a reason. The quarrel in Moravia was not about a script, it was about power. The contestants were, on the one hand, the alliance between the Teutonic princes and Rome and, on the other hand, the Slavonic principalities, leaning on Byzantium. It is more than doubtful that the powers which threw the Slavonic liturgy out of Moravia and expelled the disciples of Methodius could have been so foolish as to allow the use of these

books in the churches again, only because they were now written with different characters. Or, are we to suppose that the Teutonic princes and their bishops were so unintelligent as not to notice what Karskiy finds so obvious, namely that the Glagolitic letters are no more than the Cyrillic ones ‘turned around’, or embellished with wiggles and curlicues? Besides, who persuaded the Teutonic princes to change their attitude? When and how did he do this selling job? We have not a clue. One must ask, therefore: is it not more realistic to assume that where the Slavonic books survived, they did so – in either script – because the balance of power favoured them, not because the Slavonic scribes outwitted the antagonistic Moravian bishops, or the Teutonic barons and their henchmen?

There are other points, too, which require careful consideration.

a)     Karskiy’s opinion regarding the shapes of the Glagolitic characters is not convincing. It is just not true that ‘all that was required was to turn around some of the Cyrillic letters, while sup­plying others with curlicues, instead of dashes or dots’, in order to get the Glagolitic shapes. Of the entire alphabet, only the letters ш and ωare identical in both the Cyrillic and Glagolitic alphabets, and only è represents a ‘turned-around’ ã, The letter ь is identical in both alphabets, but its phonetic value is different. One is forced, thus, to ask: if the Glagolitic script was invented after the Cyrillic one, and if the purpose of the inventor was to camouflage its origin, why did he leave these clues?   Why did he not invent entirely new symbols for the four characters ш, ω, è, and ь?

b)     It is also hard to believe that the supposed inventor of the supposedly new Glagolitic script – had his purpose really been to placate the Teutonic priesthood in Moravia, or to ingratiate him­self with it – would have failed to hit upon an easier formula than the invention of an entirely new alphabet of complicated design. Cernorizets Hrabri tells us that, before the time of Saint Cyril, the Slavonic people used to read and calculate by means of scratches and cuts, so long as they were heathens. He adds that, after their conversion to Christianity, the Slavonic people tried to write with Greek and Latin characters, as best they could (bez ustroenia). The Freising Fragments show how this was done and, indeed, some Slavonic peoples were compelled to use the Latin alphabet and are still using it today. Why, then, did the supposed inventor of the Glagolitic script not use a modified Latin alphabet, but went to the trouble of putting together these elaborate characters?   Hecould also have decided in favour of another alphabet of simpler design, like the one borrowed by the men who produced the Codex Argenteus. The Ravenna manuscript, now in the National Library at Naples, shows that this fine script was available in Slavonic Europe as far back as the 6th century. The supposed inventor of the Glagolitsa thus seems to have gone out of his way to make things complicated and yet, complexity in writing does not suggest novelty. Rather the opposite is true: the more compli­cated a script is, the more ancient it tends to be. Linear B is the most complicated script found so far on the European mainland, and also the most archaic. The Glagolitsa definitely looks more like an ancient alphabet than like a modern one.

c) Finally, the numerals based on the Glagolitic alphabet are also indicative of great age, rather than of novelty. We must remember that towards the end of the 9th century A.D., when the Glagolitic script is supposed to have come into being, the Arabic numerals were making their way into Europe. Shall we assume that the inventor of the Glagolitsa was unaware of the Arabic numerals? Granted that he was, why did he not adopt the well–known Greek numerals? Supposing that he shunned the Greek numerals, for the same reasons for which he shunned the Greek letters, why could he not have adopted the Roman numerals? The answer is that he could have done any of these things, if he had existed. Yet, it is possible that there was no need for anyone to invent new numerals in the 9th century A.D., if it turns out that the Glagolitic script, as well as the numerals which are based on it, were already known and, thus, did not have to be invented at all, when Saint Cyril began the translation of the liturgical books into Slavonic.

To recapitulate, we believe that:

a)     the Cyrillic alphabet was invented by no one other than Saint Cyril himself;

b)     the reason advanced by Istrin for the invention of the Glagolitsa in Moravia lacks conviction;

c)     the  suspicion  that  the  Glagolitic  script antedates the Cyrillic one is strengthened by the existence of distinct Glagolitic numerals,  a fact hardly compatible with the more attractive choices which lay open to any alphabetmaker living towards the end of the 9th century A.D.



Can the Glagolitic characters tell us something about the probable time at which they may have appeared on the historical scene? They might be able to if we could study attentively: (a) the shapes of the individual characters; (b) their phonetic values; (c) their positions within the alphabet; (d) their names; (e) the numerals based on them. It would be impossible to cover all these points adequately within the limits of a short paper. It is, however, possible to touch upon each one of them briefly. We shall take the numerals first.


The complete Glagolitic alphabet is given on page 10. It is sufficient to give here the first 13 characters, with the corres­ponding numerals:


                            Glagolitic                                       Cyrillic






















































It is apparent that while in the Glagolitsa the letters and the numerals are in accord, this is not so in the Kirillitsa. The reason for the discrepancy between letters and numerals in the Kirillitsa stems from the fact that Saint Cyril simply took over the Greek numerals as they were – and these numerals were not in accord with his new alphabet.

Inevitably we are faced with the question: why did Saint Cyril not arrange the letters in his own alphabet so as to harmonize them with the numerals, and avoid the inconvenience resulting from the discrepancy? He maintained the Greek numerals because he must have realized the practical importance of the generalized use of the same numerals, as we recognize today the enormous advantage resulting from the general use of the Arabic numerals. Saint Cyril undoubtedly gave thought to letters such as q , z, y, and U, which exist in the Greek alphabet, but have no counterpart in Slavonic phonetics. These letters Saint Cyril adopted in his alphabet, but placed them at the end of it, decreeing that they should be used to represent numerals (q = 9, z = 60, y = 700, U = 400) and in Greek names such as QOMA and ALExAHDРЪ

However, if Saint Cyril did place uncommon symbols at the end of his alphabet, why did he not do the same with the new Slavonic letters Б and Ж, which do not exist in Greek? What induced him to locate them so close to the top of his alphabet? We do not know why he did this. We can only surmise that this is exactly what he would have done, if he had had the Glago­litic alphabet in front of him to use as a guide. Thus, he took over the Greek numerals almost without change because, he must have reasoned, this was the most practical thing to do. Likewise, he must have adopted the order of the Glagolitic alphabet because this, too, was the most practical thing to do, especially if the names of the letters had already proved their usefulness as a mnemotechnical device. Nobody will deny that, in this respect, the names of the Slavonic letters may be said to be a product of genius. It is infinitely easier to memorize the Slavonic alphabet by means of the story: az bouki vedi, glagol dobro est zhiveti zdelom zemli i kako liudi misleti, etc., than to try to remember the rather colourless Greek names alpha, beta, etc., or the mean­ingless Roman ah, beh, ceh, etc.


The presence of the letter dierv, ç(Cyrillic ћ) in the twelfth place of the Glagolitic alphabet, is also revealing. Better than any other argument, the position and numerical value of this letter prove that the Glagolitic script antedates the Cyrillic, and that it was invented in the heart of the Balkan Peninsula, and nowhere else. The sound which this letter represents – a palatalized ch – does not exist in Greek. Neither does it exist in Czech, Slovak, Russian, Ukrainian, or even in the dialect spoken in eastern Bulgaria. It is perceptible in western Bulgaria and Macedonia, but it is much clearer in Serbia, Montenegro, and Croatia. It is true that, had the Glagolitic alphabet been invented in Moravia, such a character could have been added to it at a later date by scribes who lived in Serbia and/or Macedonia. In such an eventuality, however, one would expect to find it near the end of the alphabet, perhaps somewhere after the yat. Instead of that, we find it in the upper half of the alphabet, among the basic sounds, and with the numerical value of  30, corresponding exactly to its position.



We are told by some knowledgeable men that, prior to the 9th century A.D., the Slavonic peoples did not write down their language.10 A study of the design of the Glagolitic letters shows this opinion to be untenable.

The Glagolitc alphabet is said to contain 40 characters. This number is both accurate and inaccurate. The positions, names, and configurations of the last eleven characters except ‘yat’ suggest that these characters were likely grafted on to the original stock at a relatively late date. This, needless to say, has happened to other alphabets. For centuries, the Greeks wrote with sixteen letters. Only gradually was this number increased to twenty–four. In the case of the Latin alphabet, we know that W, Y, X and Z are late additions. Originally, the Glagolitic characters also may have been – must have been – fewer than fourty.

One look at the last two characters of the alphabet (fita and ijitsa) is enough to convince the reader that they have been taken over from Greek, or Cyrillic, at a late date, for diplomatic reasons. Another reason for believing that the number of characters in the Glagolitsa was originally less than fourty is provided by the symbolÌ. This symbol stands for both the big and the small ier, the hard and the soft znak in modern Russian, and is thus counted twice in the alphabet.

Finally, in the total of fourty are included several obvious derivatives, such as shta and er–i (щ and yrespectively), which may be included in the total, or excluded, depending on the purpose for which the count is made. The same applies to the diphthong you and to the family of four youses (nasals).

By subtracting from the total of fourty the ten characters mentioned above, we get thirty basic Glagolitic characters. What does the study of these basic characters reveal? The first thing we notice is that of the thirty, only four can be assimilated to Cyrillic or Greek equivalents, and that not without some pulling and stretching. The four are: e, j, n,f (est, zemlya, pokoi, and fert.) The study of these four characters is, indeed, illuminating:

The most striking case is that of the letter e. It is clearly seen that this is the Cyrillic, or Greek, letter e, turned arounc or so Istrin and Karskiy obviously think. Yet nothing in life is simple, not even the case of the letter e. The truth of the matter is that at the dawn of Greek literacy, when the Greeks wrote boustrophedon, this letter faced right or left, defending on the direction in which the scribe was writing. Whenme Greeks started to write consistently from left to right, theletter е was made to face right. However, in the Etruscan alphabet e contin­ued to face left. It faces left also in the name of the Thracian tribe Derrones, stamped from right to left on two extant dekadrachmas. The inference is, of course, that the inventor of the Glagolitic alphabet may not have had to ‘turn around’ the Cyrillic е . He may have been acquainted with the left–facing e from inscriptions, or coins struck in his own land, parti­cularly if he happened to be living not in the 9th century A.D. but in the 6th century B.C., or earlier.

Likewise, the letter pokoi (n) at first glance looks as if it might have been borrowed from Cyrillic or from the Greek. Closer inspection shows that the shape of the letter is not the one it has in 9th century Greek documents. It is more like the shape of the r found in the inscription of Melos dating, as it is believed, from the 7th century B.C. As to fert (f), it looks it is true, like a Greek theta with ears hanging down from the sides. One must observe, in this connection, the fact that in Etruscan, for instance, the earliest symbol for theta was Å from which both q and f have likely evolved. Therefore, the possibility cannot be ruled out that the inventor of the Glagolitic fert’ (f) may have been a man who lived centuries before Saint Cyril, that is to say when the Greek G had the shape of the Glagolitic letter ‘pokoi’, when the shape of ‘fert’ had not yet been fixed, and when the books of the Etruscans had not yet been wiped out.

As to the Glagolitic letter ‘zemlya’ (j), one must admit that it does look like a Greek ‘theta’ with a curlicue. Of course, a curlicue is as easily added as it is dropped. Suppose that ‘zemlya’ was a very ancient, perhaps hieroglyphic, character, expressing the sound ‘z’. One could ask a question apt to jolt many a complacent spirit. Why could this symbol not have survived unchanged, with the same phonetic value, among the people who invented and used the Glagolitic script, while the Greeks snipped the letter’s curled tail, and used the truncated form to express the th sound, typical of the Greek language but absent from the Slavonic speech? The one possibility has as much merit as the other.

The four characters discussed in the preceding paragraph are all there is, when it comes to direct borrowing from Cyrillic, or from the Greek, with or without wiggles and curlicues. However, there is another case – that of the letter ь – which is very significant in this connection. Each individual letter in the Slavonic alphabets, Cyrillic as well as Glagolitic, has its special name which includes the phonetic value of that particular letter, Thus, az includes the sound ‘a’, buki the sound ‘b’, vedi the sound ‘v’, and so on. The name of the Cyrillic letter ь represents an exception of great importance. This letter (ь) is called  ier, yet is does not stand for either i, e, or ‘r’. It stands for the sound ‘u’, as in the English word ‘but’. Why, then, was it called ‘ier’? It was, and is, called ier because in Glagolitic it represents the round ‘r’. We have, here, a strong indication that the Glagolitic alphabet is older than the Cyrillic one. It is more than likely that after Saint Cyril had adopted the Greek letter p to denote the sound ‘r’, he noticed the simple and elegant form of the Glagolitic letter ь, which he decided to include in his new (Cyrillic) alphabet with a new, typically Slavonic, phonetic value.

There are other interesting things about the Glagolitic alphabet: it includes five characters which very much resemble Hebrew (Phoenician) equivalents. These five characters are:


        Hebrew11                                                                                                                                        Glagolitic



























Phonetic value







The resemblance between the two sets of characters isobvious, with the exception of Ain/On, where the Hebrew (Phoenician) letter seems ot have been rotated by 90 degrees. It will be observed, further, that the Glagoliticoresembles the Hebrew prototype more closely than does the Greek O. As to the Glagolitic letter ‘shta’ (q), not represented above, it is clearly no more than a sha (ш) sitting on top of a onelegged tverdo (o).

Of equal interest are the Glagolitic characters bouki, vedi, and glagol. Each of these corresponds closely to some character found in one or the other of the alphabets used in high antiquity, as may be seen from the following comparison:12

a) The Glagolitic bouki (b) resembles the Cretan linear

which stands for the word beth (house), or the sound B.

b)     The Glagolitic vedi (v) resembles its Carian counterpart V (U).

c)     The Glagolitic letter glagol (g) resembles the Etruscan 7, which stands for the sound G (as in God). This leftfacing G, incidentally, is also found in the oldest Greek inscriptions.

Thus, of the thirty basic Glagolitic characters, eight, that is one of every four, can be traced to ancient alphabets and need not have been borrowed or adapted from either Cyrillic or 9th–centuryGreek. The most interesting comparison, however, is the one that can be made between the Glagolitsa and the Linear B syllabary compiled by Ventris and Chadwick.13 At least seven Glagolitic characters resemble so closely symbols found in the syllabary that one is compelled to take notice of the fact. This means a ratio of 1 to 5, even if we ignore the probability that, in its initial composition, the Glagolitsa, like the Greek alphabet, may have had fewer letters, perhaps only sixteen or nineteen.

Linear B                                                                         Glagolitic

With regard to the symbol 

Untitled5, to which the syllabary assings the phonetic value de, it should be pointed out that in the earliest Greek monuments the letter ¯is used to represent any one of the sound z, dz, or zd. Thus, we find DANKLE, standing for ZANKLE. Since the Greeks could not pronounce the sound ch, it is understandable, perhaps, how the same symbol could appear as de in the syllabary, while representing the sound ch in Glagolitic. (Another interesting fact is that


, has been assigned the phonetic value ke. This is relevant, for the Greeks, unable to pronounce ch, approximated it as best they could by a palatalized k. Thus, the word chinya [to make] gives cinguo in Latin, KINEW in Greek. The word chark (cog) becomes circus in Latin, KYKLOS in Greek. Faced with a name like Chichero [Cicero], the Greeks pronounced, and wrote, KIKEPO.) With regard to No. 40 (æ ), one should note that there are digammated words in archaic Greek – i.e. words beginning with ‘Wa–’ or ‘Fa–’ – which feature an initial diphthong in the corresponding Slavonic forms. As examples one may quote the Greek words FANAX and FASGANON, the Slavonic counter­parts of which are IOUNAK (hero) and IATAGAN (curved sword). Recapitulating, we find that of a total of thirty basic characters in the Glagolitic alphabet, there are only four which might, but need not have been borrowed from Cyrillic or Greek, and from archaic Greek at that. Another five may have come directly from the Hebrew (Phoenician) alphabet, three more seem to have made their way into Glagolitic via other ancient alphabets, and another seven from Linear B prototypes, as follows:


What does this analysis prove? It proves that if we allow the inventor of the Glagolitic alphabet to have used only his own imagination, instead of imitating or adapting symbols found in very ancient alphabets, strange coincidences must have been at work, in order to produce the above results affecting over one half of the characters involved. Does this analysis also prove that Glagolitic is a direct descendent of Linear B? It would be bold, perhaps, to hastily put forward such a claim. Yet, it does seem that the Glagolitic script may have been developed in high antiquity, possibly at the time when all Mediterranean nations – Cretans, Greeks, Etruscans, and Thracians – were busy adapting the Phoenician alphabet to the phonetics of their respective languages. This hypothesis may have to be modified, if and when the reading of Linear B becomes more certain. At the present time, Linear B syllabary consists of some ninety symbols, fifteen of which have not yet been decoded. The reading of three others is doubtful. The scantiness of sibilants in the syllabary suggests the possibility of further revisions. Out of these, new points of con­tact between Linear B and Glagolitic might emerge. The digression at the end of this article shows that certain ideograms like those standing for the syllables KO, QE/CHE, RI, RU, SE, and TI, admirably fit certain words from the speech for which the Glagolitsa was designed.    



The external evidence that the Glagolitic script antedates the Cyrillic is as follows:

a)     Excavations carried out on the site of the ancient Bulgarian capital of Preslav have yielded ceramic tiles with Cyrillic inscrip­tions on the enamelled side and Glagolitic numbers on the back. It is obvious that, once the tiles had been sealed in their proper places, nobody could see the Glagolitic numbers. These numbers, therefore, could only have been carved on the tiles with a view to helping the workers responsible for making, glazing, storing, and transporting the tiles. It follows, then, that the workers were able to read Glagolitic, although they were not able to read Cyrillic numbers. This, in its turn, indicates that the Glagolitic characters were known in the area around Preslav before the Cyrillic ones were introduced.14

b)     Another piece of evidence, to the same effect, is provided by the marginal notes to the Zographe Gospel.15This gospel is written in Glagolitic. However, the marginal notes to it are all in Cyrillic. It is obvious, then, that the scholiast, who was able to read Glagolitic, either could not, or was not allowed to, write his notes in the same script. This too suggests that Cyrillic came after Glagolitic.

c)      According to trustworthy information, there is at least one Slavonic palimpsest, at the present time in the Saltykov–Shchedrin Library in Leningrad. This palimpsest has a Glagolitic base, which is also an indication that the Cyrillic alphabet is more recent than the Glagolitic one. Needless to say, this argument would be overthrown should it turn out that there are also palimpsests with a Cyrillic base. However, so far, such cases have not been brought to the attention of this writer.  



No sooner has the antiquity of the Glagolitic alphabet been accepted, even as a working hypothesis, than several problems arise which are all soluble but which are too numerous to receive all the treatment which they deserve, within the scope of this paper. The most immediate of these problems are: (1) If a Glagolitic alphabet did exist before, say 863 A.D., why did Saint Cyril have to invent a new one? (2) If such an alphabet did exist, why is it that no Glagolitic monuments survive from the period preceding Saint Cyril? (3) If such an alphabet did exist in high antiquity, why is there no reference to it in Greek or Latin sources? There are many reasons which may have motivated Saint Cyril to design an entirely new alphabet for the Slavonic peoples who were turning in growing numbers to Christianity. In the first place, in addition to the reasons which have already been mentioned, there are other indications – which cannot be discussed here –that the Glagolitic script was developed in Macedonia, or south­western Bulgaria. The Glagolitic letters must have evolved in a way suited to the language spoken in that part of the Slavonic world, but the whole Glagolitic alphabet may not have contained all the characters required to represent the sounds of the dialects spoken as far north as the Baltic Sea, and as far west as the Sudeten. This was probably the case of the nasal e and a, as well as the case of eri. Saint Cyril, who had in mind the conversion of all Slavs, must have realized the need to add several letters on to those already in existence, and this must have suggested to him that it would not be any more difficult to design a new alphabet than to improve the one which already existed. In the second place, in the 9th century A.D. the Glagolitic alphabet must have been considered a product of paganism, which it undoubtedly was, and, therefore, sinful and to be avoided. An adaptation of the Greek alphabet had a psychological advantage over the pagan ‘scratches and cuts’. Greek had the honour of being one of the three ‘god-given’ languages, since those three – Greek, Latin, and Hebrew – were used for the famous inscription ‘Jesus Christ, King of Judaea’. Some of this glory was bound to rub off on a new, ‘entirely Christian’,Slavonic alphabet. In the third place, the Kirillitsa could be made to be more functional than the Glagolitic ‘scratches and cuts’. In the fourth place, a new system of writing, clearly based on the Greek one, was bound to draw its users closer to Constantinople than would have done an improved indigenous Glagolitsa. Last but not least, the same purpose was served by the new, Greek, system of numerals, which was introduced into the Slavonic world along with the Cyrillic script, as opposed to the native Glagolitic numbers. Turning to the absence of pre-Cyrillic Glagolitic monuments, one could explain it by asking: why have we not got any Cyrillic materials from the 9th, or early 10th century? l6 Why are the earliest Slavonic manuscripts which have come to use of a later date than that?  Obviously, the earlier manuscripts have been lost or destroyed. However, if the Christian writings of the 9th century could be totally obliterated, how much easier must it have been for the books from higher antiquity to vanish!   In those days, books  were  written  on  papyrus,  parchment,  linen  cloth  or wooden slates. Like paper, all these materials are easily destroyed by fire, moisture or pests. Thus, of the undoubtedly vast Hittite literature, not a single book has come down to us. Of the very considerable Etruscan literature – which, we know, existed as late as the 4th century A.D. – only one fragment survives, and that in the form of strips of linen cloth, used to wrap up a mummy. Nor was the deliberate destruction of books a rare event. It is reported that the great Plato once wished to burn all the writings of Democritus which he could get hold of, but was dissuaded by two friends, who pointed out to him that the books were already widely circulated.17 In his Life of Porphyrios, Marc the Deacon takes pleasure in describing how, in 402 A.D., the Christians of Gaza set fire to the temple of Zeus, and how ‘the books of the pagans suffered the fate of their gods’.18 Indeed, it is not to be expected that pious Christian monks, who delighted in such acts of faith, should have bothered to copy and preserve for posterity pre­-Cyrillic, pagan, Glagolitic oracles, poetry or chronicles. That is why it will be more than a minor miracle if we should ever find authentic documents of this kind. We are probably obligated to the shrewdness of native Slavonic scribes for the preservation of the Glagolitic script, such as we have it. One is tempted to guess that these simple men applied in their own way the rule, so popular among practical politicians: ‘If you cannot beat them, join them’. They probably realized that, as Christianity became the official religion of the Slavonic principal­ities, the Cyrillic script was also bound to be accepted, as a matter of state policy. The old native scribes must have suspected that all influence was likely to slip out of their hands, unless they were able to compete more successfully with what might have rated in their eyes as ‘the Cyrillic school’. This may have furnished the motivation to round out the Glagolitic alphabet by the introduc­tion of several new symbols such as the nasal e and a, the eri, the diphthong ‘you’ and the small and big ‘ier’, and to use the thus improved Glagolitsa for the immediate transcription of liturgical books – the more, the better. On the surface, the question as to why there is no mention of ancient Glagolitic monuments in Greek or Latin sources seems more embarrassing. How could Greek sources, for instance, make reference to Glagolitic writings at a time when the Slavonic peoples, that is to say the users of the Glagolitic script, were not supposed to have even existed? If we so wish, we may declare this question to be insoluble and give it up. Or, we may wish to look into it. If we do, we shall see that Cernorizets Hrabri does not say that the Slavonic people were illiterate before the time of Saint Cyril. He says that they wrote and calculated ‘by means of scratches and cuts’. Just how far back could Hrabri’s information on this score have gone? It is not hard to visualize the situation. Hrabri wrote some time around 890 A.D., when he may have been 30 years old. Let us assume that when he was a boy in grammar school, say around 870 A.D., he had a very old teacher. This man, who may have been 80 years old, could have told his students that, before the days of the Cyrillic alphabet, people wrote and calculated by means of scratches and cuts. Perhaps he even showed them specimens of these scratches and cuts on pieces of wood, which he had received, when he was himself a little boy of ten, that is to say in or about the year 800 A.D. However, if we assume that when Hrabri wrote his Skazanie he was not thirty but eighty years old, his information on the writing by means of ‘scratches and cuts’ would go back as far as the year 750 A.D. Nor is there any need to stop there. Hrabri’s teacher could have learned those same scratches and cuts from another old man, one that was born, perhaps, sometime in the 7th century. It is not possible to say how far back we may go in this way, unless we use a clue provided by Hrabri himself. He says that the writing by means of scratches and cuts was used by the Slavonic peoples as long as they were pagans. There are ample indications, however, that paganism was on its way out of the Balkan Peninsula as early as the 4th century A.D.19   That, then, must have been the time when, according to Hrabri, the Slavonic people began to try to write the Christian liturgical books with Greek and Latin letters, as best they could. But were there any Slavonic people in the Balkan Peninsula in the 4th century A.D.?   This question has been answered in another study.20 The answer is: yes, Slavonic speaking people did live in the Balkan Peninsula in the 4th century A.D., although they were not called, at that time, by this particular name. The name which ancient Greek writers gave to the Slavonic people in south­eastern Europe was Getae.21   This was the name of but one Thracian tribe. However, Greek writers often applied it to all Slavonic peoples as a group. It is important to remember this, and to realize the fact that the Getae, being Thracians, most likely made use of Thracian letters and books. When, in the 6th century, the Slovaks and the Slovenes became famous, their name began to designate all peoples who spoke a dialect related to theirs. This, then, is the reason why we find no mention in Greek sources of a pre-Cyrillic Slavonic script: whatever references there are to such a script, are in respect to Thracian writing. Few as they are, these references are of enormous significance. No less an authority than the famous patriarch Photius testifies that, as late as of the 7th century, Thracian books were being consulted by Greek scholars.22 Is it really fantastic to infer that these were the now lost Glagolitic books? Most of the writing in antiquity was done on birch bark or on tablets made of linden–wood. In this connection, we must remember that as far back as the 5th century B.C. Euripides spoke with respect of ‘the tablets of Thrace’.23 Sophocles, too, makes Deianeira say that Hercules left an inscribed oracular tablet with her. Hercules is made to say that he wrote down an oracle in the Thracian grove of the Selli, that is to say at Dodona, in Macedonia.24 Wooden tablets were being used 400 years before the time of Sophocles, since Hesiod himself declares: ‘I have newly written (a lay) in tablets upon my knee’.25 Tablets continued in use not for centuries, but for thousands of years. Around 950 A.D., the Arab writer Ibn Abi Jacub el Nedim reported that the Russians had writing which was carved in wood.26 Quite recently, 194 tablets have been excavated at Novgorod, with Cyrillic text on them. The earliest of these tablets dates from the 11th century, the latest from the 16th.27 It is only natural to suppose, therefore, that when the Slavonic people wrote with scratches and cuts, they used wooden tablets on which they carved Glagolitic characters.28 All these are details which are in the nature of things, but which are not likely to have attracted the attention of Greek and Latin historians sufficiently to deserve special mention in the sources which have been preserved in our libraries. It may not be out of order to point out the difficulties which arise from the rejection of the above hypotheses. If we should rule that the Thracian books of the 7th century were not the Glagolitic ones, this would mean several things. It would mean that (1) the Thracian books were completely wiped out; (2) while this wiping out was taking place, the people in Thrace and Macedonia were developing a new system of ‘scratches and cuts’ for the purpose of writing and calculating – the pagan writing mentioned by Hrabri; (3) all this went on while the Christians in those parts were also trying to write with Greek and Latin characters; (4) when Christianity was made the official religion in Thrace and Macedonia – that is to say in Bulgaria – the ‘scratches and cuts’ were quickly dropped; (5) at the same time, two com­pletely new alphabets, the Cyrillic and the Glagolitic, were developed almost simultaneously, and were actually accepted and used throughout the Slavonic world. What a multiplication of improbabilities!



What seems to emerge from the preceding analysis may be summed up as follows: 1)     There is no reason to accept that the Glagolitic alphabet was invented after the Cyrillic, either in Moravia or anywhere else. The Glagolitic alphabet, therefore, must have been in existence before the Cyrillic one; 2)     The Cyrillic alphabet was invented by no one other than Saint Cyril himself; 3)     The affinity of seven Glagolitic characters with Linear B symbols, and of another twelve characters with letters from other ancient alphabets suggests that Glagolitic could easily have been one of the several alphabets known to have existed in high anti­quity; 4)     Glagolitic was probably the alphabet in which the now lost Thracian books were written; 5)     When the Slavonic principalities accepted Christianity as the  official  religion,   Saint  Cyril  invented  the new alphabet, on order and with the encouragement of the emperor. It is likely that, when this happened, the Glagolitic scribes realized that their art was in jeopardy. This must have been sufficient inducement for them to intensify their transcribing, in Glagolitic, of the Slavonic liturgical books, in competition with the scribes newly–trained to use Cyrillic. It is quite conceivable that, at that moment, the users of Glagolitic noticed the usefulness of the additional characters invented by Saint Cyril, such as the family of‘youses’ and the big and small ier (in Russian: ‘tverdiy’ and ‘myagkiy snak’). This must have suggested to them to contrive parallel Glagolitic symbols, which were palced at the end of the alphabet. 6) The original Glagolitic characters must have been more like ‘scratches and cuts’ than the ones which have come to us. This is suggested by the fact that proto-glagolic characters must have been used predominantly for writing on wooden tablets. When paper came into general use it became possible to embellish the ‘scratches and cuts’ by flowery ‘wiggles and curlicues’, which led to the Glagolitsa as we know it, so pretty and so puzzling to 20th century scholars! There are people who will call this speculation. The question is whether this speculation agrees with known historical facts, and whether or not it brings into a harmonious whole the surviving hints and shreds of evidence. Anyone rejecting this speculation should be willing to accept the still more fantastic proposition that the Slavonic peoples, arriving on the historical scene, as it were from outer space, sometime in the 6th century, wiped out in some mysterious way millions of ancient Thracians and their literature, occupied most of Europe and a good part of Asia, and organized their several States in a matter of less than 3 centuries. Not only this, but they gave themselves, for no conceivableareason, two distinct alphabets, which they adopted almost simultaneously, and then continued to use them for the next thousand years! Among the many miracles which the Slavonic peoples may have performed, this one seems to be the hardest to believe.    






We are informed that some 3,000 clay tablets have been found, on Crete and at various sites in the Peloponnese, inscribed with what has come to be known as Linear B script. According to Messrs. Ventris and Chadwick29 who have deciphered – or who are believed to have deciphered – the Linear B script, this system comprises two parts. The first represents a syllabary of some ninety signs, of which eighteen have not yet been reliably decoded. The second part consists of 152, or so, ideograms such as && (chariot) and &&(sow). Here, too, one finds figures which have not yet been explained. The script one may observe on the clay tablets combines these ideograms with symbols from the syllabary, somewhat in the manner of Japanese writing. The reader will recall that the Japanese script combines certain Chinese ideograms, called Kanji, with indigenous syllabic sings, called Kana. The student of Linear B is struck by two things. The first is the scantiness of sibillants: the consonantssh, ch andj, are totally absent. The second is the large number of doublets. Thus, the syllabary shows three symbols for pa – if this is indeed the syllable pa. There are also two signs for each one of the syllables pu, ra, and ro. The scantiness of sybillants requires no explanation, so long as one assumes the language of Linear B to have been Greek. On the other hand, should it happen that some of the so–far–undecoded signs represent sibillants, it would be difficult to maintain the hypothesis that Linear B was a Greek script. In such a case, the question regarding the language of Linear B would have to be re-opened.  



As a matter of fact, the question regarding the language of Linear B has never been settled in a satisfactory way. As Leonard Palmer has observed ‘the occurrence of pot, robe, table in English does not serve to identify it with French’.30 The situation is analogous in the case of Linear B. Take a group such as &&& . This could be read four axes, or quatre haches, or chetire topora. Possibly in or about the year 1400 B.C., some kind of Greek was spoken on the periphery of south–eastern Europe. But other languages were also spoken in that area which could have made use of this ideographic script. That is why a wag could describe the decipherment of Linear B as a long–distance telephone conversation over a defective wire, with a party having a speech impediment. We must take another look at some of the Linear B ideograms.  






Our hypothetical inventory of Pelasgian words now consists of seven items: kour, koja, sekyra, cheroup(ka), riza, roug, and ship. Is there any language whose vocabulary would comprise these words? The answer is: yes, such a language exists. It is the language of two million Macedonians who live, today, in Yougoslavia, Greece and Bulgaria. Does this mean that the Macedonian dialect (language) is identical with the Pelasgian tongue? It means just that, even though this may come as a shock to many a scholar. It means that the present–day Macedonians are the descendants of the ancient Pelasgians, and that they continue to speak the Pelasgian tongue. In Book VII, Chapter I of his Histories, Pompeius Trogus wrote: Macedonia ante a nomine Emathionis regis, cuius prima virtutis experimenta in ittis locis extant, Emathia cognominata est.(Macedonia was once called Emathia, from the name of Emathion; one may still see there the first monuments of his valour.) And a little further: Populus Pelasgi, regio Bottia dicebatur. (The inha­bitants were called Pelasgians, the district Bottia.) The seven words which were explained above form part of the Macedonian vocabulary of today, as does also the word riba, meaning fish, whose outline


, representing the syllable ri, may be found in the Linear B syllabary, under No. 53.  



There are more surprises in store for those who care to take a look at our major sources of ancient history. Strabo speaks of a place in Macedonia whose name was Leibethrum and which was sacred to the Muses.33 According to Diodorus Siculus34 the Muses were sisters who had received an excellent education. It was the Muses who invented the letters and who first composed a poem.35 Diodorus, futher, explains that all the Phoenicians did was to change the shape of the letters; these changed letters were adopted by all peoples who called them Phoenician. However, someone will ask, if the language oftthe Linear B script was Macedonian and not Greek, and if these were two different languages, how could Messrs. Ventris and Chadwick have determined correctly the syllabic values of certain signs? They could, because it is possible to reach a correct conclusion even when one starts from a wrong premise. The following example will illustrate how this may happen. Supposing an old inscription were found reading: BILLIE CHAT COURTS BEEN. The man who postulates that the language is German will be tempted to read the inscription BILLIE HAT KURZE BEINE. At the same time, the man who postulates that the language is French would prefer the interpretation: BILLIE (LE) CHAT COURT BIEN. It is obvious that both interpretations cannot be true. Or, rather, either interpretation could be true, depending on the time and the circumstances under which the inscription was carved. The following is an example of what Messrs. Ventris and Chadwick have noticed and what they have not. One of the Linear B words, we are told, reads korito (ko–ru–to). This is interpreted as the genitive case of the Greek work KOPYΣ, meaning helmet. What the decipherers obviously did not know is that the word korito, such as it is, exists today in the Macedonian language. It means tub. A Macedonian tub, put upside down on the head of a Greek, may easily become a helmet. This, however, does not free the decipherers of Linear B from the obligation of trying the Macedonian (Pelasgian) tongue on the clay tablets since, in this exercise, Greek has so far yielded such melancholy results.  



This digression has shown that:

(a)     the most ancient script in Europe was the one invented by the Muses in Pelasgia (Macedonia);

(b)    several ideograms in Linear B – the most ancient script which has been found, so far, in south–eastern Europe – clearly represent objects whose names in the Macedonian tongue suit certain syllabic signs deciphered by Ventris and Chadwick;

(c)    since no definite date can be assigned to the invention of the Glagolitic alphabet, and since several Glagolitic letters coincide with signs occurring in the Linear B syllabary – as shown  else­where in this study – it is likely that the Glagolitic alphabet goes back all the way to the time when syllabic writing (Linear B) was used in Macedonia before being replaced by alphabetic writing.36    



11100 Let Slavyanskoy Azbouki, Moscow, 1963.  

2 Op. cit., p. 127.         

3 Ibid., p. 128.  

4 Ibid., p. 129.  

5 Ibid., p. 133.  

6 Ibid., p. 130 (Referring to E.F. Karskiy, Slavyanskaya Kirillovskaya Paleografiya Leningrad, 1928, p. 359).  

7 Ibid., p. 138.  

8 Ibid., p. 146.   

9 See Mauro Orbini, E Regno degli Slavi, Pesaro, 1601, p. 46–7, and Anselmo Banduri, Opera, 1.1, Venice, 1700. (Quoted in S. Lesnoi, Istoriya ‘Russov’v neizvrashchennom vide, Paris, 1959, fasc. 9, p. 936–8.)

10 Horace G. Lunt, Old Church Slavonic Grammar   SGravenhage (Mouton), 1959, p. 1.

11 The Hebrew letters have been taken from Fr. Thiersch, Greek Grammar (English translation), Edinburgh, 1830, p. 23–4.

12 The symbols from Cretan Linear and Carian were taken from David Diringer, The Alphabet (2nd ed.), London (Hutchinson), 1949. The Etruscan theta is given in M. Pallottino: The Etruscans, London, 1956 ed., p. 259.

13 John Chadwick, The Decipherment of Linear B, Cambridge University Press, 1958.

14 See Nikola Mavrodinov, Starobъlgarskoto izkustvo, Sofia, 1959, p. 238.

15 See Quattuor Evangeliarum Codex Glagoliticus, edited by V. Jagic, Berlin 1879, p. VIII.

16  The oldest dated Slavonic manuscript is the Ostromirovo Evangelie, from the year 1056.

17 Diog. Laert.,IX. VII. 40.

18 See Henri Gregoire & M.A. Kugener (ed.) Paris (Les Belles Lettres), 1930, ch. 71, p. 57.

19 From Eusebius’ Ecclesiastical History (V. XIX. 2) we learn that, some time before 337 A.0., the city of Debeltum (the modem Burgas) had a bishop by the name of Aelius Publius Julius. In 344 or 347, a council of 250 bishops met at Serdica (the modern Sofia). Theodoret, in his History of the Church, repeatedly refers to the Christian churches in Dalmatia, Dacia, Moesia and Macedonia.

20 See my article: Y–a–t–il eu une ecriture autochtone en terre slave avant le temps de Cyrille et Methods? Revue canadienne d’etudes slaves, Vol. I, fasc. 1, Montreal, printemps 1967, p. 79–94.

21 See Theophylact Simocatta, Historiae,  Books III. 4.  7. and  VII.  2. 5. Ed. C. de Boor, Leipzig, 1887 (Teubner).the first half

22 See The Library, cod. 117a (170). Ed. Rene Henry, Paris, 1960, vol. II, p. 162 (Les Belles Lettres).

23 Alcestis, 1.968.

24 Women of Trachis, 1. 45 and 1. 155.


26 Quoted by Ch. M. Fraehn, in Meroires de I’Academie de St. Petersburg, Sciences, politique, histoire et philologie, vol. Ill, p. 513. (See V. Gitermann, Geschichte Russlands, Zurich, 1944, vol. I, p. 232).

27 A.B. Arcihovskiy and B.I. Borkovskiy, Novgorodskie Gramoty na Breste, Moscow, 1958.

28 It is interesting to note, in this connection, that Charlemagne himself ( *814) used tablets when he was learning to write. (Eginhard, Vita Karoli Magni, 25. Ed. Louis Halphen, Paris, 1947. Les Belles Lettres.).

29 Documents in Mycenaean Greek, Cambridge University Press, 1956.

30 MycenaeansandMinoans, London, 1961, p. 58.

31 Aeschylus: The Suppliant Maidens, 1. 249.

32 Hdt., I. 57.

33 9.2.25 and 10.3.17.

34 IV.4.3. (Loeb, II, p. 351).

35 V.74. (Loeb, III, p. 297).

36 It is likely, therefore, that Saint Jerome did not invent, but only perfected, the first Slavonic alphabet. That an alphabet was used by the Veneti, a Slavonic people, in the 4th century, is attested by the Emperor Julian. (See his Discourses. Paris. Les Belles Lettres. 1932. I. 1st part, p. 143.)       Note on Slavonic palimpsests. At the last moment, I am being informed that a second palimpsest with a Glagolitic base is available in the Lenin State Library in Moscow, where it is catalogued as fund 87, M, 1960. It consists of 109 sheets with text from the Boyana gospel. – G.S.  .


Note of the author:  Tagged with & graphemes will be completed – in most cases these are signs of Mycenaean and Phoenician alphabet.



* Canadian Slavonic Papers / Revue canadienne des slavistes (CSP) was established in 1956. In 1967 it began appearing twice yearly, becoming a quarterly in 1968. The journal is the official publication of the Canadian Association of Slavists (CAS). 

** The entire study can be read here – SotiroffGlagotScript


5 коментара »

  • Pavel Serafimov said:

    В началото на 70-те години на ХХ век Георги Сотиров направи нещо уникално. Той не само показа древността на свещената българска писменост, но и с това даде доказателство за балканския произход на народа ни.

    Паралелите между глаголица и Линеар Б са доста повече, аз открих 33, просто днес има повече информация за древните егейски писмености.

    В плочиците от XIV век преди Христа са документирани типични български имена като Дейко, Мирул и др. По-важните имена от документите с Линеар Б са разбира се имената Ермаа и Вокило. Те съотвестват на старобългарските родови имена Ерми и Вокил.

  • Pavel Serafimov said:

    Ще добавя и това, че освен старобългарски родови имена, в микенските документи от XIV век преди Христа се срещат и имена като Крумен, Исевер, Кувир, Винай…те съотвестват на имената на старобългарските владетели Крум, Севар, Кубер, Винех.

    Ето и регистрационните номера (по A.Evans)на микенските документи съдържащи български имена:

    1197Ea 302(LXXXI)
    1127Ea 312
    534 a 23(XXIII)
    235 Nj 42 ( XXXVIII)
    473 Rq 61 (XCVII)
    770 Xe 45

  • Electronic Shopping Site said:

    Hello to every , for the reason that I am genuinely eager of reading this blog’s post to be updated daily.
    It consists of good stuff.

  • paul ponna said:

    Paul Ponna being a productive longtime Internet Marketer as properly as an experienced coach that has produced a lot
    of worthwhile software program packages like Income Hybrid being his latest to aid folks do a better job and preserve up to date with what is
    taking place in the World wide web Advertising globe.

  • Faustino said:

    I like this web blog very much, Its a really nice spot to read and
    find information.

Leave your response! Вашето мнение е важно, напишете го!

Add your comment below, or trackback from your own site. You can also subscribe to these comments via RSS.

Моля не ползвайте нецензурни изрази! Всеки коментар, в който има линк, ще изчака редактор да го провери за спам - забавя публикуването.

You can use these tags:
<a href="" title=""> <abbr title=""> <acronym title=""> <b> <blockquote cite=""> <cite> <code> <del datetime=""> <em> <i> <q cite=""> <s> <strike> <strong>

This is a Gravatar-enabled weblog. To get your own globally-recognized-avatar, please register at Gravatar.