Vinogradov D. A., Korobov M. Gothic graffiti from the Mangup basilica // NOWELE. 2018. Vol. 71. No. 2. P. 223-235

Gothic graffiti from the Mangup basilica Andrey Vinogradov and Maksim Korobov For more than a millennium there have been reports testifying to the presence of Goths in the Crimea. However, until a few years ago, the only evidence of a Gothic or Germanic idiom spoken in the peninsula stems from the list of words recorded between 1560 and 1562 by Ogier de Busbecq. Significant new evidence, however, has become available through the recent discovery of five Gothic graffiti scratched on two reused fragments of a cornice belonging to the early Byzantine basilica at Mangup-Qale in the Crimea. The graffiti, datable to between about 850 and the end of the 10th century, exhibit words in Gothic known from Wulfila’s Bible translation, the script used being an archaic variant of Wulfila’s alphabet and the only specimen of this alphabet attested outside Pannonia and Italy. There would seem to be evidence for assuming that, among educated Crimean Goths, Gothic served as a spoken vernacular in a triglossic situation along with a purely literary type of Gothic and with Greek in the second half of the 9th century. 1. Introductory remarks In 1938, excavations of the early Byzantine basilica at Mangup-Qale conducted by Russian archaeologist Maria Tikhanova revealed several fragments of a carved acanthus-decorated cornice lying on the pavement of the south-west chapel (s.c. baptisterium). On one of the fragments the researcher discovered several Greek graffiti. One of them was published by Tikhanova in 1953 (Tikhanova 1953: 386). She dated the graffito to the 5th or 6th century AD and remarked that it was surrounded by a number of shorter and almost illegible inscriptions. No photos or drawings of the graffiti were taken at this point, and the stone was thought to have been lost soon afterwards. The cornice was almost forgotten about after that, and only in 1987 did Valery Yailenko present his own, not quite correct re-reading of the Greek graffito published by Tikhanova (Yailenko 1987: 160–173). In 1999 the inscription was studied in absentia by Aleksandr Ajbabin who re-dated it to a later period around the 11th century (Ajbabin 1999: 123–124). https://doi.org/10.1075/nowele.00013.vin NOWELE 71:2 (2018), pp. 223–235. issn 0108-8416 | e-issn 2212-9715 © John Benjamins Publishing Company 224 Andrey Vinogradov and Maksim Korobov In September 2007 a carved cornice with the inventory number 1246/1577 was found in the lapidarium of the Bakhchisaray Museum-Preserve by Andrey Vinogradov who was working there as a part of the project “Inscriptiones orae septentrionalis Ponti Euxini”.1 It fitted the description by the earlier researchers, bearing several graffiti in Greek, yet none of the graffiti was identical to the one published by Tikhanova and Yailenko. In August 2013 Vinogradov discovered another fragment of the same cornice with the inscription published by Tikhanova at the repository of the Mangup excavations of the same museum where it was being kept without an inventory number. Finally, in March 2016 the two fragments were inspected in person by the authors of the present paper (Vinogradov and Korobov). The carvings on the cornice date it quite unambiguously to the early Byzantine period. Obviously it was once part of a large church. Remembering the monumental building programme on the Mangup mountain under Justinian I it seems probable that the only church big enough, the basilica, was erected there in the mid-6th century.2 The upper sides of both fragments of the cornice are covered with miscellaneous graffiti, drawings (circles, crosses, monograms, snakes etc) and inscriptions, some of which are heavily blurred and therefore illegible. We found six readable Greek graffiti on fragment I,3 and three on fragment II.4 They are mostly short invocations with an appeal to the Lord/God and the Saints for help, and two commemorative namelists (on stone I), which are all very characteristic of Byzantine epigraphy of the Mountainous Crimea. Noticeably, all three inscriptions on stone II were made exclusively by clerics suggesting that this fragment of the cornice was re-used in the sanctuary. Some of the texts are written in minuscule, which in fact places their terminus post quem in the early 9th century; others are in uncial, and in these the letter B has an elongated lower stroke – something that is not normally attested beyond the 9th-10th centuries. So far, the story of the cornice from Mangup can be reconstructed as follows: Around the middle of the 6th century the cornice, which was obviously larger than it is today, was inserted as an architectural detail in the Mangup basilica. Sometime before the end of the first millennium two fragments of the cornice were secondarily re-used in an outer wall, perhaps in the apse of the same basilica which had been reconstructed after the end of the Khazar dominance over Mangup and the founding of the Byzantine theme (around the year 840). After 1. 2. 3. 4. http://iospe.kcl.ac.uk/project/index.html On the time frame of the basilica, see Barmina 1995, Kirilko 1997. http://iospe.kcl.ac.uk/5.193.html http://iospe.kcl.ac.uk/5.183.html Gothic graffiti from the Mangup basilica the destruction of the basilica the fragments of the cornice were removed from the wall to the floor paving over the graves in a side-chapel of the basilica. Accordingly, the graffiti on the fragments of the cornice could have been made between the middle of the 9th and the 13th century, and the paleography of some of the Greek inscriptions dates them more precisely to the period between the second half of the 9th and the 10th century. 2. The Gothic graffiti A careful examination of the digital photos of the stones revealed that, beside the Greek graffiti, there are some more inscriptions written in a different kind of script that combine both Greek- and Latin-like characters and characters of a quite uncommon form. A combination like this could only be possible in the Gothic alphabet of Wulfila. An autopsy of the stones, which we undertook in March 2016 in the lapidarium of the Bakhchisaray Museum-Preserve, was supportive of our original guess and, moreover, it helped us revise and significantly better our first readings.5 Five inscriptions written in the Gothic alphabet have been discovered so far: four on fragment I and one on fragment II.6 It should be pointed out that on both stones these inscriptions are predominantly placed in positions most favourable for writing, that is at the top and left sides. More importantly, in every case they precede the adjacent Greek inscriptions and are occasionally covered with these like the under-text of a palimpsest. But at the same time the lines in Gothic were not always the first thing to have been engraved upon the stones, as we can see in the case of the Gothic inscription I.1 which is interrupted by a previous monogram. The forms of Gothic letters in our inscriptions are, on the whole, very close to those of extant Gothic manuscripts, the differences being mostly due to the problems of scratching them upon the stone plate. However, some minor peculiarities are to be noted. The letter J has a much longer and lower stroke compared to the J of the manuscripts, and in at least one case it resembles more the Germanic rune *jēr, with its second stroke starting from the middle of the semicircle formed by the first stroke rather than from its lower end as was the case with the “bookish” J. 5. See our preliminary publication Vinogradov & Korobov 2015. 6. A modern forgery should be ruled out for several reasons. Firstly, the two stones were kept separately all the time from the moment of their discovery. Secondly, the Gothic inscriptions were carved by different hands and are partly covered with Greek majuscule and minuscule graffiti which are still more difficult to counterfeit. Thirdly, all the graffiti show the traces of a long existence (attritions and gaps). 225 226 Andrey Vinogradov and Maksim Korobov The letter A is formed with three strokes, and the letter S has a form of Greek ε or Σ. In this as well as in several other features the alphabet of the Mangup inscriptions follows the variant of the Gothic alphabet which is sometimes referred to as “Script I” (or “Σ-type”) by researchers and is attested in a number of Gothic manuscripts, such as Codex Ambrosianus B, the recently discovered Gothic fragment from Bologna, the Deed of Naples, and the Tablet of Hács-Béndekpuszta (Pannonia). It is generally thought to be older and closer to Wulfila’s original design than the better attested variant of Codex Argenteus with its Latin-like S (Marchand 1973: 15–16; Braune/Heidermanns 2004: 19–20). The letter W appears in two forms: in inscriptions I.1 and I.3 it is identical to the ypsilon of the Greek majuscule (Υ), whereas in other instances it looks more У-shaped. In inscriptions I.2, I.4 and II.1 some letters seem to be written with a slight slant to the right. The words guþ ‘God’ and frauja ‘Lord’ are always written in abbreviated forms (gþ and fa) with a horizontal stroke set over the letters (easily visible in inscriptions I.1 and I.2). The same way of writing nomina sacra is used throughout the Gothic manuscripts, and this testifies to the fact that the writers of the graffiti were familiar with scribal conventions of literary Gothic. Below we will present our readings of the five Gothic inscriptions and provide some comments on their content and vocabulary. Every inscription is referred to by its number which consists of the number of the fragment (I or II) followed by the sequence number of the inscription on the stone (counted from left to right and from top to bottom). First comes a transliteration in Latin letters, then our reading of the texts in Gothic (a transcription), and then an English translation. In the transcription, letters that did not survive physically but can be surmised from the context with a reasonable degree of confidence are added in square brackets. Round brackets contain letters that were dropped in conventional abbreviations and several other instances. I.1 This is the longest inscription, located on the left side of fragment I where it occupies almost the whole of the surface. 1. ǶASGÞMIKILS 2. SWEGÞUNSARÞU 3. ISGÞWAURKJANDS 4. SILDA vacat7 LEIKA vacat AINN 5. USSTOÞ 6. UNDAIWINS 7. USDAUÞAIM 8. JAHIN vacat MIDJUN Gothic graffiti from the Mangup basilica 1. 2. 3. 4. 5. 6. 7. 8. ƕas g(u)þ mikils swe g(u)þ unsar þu is g(u)þ waurkjands sildaleika. ainn[s] usstoþ und aiwins us dauþaim jah in midjun[gard-] ‘Who is so great a God as our God? Thou art the God that doest wonders’. (KJV) ‘(One) rose for ever from the dead and in(to) the world…’ The inscription is clearly written by a single hand and consists of two parts. The first four lines contain the second half-verse of Psalm 76:14 and the next halfverse of Psalm 76:15. The Gothic text is completely parallel to the Greek text of the Septuagint (τίς θεὸς μέγας ὡς ὁ θεὸς ἡμῶν; σὺ εἶ ὁ θεὸς ὁ ποιῶν θαυμάσια) which it renders word-for-word, using the same Gothic translations of the Greek words as found in the extant fragments of the Gothic Bible. The only exception is the word sildaleika, chosen to render Greek θαυμάσια, the plural of θαυμάσιον ‘wonder’. A Gothic counterpart of this Greek noun is not attested elsewhere in the Gothic corpus, but in the Gothic Bible the adjective sildaleiks regularly translates Greek θαυμαστός ‘wonderful’. Interestingly enough, the initial letters in the first four lines, if read as a Greek acronym, form the sequence theta-sigma-iota-sigma, which is short for Θ(εὸ)ς Ἰ(ησοῦ)ς ‘God Jesus’ and found as such elsewhere in Byzantine epigraphy. Whether it was really intended to be so understood by the bilingual author of our graffito or just occurred accidentally, we cannot know. But what we see in the lines below seems to support the former possibility. To the right of sildaleika, at a distance, we see something like Gothic ainn-. At first we assumed that it is a leftover from the word gakannides ‘you made known’ or ‘you declared’, corresponding in the manner described above to Greek ἐγνώρισας, the next word in Psalm 76:13. But there are two objections here: first, the letter I immediately following A, and second, the absence of the other letters of the supposed Gothic word on a relatively well-preserved surface where traces of 7. The lacuna in the middle of the word sildaleika was caused by an earlier monogram. A Greek inscription runs across the monogram, most likely the word κύριε (‘Lord’ in the vocative). The Greek kappa covers the Gothic D so that one could erroneously think that it’s not D but rather Þ. However, on closer inspection, we can discern a part of the curved right stroke and the diagonal stroke of D. 227 228 Andrey Vinogradov and Maksim Korobov other graffiti are easily discernible. A better suggestion would probably be Gothic ains, that is ‘one’ or ‘single, unique’, with a dittography of N and a loss of the final S. This ainn[s] is then to be placed in the second part of the inscription. In the fifth line we read usstoþ, the third person singular preterite of the verb usstandan, which in the Gothic Bible translates Greek ἀναστῆναι ‘to rise, to stand up’. The two last letters of the word, and especially the O, are blurred and could only be recognized with confidence during the autopsy. The next line begins with a capital U, of which only two vertical strokes are clearly visible. What follows are the quite legible letters N, D and A. The remainder is again covered with a later Greek graffito (alpha with a cross on the top), and through it, one can still see something that resembles the Gothic letters I and W, positioned very close to each other. The entire line can so far be read as und aiw – an expression which translates Greek εἰς τὸν αἰῶνα ‘forever’ (literally ‘into eternity’) in Luke 1:55. The stone plate to the right of the W is noticeably damaged. One can see traces of another overlying graffito – omega with a superimposed cross, almost of the same size as the alpha. Between these two figures a vertical stroke is discernible, its size matches the size of other Gothic letters in the line. It is likely to be the left stroke of Gothic N. And if we are to assume here some further letters, we could probably think of Gothic und aiwins that would correspond to Greek εἰς τοὺς αἰῶνας, with aiwins/αἰῶνας in the accusative plural. Yet this Gothic expression is unattested: its Greek counterpart was translated by Wulfila as in aiwins in Matthew 6:13 (at the end of The Lord’s Prayer). Line seven is relatively well-preserved and reads us dauþaim ‘from the dead’. It equates to Greek ἐκ νεκρῶν in Wulfila’s translation. The following line starts with jah in ‘and in(to)’ and, after a short lacuna, midjun-. The only known Gothic word that would fit in here is midjungards ‘world, universe’. We cannot determine the grammatical case of this noun since its final part is lost, so our translation must remain ambiguous at this point: ‘in the world’ or ‘into the world’, assuming dative or accusative case, respectively. There is a third possibility after the preposition in – the genitive, but the meaning of it would be ‘because of ’, and this hardly makes sense in our context. The surface below where the Gothic text would continue is badly damaged, and besides, it bears a later Greek inscription (which is a commemoration and consists of seven personal names in the genitive) that starts to the right of the fifth line of our inscription, thus avoiding the text of the Gothic graffito. This could imply that no Gothic had been written here. So, in our opinion lines five, six, seven and eight contain a previously unattested text “(One) rose for ever from the dead and in (or: into) the world…”. This sounds very much like an Orthodox Easter hymn. Two possibilities exist: either Gothic graffiti from the Mangup basilica it was translated from Greek8 (though no such Greek text has been discovered up to the present time), or it was originally composed in Gothic, maybe even ad hoc. In favour of the latter assumption speaks our observation that the text is poetic. Notice the assonances I u ~ II u ~ III u, II ai ~ III ai, and perhaps I o ~ III au, if au had already been monophthongized in the author’s dialect, and the alliteration II d ~ III d in un-’dˬaiwins | usˬ’dau-þaim. Additionally, the three lines represent a tautogram, that is, they all start with the same letter, U. And if we include the word ain(n)s into the text, it is divisible in two longer stanzas: ains usstoþ und aiwins | us dauþaim jah in midjungard(a). One may notice the vowel alliteration ains : aiwins, and the internally rhyming pair usstoþ [us:to:θ] ~ us dauþaim [us do:θe:m], if pronounced with monophthongs (but it must be borne in mind that we do not know the date of monophthongization in Crimean Gothic nor even its precise results). We think that the choice of the texts for the inscription can be explained solely from a liturgical point of view. Psalm 76:14–15 is widespread in the Orthodox church service where it functions as the so called Great Prokeimenon. It is regarded as very solemn and is performed at Christmas and Easter vesper. So, what we have here are two texts that seem to pertain to the Easter service, and this allows us to conclude that the inscription was written by a cleric. The writing of the psalm glorifying the true God can historically be linked to the restoration of the basilica in the mid-9th century and the return of the Christian power to Mangup. I.2 This inscription consists of a single line located at the upper side of fragment I running from its middle up to the right end. †FAHILPSKAL[..]SÞE[…]SÞISWI f(rauj)a hilp skal[ki]s þe[ini]s þis w(e)i[hins ?] ‘Lord, help your servant the … (priest ?)’. The text is a formulaic invocation, which is typical of Byzantine epigraphy of all periods. A Greek inscription nearly identical in content is written to the left on the same stone.9 The capital form of the initial letter F reminds one once again of the Gothic manuscripts. In Byzantine invocations which bear the same formula it is usually followed by a personal name, more rarely by a priestly title (we have two examples with “presbyter” and one with “archdeacon”).10 But here the title, or what we think is a title, is preceded by þis, the genitive singular of 8. Hymns are attested in Crimean Greek epigraphy, see e.g. http://iospe.kcl.ac.uk/5.305.html (#2). 9. http://iospe.kcl.ac.uk/5.183.html 10. Clara Rhodos 1, 103; Jerphanion 172; Haspels 77. 229 230 Andrey Vinogradov and Maksim Korobov the demonstrative pronoun sa, which in Biblical Gothic functioned also as the definite article. The first letter is fairly large, but from the second letter onwards the letters become almost half-sized, probably because the scribe was trying to fit the text into the surface that was left for writing. The initial Þ is easily legible, followed by I and S. A dent on the stone was already there and caused a greater space between the letters Þ and I. From the next word only two initial letters survived: W and I. Remembering the situation in Greek inscriptions mentioned above, we suggest that this is the beginning of the word weiha ‘priest’ (lit. ‘the holy one’, Greek ἱερεύς) in the genitive singular weihins, with an apparent shortening of the vowel in the first syllable. Another possibility is that the word was a designation of the writer’s occupation, e.g. weinagardja ‘vinedresser’. This word as such is not attested in Wulfila’s Bible, but may be reconstructed as a derivative of weinagards (Gr. ἀμπελών) ‘vineyard’ (cf. in-gardja οἰκεῖος from gards οἶκος, οἰκία).11 Of course, the writing of -i- instead of -ei- could just as well be a scribal error (or variant), but something very similar to the same shortening in a word that we reconstruct as weinagardja is seen also in inscription I.4, on which we will comment later. If our reconstructions are correct, we could perhaps speak of an observable phonetic change in Crimean Gothic as compared to the language of Wulfila’s Bible, namely the shortening ī > ĭ after word-initial w-. I.3 Just below I.2 we see another line in Gothic, very short and consisting of only four characters. WIND The first three characters are well recognizable. The initial W is written very large and ornamentally. The final letter is not easily legible, but the surviving traces would seem to resemble Gothic D. We do not understand the meaning of this. Most likely it was a personal name, or a noun designating the profession/origin of the scribe. I.4 This inscription runs in three lines from the middle of the right side of fragment I. 1. FAHILPSKALKISÞEI 2. DAMJA[..]US vacat USWINAG 3. JAHFRAWAURTIS 1. f(rauj)a hilp skalkis þei[nis] 2. damja[na]us us w(e)inag[ardjam ?] 3. jah frawaur(h)tis 11. For Crimean Gothic Busbecq lists wingart = Lat. vitis ‘grapevine’. Gothic graffiti from the Mangup basilica ‘Lord, help your servant Damian from Winag-/of the vinedressers (?) and the sinner’. The reading of this inscription is significantly hampered by the amount of damage to the stone plate. In the first line we see again the invocation frauja hilp skalkis þeinis ‘Lord, help your servant’. The second line opens with a personal name which we read as Damjanaus in the genitive, the nominative being *Damjanus. The name is a clear borrowing from Greek (Δαμιανός), and especially interesting is the fact that this was the name of the metropolitan of Gothia mentioned in a Greek inscription dated to the year 1427.12 For the rendering of Greek -ια- with Gothic -ja- in biblical names, compare Μαρία > Marja (along with Maria), and many more examples. The genitive singular ending -aus is only partly preserved. After it, there is a short lacuna, and then we read us winag-. Us is a Gothic preposition meaning ‘from, of ’. Hence winag- could be either a place name, probably Damian’s homeplace, or his professional name. The word is strongly reminiscent of Gothic weinagards ‘vineyard’. Wilhelm Tomaschek (1881: 59) once speculated that this was the Gothic name for the entire Crimea. As for a professional name, we think it could be weinagardja ‘vinedresser’ in the dative plural: damjanaus us weinagardjam ‘Damian of the vinedressers’. Of course, we cannot be sure of this, but if our reconstruction is correct here too, this is another example of the assumed shortening of -ī- before w-. In the third line we read jah frawaurtis ‘and sinful’, with the adjective in the masculine genitive singular. The lacking h in the word frawaurtis is, compared to the form of this word in the Gothic Bible, no mere mechanical error, for it is the bible that gives us examples where h is dropped in the cluster -rht- as well as examples of the opposite case where it is added as a hypercorrection (Marchand 1973: 53–54). Both phenomena are probably indicative of the writer’s actual pronunciation, i.e. that this cluster had already been simplified. II.1 This is the only inscription on fragment II. Its beginning is absent. ISJAHFRAWAURTISDAMJAN -is jah frawaur(h)tis damjan[aus] ‘of (unworthy ?) and sinful Damian’. The surviving text starts with a short vertical stroke, most likely Gothic I. In the next letter one can recognize Gothic S, with its upper part lost in a dent. From the third letter on, we can confidently read jah frawaurtis ‘and sinful’, with a visible 12. IOSPE3 V 241: … παρὰ τοῦ π[α]νιε[ρωτάτ]ο̣υ μ(ητ)ροπολίτου πόλεως Θεοδώρου[ς] κ[αὶ] πάση[ς Γοτ]θ̣ίας κυροῦ Δαμιανοῦ (http://iospe.kcl.ac.uk/5.241.html). See also IOSPE3 V 244 (http://iospe.kcl.ac.uk/5.244.html). 231 232 Andrey Vinogradov and Maksim Korobov ending -is in the genitive. This enables us to suggest that -is of the preceding word was such an ending too, and that we are dealing here with a pair of epithets. We can further suggest that it was a self-deprecating courtesy formula, something like *unwairþis jah frawaurtis ‘unworthy and sinful’. Such formulas do occasionally appear in the Greek inscriptions of the Black Sea region.13 It must remain obscure whether both genitives were governed by the imperative hilp which we saw in two previously discussed inscriptions (i.e. whether it was another appeal for divine help). After frawaurtis four letters are badly damaged. We were able to read them only during the autopsy. They are D, A, M, and J. The J has lost much of its longer lower stroke so characteristic of this letter in our inscriptions. Final A and N at the end of the line are relatively well preserved and easily recognizable. This is the same name, Damjanus, but its ending – the genitive singular ending of u-stems – is lost. Note also the spelling of frawaurtis without -h-, just as in inscription I.4. 3. Remarks on the language of the graffiti From a purely linguistic point of view, the new Gothic texts presented here are strikingly conservative both in their vocabulary and grammar. This is probably due to the continued existence of a scribal tradition of written Gothic in the Crimea which preserved the “sanctified” literary standard of some 500 years earlier, whereas the actual spoken language of the Crimean Goths was naturally bound to change with the course of time. In our view, it is permissible to speak of triglossia in Crimean Gothia, where Greek, Wulfilan “High” Gothic, and a spoken Gothic vernacular coexisted, fulfilling different sociolinguistic roles. The influence of the vernacular on the written language is exemplified by the assumed shortening of the vowel in wei- [wi:] > wi- in inscriptions I.2 and I.4 (see above)14 and the simplification of the consonant cluster -rht- > -rt-. This influence must have been rather strong to cause the breach of standard orthographic rules. As for the phraseology of the inscriptions, the first thing to note is their obvious and strong dependence on Greek patterns. For example, the invocations 13. E.g. τὸν ἁμαρτωλὸν καὶ ἀνάξιον δοῦλον in a Greek inscription of the 9th–10th century from Cherson (http://iospe.kcl.ac.uk/5.47.html). 14. We have no independent evidence, alas, for this phonetic development in Crimean Gothic. The spelling of Crimean Gothic words by Busbecq is notoriously far from being straightforward and consistent, and besides, his Greek informant could not possibly reproduce the distinction between long and short vowels even if such a distinction still existed in Crimean Gothic (see Stearns 1978: 55). Thus, CGo. wingart is not informative. Common Slavic *vinogordъ borrowed from Gothic presupposes a long vowel in the source language, but this loan dates from a much earlier period than our graffiti. Gothic graffiti from the Mangup basilica in I.2 and I.4 are in fact literal renderings of Byzantine invocations. In this respect too, the Mangup Gothica follow the lead of Wulfila who was trying to imitate in Gothic the Greek text of the Bible as closely as possible in both syntax and choice of words. Most of his translations are adhered to in these new texts, e.g. hilpan = βοηθεῖν, not mentioning the direct quotation of a psalm in I.1. 4. Conclusions As noted previously, the graffiti cannot be dated earlier than the middle of the 9th century, and this dating accords well with the paleography of Greek inscriptions on the same stones. The newly found Gothic inscriptions thus support a prolonged survival of the Gothic language in the Mountainous Crimea up to the 9th–10th centuries as well as its coexistence with Greek. In view of the fact that there are no major differences between the Gothic of the Mangup graffiti and the language of the Gothic Bible we must take the first as a purely literary idiom which was in use for varied purposes: for writing liturgical pieces as well as for private invocations and commemorations. Both Greek and literary Gothic must have been coexistent with a Gothic vernacular as recorded by Busbecq, resulting in a situation of triglossia among educated Crimean Goths. For the first time we see the spread of Wulfila’s alphabet from the Lower Danube area not only westwards but also eastwards into the Crimea and, more generally, its presence outside Italy and Pannonia as well as its practical use after the 6th century. The quotation from the Psalter in I.1 presupposes the familiarity of the Crimean Goths with Wulfila’s Bible and, specifically, with its almost completely lost Old Testament part. The knowledge of the Gothic script and Scripture makes it plausible that the Crimean Goths of the time belonged to the Gothic cultural world; a link to the West might have been provided by the Goths of the Lower Danube area who were reported by Walahfrid Strabo to perform their church service in Gothic as late as in the 9th century. The fragments of liturgical texts in the Mangup inscriptions testify to the existence of a Gothic language church service of the Byzantine Rite, and this corresponds with the existence of the Byzantine archbishopric of Gothia whose clerics were ethnic Goths or speakers of Gothic.15 Finally, the near-identity of the invocation formulas used in the Greek and Gothic inscriptions demonstrates the closeness of the epigraphic practice of the Crimean Goths to the local Byzantine practice. 15. The archbishopric of Gothia in the Crimea was mentioned for the first time in the middle of the 8th century and must have been founded after 692, i.e. shortly before the time-frame of our graffiti. 233 234 Andrey Vinogradov and Maksim Korobov Acknowledgements The results of the project “Centres” and “Peripheries” in Medieval Europe carried out within the framework of the Basic Research Program at the National Research University Higher School of Economics (HSE) in 2017 are presented in this work. References Ajbabin, A. 1999. Etničeskaja istorija rannesrednevekovogo Kryma [Айбабин А. И. Этническая история раннесредневекового Крыма]. Simferopol: DAR. Barmina, N. 1995. Mangupskaja bazilika v svete nekotorych problem krymskogo srednevekov’ja [Бармина Н. И. Мангупская базилика в свете некоторых проблем крымского средневековья]. Antičnaja drevnost’ i srednie veka 27. 77–84. Braune, W. 2004. Gotische Grammatik mit Lesestücken und Wörterverzeichnis. 20. Aufl., neu bearb. von Frank Heidermanns (Sammlung kurzer Grammatiken germanischer Dialekte A 1). Tübingen: Max Niemeyer. https://doi.org/10.1515/9783110945089 Kirilko, V. 1997. Vizantijskaja architektura Mangupa [Кирилко В. П. Византийская архитектура Мангупа]. In Archeologija Kryma 1. 89–97. Simferopol: Yuzhnogordkie vedomosti. Marchand, J. W. 1973. The sounds and phonemes of Wulfila’s Gothic (Janua Linguarum, Series Practica, 25). The Hague: De Gruyter Mouton. https://doi.org/10.1515/9783110828054 Stearns, MacDonald Jr. 1978. Crimean Gothic: Analysis and etymology of the corpus. Saratoga: Anma Libri. Tikhanova, M. 1953. Bazilika [Тиханова М. А. Базилика]. 334–389. Materialy po archeologii Jugo-Zapadnogo Kryma (Materialy po istorii i archeologii SSSR. Nr. 34). Tomaschek, W. 1881. Die Goten in Taurien. Wien: Alfred Hölder. Vinogradov, A. & Korobov, M. 2015. Gotskie graffiti iz Mangupskoj baziliki [Виноградов А. Ю., Коробов М. И. Готские граффити из Мангупской базилики]. Srednie veka 76:3–4. 57–75. Yailenko, V. 1987. O korpuse vizantijskich nadpisej v SSSR [Яйленко В. П. О корпусе византийских надписей в СССР]. Vizantijskij vremennik 48. 160–173. Address for correspondence Andrey Vinogradov Centre for Medieval Studies, School of History National Research University “Higher School of Economics” 21/4 Staraya Basmannaya ul. RU-105066 Moscow [email protected] Gothic graffiti from the Mangup basilica Co-author information Maksim Korobov Independent researcher [email protected] 235