A Byzantine Poetic Form in a Ninth-Century Bulgarian Poem A Byzantine Poetic Form in a Ninth-Century Bulgarian Poem Ekaterina Dikova*1 Abstract. The paper measures the extent to which the most characteristic features related to the so-called Byzantine dodecasyllable are applied in one of the earliest Old-Bulgarian poems – Azbuchna molitva (‘Alphabetic Prayer’) noted to be written in dodecasyllabic verses. This alphabetic acrostic is dated back to the very end of the ninth century and is attributed to Constantine of Preslav. In this article its text is given after its earliest copy, MS Syn. 262, as it is the only representative of the ver- sion closest to the Glagolitic archetype, now lost. The piece is studied in comparison with St Gregory the Theologian’s alphabetic acrostic (as published in PG 37) which Constantine of Preslav quotes just after the end of his poem and which is considered its rhythmical model. The main conclusions are that the Alphabetic Prayer is an early replica of the Byzantine dodecasyllable, follows its rhythmical peculiarities to an extent similar to St Gregory’s alphabetic acrostic, all the previously supposed deviations are motivated by genre peculiarities and rhetorical requirements, which reveals Byzantine schooling of the Old-Bulgarian writer. Nevertheless, the content and intention of the poem indubitably target the neophyte Slavonic audience. Keywords: metrical analysis, medieval Christian poetry, Byzantine iambic trimeter, alphabetic acrostic, Old Bulgarian versification, cultural influence in spiritual literature * Author’s address: Ekaterina Dikova, Institute of Balkan Studies and Centre of Thracology – Bulgarian Academy of Sciences, 45 Moskovska Street, Sofia 1000, Bulgaria. Email: e.dikova@ balkanstudies.bg. Studia Metrica et Poetica 9.2, 2022, 63–91 https://doi.org/10.12697/smp.2022.9.2.02 mailto:e.dikova@balkanstudies.bg mailto:e.dikova@balkanstudies.bg https://doi.org/10.12697/smp.2022.9.2.02 64 Ekaterina Dikova � Херовьскоу�м•�мꙑсль��оумъ�даждь�⁘ Ѡ�ьстьнаꙗ•�прѣсвѧтаꙗ�троце�⁘ Cherubic reasoning and mind provide me with, O, venerable you, most holy Trinity.1 (Constantine of Preslav) In the late ninth century, Constantine of Preslav, one of the most prominent earliest Old-Bulgarian writers, translated a set of catenae (in combination with homilies2) related to the liturgical Sunday Gospel readings, added his own introductions and conclusions to each of the orations as well as a whole oration of his own, to form the codex know as Didactic Gospel (Uchitel’noe Evangelie)3. In his introduction to the whole book, he also inserted a prayer to the Holy Trinity which he wrote in dodecasyllabic verses forming an alphabetic acrostic (thirty-six verses after each Glagolitic letter except for the jers4 plus a final doxology5). He named it ‘a measured prologue about Christ’ (Прологъ� о�Хѣ҃�оумѣренъ). The peculiarities of the original Byzantine dodecasyllable – known to its contemporaries as iambs (ἴαμβοι6) – according to the scholars who dedicated studies to it, are the exact number of the syllables, caesura after the fifth or the 1 Translation is mine and I tried to keep twelve syllables per verse. In Orthodox Christianity cherubs are awe-inspiring formidable creatures while on the West they usually have gentle and innocent child-like representations. Constantine of Preslav had certainly the first in mind. For an English translation of the whole poem see for example Butler 1999–2022. 2 See for instance Gorskij, Nevostruev 1859: 423–424. Kotova (2022) and Petrov (2022) find the previously unknown Greek parallels of parts of orations nineteenth and twentieth, respec- tively, in St John Christostom’s homilies. See also Mitov (2022). 3 The monument is dated usually to 889–893 (cf. Arhim. Antonij 1885: 7; Gallucci 2001: 3–4; Spasova 2005; Tihova 2012: XII; Slavova 2017: 3) and the literature cited there. 4 These are ъ�and�ь�(Glagolitic�ⱏ�and ⱐ respectively). They were pronounced in the ninth century, but later some of them were not, others transformed into other vocals. (In later times, according to the rules of the Church Slavonic grammar, the jers just marked the softness or hardness of the consonant before them). 5 On the acrostich in this poem see for instance the recent studies of MacRobert (2019) and Kojčeva (2019) and the literature cited there. 6 For the labels of poems given in their headings in Byzantine manuscripts see Rhoby 2015. Paul Maas (1903: 278), in the very first sentence of his famous study, states that the type of verse in his focus is known as Byzantine iambic trimeter. 65A Byzantine Poetic Form in a Ninth-Century Bulgarian Poem seventh syllable, particular stress patterns before the clausula and at the end of the verse as well as specific visual prosody. The purpose of the present article is to check to what extent the metri- cal principles of the so-called Byzantine dodecasyllable are followed in Constantine of Preslav’s Alphabetic Prayer. It begins with a brief review of the previous studies. Then the poem is analysed in juxtaposition to St Gregory the Theologian’s alphabetic acrostic which has been considered to be its rhyth- mical model (even though the latter is much shorter and quite different in its content and intention). Previous studies on the meter of Azbuchna molitva The meter of Azbuchna molitva has interested scholars since the very first study dedicated to the poem in 1900, even though the main focus has been on the acrostich and, initially, also on the authorship of the work. Aleksej Ivanovich Sobolevskij (1900: 314) determined it to be a twelve-syllable “politi- cal” verse with caesurae only after the fifth syllable and paroxytone clausulae. Ivan Franko (1914: 162, 163) counted various syllables per verse (from 8 to 12, which is definitely due to the fact he did not count the jers, as required by the rules of his contemporaneous Church Slavonic), and defined the meter as ten-syllable trochaic structuring of epic Slavic folk songs. Emil Georgiev (1938: 114–123), in his profound study of the prayer, saw a twelve-syllable meter with specific caesuring but underlined that it did not follow the Greek rhythmical system, that there were no paroxytone verse endings and no iamb in it.7 Rajko Nahtigal (1942: 51–52) was definite that the work was written in the Byzantine iambic trimeter in twelve-syllable verses with caesura after the fifth or the seventh syllable, but considered its prosody Slavonic.8 Kujo Kuev (1974: 119, 7 He rised the question of prosody but quickly passed it away since, as, he argued, Byzantine Greek had lost the difference between long and short vowels and there had been no such dif- ferences in Slavonic. (And yet, Czech and Serbian languages do keep such differences up to nowadays.) 8 Nahtigal’s main contribution is the reconstruction of the acrostich; he also corrected the length of some verses and some places of the caesurae. The precise quotation about prosody is, “Vendar starocerkvenoslovanski verzifikator ni ne mene dolžin in kratčin v kvantitativen oziru, ne mesta naglasa v ritmičnem metrično izrabil, to je ustvaril svojo slovansko prozodijo.” ‘However, the Old Church Slavonic poet did not use metrically neither the alternation of length and shortness in quantitative terms, nor the position of accents in rhythmical terms – he created his Slavonic prosody.’ 66 Ekaterina Dikova 128, 132), who dedicated a whole monograph to the Alphabetic Prayer, noted its twelve-syllable verses with caesura after the fifth, sixth9 or seventh syllable, underlined that its structure is related to the Byzantine dodecasyllable and had as its model St Gregory the Theologian’s alphabetic acrostic. It is worth mentioning here that all these scholars proposed their own reconstruction of the text to approximate it to their understanding of its meter. Ivan Bogdanov (1980: 60), as it seems, held rather Sobolevskij’s view in his short description of the rhythmical peculiarities of the Alphabetic Prayer as he stated twelve syllables per verse caesured only after the fifth syllable and saw all the deviations as exceptions caused by semantic reasons or copyists. Ivan Dobrev (1993: 239–257) analysed the place of the caesura to demonstrate that this type of dodecasyllable with asymmetric alternation (5+7 or 7+5) differed significantly from the dodecasyllable of the Bulgarian folk songs (6+6); he added to the analysis of his predecessors the count of accents per verse.10 The most recent scholar – and probably the most profound – of early Old Bulgarian non-liturgical poetry, Krasimir Stanchev (1986: 646, 652), summarised the view accepted in scholarship that the work was composed of twelve-syllable verses with a caesura after the fifth and rarely after the seventh syllable. He emphasised that this rhythmic peculiarity was adopted from Byzantine poetry as “a medieval modification of the ancient iambic trimeter” which meant a twelve-syllable line with a caesura after the fifth, sixth or seventh syllable (with the clarification that neither a different caesuring nor a syllable count of more than twelve syllables was a deviation from it). Metrical analysis I have decided to analyse the raw text of the earliest preserved copy (MS Syn. 262, ff. 1–3) in order to avoid the situation of drawing conclusions over a text that does not exist except as a reconstruction. Besides, this manuscript, accord- ing to text-critical studies, is the only one closest to the Glagolitic archetype, 9 The structure 6+6 is in the twelfth verse only, according to MS Syn. 262, if one does not accept R. Nahtigal suggestion that instead of Летть... (‘is flying’) the line began, as he logically suggested, with ꙉ.�лѣтъ (‘thirty years’). 10 He does not comment on the number of accents per line, but it is clearly visible there that various isocollic patterns alternate together with the alternations of meaning as is the case in rhetorical prose as well. 67A Byzantine Poetic Form in a Ninth-Century Bulgarian Poem while the rest of the copies11 (about fifty in total) have no direct relation to it.12 I juxtapose this early Bulgarian text to St Gregory the Theologian’s alphabetic acrostic (poem I.2.30 after PG 37 col. 908–910)13 which has been assumed to be the rhythmical model of Constantine’s Alphabetic prayer by the majority of the scholars, who dealt with its poetic meter, probably because it is cited14 straight after the end of the prayer. I have been wondering whether to include at least some of the reconstruc- tions but have decided not to, as they, even if convincing, remain in the sphere of hypotheses. It is also a matter of personal choice which of the several recon- structions of a verse to accept. One thing is sure – if any of those is indeed true to the archetype, the percentage of the dodecasyllabic peculiarities, given below, would be even greater. 1. Number of syllables The term Zwölfsilber coined by Maas is certainly related to the most obvious peculiarity of this most popular Byzantine meter. And the same peculiarity – prevailing twelve-syllable verses – was noted by almost all the scholars who analysed Azbuchna molitva. Here, I count, again after earlier scholars, the number of syllables per verse in Constantine’s prayer (according to its text in MS Syn. 262) and compare it with the number of syllables per verse in St Gregory’s parenetical alphabet (according to its edition in Patrologia Graeca). 11 Only Cyrillic transcriptions came down to us. 12 See Kuev’s (1974: 168) graphic representation of the text-critical relations between the manuscripts; for the textual evidence which emerged since then and their grouping see Veder (2000: 78–80) whose edition of the poem includes variant readings of the lines according to the respective groups of manuscripts, his reconstruction in Glagolitic and his translation in English (Veder 1999: 61–88). 13 There is another alphabetic acrostic, again a paraenetic one, by St Gregroy (though ascribed to Ignatios the Deacon in various manuscripts), and it, together with 32 poetic and 2 more prosaic ascetic-paraenetic alphabets of various times, is proven to be genealogically related to the same St Gregory’s acrostic which Constantine quotes. For the similarities in both form and meaning of all those Byzantine acrostics see Anastasijević (1905: 4) who also mentions other Byzantine acrostic prayers (ibid.: 3). 14 The exact quotation after MS Syn 262, f. 3r is: Добро�сть�отъ�б҃а�нанат•��до�б҃а� коньават•�ꙗкоже�рее�етеръ�б҃ословьць��гргоръ•�‘It is good from God to begin and to God to end as that theologian and Gregory said’ (translation is mine; the ‘and’ in the Old Bulgarian text points at a literary understanding of the name – ‘watchful’). 68 Ekaterina Dikova The only intervention in the manuscript text I have allowed myself, which is related to the number of syllables, is the addition of ь in square brackets in the places it was occasionally dropped out – three times in the pronoun вьсь (in lines 2, 39, 40), which otherwise is written with the jer (it is kept in the root вьс- in lines 6, 13, 22, and 34), and once in the adjective ьстьнъ15 (compared to ьсть in line 38). Table 1. Number of syllables in the Old Bulgarian and in the Byzantine acrostic Прологъ�о�Х҃ѣ�оумѣренъ� съкаꙁанꙗ�с҃тго�еваньгелꙗ•� сътворенъ�костѧнтнъмь�мже� �прѣложен�бꙑсть•�тогожде� съкаꙁанꙗ•�вангельскааго⁘�(Syn.� 262,�f.�2r–3r) No. of syll. Γρηγόριος ο Θεολόγος, Στίχων ἡ ἀκροστιχὶς τῶν πάντων στοιχείων ἑκάστου ἰάμβου τέλος παραινέσεως ἔχοντος (PG 37.908-910) No. of syll. 1 Аꙁъ словомь смъ // молю сѧ б҃оу ⁘ 12 1 ἀρχὴν ἁπάντων / καὶ τέλος ποιοῦ Θεόν. 12 2 Б҃е в[ь]сеꙗ твар //  ꙁждтелю ⁘ 12 2 βίου τὸ κέρδος, / ἐκβιοῦν καθ᾿ ἡμέραν. 12 3 Вдмꙑмъ /  невдмꙑмъ ⁘ 12 3 γίνωσκε πάντα / τῶν καλῶν τὰ δράματα. 12 4 Г҃а д҃ха / посъл жвоущааго ⁘ 13 4 δεινὸν πένεσθαι, / χεῖρον δ᾿ εὐπορεῖν κακῶς. 12 5 Да въдъхнеть / въ срьдьце м слово ⁘ 12 5 εὐεργετῶν νόμιζε // μιμεῖσθαι Θεόν. 12 6 Ѥже боудеть / на оуспѣхъ вьсѣмъ ⁘ 12 6 ζήτει Θεοῦ σοι / χρηστότητα χρηστὸς ὤν. 12 7 Жвоущмъ / въ ꙁаповѣдьхъ т ⁘ 12 7 ἡ σὰρξ κρατείσθω / καὶ δαμαζέσθω καλῶς. 12 15 I also inserted in square brackets the proper letters at the beginning of the lines, without which letter repetitions are senseless – I did this according to their sequence in the alphabetic acrostich as Nahtigal proposed – and added in square brackets the most reasonable reconstruction of the twelfth verse which is supposed to begin with ꙉ, the letter with number value ‘30’. These insertions do not affect the syllable count but are made just for clarity and in accordance with the accepted hypothesis that the earliest Glagolitic system reflected only non- iotated nasal vowels (Stankovska 2018: 411-414 and the literature cited there). The same applies to е – it was first not iotated – but I promised no other interventions in the manuscript. 69A Byzantine Poetic Form in a Ninth-Century Bulgarian Poem 8 Ꙃѣло бо сть / свѣтльнкъ жꙁн ⁘ 12 8 θυμὸν χαλίνου, / μὴ φρενῶν ἔξω πέσῃς. 12 9 Ꙁаконъ тво•/  свѣтъ стьꙁамъ⁘ 11 9 ἵστη μὲν ὄμμα, / γλῶσσα δὲ στάθμην ἔχοι. 12 10 [Ј]же щеть / ев҃нгельска слова ⁘ 12 10 κλεὶς ὠσὶ κείσθω, / μηδὲ πορνεύοι γέλως. 12 11  прость / дарꙑ твоꙗ прꙗт ⁘ 11 11 λύχνος βίου σοι / παντὸς ἡγείσθω λόγος. 12 12 Летть бо нꙑнѣ• /  словѣньско племѧ ⁘ [ꙉ. лѣтъ нꙑнѣ /  словѣньско племѧ] 12 12 μή σοι τὸ εἶναι / τῷ δοκεῖν ὑπορρέοι. 12 13 Къ крьщеню•/ обратша сѧ вьс ⁘ 12 13 νόει τὰ πάντα, / πρᾶσσε δ᾿ ἃ πράσσειν θέμις. 12 14 Люд тво•/ нарещ сѧ хотѧще ⁘ 12 14 ξένον σεαυτὸν ἴσθι, // καὶ τίμα ξένους. 12 15 Млост твоꙗ•/ б҃е просѧть ꙁѣло ⁘ 13 15 ὅτ᾿ εὐπλοεῖς, / μάλιστα μέμνησο ζάλης. 12 16 Нъ мънѣ нꙑнѣ•/ пространо слово даждь ⁘ 12 16 πάντ᾿ εὐχαρίστως / δεῖ δέχεσθαι τἀκ Θεοῦ. 12 17 О҃е с҃не•/  прѣс҃тꙑ д҃ше ⁘ 12 17 ῥάβδος δικαίου / πλεῖον ἢ τιμὴ κακοῦ. 12 18 Просѧщоуоумоу•/ помощ ѿ тебе ⁘ 12 18 σοφῶν θύρας ἔκτριβε, // πλουσίων δὲ μή. 12 19 Роуцѣ бо сво / горѣ въꙁдѣю прсно ⁘ 12 19 τὸ μικρὸν οὐ μικρόν, / ὅταν ἐκφέρῃ μέγα. 13 20 Слоу прꙗт•/  моудрость оу тебе ⁘ 12 20 ὕβριν χαλίνου, / καὶ μέγας ἔσῃ σοφός 12 21 Тꙑ бо даш•/ достономъ слоу ⁘ 12 21 φύλασσε σαυτόν, / πτῶμα δ᾿ ἄλλου μὴ γέλα. 12 22 Упостась же•/ вьсѧкоую цѣлш ⁘ 12 22 χάρις φθονεῖσθαι, / τὸ φθονεῖν δ᾿ αἶσχος μέγα. 12 23 Фараоша мѧ•/ ꙁълобꙑ ꙁбав ⁘ 11 23 ψυχὴ θύοιτο / μᾶλλον ἢ τὸ πᾶν θεῷ. 12 70 Ekaterina Dikova 24 Херовьскоу м•/ мꙑсль  оумъ даждь ⁘ 12 24 ὢ τίς φυλάξει ταῦτα // καὶ σωθήσεται. 12 25 [Ѡ]О ьст[ь]наꙗ•/ прѣст҃аꙗ троце ⁘ 12 26 Пеаль мою / на радость прѣлож ⁘ 12 27 Цѣломоудрьно•/ да наьноу пьсат ⁘ 12 28 юдеса твоꙗ•/ прѣдвьнаꙗ ꙁѣло ⁘ 12 29 Шестькрлатъ•/ слоу въспрмъ ⁘ 11 30 Шьствоую нꙑнѣ•/ по слѣдоу оутелю ⁘ 12 31 мен ю•/  дѣлоу послѣдоуꙗ ⁘ 12 32 [Ѣ]Ꙗвѣ сътворю•/ еваньгельско слово ⁘ 12 33 Хвалоу въꙁдаꙗ•/ тр҃ц въ б҃жьтвѣ ⁘ 12 34 [Ѫ]Юже поть / вьсѧкъ въꙁдрастъ ⁘ 12 35 Юнъ  старъ•/ свомь раꙁоумомь ⁘ 12 36 [Ѧ]Ꙗꙁꙑкъ новъ•/ хвалоу въꙁдаꙗ прсно ⁘ 12 37 О҃цоу с҃ноу /  прѣс҃тоуоумѹ дх҃оу ⁘ 13 38 Ѥмоуже ьсть /  дрьжава  слава ⁘ 12 39 Отъ в[ь]сеꙗ // твар  дꙑханꙗ ⁘ 12 40 Въ в[ь]сѧ вѣкꙑ•/  на вѣкꙑ амнъ ⁘ 12 Total: 33 × 12 syll. = 82.5% 4 × 11 syll. = 10% 3 × 13 syll. = 7.5% 23 × 12 syll. = 95.8% 1 × 13 syll. = 4.16% 71A Byzantine Poetic Form in a Ninth-Century Bulgarian Poem The numbers speak for themselves. In this non-reconstructed manuscript text, the number of twelve-syllable verses is over 82%. The percentage of thirteen- syllable verses is higher than in the Byzantine poem observed here, but all three such verses would easily become dodecasyllabic if uncontracted forms in them were contracted (жвоущааго, твоꙗ, прѣс҃тоуоумоѹ).16 Yet the shorter verses – the four which count just eleven syllables – do not have any counterparts in the Greek poem and are not particularly discussed by Maas. This is, probably, why various scholars make various attempts to add “reconstructed” syllables to them. But both – the shorter and the longer verses in the Bulgarian poem – have counterparts in Byzantine poetry. The devia- tions from the constant number of twelve syllables in it – between 10–11 and 13–14 – are explained as specific combinations of hemistichs of, respectively, 5/6+5/6 and 7/6+7/6.17 Here, it should be noted, however, that while verse 9 has two more com- pletely different readings (of fourteen and fifteen syllables) in the other copies of the prayer, verse 11 has just one different reading with  ‘and’ preceding прꙗт ‘to accept’ which certainly makes the twelve syllables. Verse 29 is also highly variable, besides in regard to two words – Шестькрлатъ�/�Шестокрлъ� /�Шестькрлатъ�/�Шестькрлатꙑхъ�‘six-winged’�and�въспрмъ�/�прмъ ‘having accepted’ – but in all the variants it counts eleven or even ten syllables and the same numbers apply also to verse 26, i. e. it counts always less than twelve syllables. Even though there is no way for us to be sure whether the eleven-syllable lines are due to text corruption or were intended by their author, I propose to hypothetically accept the latter and to closely look at the immediate context of the respective lines. To begin with, verses 9 and 11 are part of a tetracolon in which all four cola bear four accents each and the variation in the number of syllables of the two subordinate cola is a way to diversify the otherwise complete symmetry of the cola in this isocolon18 (the symmetry remains yet 16 Moreover, жвоущааго and твоꙗ�appear contracted – as жвѧщаго/жвꙋщаго and твоѧ respectively – in the other copies of the prayer. In relation to verse 37 (the one containing прѣс҃тоуоумѹ), all the variants except for the one in MS Syn. 262 read ѡ҃цꙋ��с҃нꙋ��с҃тмꙋ�д҃хꙋ which also counts twelve syllables (for the variant readings see Veder 1999: 64, 68, 78). 17 See, for example, Bernard 2018: 36. 18 Almost all the characteristic features of the Byzantine dodecasyllable, actually, lead to the construction of isocolon, which builds up the rhetorical (and not the metrical) rhythm: not only the semantic completeness of the separate verses/cola but also their equal length (the number of syllables), clausulae, ceasuring and a variety of features characteristic for particular types of writings or authors. On the notion that the isocolic structure relates this meter to the rhetorical 72 Ekaterina Dikova slightly modified by the alternation of cola with the same number of accents but different number of syllables, which could be visualised in the following way: 12 (4), 11 (4), 12 (4), 11 (4)). Next, the other pair of verses counting eleven syllables, namely 23 and 29, are part of parallel isocolic periods (divided by another tricolon) with a similar scheme – 11 (3), 12 (4), 12 (3) […] 11 (3), 12 (4), 12 (3) – so, here, the subtle change in rhythm is intended rather to pair, not to diversify. Nevertheless, it should be emphasised, that, on the one hand, the archetype of this poetic prayer was most probably closer to the Byzantine dodecasyllable than the text which came down to us in MS Syn. 262, and, on the other, the author did not label it iamboi (as the majority of the Byzantine alphabetic acrostics (cf. the numerous examples given in Anastasijević 1905), but just measured prologue. 2. Internal verse breaks (caesurae) Maas postulated verse structures of the type 5+7 and 7+5 as most typical for the Byzantine dodecasyllable.19 And these are the prevailing types in the two alphabetic acrostics in focus. To clearly reveal this in the table below, I fol- lowed Nahtigal’s manner to mark the pauses after the seventh syllable with a double slash (//) while the ones which come after the fifth syllable – with a single slash (/). All other positions of the caesura I mark with backslash (\). Table 2. presents the same two texts given this time with the number of syl- lables per hemistich. It is also clearly visible, that in MS Syn. 262, the caesurae are marked by dots in more than half of the verses. prose, including in the Byzantine theory, and for specific examples revealing that the Byzantines ignored the difference between poetry and prose see Lauxtermann 1998: 21–22. The short cola and the regular alternation of consonants brings about the energetic style of the dodecasyllable (Bernard 2018: 22) and they both, according to three Byzantine rhetors, are inseparable part of its versification system (Lauxtermann 1998: 28). The structuring in cola – according to the complete thoughts they convey – is characteristic also for the rhetorical prose (ibid.: 21). 19 Cf. the exact statistics in Bernard 2018: 27 and concerning other variants of hemistichs, see note 16. 73A Byzantine Poetic Form in a Ninth-Century Bulgarian Poem Table 2. The internal verse breaks in the two poems Прологъ о Х҃ѣ оумѣренъ... Cae- sura Στίχων ἡ ἀκροστιχὶς... Cae sura 1 Аꙁъ словомь смъ // молю сѧ б҃оу ⁘ 7+5 1 ἀρχὴν ἁπάντων / καὶ τέλος ποιοῦ Θεόν. 5+7 2 Б҃е в[ь]сеꙗ твар //  ꙁждтелю ⁘ 7+5 2 βίου τὸ κέρδος, / ἐκβιοῦν καθ᾿ ἡμέραν. 5+7 3 Вдмꙑмъ /  невдмꙑмъ ⁘ 5+7 3 γίνωσκε πάντα / τῶν καλῶν τὰ δράματα. 5+7 4 Г҃а д҃ха \ посъл жвоущааго ⁘ 5+8 4 δεινὸν πένεσθαι, / χεῖρον δ᾿ εὐπορεῖν κακῶς. 5+7 5 Да въдъхнеть / въ срьдьце м слово ⁘ 5+7 5 εὐεργετῶν νόμιζε // μιμεῖσθαι Θεόν. 7+5 6 Ѥже боудеть / на оуспѣхъ вьсѣмъ ⁘ 5+7 6 ζήτει Θεοῦ σοι / χρηστότητα χρηστὸς ὤν. 5+7 7 Жвоущмъ / въ ꙁаповѣдьхъ т ⁘ 5+7 7 ἡ σὰρξ κρατείσθω / καὶ δαμαζέσθω καλῶς. 5+7 8 Ꙃѣло бо сть / свѣтльнкъ жꙁн ⁘ 5+7 8 θυμὸν χαλίνου, / μὴ φρενῶν ἔξω πέσῃς. 5+7 9 Ꙁаконъ тво•/  свѣтъ стьꙁамъ⁘ 5+6 9 ἵστη μὲν ὄμμα, / γλῶσσα δὲ στάθμην ἔχοι. 5+7 10 [Ј]же щеть / ев҃нгельска слова ⁘ 5+7 10 κλεὶς ὠσὶ κείσθω, / μηδὲ πορνεύοι γέλως. 5+7 11  прость \ дарꙑ твоꙗ прꙗт ⁘ 4+7 11 λύχνος βίου σοι / παντὸς ἡγείσθω λόγος. 5+7 12 Летть бо нꙑнѣ• \  словѣньско племѧ ⁘ [ꙉ. лѣтъ нꙑнѣ /  словѣньско племѧ] 6+6/ [5+7] 12 μή σοι τὸ εἶναι / τῷ δοκεῖν ὑπορρέοι. 5+7 13 Къ крьщеню•/ обратша сѧ вьс ⁘ 5+7 13 νόει τὰ πάντα, / πρᾶσσε δ᾿ ἃ πράσσειν θέμις. 5+7 14 Люд тво•/ нарещ сѧ хотѧще ⁘ 5+7 14 ξένον σεαυτὸν ἴσθι, // καὶ τίμα ξένους. 7+5 15 Млост твоꙗ•\ б҃е просѧть ꙁѣло ⁘ 6+7 15 ὅτ᾿ εὐπλοεῖς, \ μάλιστα μέμνησο ζάλης. 4+8 74 Ekaterina Dikova 16 Нъ мънѣ нꙑнѣ•/ пространо слово даждь ⁘ 5+7 16 πάντ᾿ εὐχαρίστως / δεῖ δέχεσθαι τἀκ Θεοῦ. 5+7 17 О҃е с҃не•/  прѣс҃тꙑ д҃ше ⁘ 5+7 17 ῥάβδος δικαίου / πλεῖον ἢ τιμὴ κακοῦ. 5+7 18 Просѧщоуоумоу•/ помощ ѿ тебе ⁘ 5+7 18 σοφῶν θύρας ἔκτριβε, // πλουσίων δὲ μή. 7+5 19 Роуцѣ бо сво / горѣ въꙁдѣю прсно ⁘ 5+7 19 τὸ μικρὸν οὐ μικρόν, / ὅταν ἐκφέρῃ μέγα. 6+7 20 Слоу прꙗт•/  моудрость оу тебе ⁘ 5+7 20 ὕβριν χαλίνου, / καὶ μέγας ἔσῃ σοφός 5+7 21 Тꙑ бо даш•/ достономъ слоу ⁘ 5+7 21 φύλασσε σαυτόν, / πτῶμα δ᾿ ἄλλου μὴ γέλα. 5+7 22 Упостась же•/ вьсѧкоую цѣлш ⁘ 5+7 22 χάρις φθονεῖσθαι, / τὸ φθονεῖν δ᾿ αἶσχος μέγα. 5+7 23 Фараоша мѧ•\ ꙁълобꙑ ꙁбав ⁘ 5+6 23 ψυχὴ θύοιτο / μᾶλλον ἢ τὸ πᾶν θεῷ. 5+7 24 Херовьскоу м•/ мꙑсль  оумъ даждь ⁘ 5+7 24 ὢ τίς φυλάξει ταῦτα // καὶ σωθήσεται. 7+5 25 [Ѡ]О ьст[ь]наꙗ•/ прѣст҃аꙗ троце ⁘ 5+7 26 Пеаль мою / на радость прѣлож ⁘ 5+7 27 Цѣломоудрьно•/ да наьноу пьсат ⁘ 5+7 28 юдеса твоꙗ•/ прѣдвьнаꙗ ꙁѣло ⁘ 5+7 29 Шестькрлатъ•\ слоу въспрмъ ⁘ 5+6 30 Шьствоую нꙑнѣ•/ по слѣдоу оутелю ⁘ 5+7 31 мен ю•/  дѣлоу послѣдоуꙗ ⁘ 5+7 32 [Ѣ]Ꙗвѣ сътворю•/ еваньгельско слово ⁘ 5+7 75A Byzantine Poetic Form in a Ninth-Century Bulgarian Poem 33 Хвалоу въꙁдаꙗ•/ тр҃ц въ б҃жьтвѣ ⁘ 5+7 34 [Ѫ]Юже поть / вьсѧкъ въꙁдрастъ ⁘ 5+7 35 Юнъ  старъ•/ свомь раꙁоумомь ⁘ 5+7 36 [Ѧ]Ꙗꙁꙑкъ новъ•/ хвалоу въꙁдаꙗ прсно ⁘ 5+7 37 О҃цоу с҃ноу \  прѣс҃тоуоумѹ дх҃оу ⁘ 5+7 38 Ѥмоуже ьсть /  дрьжава  слава ⁘ 5+7 39 Отъ в[ь]сеꙗ // твар  дꙑханꙗ ⁘ 7+5 40 Въ в[ь]сѧ вѣкꙑ•/  на вѣкꙑ амнъ ⁘ 5+7 Total: 30 × (5+7) = 75% 3 × (7+5) = 7.5% others × 5 = 12.5% = 82.5%(5+7/7+5) (5+7) × 18 = 75% (7+5) × 4 = 16.7% others × 2 = 8.3% = 91.6%(5+7/7+5) Statistics shows that, in terms of the internal verse breaks, the two poetic works are a bit closer to one another than in relation to the first metrical feature observed, and that the total percentage of the structuring, which is charac- teristic for the Byzantine dodecasyllable (that is either 5+7 or 7+5) is 82.5% in the Bulgarian and 91.6% in the Greek acrostic. The deviations account to 12.5% and 8.3% respectively, so they might be related to the specific genre of the alphabetic acrostics, since the significance of the sequence of letters pre- vails over the importance of adhering to other poetic principles. Once again, the Glagolitic archetype of the Bulgarian work would have probably shown slightly different percentages. A noticeable peculiarity here is, however, the fact that Constantine of Preslav used the caesuring 7+5, this subtle alternation of rhythm, to mark both the opening (the first two verses) and the conclusion of his poem (its penultimate verse), which reveals his profound knowledge – and well- crafted skill – not only in the most used Byzantine poetic measure but also in Byzantine rhetoric. 76 Ekaterina Dikova 3. Clausulae Specialists in Byzantine poetry accept the prevailing paroxytone clausulae – that is an accent on the penultimate syllable at the end of the verse 20 – as one of the characteristic features of the meter.21 But when it comes to accentuation in Old Slavonic texts, extreme attentiveness is needed, first, because accents are systematically reflected in the graphics only after the fourteenth century (often with variations within the same manuscripts), second, because of its instability – it moves to various syllables in the forms of the same words and takes different positions in different dialects including within the same lan- guage. It is not by chance that accentuation theories related to Old Slavonic are far from being unshakable.22 I have nevertheless decided, as a kind of experiment, to underline only those last words which seem irrefutably paroxytone. These are mainly two- syllable words like бо҃у, слово, племѧ, даждь even the vocative дш҃е, and certain three-syllable words like вьсѣмъ, хотѧще. I take the risk of making some mis- takes in order to get at least some approximate idea of the extent to which this peculiarity accords with the respective characteristic feature of the Byzantine dodecasyllable. Below, I mark paroxytone words not only at the verse ends, as Sobolevskij and Georgiev did, but also the words at the ends of hemistichs because the latter are not less important than the first for the specialists in Byzantine poetry. 20 Bernard, for instance, turns our attention to the fact that a particular poem belonging to the genre of iambs on iambs quotes only paroxytone examples as final verse words (Bernard 2018: 18). 21 The actual scheme that Maas (1903: 290) derives is that the paroxytone verse endings are the rule, proparoxytone – rareness and the oxytone ones – exception. A later Byzantine rhetor (Joseph Rhacendytes) points out that for the iambic verses it is best to end with a penultimate accent as noted by Hörandner (1995: 288), and Bernard (2018: 22) comments that this same recommendation for a paroxytone on the sixth feet is a uniquely clear and explicit medieval recognition of this crucial feature of the most used meter. The rule is known even before Maas (cf. Bouvy 1886: 155–157). 22 In fact, there is only one hypothetic reconstruction particularly related to the Old Bulgar- ian accent. It tackles it in relation to a later period, to the fourteenth century, besides on the basis of even later monuments and does not exactly determine the distribution of accents (see the profound studies by the best specialist in comparative historical accentology related to Sla- vonic – Dybo 1971: 194; 1977: 93–114). (My personal opinion is that accents are not marked in the earlier manuscripts namely because of the different Slavonic accentuation and with the view to providing freedom for any local pronunciations related to word stress.) 77A Byzantine Poetic Form in a Ninth-Century Bulgarian Poem Table 3. Words bearing penultimate accents before internal and final pauses of verses. Прологъ о Х҃ѣ оумѣренъ... Στίχων ἡ ἀκροστιχὶς... 1 Аꙁъ словомь смъ23 // молю сѧ б҃оуб҃оу ⁘ 1 ἀρχὴν ἁπάντων / καὶ τέλος ποιοῦ Θεόν. 2 Б҃е в[ь]сеꙗ твартвар //  ꙁждтелю ⁘ 2 βίου τὸ κέρδος, / ἐκβιοῦν καθ᾿ ἡμέραν. 3 Вдмꙑмъ /  невдмꙑмъ ⁘ 3 γίνωσκε πάντα / τῶν καλῶν τὰ δράματα. 4 Г҃а д҃хад҃ха / посъл жвоущааго ⁘ 4 δεινὸν πένεσθαι, / χεῖρον δ᾿ εὐπορεῖν κακῶς. 5 Да въдъхнеть / въ срьдьце м словослово ⁘ 5 εὐεργετῶν νόμιζε // μιμεῖσθαι Θεόν. 6 Ѥже боудеть / на оуспѣхъ вьсѣмъвьсѣмъ ⁘ 6 ζήτει Θεοῦ σοι / χρηστότητα χρηστὸς ὤν. 7 Жвоущмъ / въ ꙁаповѣдьхъ т ⁘ 7 ἡ σὰρξ κρατείσθω / καὶ δαμαζέσθω καλῶς. 8 Ꙃѣло бо стьсть / свѣтльнкъ жꙁнжꙁн ⁘ 8 θυμὸν χαλίνου, / μὴ φρενῶν ἔξω πέσῃς. 9 Ꙁаконъ твотво•/  свѣтъ стьꙁамъ ⁘ 9 ἵστη μὲν ὄμμα, / γλῶσσα δὲ στάθμην ἔχοι. 10 [Ј]же щеть / ев҃нгельска словаслова ⁘ 10 κλεὶς ὠσὶ κείσθω, / μηδὲ πορνεύοι γέλως. 11  прость / дарꙑ твоꙗ прꙗтпрꙗт ⁘ 11 λύχνος βίου σοι / παντὸς ἡγείσθω λόγος. 12 ꙉ. лѣтъ нꙑнѣнꙑнѣ /  словѣньско племѧплемѧ ⁘ 12 μή σοι τὸ εἶναι / τῷ δοκεῖν ὑπορρέοι. 13 Къ крьщеню•/ обратша сѧ вьс ⁘ 13 νόει τὰ πάντα, / πρᾶσσε δ᾿ ἃ πράσσειν θέμις. 14 Люд тво•/ нарещ сѧ хотѧщехотѧще ⁘ 14 ξένον σεαυτὸν ἴσθι, // καὶ τίμα ξένους. 23 It is quite tempting to enlist смъ among the paroxytone words, but here it is more likely an enclitic as suggested also by the count of accents in Dobrev (1993: 241). 78 Ekaterina Dikova 15 Млост твоꙗ•/ б҃е просѧть ꙁѣло ⁘ 15 ὅτ᾿ εὐπλοεῖς, / μάλιστα μέμνησο ζάλης. 16 Нъ мънѣ нꙑнѣнꙑнѣ•/ пространо слово даждьдаждь ⁘ 16 πάντ᾿ εὐχαρίστως / δεῖ δέχεσθαι τἀκ Θεοῦ. 17 О҃е с҃нес҃не•/  прѣс҃тꙑ д҃шед҃ше ⁘ 17 ῥάβδος δικαίου / πλεῖον ἢ τιμὴ κακοῦ. 18 Просѧщоуоумоу•/ помощ ѿ тебе ⁘ 18 σοφῶν θύρας ἔκτριβε, // πλουσίων δὲ μή. 19 Роуцѣ бо сво / горѣ въꙁдѣю прснопрсно ⁘ 19 τὸ μικρὸν οὐ μικρόν, / ὅταν ἐκφέρῃ μέγα. 20 Слоу прꙗтпрꙗт•/  моудрость оу тебе ⁘ 20 ὕβριν χαλίνου, / καὶ μέγας ἔσῃ σοφός 21 Тꙑ бо дашдаш•/ достономъ слоуслоу ⁘ 21 φύλασσε σαυτόν, / πτῶμα δ᾿ ἄλλου μὴ γέλα. 22 Упостась же•/ вьсѧкоую цѣлш ⁘ 22 χάρις φθονεῖσθαι, / τὸ φθονεῖν δ᾿ αἶσχος μέγα. 23 Фараоша мѧ•/ ꙁълобꙑ ꙁбав⁘ 23 ψυχὴ θύοιτο / μᾶλλον ἢ τὸ πᾶν θεῷ. 24 Херовьскоу м•/ мꙑсль  оумъ даждьдаждь ⁘ 24 ὢ τίς φυλάξει ταῦτα // καὶ σωθήσεται. 25 [Ѡ]О ьст[ь]наꙗ•/ прѣст҃аꙗ троце ⁘ 26 Пеаль мою / на радость прѣлож ⁘ 27 Цѣломоудрьно•/ да наьноу пьсатпьсат ⁘ 28 юдеса твоꙗ•/ прѣдвьнаꙗ ꙁѣло ⁘ 29 Шестькрлатъ•/ слоу въспрмъ ⁘ 30 Шьствоую нꙑнѣнꙑнѣ•/ по слѣдоу оутелю ⁘ 79A Byzantine Poetic Form in a Ninth-Century Bulgarian Poem 31 мен ю•/  дѣлоу послѣдоуꙗ ⁘ 32 [Ѣ]Ꙗвѣ сътворю•/ еваньгельско словослово ⁘ 33 Хвалоу въꙁдаꙗ•/ тр҃ц въ б҃жьтвѣ ⁘ 34 [Ѫ]Юже поть / вьсѧкъ въꙁдрастъ ⁘ 35 Юнъ  старъстаръ•/ свомь раꙁоумомь ⁘ 36 [Ѧ]Ꙗꙁꙑкъ новъновъ•/ хвалоу въꙁдаꙗ прснопрсно ⁘ 37 О҃цоу с҃ноус҃ноу /  прѣс҃тоуоумѹ дх҃оудх҃оу ⁘ 38 Ѥмоуже ьстььсть /  дрьжава  славаслава ⁘ 39 Отъ вв[ь][ь]сеꙗсеꙗ // твар  дꙑханꙗ ⁘ 40 Въ в[ь]сѧ вѣкꙑвѣкꙑ•/  на вѣкꙑ амнъ ⁘ ...× ˊ ˘/  – 17 = 42.5% / ...× ˊ ˘  – 18 = 45% ...× ˊ ˘/  – 13 = 54% / ...× ˊ ˘  – 13 = 54% The statistical results of the analysis reveal, before all, that Sobolevskij and Georgiev were both right and not right  – there are female clausulae in Azbuchna molitva, but their percentage is neither 100, nor 0, but rather around 50, which is also the case in the Greek acrostic. Besides, even though not entirely sure, the supposed lexemes in the Bulgarian poem are probably not the only paroxytone ones and the actual situation in the archetype might have been different. Quite similar is the distribution of paroxytone words just before the clausalae in the two writings. If we turn to Maas postulates, we might suppose some diachronic conditioning, because he calculated that the most serious deviation of the rule was in earlier works and authors, and because some recent investigations in early dodecasyllables – of the third to fourth 80 Ekaterina Dikova century – prove that paroxytone endings are even the exception in them.24 So, is this just an imitation of the earlier tradition? The data in the Byzantine paraenetic alphabets reveal that the peculiarity is rather characteristic for their whole genre, no matter the time of origin (Anastasijević 1905: 20, 24, 34, 38–39, 41, 47, 48, 56, 58). It seems that Constantine of Preslav was not only fluent in iambic trends but much aware of the subtle peculiarities of the sub- genre of alphabetic acrostics or, at least, attentive enough to be able to apply all his knowledge for high poetic purposes in his own language. 4. Prosody Simply put, prosody is kept on the strong positions of the iambic trimeter (that is, on the even syllables) – the even syllables seem to be graphically long even in the latest poems labeled as ἰάμβοι. Specialists in Byzantine poetry underline the visuality of this feature and call it Augenpoesie (Rhoby 2011: 137) and Sheinprosodie (Bernard 2010: 16).25 The presence of visual prosody is accepted axiomatically by all the experts in the field26 but Lauxtermann is most specific in determining it: the iambic trimeter is a dodecasyllabic verse without resolution27 with anceps on the odd positions (their length does not matter), long even syllables and brevis in longo at the end of the verse (which 24 Another point in this respect is considered by Rhoby (2011: 140, note 134) who quotes the scholarly opinion that about 48% of the lexemes in the late Greek thesaurus were paroxytone by default. In addition, a linguistic factor that should be taken into consideration when observing the development of paroxytone verse endings is that in the late Greek language, the enclitics transfer their accent on the last syllable of their preceding word (Bernard 2018: 33). 25 See also Maas (1903: 301–303) who names it “historische Ortographie der Versification”. Lauxtermann (1998: 24, 33) emphasises that even though prosody might have been understood by the Byzantines, the ancient metrics remains abstract for them. How much more this must be true for the Old Bulgarian men of letters. 26 Special attention is paid to “mistakes in prosody” in certain works by George of Pisidia by Romano (1985: 4–6). 27 Probably because resolutions would break isosillabicity. And yet, some of the verses of 13 and 14 syllables may well be explained by the substitution of a visually long syllable with two short ones. The exceptional verse 19 of St Gregory’s poem observed here is most probably of this kind. Moreover, St Gregory the Theologian is one of the most cited earlier authors when it comes to verses with resolutions (see for example Zagklas (2019: 4 and the literature given in note 13 there)). 81A Byzantine Poetic Form in a Ninth-Century Bulgarian Poem means that even a short final syllable counts like a long one).28 One last nec- essary clarification is that the dichrona – α, ι, υ – are counted long or short depending on the syllable they belong to.29 So, how do the two alphabetic acrostics accord with these principles? St Gregory’s poem shows that 94 (83%) of its even syllables counted30 are visu- ally long (marked in bold below) and 19 (16.9%) of them (underlined below) are not. Table 4. Visual prosody in St Gregory’s poem Γρηγόριος ο Θεολόγος, Στίχων ἡ ἀκροστιχὶς... 1 ἀρχὴν ἁπάντων / καὶ τέλος ποιοῦ Θεόν. 2 βίου τὸ κέρδος, / ἐκβιοῦν καθ᾿ ἡμέραν. 3 γίνωσκε πάντα / τῶν καλῶν τὰ δράματα. 4 δεινὸν πένεσθαι, / χεῖρον δ᾿ εὐπορεῖν κακῶς. 5 εὐεργετῶν νόμιζε // μιμεῖσθαι Θεόν. 6 ζήτει Θεοῦ σοι / χρηστότητα χρηστὸς ὤν. 7 ἡ σὰρξ κρατείσθω / καὶ δαμαζέσθω καλῶς. 8 θυμὸν χαλίνου, / μὴ φρενῶν ἔξω πέσῃς. 9 ἵστη μὲν ὄμμα, / γλῶσσα δὲ στάθμην ἔχοι. 10 κλεὶς ὠσὶ κείσθω, / μηδὲ πορνεύοι γέλως. 28 Here follows the exact quotation: “The pure form, so we are told, consists of iambs in the second and the fourth positions, iambs or spondaics in the first, the third and the fifth positions, and iambs or pyrrhics in the sixth position. In other words, the pure form is the iambic verse of twelve syllables with anceps in the uneven positions and brevis in longo at the verse end” (Lauxtermann 1998: 17). To be even more precise, the compulsory short positions are just the third, the seventh and the eleventh, since, from the prosodical aspect, the iamb consists of a short and a long syllable, the spondee – of two long ones and the pyrrhic of two shorts but in this case, its second could be also long (because the sixth position could be also iamb), so such a verse has the following schematic representation: U̅ – U – U̅/ – U// – U̅ – U –. The same excerpt from the poem on iambs, attributed most often to Michael Psellos, is referred to also in Hörandner (1995: 286 including note 28) and Bernard (2018: 17). 29 Lauxtermann convincingly proves that the pure iambs (ϰαθαροὶ στιχοὶ) in Byzantine metrics is a term referring to verses without resolution and that the Byzantines used it to designate their dodecasyllable in order to distinguish it, on the one hand, from the ancient iambic trimeter, and, on the other, to emphasise the genetic connection between the two (Lauxtermann 1998: 7–19). 30 I excluded from counting the last syllables of each verse as they are long by position as well as the second hemistich of verse 19 (the only one of thirteen syllables) since I am not sure how to count syllables if indeed a resolution is applied. 82 Ekaterina Dikova 11 λύχνος βίου σοι / παντὸς ἡγείσθω λόγος. 12 μή σοι τὸ εἶναι / τῷ δοκεῖν ὑπορρέοι. 13 νόει τὰ πάντα, / πρᾶσσε δ᾿ ἃ πράσσειν θέμις. 14 ξένον σεαυτὸν ἴσθι, // καὶ τίμα ξένους. 15 ὅτ᾿ εὐπλοεῖς, /μάλιστα μέμνησο ζάλης. 16 πάντ᾿ εὐχαρίστως / δεῖ δέχεσθαι τἀκ Θεοῦ. 17 ῥάβδος δικαίου / πλεῖον ἢ τιμὴ κακοῦ. 18 σοφῶν θύρας ἔκτριβε, // πλουσίων δὲ μή. 19 τὸ μικρὸν οὐ μικρόν, / ὅταν ἐκφέρῃ μέγα. 20 ὕβριν χαλίνου, / καὶ μέγας ἔσῃ σοφός 21 φύλασσε σαυτόν, / πτῶμα δ᾿ ἄλλου μὴ γέλα. 22 χάρις φθονεῖσθαι, / τὸ φθονεῖν δ᾿ αἶσχος μέγα. 23 ψυχὴ θύοιτο / μᾶλλον ἢ τὸ πᾶν θεῷ. 24 ὢ τίς φυλάξει ταῦτα // καὶ σωθήσεται. The analysis of the Slavonic piece runs into two obstacles. First, the manu- script text has only one of the letters for o and i – the pairs о/ѡ and /і (and the respective Glagolitic ⱁ/ⱉ and ⰻ/ⰹ) are not preserved in it.31 Second, the Old Bulgarian language just emerged as a literary one in the ninth century and did not have the long tradition of the Greek including in poetry. Could those obstacles be possibly overcome? I have checked a later and shorter translated poetic text first – a random Synaxarion verse unit32 – with the hope that it would suggest what to look for in the alphabetic prayer. Here, the jers should not be counted as they no longer sounded in the fourteenth century. 31 Even if they were present in the original, the manuscript copying procedure, especially if combined with dictation, would quickly erase the difference between the members of the opposition. 32 The Synaxarion verses are in fact Christopher of Mytilene’s dodecasyllabic calendar (of the eleventh century) which is a model for all the peculiarities of the Byzantine dodecasyllable (cf. Eftymiadis 2014: 163–165). They have two South Slavonic translations dated to the fourteenth century (cf. Taseva 2006: 170–171). 83A Byzantine Poetic Form in a Ninth-Century Bulgarian Poem Table 5. Visual prosody in Synaxar verses Christopher of Mytilene’s dodeca- syllabic calendar (eleventh century) Eustratiades 154 Bulgarian translation, MS Zogr 80, f. 2v (fourteenth century) Ὁ χρηστὸς ἡμῖν Εὐτρόπιος τοὺς τρόπους, хоⷭ ҇вашїїмъ нравомъ бл҃гы нравъ• 4–5, 9–10 Ἐφεῦρε Χριστὸν καὶ τέλους διὰ ξίφους. брѣте х҃а ̇ менѫѫѫѫ коннѫ• 8–9 Καὶ Κλεόνικος εὐκλεᾶ νίκην ἔχει,  клеоео̀нкъ бл҃гославнѫѫѫѫ побѣдѫ ма•ⷮ 2–3, 8–9 Σταυρῷ κρεμασθείς, ὡς Χριστός μου πάλαι. (11 syll.) на кртⷭ҇ѣ повѣшенъ ꙗко же х҃с моо древле• 12–13 Εἱρκτὴν τὸ σῶμα καὶ πρὸ τῆς εἱρκτῆς ἔχων, тъмнцѫ тѣлѡ ͗ прѣжⷣе тъмнцѫ ̀мѣѫ• Εἱρκτῶν λυτροῦται Βασιλίσκος ἐκ δύο. тъмнцѧ ͗ꙁбав сѧ васлскъ обоохъ• 12–13 The Greek source text shows a slightly lower percentage of deviations (the underlined vowels in Κλεόνικος, Χριστός, Βασιλίσκος ἐκ) – just four visually short syllables on even positions, besides, three of them in personal names. The Bulgarian text, again, does not show the desired pairs о/ѡ and /і, but what stands out even at first glance is that all, but one (бл҃гы),33 adjacent vowels appear on the borders between even and odd syllables. Could this peculiar- ity – the exclusive use of adjacent vowels on the borders between even and odd syllables – be some graphic imitation of the Byzantine visual prosody? The analysis of Azbuchna molitva does not give a definite answer in this respect, as the percentage of the adjacent vowels beginning with an even syl- lable there is 57.1% (20 uses out of 35) while the ones with initial odd syllable account to 42.8%. In addition, the positions of the first, fifth and nineth sylla- bles is considered anceps for the Byzantine dodecasyllable (see note 28 above), so it is more accurate to count 20 out of 29 uses on even positions (68.9%) and 9 appearances starting on the odd positions which definitely need to be short (31%). 33 The exceptional one falls on an anceps position (see note 28). 84 Ekaterina Dikova Table 6. Visual prosody in Constantine’s poem 1� Аꙁъ�словомь�смъ�//�молю�сѧ�б҃оу�⁘ 2� Б҃е�в[ь]сеꙗеꙗ�твар�//��ꙁждтелю�⁘�����������������4–5 3� Вдмꙑмъ�/��невдмꙑꙑмъ�⁘������������������3–4, 10–11 4� Г҃а�д҃ха�/�посъл�жвоущаго�⁘ 5� Да�въдъхнеть�/�въ�срьдьце�м�слово�⁘ 6� Ѥже�боудеть�/�на�оуспѣхъ�вьсѣмъ�⁘ 7� Жвоущмъ�/�въ�ꙁаповѣдьхъ�т�⁘���������������3–4 8� Ꙃѣло�бо�сть�/�свѣтльнкъ�жꙁн�⁘ 9� Ꙁаконъ�тво•/��свѣтъ�стьꙁамъ�⁘ 10 [Ј]же�щеть�/�ев҃нгельска�слова�⁘ 11� �прость�/ дарꙑ�твоꙗ�прꙗт�⁘����� 7–8, 9–10 (11 syll.) 12 [ꙉ.�лѣтъ�нꙑнѣ�/ �словѣньско�племѧ] 13� Къ�крьщенюю•/�обратша�сѧ�вьс�⁘����������������4–5 14� Люд�твоо•/�нарещ�сѧ�хотѧще�⁘����������������2–3, 4–5 15� Млост�твоꙗоꙗ•/�б҃е�просѧть�ꙁѣло�⁘� 4–5 16� Нъ�мънѣ�нꙑнѣ•/�пространо�слово�даждь�⁘ 17� О҃е�с҃не•/��прѣс҃тꙑꙑ�д҃ше�⁘���������������������������8–9 18� Просѧщоуоумоу•/�помощ�ѿ�тебе�⁘����������������3–4 19� Роуцѣ�бо�своо�/�горѣ�въꙁдѣю�прсно�⁘������������4–5, 9–10 20� Слоу�прꙗт•/��моудрость�оу�тебе�⁘�����������3–4 21� Тꙑ�бо�даш•/�достономъ�слоу�⁘ 3–4 22� Упостась�же•/�вьсѧкоуюоую�цѣлш�⁘ 8–9 23� Фараоша мѧ•/�ꙁълобꙑ ꙁбав �⁘ 24� Херовьскоу�м•/�мꙑсль��оумъ�даждь�⁘ 25 [Ѡ]О�ьст[ь]наꙗаꙗ•/�прѣст҃аꙗаꙗ�трооце�⁘ 4–5, 8–9, 10–11 26� Пеаль�моюою�/�на�радость�прѣлож�⁘ 4–5 27� Цѣломоудрьно•/�да�наьноу�пьсат�⁘ 28� юдеса�твоꙗоꙗ•/�прѣдвьнаꙗ ꙁѣло�⁘ 4–5, 9–10 29� Шестькрлатъ•/�слоу�въспрмъ�⁘ 9–10 (11 syll.) 85A Byzantine Poetic Form in a Ninth-Century Bulgarian Poem 30� Шьствоуюоую�нꙑнѣ•/�по�слѣдоу�оутелю�⁘��������2–3 31� мен�юю•/��дѣлоу�послѣдоуꙗ�⁘�����������������4–5, 11–12 32 [Ѣ]Ꙗвѣ�сътворю•/�еваньгельско�слово�⁘ 33� Хвалоу�въꙁдаꙗаꙗ•/�тр҃ц�въ�б҃жьтвѣ�⁘ 4–5, 6–7 34 [Ѫ]Юже�поть�/�вьсѧкъъ�въꙁдрастъ�⁘ 3–4, 8–9 35� Юнъ��старъ•/�свомь�раꙁоумомь�⁘ 5–6 36 [Ѧ]Ꙗꙁꙑкъ�новъ•/�хвалоу�въꙁдаꙗ�прсно�⁘ 9–10 37� О҃цоу�с҃ноу�/��прѣс҃тоумѹ�дх҃оу�⁘ 38� Ѥмоуже�ьсть�/��дрьжава��слава�⁘ 39� Отъ�в[ь]сеꙗеꙗ�//�твар �дꙑханꙗ�⁘ 4–5, 11–12 40� Въ�в[ь]сѧ�вѣкꙑ•/��на�вѣкꙑ�амнъ�⁘ In addition, 6 of the odd-syllable positions (further 20% of the whole) are between syllables 3–4, while there is none positioned between syllables 2–3. Are the cases 3–4 compensating for the lack of cases 2–3? Indubitably, it is difficult to compose such a combination from the very second syllable when the poet’s main concern is to select a word beginning with a particular let- ter to convey particular thought in the coherent text of a prayer. As already demonstrated above, the importance of the acrostich prevails over the iambic peculiarities. The modern translations of Azbuchna molitva further exemplify that either meter or acrostich are easily lost in translation in the attempt to convey the message of the source text.34 We should also bear in mind that the proper prosodic versification was considered the most difficult part for the poets in Byzantium and they (at least the better-educated ones) strived at keeping it according to their knowledge 34 There are three translations in modern Bulgarian. The earlier one, by Kyril Hristov of 1922 alternates ten-syllable and eleven-syllable rhymed verses with caesura after the fourth syllable from the beginning, or before the fourth from the end, has no acrostich but is the best in con- veying the emotion of the original poem. The second one, of Emanuil Popdimitrov of 1933, offers isosyllabic ten-syllable verses without attention to caesuring but with acrostich, and the last one of 1997, by Stojan Shishkov, is pretty much the same. And all this is for translations to the same language yet of different time. In the English translation mentioned at the beginning, both isosyllabicity and ceasuring are lost together with the very acrostich of the poem for the sake of conveying the message of the source text. 86 Ekaterina Dikova and skills.35 How much more difficult would it be to imitate prosody in a young literary language which knows nothing of poetic alternation between long and short vowels? And if we could only hypothesise on the possibility that Constantine of Preslav was searching for a way to imitate the visual prosody of the Byzantine iamb – and perhaps that some of this imitation, like the alter- nation of о/ѡ and /і, were lost in the transliteration and manuscript copying process – his other achievements are out of doubt. First and foremost, his prayer accords with the principles of both Byzantine poetry and rhetoric, besides to a degree that not only makes him able to pre- serve their requirements in his target language, but also allows him to invent new ways to respect them – even by exceeding models and breaking rules. I mean that Constantine allowed adjacent vowels in his poem which are oth- erwise prohibited by the rhetorical principles, all the rest of which he strictly follows, particularly the principles related to the rhythm of the dodecasyl- lable – apart from the hiatus, these are: compact structuring of verses/cola, avoidance of pleonasm, of tautology, and of enjambment as well as an isocolic arrangement.36 Conclusions The results from the statistic study presented in this article prove that all the Byzantine metrical requirements are met, as far as possible, which is visible even in the transliterated text of MS Syn. 262, which has most probably lost many of the features of its archetype (not only because of the transliteration from Glagolitic to Cyrillic script, but also in the process of multiple manuscript 35 See Bernard (2010: 108) as well as Hilberg (1900) who distinguish among Byzantine iam- bographers classic representatives (as George of Pisidia), epigones (as Theodore Prodromos), and amateurs (as the author of Χριστὸς πάσχων) – this division is mentioned numerous times by many modern scholars. 36 These rhetorical principles, contributing to the specific rhythm of the Byzantine dodecasyl- lable, are related to the ideas of εὐρυθμία and γοργότης. They are not characteristic just for this type of poetry (and are applied also in non-poetic genres) but are part of the complex of peculiar triggers of rhythm in it according to the Byzantine rhetors – mainly, yet not exclusively, in the Synopsis of Joseph Racendytes in the chapter dedicated to iambic verses (Walz 1832: 559–562). This specific peculiarity – rhetorical principles related to εὐρυθμία and γοργότης in the Byz- antine dodecasyllable – is studied in detail by Hörandner (1995: 287–290) and Lauxtermann (1998: 19–33). 87A Byzantine Poetic Form in a Ninth-Century Bulgarian Poem copying, besides in a foreign Slavonic land where local language features have indubitably left traces). Azbuchna molitva convincingly fits the tradition of the Byzantine dodeca- syllable and more precisely in its branch of paraenetic alphabets. What has been previously seen as deviation in this poem actually accords with its genre. The conclusion to be drawn here is that Maas’s postulates should be better not taken as a fixed system for all the Byzantine iambographers (and their non- Byzantine followers37) but rather as a stable basis requiring further elaborations in relation to material explored after him. Systematisation of specific features needs to continue, besides, not just from the chronological aspect (not just seen in their development) but also from the perspective of the genre (some specifics might be peculiar for a genre, apart from personal styles). The Bulgarian poem, like the Byzantine paraenetic dodecasyllabic alpha- bets, aims at presenting the sequence of the letters in an attractive memorisable way. But it is also a kind of micro catechesis as it teaches neophytes to the basic notions of the Christian faith, besides, to the extent of subtle philological details like, for instance, in verse 22, the proper pronunciation of the realia упостась (hypostasis)38 and its literal meaning (which would ensure a better understanding of the notion by the neophytes at a later stage). The most specific feature of the Byzantine dodecasyllable – its visual pros- ody – is probably impossible to fully reproduce in a Bulgarian versification. And, perhaps, for this reason, or just with the idea that it would uselessly burden his perceivers, Constantine of Preslav named his poem not iambs, but just measured prologue. There is certainly much more about this nineth- century masterpiece that we could not possibly notice or understand. It should, however, never fall in oblivion – because of its mastery and because of its deep meaning.39 37 On “iambico” in the earlier Georgian poetry see Lomidze 2021: 46 and the literature quoted there. 38 On the reconstruction of a mid-eleventh-century pronunciation of hypsilon in Byzantium see Lauritzen (2009). 39 The article is written within the frame of the project The Vocabulary of Constantine of Preslav’s Uchitel’noe evangelie (“Didactic Gospel”): Old Bulgarian-Greek and Greek-Old Bulgarian Word Indices financed by the Bulgarian National Science Fund (contract КП-06-Н50/2 of 30.11.2020). 88 Ekaterina Dikova References Anastasijević, Dragutin 1905. Paränetischen Alphabete in der griechischen Literatur. München: C. Wolf & Sohn. Arhim. Antonij (Vadkovskij) 1885. Iz istorii drevnebolgarskoj tserkovnoj propovedi. Konstantin, episkop Bolgarskij i ego Uchitel’noe evangelie. Kazan’: Tipografija Imperatorskogo Universiteta. Bernard, Floris 2010. The beats of the pen: social contexts of reading and writing poetry in 11th-century Constantinople. Ghent: Ghent University. Bernard, Floris 2018. Rhythm in the Byzantine Dodecasyllable: Practices and Perceptions. In: Rhoby, Andreas; Zagklas, Nikos (eds.), Middle and Late Byzantine Poetry: Texts and Contexts. Turnhout, Belgium: Brepols, 13–41. Bogdanov, Ivan 1980. Bezsm”rtni slova: komentirani literaturni pametnici. Sofija: Otechestven front. Bouvy, Edmond 1886. Poètes et mélodies: Étude sur les origines du rythme tonique dans l’hymnographie de l’église grecque. Nimes: Lafare frères. Butler, Thomas 1999–2022. Alphabet prayer [by] Constantine Preslavsky translated by Thomas Butler. In: Slovoto. https://www.slovo.bg/showwork.php3?AuID=48&WorkID=671&Level=1 Dobrev, Ivan 1993. Za starob”lgarskata merena rech i za razmerenostta na b”lgarskata narodna pesen. In: Sbornik ot nauchni trudove, posveten na sedemdesetgodishninata na profesor Miroslav Janakiev. Sofija: Jur. Ivanova, 239–257. Dybo, Vladimir 1971. Zakon Vasil’eva-Dolobko i aktsentuatsija form glagola v drevnerusskom i srednebolgarskom. In: Voprosy jazykoznanija 2, 93–114. Dybo, Vladimir 1977. Imennoe udarenie v srednebolgarskom i zakon Vasil’eva- Dolobko. In: Slavjanskoe i balkanskoe jazykoznanie. Antichnaja balkanistika i sravnitel’naja grammatika. Moscow: Nauka, 189–272. Eustratiadis, Sofronios 1961. Agiologion tis Orthodoxou Ekklisias. [Athens]: Apostoliki Diakonia tis Ekklisias tis Ellados. Eftymiadis, Stephanos 2014. Greek Byzantine Hagiography in Verse. In: The Ashgate Research Companion to Byzantine Hagiography, vol. II: Genres and Contexts. Farnham & Burlington: Ashgate, 161–179. Franko, Ivan 1914. Kleine Beiträge zur Geschichte der kirchenslavischen Literatur [I. Konstantins “Alphabetisches Gebet”]. In: Archiv für slavische Philologie 35, 150–179. https://www.slovo.bg/showwork.php3?AuID=48&WorkID=671&Level=1 89A Byzantine Poetic Form in a Ninth-Century Bulgarian Poem Gallucci, Eleonora 2001. Uchitel’noe evangelie Konstantina Preslavskogo i posledovatel’nost voskresnykh evangel’skikh chtenij tserkovnogo goda. In: Palaeobulgarica 25(1), 3–20. Georgiev, Emil 1938. Dve proizvedenija na Sv. Kirila. K”m v”prosa za knizhovnata dejnost na slavjanskija p”rvouchitel. (= Studia historico-philologica serdicensia. Supplementi 2). Sofija: Kultura. Gorskij, Aleksandr; Nevostruev, Kapiton 1859. Opisanie slavjanskikh rukopisej Moskovskoj sinodal’noj biblioteki. Otdel 2. Pisanija svjatykh ottsev 2. Pisanija dogmaticheskija i dukhovno-nravstvennyja. Moskva: Sinodalʹnaja tipografijа. Hilberg, Isidor 1900. Über die Accentuation der Versausgänge in den jambischen Trimetern des Georgios Pisides. In: Festschrift Johannes Vahlen. Berlin: G. Reimer, 149–172. Hörandner, Wolfram 1995. Beobachtungen zur Literarästhetik der Byzantiner. Einige byzantinische Zeugnisse zu Metrik und Rhythmik. In: Byzantinoslavica 56(2), 279–290. Kojčeva, Regina 2019. Naučnye vozzrenija Rajko Nahtigala v svete paleoslavističeskoj gimnologii. In: Stankovska, Petra; Derganc, Aleksandra; Šivic-Dular, Alenka (eds.), Rajko Nahtigal in 100 let slavistike na Universi v Ljubljani, Ljubljana: Znanstvena založba Filozofske fakultete Univerze v Ljubljani, 101–110. Kotova, Dobriela 2022. Slovo 19 ot Uchitelnoto evangelie i negovite gr”cki iztochnici. In: Palaeobulgarica 46(1), 3–28. Kuev, Kujo 1974. Azbuchnata molitva v slavjanskite literaturi. Sofija: Izdatelstvo na B”lgarskata akademija na naukite. Lauxtermann, Marc 1998. The velocity of pure jambs: Byzantine observations on the metre and rhythm of the dodecasyllable. In: Jahrbuch der Österreichischen Byzantinistik 48, 9–33. Lauritzen, Frederick 2009. Michael the grammarian's irony about hypsilon. A step towards recostructing Byzantine pronunciation. In: Byzantinoslavica 67, 161–168. Lomidze, Tamar 2021. On the Character of Georgian Verse. In: Studia Metrica et Poetica 8(2), 45–57. https://doi.org/10.12697/smp.2021.8.2.03 Maas, Paul 1903. Der byzantinische Zwölfsilber. In: Byzantinische Zeitschrift 12, 278–323. https://doi.org/10.1515/byzs.1903.12.1.278 https://doi.org/10.12697/smp.2021.8.2.03 https://doi.org/10.1515/byzs.1903.12.1.278 90 Ekaterina Dikova MacRobert, Catherine Mary 2019. Methodological Implications of Nahtigalʼs Remarks on the Acrostic Prayer. In: Stankovska, Petra; Derganc, Aleksandra; Šivic-Dular, Alenka (eds.), Rajko Nahtigal in 100 let slavistike na Universi v Ljubljani. Ljubljana: Znanstvena založba Filozofske fakultete Univerze v Ljubljani, 111–123. Mitov, Georgi 2022. Vizantijskite novozavetni kateni i Uchitelnoto evangelie na Konstantin Preslavski: ot katenata k”m homilijata (Predvaritelni nabljudenija). In: Palaeobulgarica 46(2), 81–93. MS Syn. 262 / Evangelie uchitel’noe Konstantina Bolgarskogo. 1-ja chetvert’ XII v. [Moskva]: Gosudarstvennyj istoricheskij muzej. https://catalog.shm.ru/entity/ OBJECT/164766 MS Zogr. 80. Verse Prolog for March – August, 1345–1360. Monastery of Zogrpaf, Holy Mount Athos. Nahtigal, Rajko 1942. Rekonstrukcija treh starocerkvenoslovanskih izvirnih pesnitev. I. Azbučna pesnitev. In: Razprave Akademije znanosti in umetnosti v Ljubljani, filoz.-filol.-hist. razred 1, 43–72. Petrov, Ivan P. 2022. The Greek Sources of Učitel’noe Evangelie Revisited: Sermon 20. In: Palaeobulgarica 46(2), 3–29. PG 37 1857. Patrologiae Cursus Completus, Series Graeca. T. 37: S. Gregorius Theologus, Archiepiscopus Constantinopolitanius. Migne, Jacques-Paul (ed.). Paris: Garnier. Rhoby, Andreas 2011. Vom jambischen Trimeter zum byzantinischen Zwölfsilber. Beobachtung zur Metrik des spätantiken und byzantinischen Epigramms. In: Wiener Studien 124, 117–142. https://doi.org/10.1553/wst124s117 Rhoby, Andreas 2015. Labeling Poetry in the Middle and Late Byzantine Period. In: Byzantion 85, 259–283. Romano, Roberto 1985. Teoria e prassi della versificazione: Il dodecasillabo nei panegirici epici di Giorgio di Pisidia. In: Byzantinische Zeitschrift 78, 1–22. https://doi.org/10.1515/byzs.1985.78.1.1 Slavova, Tatjana 2017. Za ezika na Uchitelnoto evangelie na prezviter Konstantin (lichni i nelichni glagolni formi). In: Palaeobulgarica 41(2), 3–21. Sobolevskij, Aleksej 1900. Cherkovno-slavjanskite stihotvorenija ot IX–X v. i tjahnoto znachenie za izuchavaneto na cherkovnoslavjanskija ezik. In: Sbornik Narodni Umotvorenija 16/17, 314–320. https://catalog.shm.ru/entity/OBJECT/164766 https://catalog.shm.ru/entity/OBJECT/164766 https://doi.org/10.1553/wst124s117 https://doi.org/10.1515/byzs.1985.78.1.1 91A Byzantine Poetic Form in a Ninth-Century Bulgarian Poem Spasova, Marija 2005. Istoricheski i kvaziistoricheski podhod pri t”lkuvaneto na svedenijata za prevodacheskata dejnost na slavjanskite p”rvouchiteli po vreme na moravsko-panonskata im misija. In: Totomanova, Anna-Maria; Slavova, Tatiana (eds.), Nest” uchenik” nad” uchitelem’ svoim’. Sbornik v chest na prof. dfn Ivan Dobrev, chlen-korespondent na BAN i uchitel. Sofija: Universitetsko izdatelstvo „Sv. Kliment Ohridski“, 106–144. Stanchev, Krasimir 1986. Ritmichnata struktura na Kirilovija “Proglas k”m Evangelieto” i na proizvedenijata ot Preslavskija stihotvoren cik”l (Starob”lgarskijat izosilabiz”m). In: Studia Slavica medievalia et humanistica Riccardo Picchio dicata 2. Roma: Edizioni dell’Ateneo, 647–652. Stankovska, Petra 2018. Predpolagaemaja reforma Klimenta Ohridskogo i dal’nejšee razvitie glagolicy: grafemy dlja napisanija nosovogo perednego rjada i ery. In: Sv. Kliment Ohridski v kulturata na Evropa. Sofija: Kirilo-Metodievski naučen Cent’’r, 402-421. Taseva, Lora 2006. Parallel’nye juzhnoslavjanskie perevody Stishnogo prologa i triodnykh sinaksarej. In: Byzantinoslavica 64, 169–185. Tihova, Marija 2012. Starob”lgarskoto Uchitelno evangelie na Konstantin Preslavski po naj-starija prepis (GIM, Sin. 262). (Monumenta linguae slavicae dialecti veteris 58). Freiburg i. Br. Veder, William 1999. Utrum in alterum abiturum erat? A Study of the Beginnings of Text Transmission in Church Slavic. Bloomington, IN: Slavica. Veder, William 2000. Das glagolitische Alphabet der Azbučna Molitva. In: Miklas, Heinz; Richter, Sylvia; Sadovski, Velizar (eds.), Glagolitica. Zum Ursprung der slavischen Schriftkultur. Wien: Verlag der Österreichischen Akademie der Wissenschaften, 77–87. Walz, Christian 1832. Rhetores Graeci, Band 3. Stuttgart: Stuttgartiae Sumtibus J. G. Cottae. Zagklas, Nikos 2019. Byzantine Poetry: An Introduction. In: Hörandner, Wolfram; Rhoby, Andreas; Zagklas, Nikos (eds.) A Companion to Byzantine Poetry. (Brill’s Companions to the Byzantine World 4). Leiden, Boston: Brill, 1–19. https://doi.org/10.1163/9789004392885_002 https://doi.org/10.1163/9789004392885_002