Written for the ATTEMT Workshop, King’s College, London, December 20–21, 2013 THE QUAESTIONES AD ANTIOCHUM: TEXT PRODUCTION AND TRANSMISSION IN SLAVONIC WILLIAM R. VEDER Laws, like sausages, cease to inspire respect in proportion as we know how they are made John Godfrey Saxe Ilse de Vos and Olga Grinchenko have produced a meticulous analysis of the Slavonic witnesses to the Quaestiones ad Antiochum available to them (de Vos 2014), and their phylogenetic software has led them to classify them in two branches, viz. a homogeneous translation and a heterogeneous translation. Unfortunately, they did not verify their analysis by spelling out those ‘translations’ and collating them with the witnesses. Homogeneous Translation The homogeneous translation is only partially homogeneous: it must be reconstructed as two translations a = codd.1 HT22T50T66T90 and a' = codd. Pr and P (ed. Porfir’ev 1890), made by the same translator from two different Greek exemplars: a 1-16 18-23 25-51 55-77 79-101 103-105 107-111 115-124 126-132 a' 1-16 18-23 24 25-51 52-54 55-77 78 79-101 102 103-105 106 107-111 112 115-124 125 126-132 133-135 1: Inventory of Versions a and a' Translation a’s structure fully parallels that of cod. Oxford Bodleianus Auct. F.4–7 (12 questions lacking, Q.25 after 26 and Q.120 after 130). Translation a' is based on a Greek exemplar different not only in structure, but also in many readings, e.g. Q.27 οὐσία : + τῆς οὐσίας τῶν ψυχῶν = сѫщьство + отъ сѫщьства дѹшевьнаѥго a'; Q.28 µετὰ ταυτὴν τὴν ἐπιφάνειαν : µετ' αὐτὴν τὴν ἐπιφάνειαν τοῦ Χριστοῦ = по семь ꙗвлении : по самомь ꙗвлении христовѣ a', δύναται : + ὥστε οὐδεὶς Θεοῦ οὐσίαν γυµνὴν ἰδεῖν οὐδαµοῦ = можетъ + тѣмьже никътоже божиѥ сѫщьство наго никътоже видѣ никакоже a'; ἀσθένειαν : add τῶν ὀφθαλµῶν = чловѣчьскꙑѩ же ради немощи : противѫ чловѣчьскою очию немощи a'. In addition, translation a' shows marked efforts to improve the grammatical correctness of translation a, e.g. Q.32 въ адѣ сѫщиихъ ни бога поминаѭтъ → сѫщѧ (Npl), Q.33 дѹша и тѣло раꙁлѫчивъша сѧ → раꙁлѫчивъши сѧ (du), Q.35 41 98 чьто ради → чесо (G), Q.45 бꙑстъ прьвоѥ отъкръвениѥ повелѣваѭще → повелѣваѩ and Q.47 съвѣдѣтельствѹѥтъ писаниѥ глаголѭщеѥ → глаголѧ (Nsg), Q.64 благословлѧѩ сꙑнꙑ іѡсифовꙑ → дъва сꙑна іѡсифова (du). As the body of translations a and a' is identical in text, and as neither translation technique, nor lexicon, nor morphosyntax vary significantly, translation a' must be identified as additions to translation a, made by the author of a from a different Greek exemplar. Heterogeneous Translation 1 The non–bold sigla are those of de Vos 2014; to them can now be added A – Sbornik, Moskva RGB F.173.I nr.50, f.1–43v. Written for the ATTEMT Workshop, King’s College, London, December 20–21, 2013 The heterogeneous translation is heterogeneous to the power of three: it must be reconstructed as one translation (b = codd. L and T4) and two series of selections from it (c = codd. T12M and Z, and d = codd.2 T91T2T3). a' 1-16 18-48 49 50-77 78 79-93 94 95-112 115 116 117 118-119 120-123 124-131 132 133-135 b 1-16 17 18-48 50-77 79-93 95-112 113-114 116 118-119 124-131 133-135 137 2: Inventory of Versions a' and b Translation b is a revision of translation a', based on a third Greek exemplar with 127 questions and numerous readings different from the exemplars of a and a'. The revision is evident from the retention of readings of a', e.g. Q.14 οὐκ ἔστιν ῥῆµα ἐκφοβοῦν καὶ διασκορπίζων → нѣстъ глаголъ ѹстрашаѩ и растачаѩ a' → нѣстъ глагола ѹстрашаѭща и растачаѩ (not G), Q.15 λογισµός· ἀπαλλάττεται δὲ τοῦ τοιούτου → помꙑслъ· свобаждаѥтъ сѧ таковаго a' → мꙑсль· гонеꙁаѥтъ же ѥго (not f), Q.24 ἅπαντα τὰ ἐπισυµβάντα → вьсꙗ сълѹчьшаꙗ сѧ a' → вьсе събꙑвъша сѧ (not sg), Q.33 ἡ λύρα, ἐὰν µὴ ἔχῃ τὸν κρούοντα → цѣвьница аще не иматъ ѹдарꙗѭщаѥго a → цѣвьница аще не иматъ ѹдарꙗѭщаѥго ѭ a' → гѫсли аще не имѫтъ ѹдарꙗѭщаѥго въ нѧ (from addition in a'), Q.39 ὡς ξύλον ἀργὸν τήν ποτε εἰκόνα → ꙗко дрѣво просто иже нѣкогда иконѫ a' → ꙗко просто дрѣво и праꙁдьно бꙑвъшѫѭ древле иконѫ (retention from a' + retranslation), Q.54 εἰς νύκτα καὶ εἰς σκότος → въ нощи и тьмѣ a' → въ нощи и въ тьмѫ (not A), Q.103 πόθεν γίνονται θανατικά; → отъкѫдѹ бꙑваѭтъ съмрьтьнаꙗ нахождениꙗ рекъше моръ a' → отъкѫдѹ бꙑваѭтъ мори (from gloss in a'); it is also evident from systematic correction, e.g. λοιπόν → 40x прочеѥ a': of these, b strikes 20 and for the others enters occasional replacements, of which only the three instances of къ томѹ (67 82 135) are adequate to the meaning of the text. Translation b is, in fact, a partial translation, accounting for different wording and additions to Greek text units and adding 4 questions (as well as deleting 10); its main thrust is to archaise and enhance the Slavonic of a'. b 1-4 5 6 7 8-9 10 11 12 13 14 15-18 19-20 21 22-23 24 25-26 27-31 32 33 34 35 36 37-67 68 69 70 71 72 73-77 79 80 c 5 7 10 11 13 19-20 21 22-23 24 25-26 32 35 36 68 69 73-77 79 d 11 15-18 19-20 22-23 25-26 32 33 35 69 71 b 81 82 83 84 85-86 87-89 90-92 93 95-97 98 99-112 79-93 113 114 116 118-119 124-127 128 129–131 133-135 137 c 81 83 85-86 90-92 98 128 d 81 83 87 90-91 113 129 3: Inventory of Versions b, c and d Versions c (34 questions) and d (23 questions) are not translations, but excerpt versions from translation b, their Slavonic text edited without recourse to Greek exemplars. Version d is a hybrid: 11 questions are excerpted from c (lightly edited) and 11 more from b (heavily edited). The sequence b → c → d can be illustrated by the following collations: Title: d retains the simplification of c but obfuscates authorship (see Veder 2014b) b Свѧтаѥго аѳанасиꙗ архиепископа алеѯанъдрьскаѥго· къ анътиохѹ кънѧꙃѹ· о мъножаишихъ въꙁисканиихъ· въ божьствьнꙑихъ писаниихъ недоѹмѣѥмꙑихъ... c Въпрошениѥ анътиоха кънѧꙃа· и отвѣти блаженаѥго аѳанасиꙗ аръхиепископа⁘ d Въпроси прѣподобьнаѥго анътиоха кънѧꙃа· отъвѣти свѧтаѥго епифаниꙗ⁘ Q.11: the devil is demoted from subject to object in two steps b тѣмьже и диꙗволъ виновьнъ побѣдѣ и вѣньцемъ бꙑваѥтъ любѧщиимъ бога⁘ c да ѥмѹже не ѹдолѣѥтъ диꙗволъ· вѣньць прииметь· ꙁане въꙁлюби бога⁘ 2 To them can now be added A1 – Izmaragd, Moskva RGB F.173.I nr.46, f.132–136, 183–184, 267v– 270. Written for the ATTEMT Workshop, King’s College, London, December 20–21, 2013 тако иже къто диꙗвола прѣборетъ· тъ побѣдьнꙑи вѣньць прииметъ отъ бога⁘ Q.91: d develops the text of c c аще ли живъ бꙑлъ немилостивъ· то ꙁатвори богъ о немь· d аще ли живъ немилостивъ бꙑлъ· то неприѩтьна того по съмрьти милостꙑни· бѫдꙑ бо живъ· не ѹ болѣаше по своѥи дѹши даꙗти самъ· c обаче лѣпо намъ ѥстъ сьде пещи сѧ своими дѹшами· а не надѣꙗти сѧ прощеномъ бꙑти d обаче лѣпо нꙑ ѥстъ въ животѣ комѹжьдо о своѥи дѹши печаловати· а не надѣꙗти сѧ по съмрьти прощенѹ бꙑти c щѹждиими приносꙑ⁘ d щѹждꙑими приносꙑ· отънѫдь бо мало дѹши пособиѥ скѫпаѥго· по немь даѥмꙑими· а иже въ животѣ къто милостивъ и подативъ· и по съмрьти того дѹши мъного пособиѥ· раꙁдаваѥмꙑими о немь· того ради апостолъ рече· ни отъ скръби· ни отъ тѫгꙑ· тиха дателꙗ любитъ богъ· рекъше ѥгда въ съдравии и въ обилии живѫще· даꙗти о съпасении дѹшѧ своѥѩ· оскръбѣти· каѭще сѧ о грѣсѣхъ· то велико ѥстъ прѣдъ богомь· о томь апостолъ рече· а не ꙗко въ бѣдѣ и въ напасти даꙗти· о христѣ іисѹсѣ· вьсегда и нꙑнѣ и присно⁘ d Dating the Slavonic Texts As we see, the Slavonic translation of the Quaestiones ad Antiochum has not one text but five. Their sequence a → a' → b predates ca. 930, when 17 questions were excerpted from b into the Knjažij Izbornik, a Slavonic edificatory florilegium (ed. Veder 2008). The abridged versions c and d should not be dated much later, because they presuppose intimate knowledge of the text, which cannot be acquired by casual reading; such versions normally are by–products of the full text (if not translating or editing it, then at least producing the fair copy). Version d was made for the Izmaragd in 164 Chapters, a rambling compilation from a variety of Slavonic sources, among which the Knjažij Izbornik (see Veder 2014b): the 11 questions from c form its chapter 80, the 12 from b fill its chapters 78, 81, 112 and 145. This compilation has been mistakenly attributed to 14th century Russia (Jakovlev 1893, Pudalov 1996). The Transmission of Text a The witnesses HT22T66T50T90 (and A, see note 1) are descended from a Glagolitic protograph (ⰰ), written rather inattentively, as can be deduced from faulty readings transmitted in all copies: confusion of P → ⱔ in милостꙑнѭ → милостꙑнѧ (832 88) винѫ сиѭ → винѫ сиѩ (103), вещь ѭже → вещь ѩже (131); P → ⱔ and ⱐ → ⱁ in глаголѭтъ → глаголѧ то (40); P → ⰵ in бѫдѫтъ → бѫдетъ (79); P → ⰵ and ⱔ → ⰵ in подобаѭтъ ѩже → подобаѥтъ ѥже (120); ⰵ → P in вънидетъ → вънидѫтъ (21); ⰲ → ⱅ in похѹливъ → похѹлитъ (73); ⱈ → ⰳ in драхмъ → драгмъ (6); ⰽ → ⱈ in вьсꙗкъ → вьсѣхъ (108); ⰾ → ⱅⱃ and ⱒ → ⰿ in сластьхъ → страстьмъ (11); ⰿ → ⱒ in прѣдъварьшиимъ дѹшамъ → прѣдъварьшиихъ дѹшь (33); ⱅ → ⰿ in хощетъ молѧи сѧ → хощемъ молѧи сѧ (122); ⱐ → ⱁ in длъжьнъ → длъжьно (130); ⱐ → ⱔ in живꙑи → живѧи (124). The extant witnesses are copied not from the protograph itself, but from two Glagolitic copies of it, viz. ⰰ1 (AT66) and ⰰ2 (HT50T90 and T22, contaminated with copy ⰱ of a'), which differ in layout (ⰰ1 has two syntagms in marginal position, retained by AT66), text length (ⰰ2 has three sizeable sauts du même au même in 35–36, 84 and 119, both have different shorter omissions), and – moderately – in lexicon and grammar. The fact that both copies were written in Glagolitic can be deduced from faulty readings opposing both copies: confusion of ⱔ → ⰵ in ꙗвлꙗѥтъ сѧ, обрѣтаѥтъ сѧ ⰰ1 : Written for the ATTEMT Workshop, King’s College, London, December 20–21, 2013 ꙗвлꙗѩ сѧ, обрѣтаѩ сѧ ⰰ2 (108); ⰶ → ⱋ in и мѫжи ⰰ1 : имѫщи ⰰ2 (98); ⱄ → ⰻ in а҃· исарии ⰰ1: исарии ⰰ2 (for ассарии, 103); ⱁ → ⰵ в ѡкрадаєми ⰰ1 : ѡкрадеми ⰰ2 (for окрадоми, 126); ⱛ → ⰻ in юношьскꙑими ⰰ1 : иночьскꙑими ⰰ2 (29); anagram ⱃⰰⱁ → ⰵⱃⰵ in еретичьскаѥго ⰰ1 : праотьчьскаѥго ⰰ2 (67), and tautograms ⰲⱁ → ⰲⱁⱅⱁ in цѣсарьство ⰰ1 : цѣсарьство то ⰰ2 (42), ⰽⱁ → ⰽⱁⰳⱁ в вьсꙗкого ⰰ1 : вьсꙗко ⰰ2 (26), ⰿⱐ → ⰲⱖⰿⱐ in ѹмьнѣ ⰰ1 : ѹвꙑ мьнѣ ⰰ2 (122), ⱀⱁ → ⱀⱁⰳⱁ в ѥдиного начѧлиѥ ⰰ1 : ѥдиноначѧлиѥ ⰰ2 (1) and ⱅⱁ → ⱅⱁⱀⱁ and P → ⱔ in кто сѫть ⰰ1 : что носѧть ⰰ2 (58). All six witnesses are copied from these two exemplars: they retain their separative readings. That AT66 are direct copies from Glagolitic is proven by their individual retention of Glagolitic letters: архїепсⱂа T66 (Title) and ⰴо A (97). Тhat the others are, is proven by their individual faulty readings3. In T50, jotation is lacking of ⰰ and ⱛ in ꙗкова любо → какова ѹбо (73); confusion of P → ⱔ in божиѭ → бж҃їа (42), глаголѭ → гл҃а (45) and гл҃ѧ←ю (16); ⱅ → ⰴ in потъщашѧ сѧ → подщаша сѧ, подъщаша сѧ (12), ⱅ → ⰸ in навата → наваꙁа (72), ⱂ → ⱀ in прьвꙑи → нръвыи (3); ⰵ → ⰰ in ѥдиноѩ → е←адиноꙗ (98), ⰵ → ⱁ in въꙁищемъ → вmꙁыщом (123). In T90, jotation is lacking of ⰰ in ꙗдрѣхъ → адрѣхъ (28); confusion of P → ⱔ in глаголѭ → гл҃ѧ (16); ⰿ → ⰴ in малѣхъ → далѣхъ (132), ⱍ → ⱋ in тѧжьчаишии → тѧжщашє (76). In H (Q.107 only), jotation is lacking of ⰰ in таковꙑѩ → таковыа; confusion of ⰾ → ⰱ in о тѣлѣ → ѻ тебѣ; ⰰ ↔ ⰵ in нѣкоѥго растворениꙗ → нѣкоего растворенїе and сѫще → сꙋща; ⱑ → ⰰ in вѣдѣти → вѣдати; tautogram ⱂⱁ → ⱒⱐⱂⱁ and ⰿ → ⱅ in таковꙑи попѹщаѥмъ → таковыхъ попꙋщаетъ and haplogram ⰱⱁⱀⰰ → ⱀⰰ, tautogram ⱄⱅ → ⱄⱏⱄⱅ and P → ⱁ in ѹбо на страсть блѫдьнѫѭ → ѹ насъ страстїю блꙋдною. In T22, individual faulty readings are as numerous as in the five other witnesses together (and as colourful, see its ꙗкова любо → ѡкова лꙋбо (73), where T50 above reads какова ѹбо): this is because the scribe has to cope with two Glagolitic exemplars and is, consequently, very much distracted. The Transmission of Text a' The witnesses Pr and P are descended from a Glagolitic protograph ⰱ, but it is not clear whether directly or indirectly (as in the case of ⰰ above). More witnesses are needed to decide whether separative readings are individual or inherited. Still, Pr and P can be shown to be direct copies from Glagolitic exemplars. P follows ⰱ, but adds from b 3 questions (17, 113–114), 2 clauses (in 1 and 18) and words or collocations in 51 places.4 In P, jotation is lacking of ⰰ in іꙗкѡв- → іаков(39 64), христиꙗн- → хрстіан- (up to 42, thereafter хрстіян-), диꙗвол- → діавол- (up to 57, thereafter діявол-) and бꙑваѭтъ → бываѫть (127), and is excessive in ѫроде → юроде (722), отънѫдь → ѿнюдъ (113), тѣлеса → тѣлеся (114); confusion of P → ⱔ in тꙑсѫщ- → тысѧщ- (26 118) and желаѭтъ → желая (119); ⱔ → P in милостꙑнѧ → млстню (83), пѹстꙑнѧ → пустыню (98); P → ⰵ in вънидѫтъ → внидетъ (101), бꙑваѭтъ → бываетъ (103); P → nV в мѫжиѥ → множае (69): ⱔ → ⰵ в сѫщѧ → сꙋще (1), ѩже → еже (35 39), июдѣѩ → іюдее (43), приближаѭщѧ сѧ → приближающе ся (47), сиѩ → сие (56), иꙁдꙑхаѭщѧ → издыхающе (105), съдѣловаѩ → содѣлѡваетъ (131); Vm → ⱔ in имѫщемъ → имꙋща (802); ⰳ ↔ ⱈ in господѹ → хрстосъ (124), сьдешьнꙗѥго → здешнихъ (101); ⰴ → ⰿ in дъвоихъ → многихъ (26); ⰿ → ⰲ in 3 In the survey of faulty readings in the witnesses, the sign ← marks autocorrections and the sign ↓ erasures. Full–text collations of all witnesses are available in a dossier ‘Pseudo–Athanasius’, which can be requested from <hilandar@osu.edu>. 4 P is an example of very early textual scholarship in Slavonic. Its author annotates the text of the questions with references to parallel readings in a vast range of patristic works in Slavonic translation. Porfir’ev 1890 deserves praise for choosing this witness for the editio princeps of the Slavonic translation. Written for the ATTEMT Workshop, King’s College, London, December 20–21, 2013 влькомъ → волковъ (42), прѣмѣньноѥ → превратное (94), ⰿ ↔ ⱅ in съматрꙗѭщи → сотворяюще (93) and ѥстъ → есмь (1), поꙁнаваѥтъ → познаваемъ (2), приѥмлетъ → приемлемъ (93), хощетъ → хощемъ (132); ⰿ ↔ ⱒ в съмотрительнѣ → сходнѣ (28), вѣрьнѣишиимъ → вѣрнѣишихъ (44) and свѧтꙑихъ → ст҃ымъ (39), большиихъ даровъ → болшимъ даромъ (70), реченꙑими → реченныхъ (70); ⱀ → ⰲ in ини же → и вси иже (15), ⱀ → ⰾ in и не могꙑ → или могіи (118); ⱂ → ⱀ in покаꙁани → наказани (702), ⱅ → ⱂ in трислъньчьнѣѥмь → присл҃нчномъ (72), истачаѭтъ → испущаютъ (80); ⰰ ↔ ⰵ in раꙁличиꙗ → различіе (31), ꙗвѣ → евѣ (47), ꙁнамениꙗ → знамение (111) and слѹжениѥ → служеніѧ (26), малакиѥ → малакия (72), благотворениѥ → бл҃готворенія (90), съгрѣшениѥ → согрѣшения (107); ⰰ → і in блажении → ближния (72); ⰵ ↔ ⱁ in июдѣѥмъ → іюдеѡмъ (37) and прообраꙁи → преобрази (63); ⰹ ↔ ⱓ in врѣмени → временю (113); ⱁ → ⱃ in томѹ же → трие же (98), ⱁ ↔ ⱐ в вьсꙗко → всѧкъ (2) and равьнъ → равно (129); ⱛ → ⰻ in пророкѹ → прорки (92), ⱛ → ⱁ in навѹходоносора → новоходоносора (88); ⱐ → ⰵ in тѣлесъ → тѣлесе (58), сь → се (92), нъ въ → не во (108); ⱑ → ⱖ in славѣ → славы (48), овѣхъ → ѡвыхъ (982); anagrams жльтость → же толстъ (24), своꙗ → вся (69), tautograms ⱄPⱅⱐ → ⱄPⱅⱐⱄⰲⱔⱅⱐ in сѫтъ анъгели → сꙋть ст҃іи агг҃ли (12), ⱀⰹ → ⱀⱐⱀⰹ in ни → но ни (21), ⰾⰹ → ⰾⰹⰾⰹ in анъгели → агг҃ли ли (29), ⱄⰹⱃ → ⰿⰹⱃⱄⰹⱃ in сирѣчь → миръ сирѣчь (37), ⰵⰿ → ⱐⱀⱁⰿ in ѹбоꙃѣѥмь → убо земномъ (85), ⰶⱃⱐ → ⰶⰵⰶⱃⱐ in жрьтвꙑ → же жертвы (91), ⱁⱀ → ⱁⱀⰵⱀ in монтанитомъ → манентанитомъ (116), ⱋⱐ → ⱋⱐⱍⱐ в тꙑсѫщь свѧтꙑихъ → тысящ чстыхъ (118), ⰾⰹ → ⰾⱐⰾⰹ in свободи ли сѧ → свободилъ ли ся (129). Pr faithfully renders ⰱ, but after completion adds from a copy of b Q.17 and a clause to Q.18, as well as a note to reverse the sequence 26–25 (all marginal). In Pr, individual faulty readings exceed those of P by a factor 8: jotation is consistent only in initial position (save 12x їꙋд-) and after prefixes, elsewhere it is most often lacking, e.g. моѥю → моеꙋ (but моею, 130), ѫ is never jotated, but usually replaced by ѧ (e.g. бꙑваѭ- → бываѧ- 19 26 34 48 72 82 84 123, against бываѫ- 100 1032 105, бываю- 29), it is excessive in трава → трѣва (25), праотьчьскаѥго → прѣотечьскаго (67), четꙑре → четырїе (123); palatality is not marked in лаꙁарева → лаꙁарова (21), авелꙗ → авела (57 67), кесарева → кесарова (87), кесарю → кесарꙋ (872); confusion of P → ⱔ in вън ѫтрь → вън ѧтръ (34), ѫтробѣ → ѧтробѣ (55), ѫроде → ѧроде (722); ⱔ → P in сѫщѧ → сѫщѫ (1), отъстоѩще → ѿстоѫще (1), потъщашѧ сѧ → потъщашѫ с (12), вьсеѩ → въсеѫ (5), ѥдиноѩ → единоѫ (23), раꙁрѹшѧтъ сѧ → раꙁдрꙋшаѫт сѧ, раꙁливаѩ сѧ → раꙁливаѫ сѧ (24), прѣдъварьшѧѩ → прѣдваршѧѫ (25), дѹшѧ → дш҃ѫ (35), владꙑчицѧ богородицѧ → влдчцѫ бц҃ѫ (39), повелѣваѩ → повелѣваѫ (45), благословлꙗѩ → бл҃гсвлѣѫ (64), твоѩ → твоѫ (69), сѫщѧѩ → сѫщѫѧ, стоѩтъ → стоѫть (69), щѧдитъ сѧ → щѫдит сѧ (70), чьтѧи → чътѫи, обещьщаѩи → ѡбеꙁчъщаѫи (72), съгрѣшаѩи → съгрѣшаѫи (78), нѫждѧ → нѫждѫ (83 119), въждѧдахъ → въꙁжѫдах (86), истъщѧтъ сѧ → истъщѫт сѧ (87), помꙑшлꙗѩ → помышлѣѫи (93), дѹшеборьствѹѭщѧ → дш҃ѧ раꙁдѣлѣѫщѫѫ (105), настоѩщаѥго → настоѫщаго (106), живѧи → живѫи (124), ꙁьрѧщѧѩ → ꙁрѧщѫѫ (125), троицѧ → троцѫ (13 72); P → ⰵ in прѣльщаѭтъ → прѣльщает (125); P → ⱐ in сѫмьнѣни- → съмнѣнї- (16 532 1082); ⱔ → ⰵ in сѫщѧ → сѫще (47); V → ⱔ in блиꙁньца → блиꙁньцѧ (63), ѥже → ѧже (134); ⰾ → ⱀ in ꙁемлетрѹдьникъ → ꙁемнотрꙋдникь (117); ⰿ ↔ ⱒ in мрьтвꙑимъ → мр҃твыхь (24), по ꙁавѣтомъ → по ꙁавѣтох (69) and съкрѹшенꙑихъ → съкрꙋшеным (92); ⱅ → ⰴ in анътиохѹ → андїѡхꙋ (Title), анътиохии → андїѡхїи (45); ⰰ ↔ ⰵ in вьсꙗ → въсе (97) and имѫще → имѧща (107); ⰵ → ⱁ in садъдѹкѣѥмъ → садꙋкеѡм (24), писаниѥмъ → писанїѡм (44), въꙁдꙑмлениѥмъ → въꙁдымленїѡм (103), манихѣѥмъ → манихеѡм (116); ⰵ ↔ ⱐ in чесо→ чъсо- (55 61), чесо → чьсо (69 98) and съжьгѫтъ → съжегѫт (89), прѣмѣньноѥ → Written for the ATTEMT Workshop, King’s College, London, December 20–21, 2013 прѣмѣненное (94), полѹдьнь → полꙋдн҃е (123); ⱛ → ⱐ in имѫщѹ → имѧщъ (118); ⱐ → ⱁ in малъ → мало (105); ⱐⰹ → ы in женьскꙑи → женmскы (24), ратьничьскꙑи → ратничьскы (31); ⱖ ↔ ⱑ in христовꙑихъ → хв҃ѣх (44), ꙁълꙑмъ → ꙁлѣмь (121) and инѣхъ → иныих (129); epenthesis in раꙁрѹшѧтъ сѧ → раꙁдрꙋшаѫт сѧ (24), раꙁрѣшение → раꙁдрѣшенїе (72); haplograms ⰱⱁⰱⰻ → ⰱⱁ in бо би божии → бо бж҃їи (103), ⰳⰰⱃⰰ → ⰳⰰ в дрѹга раꙁлѫчаѥма → дрꙋга ꙁлѫчаема (119), ⱛⱛ → ⱛ in съпасению ѹспѣшьнаꙗ → спсенїꙋ спѣшнаа (131), ⰹⰿⱐⰹⱒⱐ → ⰹⰿⱐ in облагодѣтивъшиимъ ихъ → ѡ блгдти бывшїим (117). The Transmission of Text b The witnesses L and T4 are descended from a Glagolitic protograph ⰲ, which is marked by an anagram ⱁⰲⱑ → ⱑⰲⱁ in провѣдѧ → прѣводѧ (79) and a tautogram ⰹⱀⱁⰳⱁ → ⰹⱀⱁⰳⱁⰹⱀⱀⱁⰽ in ѥдиного тъкъмо → единого инока (1), but it is not clear whether directly or indirectly (as in the case of ⰰ above). More witnesses are needed to decide whether separative readings are individual or inherited. Still, L and T4 can be shown to be direct copies from Glagolitic exemplars. L lacks Q.91–92 as well as the answer to 110 and the question to 111. In L, individual faulty readings are as numerous as in Pr: jotation is consistent only in initial position (save 10x іѹд-) and after prefixes, elsewhere it is often lacking, e.g. мѡѵсею → мѡусеѹ, мѡисеѹ, 59 (but моею, 130), ѫ is never jotated, but usually replaced by ѧ (e.g. 64х глаголѭ- → гл҃ѧ- against 18x бꙑваѭ- → бываѫ-), while ѧ is replaced over 200 times by ѫ (e.g. 16 приѩ- → приѫ-), it is excessive in трава → трѣва (25), and дѣлателе → дѣлателие (111); palatality is not marked in авелꙗ → авела (57 67), лаꙁарь → лаꙁаръ (69), кесарю → кесарѹ (87); confusion of P → ⰵ in раждаѭтъ сѧ, ѹмираѭтъ → раждаетm сѧ, ѹмираеть (106), P → ⱐ in лѫща → льща (412), инакѫ → инакъ (118); ⱔ → ⰵ in страждѫщѧ → страждѫще (105), покрꙑваѭщѧ → покрываѫще (133), ⱔ → ⱐ in осьлѧ → оселъ (40); ⰲ → ⰿ in въ мрацѣ → мъ мрацѣ (60), ⰲ → ⱒ in прѣводѧ → прѣходѧ (79); ⰳ → ⰽ in льгъчаѥ → лекmчае (101 107, against легmчаи, 116); ⰴ → ⰱ in добрѣ → бобрѣ (69), ⰴ → ⰲ in ѱифади → псифави (125); ⰶ → ⰸ in дрьжитъ → дрьꙁить (93); ⰿ ↔ ⱒ in некрьщенꙑимъ → некръщеныхь (67) and длъжьнѹѥмꙑихъ → дльжнѹемомъ (Title), ѹбогꙑимъ → ѹбогыхь (70) and похотиѭ → помышлѣѫ (98); ⱀ → ⰿ in тварьноѥ → твармое (52); ⱅ ↔ ⰴ in анътиохѹ → андиохѹ (Title) and анътихристъ → андихристь (109 125, against 5x анти-), дъва → та (36), ⱅ ↔ ⱀ in имѫтъ → имѧнm (69), ни → ти (3); ⱇ → ⱅ in наꙁареѳа → наꙁарета (44); ⱌ → ⱍ in слъньце → слънmче (114); ⱍ → ⰳ in чловѣча → гл҃а (16); ⰰ ↔ ⰵ in небеса → нб҃се (12), бещьствѹѭща → беꙁчьствѹѫще (15), аще → єже (21), вьсꙗ → въсе (109) and словесе → словеса (53); ⰵ ↔ ⱁ in словесемь → словосемъ (1), четверѣ → четворѣ (18), июдѣѥмъ → іѹдеомь (373 72), мѡѵсеѥвꙑимь → мѡисеовѣмь (38), мѡѵсеѥвѹ → мѡусеовѹ (53), седмер- → седмор- (522), срьдьцевѣдьць → сръдцовѣдець (100), въскѹрениѥмъ → въскѹрениомъ (103), манихѣѥмъ → манихеѡмъ (116) and прообраꙁова → преѡбраꙁова (44), ⰵ → ⱛ in движению → движение (96), ⰵ ↔ ⱐ in обрѣꙁаѥте сѧ → ѡбрѣꙁает сѧ (38), потрѣблена → потрѣбна (42) and съжьгѫтъ → съжегѫтm (89), вьси → веси (112), тъчиѭ → течиѫ (98); ⱁ ↔ ⰻ in добровоньна → добрѡвина (47) and прилѣплꙗти сѧ → пролѣплѣти сѧ (58), ⱁ → ⱛ in послѹшьство → послѹшьствѹ (89), ⱁ ↔ ⱐ в само → самъ (42) and сѫдьбъ → сѫдѡбъ (69 105), тъ → то (88), никꙑиже → никоиже (47), нѣкꙑимь → нѣкоимъ (114); ⱐ ↔ ⰻ in плъть ѭже → пльтиѫ же (114), тьмами → тьмамъ (88), ⱐⰹ → ъи in триѵпостасьнꙑи → триипостасенъи (1), ⱐⰹ → ы in ѥдинꙑи → едины (37), ракъ и → ракы (39), сѫботьнꙑи → сѫботны (53), четврьтꙑи → четвръты (67), кꙑи → нѣкы (71), сластьнꙑи и тѣлесьнꙑи → сластны и тѣлесны (87), могꙑи → могы (88), Written for the ATTEMT Workshop, King’s College, London, December 20–21, 2013 тьлетворьнъ и → тлетворны (103), съмрьтьнꙑи → съмрmтны (105), таковꙑи → таковы (108), скѵѳьскꙑи → скиѳскы (109), женьскꙑи → женmскы (116), съвѧꙁавꙑи → свѧꙁавы (118); ⱖ ↔ ѫ in слѹгꙑ → слѹгѫ (113); ⱑ ↔ ⱖ in сътьрѣ сѧ → сътры сѧ (24) and мѡѵсеѥвꙑимь → мѡисеовѣмь (38), ⱑ → ⰰ in имѣниꙗ → иманиа (90); epenthesis in живѧщиꙗ → живлѣѫщиа (18), любьникꙑ → любленикы (20), раꙁрѹш- → раꙁдрѹш- (242 98), ꙁемли → ꙁемлы, ꙁемли (43 114 137, against 8x ꙁеми), ꙁемлѭ → ꙁемлѫ (62, against 4x ꙁемѧ), недѫживи → недѫжmливи (113), negative in праꙁдьн- → праꙁн- (39 72); anagram ⰽⰲⱃⱐ → ⰽⱃⱐⰲ in бескврьньни → бескръвны (105); haplograms ⰲⱁⱄⰲⱁ → ⰲⱁ in чловѣчьство своѥ → чловѣчьство ѥ (72), ⰼⱔⰴⰰ → ⰼⱔ in въждѧдахъ сѧ → въждѫх сѧ (86), ⱀⰰⰿⱐⱀⱖ → ⱀⱖ in намъ нꙑнѣ → нинѣ (72), ⱀⰵⰲⱐ → ⰲⱐ in невъꙁможьна → въꙁможна (114), ⱀⰵⱃⰵ → ⱃⰵ in ꙗко не рече → ꙗко рече (72); tautograms ⰲⱁ → ⰲⱁⰲⰵ in дъвоѥ → двоеве (71), ⰶⱔ → ⰶⱔⱀⱔ in лежѧщаѥго → лежѫнѧщаго (90), ⱀⰰ → ⱀⰰⱀⰰ in нашихъ → нанаших (1), ⱑ→ ⱑⰰ in видѣти → видѣати (28). Т4 omits questions 27–31, 38–39, 42 and breaks off after Q.78 (perhaps owing to a defect in the exemplar). Had its text been complete, the individual faulty readings in Т4 would have outnumbered those of Pr and L; as it is, they surely exceed them in variety, which points to lack of competence on the part of the scribe. Jotation is lacking in вьсѭ → всꙋ (5), юг- → ѹг- (372), южьнꙑѩ → ѹжныѧ (37), иноѩ → иноа (43), отънѭдѹже → ѿнꙋдѹже, ꙗко → аки (44), исходѧщиꙗ → исходѧщиа (47), еѵанъгелиꙗ → єуаглїа (54), мѣнѭ → мнꙋ (63), іꙗкѡвъ → иаковъ (64); confusion P ↔ ⱔ in блѧдѫтъ → блꙋдѧтъ (10), простꙑнѭ → простынѧ (72) and живѧщиꙗ → жївꙋщаѧ (18); ⱔ ↔ nV in не тьлѣѩ → нетлѣнїѧ (242) and начинати → начѧти (53 542); P ↔ ⰵ in чѹѭтъ → чюєтъ (34), съкаꙁаѭтъ → скаꙁаєтъ and бѫдетъ → бѹдꙋт (72); ⱔ ↔ ⰵ in божиѩ → божиє (43), имъ даѩ → єгда есть (69), ѹтѣшѧтъ сѧ → ѹтѣшаєть сѧ (69) and ꙁнамении → ꙁнамѧнїи (24); ⰱ → ⰳ in благаꙗ → гл҃гаѧ (69), ⰱ → ⱂ in бо вѣдѣти → поведаєтъ (73), небрѣгꙑ → непрегыи (36), боѩтъ сѧ → поѧтъ сѧ (40), ⰱ → ⱎ in дрьꙁнѫ бо → дерьꙁнꙋша (72); ⰲ → ⰱ в навата → брата (72), ⰲ → ⰳ in іѡвовꙑ → єговы (71), ⰲ → ⰾ in вѣцѣ → в↓члѣцѣ (35); ⰳ → ⱈ в грѣха → г←х)ѣха (72) and христовѣ → гсни (19); ⰴ → ⰲ in дивьно → виновно (16), ⰴ → ⰳ in дѣлѣ → годѣ (21), ⰴ → ⰾ in доньдеже → донележе (20) and дѣлѣ → годѣ (21), ⰴ → ⱂ in плодъ → плод←пь (2), ⰴ ↔ ⱅ in дъвѣ → двѣ(←три) (9), дъва → ты (36) and творимъ → дворим (41), тако → т←дако (72); ⰶ → ⰸ in жльчь → ꙁолчь (182); ⰸ → ⰿ in раꙁарꙗѭщѫ → рамраꙁорѧщꙋ (43); ⰽ → ⱈ and ⰼ → ⰾ in покѹждениѥ → похꙋлениє, покѹждаѭще → похꙋлѧюще (72) and христиꙗнина → крщнїѧ (2), христиꙗнѣхъ → кр҃щеныхъ (69), 11x христиꙗн- → крстиꙗн-; ⰾ → ⱅ in польꙃевало би → полꙁовати (36), послаби → постаби (44); ⰿ → ⰲ in оградомъ и овощемъ → ѻвощовъ и ѻградовъ (48), ⰿ → ⰳ in имъ даѩ → єгда есть (69), ⰿ ↔ ⱅ in поꙁнаѥмъ → поꙁнаєть (23), отвѣщаѥмъ → ѿвѣщаєтъ (26), дѣиствѹѥмꙑ → дѣиствꙋєтъ (33) and обрѧщетъ сѧ → ѻбрѧщемъ сѧ (53), омꙑѭтъ и очистѧтъ → ѻмываємъ и чистимъ (70) , проститъ → прости имъ (77), ⰿ ↔ ⱒ in мрьтвꙑимъ → мр҃твыхъ (24), бꙑваѭщамъ молитвамъ и милостꙑнꙗмъ → бывающих мл҃твах и млстнѧмъ (34) and прѣображаѭщиихъ сѧ → преѻбражающих←мъ сѧ (26), ни женахъ → намъ (58); ⱀ → ⰲ and ⱄ → ⰴ in принесошѧ → приведѡша (70), ⱀ → ⰳ in ни → г҃ь (1), ⱀ ↔ ⰴ in не бо нъ → не бѹди (10), ⱀ → ⰿ in и намъ → имамъ (54), ⱀ → ⱅ in ни → то (50), не → н←тѣ (54), плъньства → плотьства (72); ⱄ → ⱍ in съто → что (36), сего → чего (44); ⱅ → ⰲ in петръ → петръ и павелъ (73), ⱅ → ⱋ in ѹмираѭтъ → ѹмирающ←тѣ (71); ⱇ → ⱅ in ѳарсисъ → тарсисъ (66); ⱒ → ⰲ in въсходитъ → бъꙁводитъ (72); ⱋ → ⱍ лѫщѧ → алчьбы (41); ⱌ → ⱈ in цѣсареви → хсви (11); ⰰ ↔ ⰵ in аще → еже (21), съмотрениꙗ → смотрениє (60), прочаꙗ → прочеє (62), сѫпротивѧща сѧ → сꙋпротивѧще сѧ, писаниꙗ → писаниє (72) and ѥще → аще (69), хѹлениѥ → хѹлениѧ, бещьстиѥ → бещестиѧ, поклоньше сѧ → Written for the ATTEMT Workshop, King’s College, London, December 20–21, 2013 поклониша сѧ, веельꙁевѹломь → вельꙁавꙋломъ (72), ⰰ → і in дѹша истаѥго → дш҃а ст҃го (2); ⰵ ↔ ⱁ in отъ небꙑтиꙗ въ бꙑтиѥ → ѿ небытиѧ· (↓но) в бытїє (3), вѣньцемъ → вѣньцомъ (11), тѣлеси → тѣло (16), жен- → жон- (16 24 51, against 3x жен-), ѥще → ещо (21), младеньцемь → младенцомъ (24), исплъненоѥ → исполненоѻ (52), бѫдѫщемь → бѹдѹщомъ (722), себѣ → собѣ and словесе → словѣ←осе (72); ⱁ ↔ ⱛ in бещиньно → бес чинꙋ (54) and ѹдолѣѭща → ѻдолѣюща (73), ⱁ ↔ ⱐ in вьсꙗко → всѧкъ (2), воюѥтъ → въюєтъ (43), плодъ тъ → плодо тъ (50), отъ → ѿѻ (53), раꙁѹмѣѥмо → раꙁꙋмѣємъ, собьства → събьства (72) and иꙁгънанъ → иꙁгнанно (47), ꙁабъвениѥ → ꙁабовенїѧ (75); ⱐ → ⰵ in сѫщь → сꙋще (47 53), таиньство → таинество (60), ⱐ → ⱛ in ѥмь → ємꙋ (13), вѣрьнѣишиимъ → вѣрнѣиша ємѹ (44); ⱛ → ⱔ в іюда → ѧда (75); ⱐⰹ → ы in ѥдинꙑи → едины (37), нъ и → нынѧ (70); чѧдь и → чѧды (69); ⱑ → ⱖ in съмотрьнѣ → смотри ны (33), кожанѣ → кожаны (56); epenthesis only in коꙁиꙗхъ → коꙁлѧхъ (70), lacking in ꙗвлꙗѭтъ сѧ → ꙗвѧтъ сѧ (11) and приѥмлѭтъ → прїимꙋтъ (70); angrams ⱅⱑⱀ → ⱅⰲⱑ in свои стѣнь → своиствѣ (1), ⱃⱁ → ⱁⱃ in ѹкромѧще сѧ → ѹкормѧще сѧ (1), ⱂⱄⰰ → ⰰⱂⱄ in псалмѣ → апслмѣ (19), ⱍⱑⱄⱅ → ⱅⱐⱍ чѧстьно → точьно (20), ⱃⱐⰲ → ⰲⱐⱃ in бескръвьнꙑѩ → бесъкверны (34), ⰹⰾⱁ → ⱁⰾⰹ in милотьхъ → молитвах (70), ⱎⱐⱀ → ⱀⱐⱎ in сьдешьниихъ → ꙁденышних (70), ⱃⱐⰴ → ⰴⱃⱐ in милосрьдию → мдрсть (73); haplograms ⱀⰰⱄⱀⰰⱀⰹ → ⱀⰰⰹ in творена сн҃а ни → творена и (1), ⱛⱛ → ⱛ in вьсемѹ ѹѥдиненꙑ → всемꙋ єдины (22), ⰸⱐⰾⰰⰰ → ⰸⰰ in ꙁълаꙗ → ꙁа (69), ⰴⰰⰶⰵⱀⰵⰲⱑ → ⰲⱐⰶⰹⰲⱁⱅⱑ in даже не вѣстъ → в животѣ (71), ⱀⰰⰿⱐⱀⱖ → ⱀⰰⰿⱐⱀⰵ in намъ нꙑнѣ → намъ не (72), ⰱⱁⰱⱁ → ⰱⱁ in ѹбо бога → ѹ бг҃а (72), ⰴⱏⱁⰾⱏ → ⰾⱐ in прѣдъ олътаремь → пре лтаремь (74) ; tautograms ⰿⱐⰹⱒⱐ → ⰿⱐⰹⰿⱐⰹⰿⱐ in длъжьнѹѥмꙑихъ → долъжьнꙋємъ· и мы ѻ томъ (Title), Pⰱ → Pⱅⱐⰱ in глѫбити → гл҃итъ быти (1), ⱅⰹ → ⱅⰹⰴⰹ in десѧти драхмъ → десѧти· дидрагmма (6), ⱅⱐ → ⱅⱐⱅⱁ in бꙑваѥтъ → бываєтъ· то (11), ⱀⱁ → ⱀⱁⰱⱁ in писано → писано бо (12), ⰿⰰ → ⰿⰵⱀⰵ in имамъ → именемъ (24), ⱅⱐ → ⱀⰵⱅⱐ in тьлѣниѥ → нетлѣниє (24), ⰲⰰⰵ → ⰲⰰPⱋⰵ in отвѣщаѥмъ → ѿвѣщавающѣмъ (26), ⱃⰰⰸ → ⱃⰰⰿⱃⰰⰸ in раꙁарꙗѭщѫ → рамраꙁорѧщꙋ (43), ⰻⱁ → ⰻⱁⰻⰵ in іоръдана → и ѻ їерданѣ (44), ⱁⱅⱁⰿ → ⱁⱅⱁⰿⱀⱁ in о томь → ѿ многых (69), ⰲⱁ → ⱀⱁⰲⱁ in волꙗ → но волѧ (69), ⱅⱁ → ⱅⱁⱂⱁ in то не → то поне (71), ⰵ → ⰵⱁ in оригенъ → ѻригеѻнъ (72), ⱀⰵ → ⱀⰵⱀⰵ in коньць → коненець (72), ⰽ → ⰽⰳ in вьсѧ грѣхꙑ → всѧкъ грѣхъ (73), ⰶⱐⰹ → ⰶⰵⰾⰵ in тѧжии вьсѣхъ → тѧжеле всѣхъ (76). The Transmission of Text c The witnesses T12, M and Z are descended from a Glagolitic protograph ⰳ, but it is not clear whether directly or indirectly (as in the case of ⰰ above). More witnesses are needed to decide whether separative readings are individual or inherited. Still, T12, M and Z can be shown to be direct copies from Glagolitic exemplars. T12 contains only the first five questions of the selection. Its individual faulty readings are excessive jotation in ѹне → ѹнѥ (5), вꙑшьнюѥмѹ → вышьнѥмꙋ, глаголетъ → гл҃ѥть (10), ꙁане → ꙁанѥ (11); confusion ⱂ → ⰲ and ⱁ → ⱐ in попѹщена → въпѹщена (13); ⱁ → ⱐ in когда → къгда, прѣꙁорьствомь → преꙁорьствъмь (10); and lack of epenthesis in ꙗвлꙗѭтъ сѧ → ꙗвѧть сѧ (11, as in M). M contains 31 of the 34 questions. Its individual faulty readings are lack of jotation of non–initial vowel except ю (non-initial ѧ is often written as ꙗ); confusion of P → ⱔ in глаголѭ- → гл҃ꙗ- (19 68 73 85), имѫ- → имꙗ- (20 692), тꙑсѫщи → тысѧщи (26), въсприѥмлѭтъ → въспрїемлꙗтъ (69), милостꙑнѭ → милостынѧ (83) and млстынꙗ (90), P → ⰵ in раꙁрѹшаѭтъ сѧ → раꙁдрꙋшаетm сꙗ (24), бꙑваѭтъ → бываеть (26); ⱔ ↔ ⰵ in сѧ → се (11), ѹтѣшѧтъ сѧ → ѹтѣшает сꙗ (69) and оскврьнивъше сѧ → осквръниша сꙗ (73), ⱁ → P in коѥ → кꙋю (11), ⱛ → ⱔ in ѹмѹ чловѣчѹ → ѹмꙋ чл҃чꙗ (81); confusion of Written for the ATTEMT Workshop, King’s College, London, December 20–21, 2013 ⰲ → ⰾ in вѣцѣ → чл҃цѣхъ (35); ⰴ → ⰿ in дрѹꙃии → мнѡѕи (10), длъга → мнѡга (69), ⰴ → ⱅ in дѣмоньскꙑи → теменъ (7); ⰾ → ⰳ in больнъ → богатыи и боленъ (92), ⰾ → ⱀ in раꙁличима → раꙁно ꙫчима (24), ⰾ → ⱅ in славлениꙗ → съставленїа (26); ⰿ → ⰲ in грѣхомъ → грѣховь 92b:2, ⰿ → ⱒ in имъ → ихъ 32b, добрꙑихъ дѣлъ → добрымъ дѣломъ (69), бранимо → браних (23); ⱄ → ⰾ in исаиꙗ → іліа (10); ⱅ → ⰿ in ꙁнаѥмъ → ꙁнаетъ (22) and бѫдетъ → бꙋдемъ (24); ⱋ → ⱍ in просѧщиихъ → прочихъ (85); ⰰ → ⰵ in сѫща → сꙋще (24) and съкрꙑвьше сѧ → скриша сѧ (81); ⰵ → ⱛ in ѥже → ѹже (69); ⱁ → ⰵ in ѹдолѣѥтъ → ꙋделѣеть (11), ѹдолѣлъ → ꙋделѣлъ (73); ⱛ → ⱁ in дѹши → дѡщи (26); ⱐ → ⱁ in грѣхъ → грѣхом (77), ⱐ → ⱛ in сѫщемъ → сꙋщемꙋ (23); epenthesis in раꙁрѹшаѭтъ сѧ → раꙁдрꙋшаетm сꙗ (24), ꙁемьнꙑимъ → ꙁемлънымъ (26), lacking in ꙗвлꙗѭтъ сѧ → ꙗвѧт се (11); convoluted anagrams in бранимо ли ѥстъ → браних рани мл҃енїе (23) and ѹже сꙑнъ мьнии → сн҃омь ѹже ни (73); haplograms ⱀⱁⱀⰰ → ⱀⰰ in чѧстьно наслаждениѥ → часть наслажденїа (20), ⱀⰰⰿⰵ → ⱀⰰ in ꙁнамениꙗ → ꙁнанїа (22); tautogram ⰿ → ⰿⰿ in раꙁѹмѣи → раꙁꙋмьмѣи (74). Z contains a selection of 29 questions, of which 36, 86 and 98 are not included in M. Its individual faulty readings are lack of jotation in диꙗвол- → дїавол- (102 11), вьсꙗ → вьса (36), олътарю → ѡлmтарꙋ (74), excessive jotation in вꙑшьнюѥмѹ → вышнѥмꙋ (10), глаголетъ → гл҃ѥт (91); confusion ⱔ → P in дѹшѧ → дш҃ꙋ (69), обличѧтъ сѧ → обличають сѧ (11), въсхꙑтѧтъ → вьсхыщают (83), милостꙑнѧ → млстню (85), лежѧщаѥго → лежꙋщаго (90); P → ⰵ in раꙁдаѭтъ → раꙁдаеть (69)\, P → ⱐ in глѫбоцѣ → гльбоцѣ (69); ⰱ → ⱂ in бꙑла → прибила (35); ⰳ → ⱈ in господь → х҃с (74); ⰴ ↔ ⱅ in прѣгадаѭтъ сѧ → прѣгатают се (10), дѣла → тѣла (35) and то → да (69); ⰼ → ⰴ in прѣжде вѣдѣлъ → прѣдвѣдль (36); ⰾ → ⱃ in дѣла → дѣра (35); ⰿ → ⱒ in положимъ → положихѡм (90); ⱅ → ⰲ in обратитъ → ѡбративь (79); ⱋ → ⱅ in раꙁѹмѣѭще → раꙁꙋмѣют (85); ⰵ → ⰰ in ꙁабъвениѥ → ꙁабmвенїа (75), ⰵ → ⱐ in тѣмьже → тѣмьжь (69). The Transmission of Text d The witnesses T91T2T3 (and A1, see note 2) are descended from protograph ⰴ (the Izmaragd in 164 Chapters). They are all copies of one and the same exemplar, written in 2–column layout (retained in T91), which explains the frequent transpositions of segments of text in A1T2. That A1 is a direct copy from Glagolitic is proven by its retention of the Glagolitic letter ⰴ in ⰴш҃а (16 33). T91 renders the Izmaragd only selectively; it omits 5 questions between 20 and 33, and three between 33 and 81. Its individual faulty readings are lack of jotation in диꙗвол- → дїавол- (153), ѥѩ → еа, величаниꙗ → величанїа (15), житиꙗ → житїа (87 113), пиꙗньчивъ → пїанmчивъ, коньчаниꙗ → кончанїа (113); confusion ⱔ → P in летѣвъшѧ → летѣвшꙋ (71); P → ⱁ in нѫдитъ сѧ, нѫдѧщии сѧ → нодит сѧ, нодѧщїи сѧ (83), P → ⱐ in нѫждьнꙑ → ньжны (83); ⱛ → P in пронѹрити → пронѫрити, иꙁнѹритъ → иꙁнѫрить (87); ⱔ → ⰵ in обличѧтъ сѧ → обелечет сѧ (11), ошьдъшѧѩ → ѿшедше (19), дѹшѧ → дѡще (22); ⰳ → ⱂ in глѣнъ → плѣнь (18), ⰾ → ⰿ in ли → л(←м)и (25); ⰰ ↔ ⱖ in невѣдома → невѣдомы (71) and бѫдꙑ → бѹда (91); ⰵ → ⰰ в съставлꙗѥтъ сѧ → съставлѧа (16); ⱛ → ⱁ in дѹшѧ → дѡще (22), не ѹ → не ѡ (91); ⱐ → ⰵ in очищьше сѧ → ѡчищеши сѧ (71), прошьшемъ → прошешимъ (113); ⱐⰹ → ы in сельнꙑи → селны (25), прѣдавꙑи → предавы (113); anagrams ⱄⱅPⱂ → ⱂⱛⱄⱅ в отъстѫпѧтъ → ѿпѹстѧть (18), ⱀⱖ → ⰹⰿ в нꙑ → имъ (113), ⰴⰰⰴⱐ → ⰴⱐⰴⰰ in вилдадъ → валдъда (113); haplograms ⱀⰰⰴⰵ → ⰴⰵ in осмь на десѧте → п҃ (71), ⱁⱋⰵⰿⱐⱃⰵⱋⰹ → ⱁⱋⰵⰿⱐ in хощемъ рещи → хощемъ (113). Written for the ATTEMT Workshop, King’s College, London, December 20–21, 2013 In Т3, individual faulty readings are lack of jotation in диꙗволъ → дїавол- (112 15 ), іюда → иѹда (113); confusion ⱔ → P in летѣвъшѧ → летѣвшꙋ (71), P → ⰵ in глаголѭтъ → гл҃еть (19); haplogram ⱀⰰⰴⰵ ↔ ⰴⰵ in осмь на десѧте → п҃ (71) ; tautogram ⱂⱁ ↔ ⱀⰵⱂⱁ in похѹлѭ → не похѹлю (15). In Т2, jotation is consistent; confusion ⱔ ↔ ⰵ in ѹмираѭщѧ → ѹмирающе (71); ⰵ → P in бꙑваѥтъ → бываютъ (15); ⰶ → ⱋ in глаголѭ же → гл҃юще (18); ⰸ → ⱀ in прѣꙁорьства → пронырства (15); ⰾ → ⱁ in слѹꙁъ → соѹꙁъ (18); ⰿ ↔ ⱒ in грѣшьнꙑимъ → грѣшныхъ (35) and грѣшьнꙑихъ → грѣшнымъ (23); ⱀ → ⰲ in вънидѫтъ → ввѡдитъ (69), ⱀ → ⰿ in раꙁлѫчена → раꙁлꙋчаема (33); ⱀ → ⱃ in тъности → трости (16); ⱅ → ⰴ in прѣтварꙗѭтъ сѧ → водворѧют сѧ (26), ⱅ → ⱀ in поꙁнатиѥ → поꙁнание, поꙁнатиꙗ → поꙁнанїѧ (23); ⱃ ↔ ⰴ in пронѹрити → принꙋдити (87); ⱋ ↔ ⱅ in причѧщаѭще сѧ → причащают сѧ (113); ⱖ → ѫ in бѫдꙑ → бꙋдꙋще (91); ⰰ ↔ ⰵ in раꙁлѫчаѥтъ сѧ → раꙁлꙋчение (16), голѫбица → голꙋбице (18) and ѥще → аще (69); ⱁ ↔ ⰵ in рока → река (113); excessive epenthesis in неподативи → неподатливы (91); anagram ⰰⰶⰵ → ⰶⱐⰰ in бо ꙗже → бж҃їѧ (33); haplograms ⰿⱐⰿP → ⰿP in съвъкѹплениѥмь мѫжа → совокꙋпленїе мꙋжа (16), ⱀⰰⰴⰵ → ⰴⰵ in осмь на десѧте → осмь десѧть (71), ⱀⰵⱍⱐ → ⱍⱐ in нечьстиви → честивїи (71); tautograms ⱂⱁ ↔ ⱀⰵⱂⱁ in похѹлѭ → не похꙋлю (15), ⰵ ↔ ⰵⱁ in оригенъ → ѻрегеѡнъ (16). 3 Exploring the Glagolitic → Cyrillic Interface Copies from Glagolitic inevitably retain traces of their exemplars, be it 1 Glagolitic letters, 2 problems with Glagolitic numerical values (not attested in the transmission of the Quaestiones), 3 problems with jotation of vowels and marking of palatality of consonants, 4 confusion of nasals, 5 confusion of consonants, 6 confusion of vowels, 7 epenthesis of consonants and vowels, 8 parablepses (anagrams, haplograms and tautograms)5. This is because cyrillisation of a text written in Glagolitic requires transliteration and transcription simultaneously under context–sensitive rules6. This daunting task is not achieved easily: the incidence of parablepses in almost all witnesses shows how challenging the morpheme–by–morpheme (and sometimes letter– by–letter) process of decryption and transliteration + transcription can be. Of the 18 Slavonic witnesses to the Quaestiones none is perfect, and the prayers of the scribes to the readers not to judge errors harshly, but to correct them, are not vain. They are, in fact, no more than students’ copies, learning both to handle Cyrillic spelling without the benefit of a model to follow and acquiring a new text and its language. Researchers should, like schoolmasters, first grade them (with passing grades for AHT3T12T50T66T90 and failing for LPrT4T22) before allowing them to testify to anything but their respective exemplars. As with any classwork, coincidence in error is rife, but its diagnostic value (for copying from elsewhere) is not high: in the transmission of text a faults coinciding in two or more witnesses cannot be linked to any of the two exemplars in 86% of cases, and of the remaining 14% most concern features 3 and 7, which cannot be attributed to Glagolitic, and 5–6, which should be attributed to 5 The order of these features is observed in the layout of the survey of faulty readings above. Consonants must be transliterated, save ⰱ ⰲ ⰿ ⱂ when palatalised and ⱇ when used for Greek φ, vowels must be transcribed when jotated or following a palatal consonant, otherwise transliterated. The rules are complicated by the fact that ⱛ (ѹ), even when not following a palatal consonant, and ⰼ (жд) must be transcribed by digraphs, while ⰽⱄ (ѯ) and ⱂⱄ (ѱ) must be transcribed by monographs, and that ⱐ (ь) may not be transliterated after non–palatal consonants, but must be transcribed by ъ. Note that confusion ь ↔ ъ is not included in the survey above: it is so widespread that it would obscure all other data. 6 Written for the ATTEMT Workshop, King’s College, London, December 20–21, 2013 Glagolitic only with serious adstruction). The probability of coincidental variants with these features being conjunctive is negligible. That these features have not been noticed over the past century of Slavonic studies is not surprising. They are easily overlooked when reading just one witness at a time, because they need full–text lineated collation of all available witnesses to come to the fore, and full–text collation has not yet found its place in the slavist’s toolkit7. Strange as it may seem, the scholar who most systematically applied the technique in his work, Nicolaas van Wijk, did so only for Middle Dutch texts8, not the Slavonic he worked on. A contributing factor, of course, is that Slavonic studies is not all that interested in the Slavonic of the texts, but rather in what appears to be not Slavonic in them: it seeks dialect traits as signs of development of the language. For instance, historical linguistics would treat the three occurrences of о in Т4 жона, жонꙋ, жоны as indubitable signs of the dialect of the scribe and would treat his three spellings with е as mere calques from the exemplar, which would be fine if the exemplar were written in Cyrillic; but this particular exemplar was written in Glagolitic, wherefore the three occurrences of о may neither be dissociated from the 10 occurrences of о → е, nor from any other of the faulty readings of Т4 (from the point of view of text transmission, they are merely unintended accidents). A third factor is that Slavonic studies has imprudently convinced itself that Glagolitic writing was a phenomenon restricted to the first 30 years of Slavic literacy, whereas new texts can be shown to have been produced in Glagolitic for at least another two centuries (see Veder 2014a). Concepts and Conclusions The Slavonic texts of the Quaestiones ad Antiochum present an unequivocal sequence of translation → enhancement → revision → selection with concomitant changes in text. In the same sequence are related the Slavonic texts of the Scala Paradisi, the translation9, its enhancement10 and its revision11 based on three different Greek exemplars12, the selections13 made without recourse to Greek. The Quaestiones and the Scala are not only related typologically, they share the same revisor: the corrections, deletions and substitutions he applies to the translator’s text are largely identical in both. It is he who gives both the status of Slavonic texts (of sorts) only in third instance. But there is a problem in the way the Scala Paradisi has been treated by Slavonic studies. The original translation, which still awaits to be dissociated from its en7 The only example of a Slavonic collation I know from the past century are the four pages Miroslav Janakiev inserted in Stojanov, Janakiev 1976: 109–112. From the past decade, we have the impressive work of Ostrowski, Birnbaum, Lunt 2004, but it has as yet no follow–up. 8 Van Wijk’s collations (all in longhand) were still consulted at Leiden University Library in 1968, when I there started working on my dissertation. 9 E.g. codd. Moscow RGB F.173.I nr. 152, F.304 nr. 155, 158, 161, 167, 168, 169, 170, 183, 184, 185, all copied from the same Glagolitic exemplar (see Veder 2015). Full–text collations of all codd. here and in notes 10–11 and 13–14 are available in a dossier ‘Lestvica’, which can be requested from <hilandar@osu.edu>. 10 E.g. codd. Moscow RGB F.304 nr. 156, 157, 159, 160, 162, 171, copied from the same Glagolitic exemplar. 11 E.g. codd. Moscow RGB F.256 nr. 198 and 199, F.304 nr. 10, copied from the same Glagolitic exemplar. 12 The translation generally agrees with the edition of Sophronius 1883 and the Latin translation by St Ambrose of Camaldoli 1531); the Greek of the enhanced version is unedited, but can be read in excerpts on many websites of the Greek Orthodox Church; the revision generally agrees with the edition of Rader 1633. 13 E.g. cod. St–Peterburg RNB Pogodin 1032 (15th c.), f.184–188. Written for the ATTEMT Workshop, King’s College, London, December 20–21, 2013 hancement, has been identified as ‘Athonite translation of the 14th century’ (Popova 2011: 180), or more narrowly ‘Athos, ca. 1370’ (Sobolevskij 1903: 11). But its revision, known from the earliest extant copy (nr. 198, ca. 1150–1200), has for the past century and a half been identified as ‘the first translation’ (Popova 2011: 135), made at ‘Preslav, 10th century’ (Sobolevskij 1908: 11). The selections from it have not been studied, but one of them was used already to compile the Scaliger Paterikon before ca. 950 (see Veder 2005: 110–11914). Slavic studies has inversed the relation between translation and revision. Why? First, the lack of full–text collation has prevented the formation of any clear concept of textual relationships other than translation and non–translation (usually termed ‘redaction’). Second, Slavic linguistics operates under the assumption that Slavonic is a language that develops like any living language, and it has never stopped to test the validity of that premise. Consequently, any text that is not in conformity with the ‘canon of Slavonic manuscripts’ (not texts, see Schaeken, Birnbaum 1999: 93–127) up to the early 12th century must be dated later (both the translation and the enhancement of the Quaestiones would have been similarly postdated, if Slavic linguistics had got hold of them). Finally, the confusion of the concepts Slavonic and Old Bulgarian, especially in Bulgarian scholarship, leaves no room to distinguish less Bulgarian texts (the revisions and selections) from less Slavonic (the translations and their enhancements). There is need for such distinction: not only are the translations and enhancements of the Quaestiones ad Antiochum and the Scala Paradisi not in conformity with the ‘canon’, but so are the lapidary inscriptions (Schaeken, Birnbaum 1999: 127–130) and the so–called ‘folk’ vitae of Sts John of Rila15, Proxor of Pčinja, Gabriel of Lesnovo and Joachim of Osogovo (Gergova 1996: 43–62), which apparently never got the benefit of a thorough revision. Also there are texts assigned to remote dates or places, which evidently do not belong there because their transmission, like that of the Quaestiones and the Scala, is based on Glagolitic exemplars, e.g. the ‘Middle Bulgarian’ (= ‘14th c.’) translation of the Epistle of Patriarch Photius to Prince Boris– Michael (Slavova 2013), the Works of Dorotheus of Gaza (Dimitrov 2013) and the Kniga o Troici i o věrě (Velinova 2015), or the ‘Russian’ Pčela (Pičxadze 2009) and Prolog (Prokopenko 2010). One wonders, what full–text collations and their linguistically unbiased analysis could bring to light about other translations assigned to the 14th century (see recently Taseva, Jovčeva 2004) or to Russia (see Veder 2012). All in all, Slavonic studies has a long way to go before it can adequately explain its texts. References Ambrose of Camaldoli, Saint 1531 Ioannis Climaci Abbatis Montis Synai Scala Paradisi. Venezia, repr. Köln 1583, repr. PG 88. Dimitrov, Kamen 2010 Rečnik–indeks na Slovata na avva Dorotej (po răkopis 1054 ot sbirkata na M.P. Pogodin). Veliko Tărnovo. 14 Cod. Leiden UB Scal. 74 (13th c.). The earliest vita of St John is usually dated ‘before 1183’, because it ignores the translation of his relics to Hungary, but it is by no means excluded that it was composed shortly after his death in 946. 15 Written for the ATTEMT Workshop, King’s College, London, December 20–21, 2013 2010 Avva Dorotej. Slova, srednobălgarski prevod. Grăcko–bălgarski slovoukazatel. Veliko Tărnovo. Gergova, Emilia 1996 Zapadnobălgarskite anahoreti v starobălgarskata literatura. Sofia (Dissertation). Jakovlev, Vladimir A. 1893 K literaturnoj istorii drevnerusskix sbornikov. Opyt issledovanija Izmaragda. Odessa, repr. Leipzig 1974. Ostrowski, Donald, David Birnbaum, Horace G. Lunt 2004 The Povest’ vremennykh let. An Interlinear Collation and Paradosis. Cambridge, MA. Pičxadze, Anna A. et al. (eds) 2008 Pčela. Drevnerusskij perevod, 1–2. Moscow. Popova, Tat’jana G. 2011 Slavjanskaja rukopisnaja tradicija Lestvicy Ioanna Sinajskogo. Saarbrücken. Porfir’ev, Ivan Ja. 1890 ‘Apokrifičeskie skazanija o novozavetnyx licax i sobytijax’. Sbornik ORJaS 52. Prokopenko, Larisa et al. (eds.) 2010 Slavjano–russkij Prolog po drevnejšim spiskam: Sinaksar’, Sentjabr’– fevral’,1–2. Moscow. Pudalov, Boris M. 1996 Sbornik „Izmaragd‟ v drevnerusskoj literature. S.–Pb. (Dissertation). Rader, Matthäus 1633 S. Patris Joannis Climaci opera omnia. Paris, repr. Patrologia Graeca 88. Schaeken, Jos, Henrik Birnbaum 1999 Die altkirchenslavische Schriftkultur. München 1999 (= Slavistische Beiträge 382) Slavova, Tatjana 2013 Slavjanskijat prevod na poslanieto na patriarh Fotij do knjaz Boris–Mihail. Sofia. Sobolevskij, Andrej I. 1903 ‘Perevodnaja literatura Moskovskoj Rusi XIV–XVII vv.’ Sbornik Otdelenija russkogo jazyka i slovesnosti 74/1, repr. Leipzig 1989. 1908 Drevnjaja cerkovnoslavjanskaja literatura i ee značenie. Xar’kov. Sophronios: Σωφρόνιος ὁ Ἐρηµήτης 1883 Κλίµαξ τοῦ ὁσίου πατρὸς ἡµῶν Ιωάννου. Constantinople, repr. Athens 1959. Stojanov, Stojan, Miroslav Janakiev 1976 Starobălgaski ezik. Tekstove i rečnik. Sofia (1st ed. 1956). Taseva, Lora, Marija Jovčeva et al. (eds.) 2004 Prevodite prez XIV stoletie na Balkanite. Sofia. Veder, William R. 2005 Hiljada godini kato edin den. Sofia. 2008 Knjažij Izbornik za văzpitanie na kanartikina, t. 1–2. Veliko Tărnovo. 2012 ‘Glagolitic Books in Rus’’ in: Brian J. Boeck et al. (eds.) Dubitando. Studies in History and Culture in Honor of Donald Ostrowski. Bloomington: 315–334. 2014a ‘Why Wish Away Glagolitic?’ Ricerche slavistiche 12(58): 373–385. 2014b ‘Gennadius Slavicus.’ In: Srednovekovnijat čovek i negovijat svjat. Veliko Tărnovo: 273–285. Written for the ATTEMT Workshop, King’s College, London, December 20–21, 2013 2015 Glagolitic Manuscripts in the Trinity–St Sergius Laura. In: Cristiano Diddi (ed.) Πολυίστορ. Scripta slavica Mario Capaldo dicata. Moskva: 351–364. Velinova, Vasja 2015 ‘Ošte vednăž za Kniga na vjarata’. In: Sbornik, posveten na 100–godišninata ot roždenieto na akademik P. Dinekov. Sofia, forthcoming. de Vos, Ilse, Olga Grinchenko 2014 ‘The Quaestiones ad Antiochum Ducem. Exploring the Slavonic Material.’ Byzantion 84: 105–143.