The Arabic Life of John the Merciful and its relationship with the Syriac version and the Greek text

The Greek text

The Greek tradition of the Life has received noteworthy attention from the group led by P. A. Cavallero at the University of Buenos Aires. He has dedicated the last fifteen years to the literary production of Leontios of Neapolis. In addition to the Spanish edition of the Life,2 which has substantially improved the edition of A. J. Festugière,3 these developments include studies of the multiple versions of the Greek text.4 Cavallero’s group is also studying the fragments and the unedited recensions of the Life and is working on a new critical edition, which will be published soon. This edition will consider a new collation of the Greek manuscripts, the Syriac version, and the Latin versions.5

The Latin version (BHL 4392)

Regarding the ancient translations of the Life, many discoveries have been added to the information collected in the study of V. Déroche, which provided scholars with important starting points for studying the Latin, Syriac, and Arabic versions.6 The critical edition of the anonymous Latin version known as BHL 4392 has recently been published, which had remained the only unedited version after the one by John of Amalfi.7 The edition provides a text of extreme interest because of its significant affinities with the Syriac version.8

The Syriac version

In 2020, I published a new critical edition of the Syriac version, using recensions, manuscripts, and indirect witnesses unknown to P. Bedjan, who had published the text for the first time.9 This edition details the manuscript tradition, a stemma codicum, and an Italian translation.10 More recently, I published other studies on the subject: a critical edition of all the existing recensions of the Syriac text, with an explanation of their mutual relation; a study of the ecdotic link between the Syriac version and the Greek manuscript tradition; and a study of the contribution that the Syriac provides to the reconstruction of the Greek original.11

The Arabic version

New data concerning the Arabic version of the Life have emerged recently. The Arabic version, which is almost entirely unedited,12 has been recently studied by André Binggeli, who in 2019 presented his work in Leuven at a conference dedicated to the Lavra of Mar Saba. The study, which will be published soon in the Proceedings of the conference, considers cases of several hagiographical works with a tradition in Arabic and Syriac, among which is the Life of John the Merciful.13

Breydy’s remarks on three excerpts of the Arabic version

The Arabic version of the Life of John the Merciful, preserved in several manuscripts of different eras,14 is still inaccessible to scholars who do not read Arabic. An exception regards the three excerpts quoted in the Annals of Eutychius, edited and translated into German by M. Breydy, reproducing the content of chapters 6, 18, and 45 of the Life.15 Breydy observed that the source of Eutychius must have been a Sinaitic copy of the Arabic Life: the three excerpts preserve a high-quality text, reproducing more accurately the Greek than the Arabic manuscripts copied in Mar Saba do.16 Apart from the ecdotic relationship between the Annals and the Life of John the Merciful,17 it is noteworthy that Breydy had already realized the significant differences, even in the earliest manuscripts, between the Greek text and the Arabic version “réalisée dans les milieux melchites palestiniens, qui prétend traduire directement d'un original grec, mais dont la correction stylistique laisse beaucoup à desirer.18

The study of Vincent Déroche

In his monograph on Leontios of Neapolis, V. Déroche highlighted the most relevant data provided by Breydy's study. First, he remarked that the earliest manuscripts, such as the Syriac version, presented the story in 48 chapters.19 Secondly, the Arabic version indirectly reflects the Greek text's longer and more complete recension, the so-called Vie Longue.20 Moreover, due to insufficient evidence, Déroche does not take a position on the Greek or Syriac derivation of the Arabic version.21

Affinities between Greek, Syriac, and Arabic in one of Breydy’s extracts

In the preface of my edition of the Syriac version, I inform the reader that the Arabic version —in the form preserved by the Annals of Eutychius, the only text accessible to me at that time— preserves original details of the Greek, which are absent in the Syriac text.22 In the same paragraph, I suggest that the Arabic and the Syriac versions may depend on the same branch, considering the subdivision of the work into 48 chapters, which is certainly not original.23 These two remarks were based on the analysis of a passage that the Annals had taken from chapter XVIII of Life of John the Merciful, containing the list of provisions sent by the Patriarch to Modestus of Jerusalem, who was rebuilding the city after its destruction by the Persians in 614.24 On the one hand, the correct translation in Arabic of the term μαινομένη, “dried fish”, absent in Syriac, seems to support the translator’s direct knowledge of the Greek original; on the other hand, the order of the provisions is different in the Greek text compared to the Syriac and Arabic versions,25 suggesting a close link between the two latter.26 Without directly examining the Arabic manuscripts, I did not take a position on the Greek or Syriac model of the Arabic translation, just stating that Syriac and Arabic belong to the same branch of the manuscript tradition. In theory, it could not be excluded that the Arabic version was translated from a Greek model similar to the one used by the Syriac translator.

Two new hypotheses on the Arabic Life’s origin

The recent analysis of the Arabic version by André Binggeli, based on the earliest manuscripts (Sinai Ar. 428 and Sinai Ar. 431), convincingly demonstrates a close relationship between Syriac and Arabic versions. First, Binggeli provides a table showing the almost perfect correspondence between the chapter numbering in Syriac and Arabic: the Arabic version has the same omissions and reorganization of chapters we find in Syriac.27 He then offers a table that compares pieces from the last chapter: the Arabic version presents the same omissions, additions, and reformulations that distinguish the Syriac version from the Greek original.28 The Arabic version is much more condensed than the Greek text: it follows the Syriac (almost ad verbum) in its drastic reduction of rhetorical and descriptive sections. In conclusion, Arabic and Syriac create a different narrative fabric from the one conceived by Leontios of Neapolis. Nevertheless, the Arabic version keeps one original detail of the Greek which is not present in Syriac: the mention of orphans and widows, which could reveal a direct knowledge of the Greek original. Binggeli concludes by formulating two alternative hypotheses: a) Syriac and Arabic were translated from the same lost Greek model; b) the Arabic version was translated from Syriac and eventually revised according to the Greek original. To support the claim that the Arabic translator knew the Greek original, Binggeli adduces the colophon of Sinai Ar. 428, which informs that the Life of John the Merciful has been translated from Greek (rūmī) into Arabic at Mar Saba.29 He parallels the Martyrdom of the Forty Monks of Sinai and Raithu, a hagiographic text translated from Syriac into Arabic and later revised according to the Greek original, as A. Treiger convincingly demonstrated.30

In my analysis of the Arabic manuscripts, made possible thanks to a collaboration with Bishara Ebeid, I have found many other agreements with the Syriac version against the Greek original.31 I mention here, for example the number of poor people in Alexandria, 8500 instead of 7500 (Festugière, 348.39; Venturini, 7.1-2); the last sentence of chapter VII, which was not part of the Greek original (Festugière, 353, 19; Venturini, 20.7-8); the different ending of chapter VIII (Festugière, 354, 79-85; Venturini, 24.16-25.10); the peculiar conclusion of chapter XIX and beginning of chapter XX (Festugière, 368.2-6; Venturini, 50.7-13); the price of a slave, thirty coins instead of fifty coins (Festugière, 370, 114; Venturini, 55.8); and the suppression of the chapters concerning the heretics in Alexandria.32

I will go back to the two hypotheses above: do these innovations point to a Greek common ancestor, or do they directly reflect the Syriac text? I will extend my view to the whole manuscript tradition of the Life of John the Merciful, looking for other helpful data.

The Latin version BHL 4392: peculiarities and common features to the Syriac version

The possible dependence of Arabic and Syriac on a common Greek model was suggested by the discovery of the surprising features of the anonymous Latin version known as BHL 4392.33 This late translation, whose origin is still unknown,34 reproduces almost all the loci in which the Syriac version differs from the Greek original (including those common to the Arabic version, listed above). As it is challenging to assume a direct relationship between Syriac and Latin, I first supposed that the two versions depended on the same lost Greek model35. I did not exclude that this unknown recension might have been the common model of the Syriac, Latin, Arabic, and Georgian versions.36

The Latin version offers a complete reinterpretation of the structure of the saint's biography, achieved through a new ordering of the chapters, which remains without any parallel in the manuscript tradition. The chapters are arranged in the following order (reference is made to Festugière’s numeration): 6 (1^st part)+47, 16+46, 45, 4+5, 2, 3, 1, 10, 9, 6 (2^nd part), 8, 27, 18, 20+21, 22, 23, 38, 19, 40, 51, 11, 24, 25, 7, 31, 35, 30+32, 12.37 The rhetorical sections are suppressed, as any other author’s digression.38 The Latin even includes some new episodes, such as the account of John’s youth and a lengthy excursus concerning the story of Epiphanius, which is not reproduced by any other version of the work.39 The ending is absent, including the account of the patriarch’s death, a possible consequence of physical damage to the antigraph.40 As in Syriac, the “doctrinal chapters” have disappeared, albeit with many others (30% of the original chapters is missing). In addition to these elements, which suggest that we are dealing with a late and inaccurate translation, we find a long list of agreements with Syriac against the Greek original, which I provide below.

Bohdziewicz-Warburg, p. 310, chap. 8, §7: Fuitque descriptorum numerus octo milia quingenti

Venturini, 7.1-2: ܗ̇ܘܐ ܗܘܐ ܕܝܢ ܡܢܝܢܗܘܢ ܝܬܝܪ ܡܢ ܬܡܢܝܐ ܐܠܦܝ̈ܢ ܘܚܡܫܡܐܐ

Festugière, 348.39: πλείους δὲ ἦσαν τῶν ἑπτὰ ἥμισυ χιλιάδων

Bohdziewicz-Warburg, p. 309, chap. 6, §2: Mensura et mensura, pondus et pondus abhominabilia sunt apud Deum

Venturini, 9.3-4: ܕܣܢܐ ܐܠܗܐ

Festugière, 348, 25: στάθμιον μέγα καὶ μικρὸν ἐμίσησεν ὁ θεός

Bohdziewicz-Warburg, p. 312, chap. 12, §28-30: Super quo gratias egerunt altissimo ac sic ulterius prospere nauigantes uenerunt Alexandriam. Et narratis omnibus que fuerant gesta sancto patriarche, optulit ei argentum memoratum. Qui totum respuens uniuersa cum uase ipso donauit mercatori.

Venturini, 24.16-25.10: ܘܝܗܒܘ ܬܫܒܘܚܬܐ ܠܐܠܗܐ . ܗ̇ܘ ܕܥܒܕ ܨܒܝܢܐ ܕܕܚܠܘ̈ܗܝ ܀ ܘܗܘܐ ܕܟܕ ܐܬܡ̇ܢܥܘ ܠܐܠܟܣܢܕܪܝܐ . ܣܠܩ ܡܠܚܐ ܗ̇ܘ ܠܘܬ ܝܘܚܢܢ ܪܝܫ ܐܦܝܣ̈ܩܘܦܐ . ܘܐܫܬ̇ܥܝ ܠܗ ܒܗ̇ܘ ܡܕܡ ܕܛܝܒ ܠܗܘܢ ܐܠܗܐ ܒܐܘܪܚܗܘܢ . ܘܐܦܝܣܗ ܕܢܣܒ ܡܢܗ ܗ̇ܘ ܡܕܡ ܕܝܗܒ ܠܗ . ܘܠܐ ܨܒܐ ܛܘܒܢܐ ܕܢܣܒ ܡܢܗ ܡܕܡ . ܐܠܐ ܐܡܪ ܠܗ . ܒܪܝ . ܗ̇ܘ ܡܕܡ ܕܢܣܒܬ ܡܢܢ . ܕܐܠܗܐ ܐܝܬܘܗܝ . ܘܗܫܐ ܙܠ ܗܘܝܬ ܡܬܥܗܕ ܠܡܣ̈ܟܢܐ . ܐܝܟܢܐ ܕܢܫܪܬܚ ܠܟ ܡܪܢ ܡܢ ܓܙܘ̈ܗܝ ܕܠܐ ܡܘܦܝܢ . ܘܐܙܠ ܓܒܪܐ ܡܢ ܠܘܬܗ . ܟܕ ܡܫ̇ܒܚ ܠܐܠܗܐ . ܘܡܩ̇ܠܣ ܠܛܘܒܢܐ . ܘܥ̇ܒܕ ܗܘܐ ܐܡܝܢܐܝܬ ܙܕܩ̈ܬܐ . ܘܐܠܗܐ ܡܫܪܬܚ ܗܘܐ ܒܝܬ ܐܝܕܘ̈ܗܝ ܐܡܝܢܐܝܬ ܟܠ ܝܘ̈ܡܝ ܚܝܘ̈ܗܝ .

Festugière, 354, 79-85: Καὶ οὐ θαῦμα, ὦ φιλόχριστοι. ὁ γὰρ τοὺς πέντε ἄρτους πληθύνας καὶ πάλιν τὸ ὕδωρ Αἰγύπτου εἰς αἷμα μεταποιήσας καὶ τὴν ῥάβδον εἰς ὄφιν μετασχηματίσας καὶ τὴν φλόγα εἰς δρόσον μετενέγκας ἀκαμάτως, καὶ τοῦτο τὸ παράδοξον ἐποίησεν ἵνα καὶ τὸν ἑαυτοῦ θεράποντα πλουτίσῃ καὶ τὸν ναύκληρον ἐλεήσῃ.

Bohdziewicz-Warburg, p. 310, chap. 9, §14: scias in omnibus aliis amphoris mel huic simile contineri.

Venturini, 31.1-2: ܕܟܠܗܝܢ ܗܠܝܢ ܩܘ̈ܠܐ ܕܚܙܝܬ . ܗܟܢܐ ܡ̈ܠܝܢ ܡܢܗ ܕܗܢܐ ܕܒܫܐ

Festugière, 357, 45-46: «Ὅλα τὰ ἐθεάσω ἀνερχόμενα χρήματα ἀντὶ μέλιτος γέμουσιν»

Bohdziewicz-Warburg, p. 313, chap. 14, §1. The “thousand pounds of iron” (mille rotas ferri) are mentionned in the sixth position of the list sent by John to Modestus, as in Syriac (cf. Venturini, 46.4-7)

Bohdziewicz-Warburg, p. 318, chap. 19, §14: uidebimus quis preualebit

Venturini, 49.11-12:ܡ̇ܢ ܙ̇ܟܐ ܢܚܙܐ

Festugière, 367, 64: Ἴδωμεν τίς περικακεῖ

Bohdziewicz-Warburg, p. 314, chap. 15, §39: precio bisantiorum triginta

Venturini, 55.8: ܛܝܡܘ̈ܗܝ ܬܠܬܝܢ ܕܝܢܪ̈ܐ

Festugière, 370, 114: νομισμάτων πεντήκοντα

Bohdziewicz-Warburg, p. 315, chap. 15, §70: et qui audierunt hec, glorificauerunt Deum. Amen.

Venturini, 58.12-13:ܠܐܠܗܐ ܫܘܒܚܐ . ܘܥܠܝܢ ܪ̈ܚܡܘܗܝ ܠܥܠܡܝܢ . ܐܡܝܢ

Festugière, 371.180: om.

Bohdziewicz-Warburg, p. 318, chap. 19, §27: Narrauit similiter eidem diuiti exemplum de quodam seruiente quem olim habuerat in Cypro dicente sibi: «Quotiens hylariter largior et habundanter pauperibus, habundant et uestra, et si quando parce, minorantur et uestra».

Venturini, 50.7-13: ܬܘܒ ܬܚܘܝܬܐ . ܕܢܩܦܐ ܘܠܚܡܐ ܠܫܪܒܐ ܗܢܐ ܕܩܕܝܡ ܣܝܡ . ܡܫܬܥܐ ܗܘܐ ܩܕܝܫܐ ܝܘܚܢܢ ܘܐܡܪ ܗܟܢܐ . ܕܐܝܬ ܗܘܐ ܠܝ ܠܡ ܐܢܫ ܡܗܝܡܢܐ ܥܠ ܒܝܬ ܡܕܝ̈ܐ ܕܝܠܝ ܒܩܘܦܪܘܣ . ܘܐܝܬܘܗܝ ܗܘܐ ܒܬܘܠܐ ܘܢܟܦܐ ܟܠܗܘܢ ܝܘ̈ܡܬܗ . ܘܟܠ ܟܡܐ ܕܝܗܒ ܗܘܐ ܒܠܒܐ ܛܒܐ . ܝܬܝܪ ܡܬܒܪܟ ܗܘܐ ܘܣܓܐ ܗ̇ܘ ܡܐ ܕܐܝܬܘܗܝ ܗܘܐ ܒܝܬ ܐܝܕܘ̈ܗܝ . ܘܡܐ ܕܒܚܘܣܟܐ ܝܗܒ ܗܘܐ . ܚܣܪ ܗܘܐ ܠܗ .

Festugière, 368.2-6: Ἄξιον δὲ καὶ ἁρμόζον τῷ προειρημένῳ κεφαλαίῳ ἐξηγεῖτο ἐν μιᾷ ἐπὶ πάντων ὁ ὅσιος. Εἶχον γάρ, φησί, τινα παραμονήτην ἐν Κύπρῳ εἰς τὴν ἐμὴν ἀποθήκην, πιστὸν πάνυ καὶ παρθένον ἕως τελευτῆς. καὶ ἐξηγεῖτό μοι μεθ’ ὅρκων ὅτι∙ «Ἐν Ἀφρικῇ ἐμοῦ ὄντος [follows the Story of Peter the Tax Collector]

Bohdziewicz-Warburg, p. 315, chap. 17, §3: precepit eos capi et separatim carceri mancipari

Venturini, 61.6-7: ܘܦܩܕ ܛܘܒܢܐ ܕܢܬܠܒܟܘܢ ܘܢܬܦܪܫܘܢ ܡܢ ܚܕܕ̈ܐ

Festugière, 373, 17-18: ἐπέτρεψεν τὴν μὲν γυναῖκα πτερνισθῆναι καὶ ἀποχωρισθῆναι ἐξ αὐτοῦ, αὐτὸν δὲ λωρισθῆναι καὶ εἰς φυλακὴν ἰδιάζουσαν ἀποκλεισθῆναι

Bohdziewicz-Warburg, p. 308, chap. 2, §8: disparuit

Venturini, 61.12: ܐܬܓܢܙ

Festugière, 373, 25: ἀνεχώρησεν

Bohdziewicz-Warburg, p. 315, chap. 17, §21: gratias agebat deo

Venturini, 63.10: ܫܘܒܚܐ ܠܗ ܠܐܠܗܐ

Festugière, 375, 91: βαβαῖ

Festugière, 375. 1: chap. XXIV = On the utility of the prayers for the dead + Story of a prisoner in Persia

Venturini, 65: chap. XXIII = On the utility of the prayers for the dead; chap. XXIV = Story of a prisoner in Persia Bohdziewicz-Warburg, p. 320: chap. XXIII = Story of a prisoner in Persia

Festugière, 381-382: chap. XXX = A ruler procrastinates a loan; chap. XXXII = Punishment of the ruler

Venturini, 79-81: chap. XXIX = A ruler procrastinates a loan + Punishment of the ruler

Bohdziewicz-Warburg, p. 322: chap. XXVIII = A ruler procrastinates a loan + Punishment of the ruler

Bohdziewicz-Warburg, p. 322, chap. 28, §1: Eo tempore erat in partibus Alexandrie uir quidam habens casale ualde bonum, sed fructum non fecerat domino suo solitum propter siccitatem et aque penuriam.

Venturini, 79.2-4:ܐܢܫ ܡܢ ܒܢ̈ܝ ܐܬܪܐ ܐܝܬ ܗܘܐ ܠܗ ܦܪܣܛܝܘܢ ܪܒܬܐ ܘܐܬܠܒܟ ܒܕܝܡܘܣܝܐ ܕܩܪܝܬܗ . ܗ̇ܝ ܕܒܫܢܬܐ ܗ̇ܝ ܡܕܡ ܠܐ ܐܥܠܬ ܠܗ ܡܢ ܥܠܠܬܐ . ܒܝܕ ܒܨܝܪܘܬ ܡ̈ܝܐ.

Festugière, 381, 1-3: Ἄλλοτε δὲ ἄλλος ἐπειγόμενος ὑπὸ τῶν τὰ δημόσια εἰσπραττόντων καὶ μὴ εὐπορῶν ἀποδοῦναι—ἦν γὰρ ἀστοχήσασα ἡ χώρα διὰ τὴν λειψυδρίαν τοῦ Νείλου—

Bohdziewicz-Warburg, p. 316, chap. 18, §9-10: Et sic per diem sequentem laborans, ad aliam meretricem per noctem similiter declinabat. Et sic diebus singulis apud singulas similia faciens circuibat.

Venturini, 88.1-5: ܘܐ̇ܙܠ ܗܘܐ ܠܥܒ̇ܕܗ ܘܠܦܘܠܚܢܗ ܀ ܘܗܟܢܐ ܥ̇ܒܕ ܗܘܐ ܟܠܝܘܡܐ . ܟܕ ܐ̇ܙܠ ܠܘܬ ܚܕܐ ܚܕܐ ܡܢܗܝܢ ܟܠ ܠܠܝܐ . ܘܝ̇ܗܒ ܗܘܐ ܠܗ̇ ܐܓܪܗ . ܘܡܢ̇ܛܪ ܗܘܐ ܠܗ̇ ܡܢ ܚܛܝܬܐ . ܟܕ ܡܘܡܐ ܠܗܝܢ ܕܠܐ ܢܘ̈ܕܥܢ ܠܐܢܫ ܣܘܥܪܢܗ

Festugière, 387, 20: om.

Bohdziewicz-Warburg, p. 317, chap. 18, §44: et suffocans ipsum

Venturini, 91.12-13: ܘܫܪܝ ܚܢܩ ܠܗ

Festugière, 390, 143-144: καὶ διέρρηξεν αὐτόν

Bohdziewicz-Warburg, p. 318, chap. 20, §5: duro fueram corde

Venturini, 95.13: ܩܫܐ ܠܒܐ

Festugière, 392, 19: ἄσπλαγχνος καὶ ἀσυμπαθής

Bohdziewicz-Warburg, p. 318, chap. 20, §9: largiri

Venturini, 96.6-7: ܘܬܘܒ ܝܗܒ ܗܘܐ

Festugière, 392, 32: κλέπτειν

Bohdziewicz-Warburg, p. 318, chap. 20, §10: tres dragmas

Venturini, 96.8: ܡܬܩ̈ܠܐ ܬܠܬܐ

Festugière, 392, 33: τριμίσια

Bohdziewicz-Warburg, p. 309, chap. 4, §1: ciues alexandrini

Venturini, 103.2: ܠܒܢ̈ܝ ܐܠܟܣܢܕܪܝܐ

Festugière, 397, 2: τοὺς πολλούς

Bohdziewicz-Warburg, p. 319, chap. 21, §4: qui et peritior et magis intentus operi

Venturini, 110.14-15:ܟܕ ܐܡܝܢ ܐܢܐ ܒܦܘܠܚܢܐ ܘܐܘܡܢ ܐܢܐ

Festugière, 401, 10: κἀγὼ πλέον σου σχολάζων

Bohdziewicz-Warburg, p. 319, chap. 21, §7: meis acquiesce consiliis

Venturini, 111.2: ܬܫܡܥ ܒܩܠܝ

Festugière, 402.15: om.

Although the agreement between Syriac and Latin is not always literal, it is undeniable that they preserve a significant number of common variants, which are absent in the huge manuscript tradition of the Greek text. Moreover, the other two Latin versions, respectively of Anastasius and John of Amalfi, do not have any ecdotic connection with the anonymous version. While the first one reflects the so-called Vie courte41, the “Amalfitan version” seems to depend on the same branch of manuscript Vat. gr. 1669.42

An alternative hypothesis: Arabic as the missing link between the Syriac and the Latin versions

Excluding that the Latin translator knew the Syriac text, there is an alternative to the theory of a lost Greek Vorlage: the “Syriac form” of the Life may have reached the Latin West through the mediation of Arabic.43 After the publication of the Latin text and an examination of the Arabic manuscripts, it is now possible to propose this theory as the most convincing one44. One of the most eccentric features of this version is the lexicon: common and proper names, some of which have a Latin origin, are abnormally altered. A possible explanation for these deformations is that they came into Latin through Arabic instead of Greek. The following table compares the lexicon used to translate the same Greek words by John of Amalfi and the anonymous compiler of BHL 4392.

Greek text (ed. Festugière)	John of Amalfi’s Latin version (ed. Chiesa)	Anonymous Latin version (ed. Bohdziewicz-Warburg)	Arabic (Garshuni) version (Vatican Syr. 202; Dayr Mar Mattay, ms. 138)
οἴνου	Vinum	Syrupo	ܫܪܐܒ (sharāb)
κεντηνάρια	Centenaria	Quintaria	ܩܢܐܛܝܪ/ܩܢܛܐܪ (qinṭār)
λίτρας	Libras	Rotae	ܪܛܠ (raṭil)
τριμίσιον	tertia pars numismatis	Tres dragmas	ܬܠܬܗ̈ ܕܪܐܗܡ (3 darāhim)
Βιτάλιος	Vitalis	Petalius	ܒܛܐܠܝ (proper noun)
Τρωΐλος	Troilus	Trophila	ܛܪܘܒܝܠܐ (proper noun)

While John of Amalfi translates the Greek words accurately, even using the calque ceratim instead of the Latin equivalent siliqua, the Anonymous deforms them abnormally, including those of Latin etymology, such as κεντηνάρια and the name Βιτλάιος. In the latter, the initial b is changed into p (Petalius), as commonly happens in the passage from Arabic into Latin. The name of the bishop Troilus is changed into Trophila, which reflects the spelling of the Arabic manuscripts.

The Syriac origin of the Arabic version: a revealing numismatic word

The most interesting case is the translation of the term τριμίσιον, used in chapter XL (ar. 36; lat. 20; syr. 35), telling the story of an astute servant, who stole from his master to give to the poor people.45 The Latin version certainly depends on Arabic, since it reproduces, in addition to the Arabic word dìrham, an evident translation error. The τριμίσιον, which was a third of a solidus at the time of the author, is wrongly translated by both the Arabic and the Latin translators as “three drachmas” (ܬܠܬܗ̈ ܕܪܐܗܡ f. 120ra; tres dragmas, Bohdziewicz-Warburg, p. 318, chap. 20, §10). This is a significant error, unlikely polygenetic, caused by the translator’s misunderstanding of the obsolete monetary term τριμίσιον. But which translator are we talking about? The Syriac version presents the same kind of error, translating ܬܠܬܐ ܡܬܩ̈ܠܐ “three coins”.46 This situation occurs in two other passages of the Syriac version: in the prologue, when the salaried servant says that he received ܕܝܢܪ̈ܐ ܬܠܬܐ ܒܟܠ ܫܢܐ “three coins every year” from his master (Venturini, 2.12); and in chapter XLIV, when the patriarch calculates the inheritance he has left to the Church of Alexandria, repeating thrice the amount of ܕܝܢܪ̈ܐ ܬܠܬܐ , “three dinars” (Venturini, 114.7; 114.9; 114.14). In all these cases the Arabic text, which in chapter 36 had used the term ܕܪܐܗܡ for the Syriac ܡܬܩ̈ܠܐ, reproduces the Syriac literally, translating ܬܠܬܗ̈ ܕܢܐܢܝܪ , “three dinars” (Vatican Syr. 202, f. 122rb; Dayr Mar Mattay, ms. 138, f. 107rb).47 It is challenging to explain how such an error could have been committed in the Greek manuscript tradition: it seems closely connected to a precise choice of the Syriac translator, which was later spread among the manuscript tradition, including the Latin version.48 The Arabic translator certainly read the Syriac text to propose such a translation, not the Greek original.49 Finally, it is noteworthy that the Georgian version in the first passage also preserves the reading “three drachmas”, which suggests its dependence on Arabic.50 The theory that all the versions derive from the same (and inexplicably lost) Greek model is now even more unlikely. The most realistic solution is also the simplest: a direct link between the three versions, among which the Arabic assumes the role of cultural mediator.

The multifaceted shape of the Syriac tradition

The manuscript tradition of the Syriac version of the Life of John the Merciful is complex, as its text was transmitted in several different recensions.

α = this siglum identifies the best-known recension of the Syriac Life. Its main feature is that it is shorter and less detailed compared to the Greek text.51 It is transmitted by three manuscripts, depending on the same archetype,52 while a Sinaitic fragment represents an independent branch.53 P. Bedjan edited this recension based on the manuscript Paris BnF Syr. 235, comparing its text with manuscript London British Library Add. 14645.54

Σ = I use this siglum to refer to the Syriac version's most complete and detailed form, the closest to the original. We know about its existence only indirectly, thanks to some incomplete witnesses (β and π). It has been almost entirely neglected by the manuscript tradition to the advantage of the abridged version. However, both forms of the text probably circulated together for some time after the translation from Greek into Syriac.55

β = this siglum identifies a long extract from the Life, which contains the Story of Peter the Tax Collector (chapter 20 of the Syriac version). This edifying story detached quite early from the Syriac Life and was transmitted as a separate work in several hagiographic collections.56 It preserves much more details of the Greek than the other sources do, and it represents the most direct (albeit incomplete) witness of the original Syriac translation (Σ).57

π = this recension, which I called versio syra minor, is preserved by manuscript Paris BnF Syr. 234, f. 104r-127v [13^th century], and was recently published by the writer.58 It is a metaphrase of the Syriac version, which presents a further reduced form if compared to α. This abrégé was performed on Paris BnF Syr. 235 (or more likely on a closely related manuscript), but it was retouched with the help of an unknown witness of Σ.59 The definition versio syra mixta would be equally suitable to this peculiar text, considered the concurrence of ancient and modern textual features, blended and reinterpreted as in a new work.60

Arabic’s close relation to α: common errors to both versions

I will start my analysis by comparing the Arabic version to π on one hand, and to α on the other. It is pretty easy to exclude a direct dependence of the Arabic text on π: it does not present the highly shortened form of π and its peculiar innovations, a unicum in the manuscript tradition of the Life of John the Merciful.61 Furthermore, the Arabic version does not reproduce the common errors to π and its probable antigraph, Paris BnF Syr. 235.62 However, comparing the Arabic text with the original readings that π has derived from its secondary source, Σ, is interesting for this study. The first example I have chosen is significant, as π is the only one that correctly reflects the Greek original's meaning, while the rest of the Syriac tradition presents an evident misunderstanding.

Conspectus siglorum

• V = Vatican Syr. 202 [17^th century]
• D = Dayr Mar Mattay, ms. 138 [n. d.]
• A = Sinai Ar. 428 [9^th century]
• Σ = original Syriac version
• α = versio syra = mss, LPM
• π = versio syra minor = Paris BnF Syr. 234 [13^th century]
• β = Story of Peter the Tax Collector
• L = London British Library Add. 14645 [10^th century]
• P = Paris BnF Syr. 235 [12^th-13^th century]
• M = Dayr Mar Mattay, ms. 142 (27) [15^th century]

S = Sinai Syr. 24, ff. 199r-200v [9^th century]

Greek text = Festugière, 357.70-76	Syriac rec. α = Bedjan, 325.6-11	Syriac rec. π = Venturini (2022), 18.10-11	Arabic version = Vat. Syr. 202, f. 15r; DMMM 138, f. 110r
θαυμάσαντος δὲ καὶ τοῦ πάπα τὴν ὀξεῖαν τοῦ ἀνδρὸς μεταβολήν, οὐδὲν περὶ τῆς ἐγχειρήσεως ἐνεκάλεσεν, ἀλλὰ μᾶλλον καὶ παρεμυθήσατο αὐτὸν λόγοις παρακλητικοῖς. τοιαύτη δὲ ἀμφοτέρων συνεδέθη ἔκτοτε ἐκ θεοῦ ἀγάπη, ὡς καὶ σύντεκνον γενέσθαι αὐτὸν τοῦ πολλάκις εἰρημένου λαμπροτάτου ἀνδρός.	ܘܟܕ ܚܙܝܗܝ ܛܘܒܢܐ ܠܫܘܚܠܦܐ ܚܪܝܦܐ ܕܗܘܐ ܠܗ ܬܡܗ: ܘܠܐ ܩܪܒ ܠܗ ܡܪܫܘܬܐ ܡܕܡ ܠܦܛܪܝܩܝܘܣ ܗ̇ܘ: ܥܠ ܣܥܝܐ ܗ̇ܘ ܠܐ ܢܡܘܣܝܐ: ܐܠܐ ܡܠܘܢ ܘܝܬܝܪܐܝܬ ܡܒܝܐ ܗܘܐ ܠܗ ܒܡ̈ܠܐ ܒܣ̈ܝܡܬܐ. ܘܒܕܐܝܟ ܗܕܐ ܪܚܡܘܬܐ ܥܡ ܚܕܕ̈ܐ ܢܬܐܣܪܘܢ: ܐܝܟܢܐ ܕܐܦ ܐܪܙܐ ܩܕܝܫܐ ܕܡܥܡܘܕܝܬܐ ܢܩܒܠ ܡܢܗ ܕܛܘܒܢܐ63.	ܘܟܕ ܚ̣ܙܐ ܩܕܝܫܐ ܠܣܓܝܐܘܬܐ ܕܡܬܬܘܝܢܘܬܗ . ܠܐ ܐܟܣܗ . ܐܠܐ ܩܒܠܗ ܘܐܢܝܚ̣ܗ ܒܚܕܘܬܐ ܘܩ̇ܒ̣ܠ ܝܠܕܘܗ̈ܝ ܡܢ ܡܥܡܘܕܝܬܐ .	ܦܠܡܐ ܪܐܝ ܐܠܒܛܪܝܪܟ ܟܬܪܗ̈ ܢܕܐܡܬܗ̇ ܘܚܪܐܪܗ̈ ܐܡܐܢܬܗ̇ ܥܔܒ܆ ܘܠܡ ܝܠܘܡܗ̇ ܥܠܝ ܡܐܦܥܠ64 ܡܢ ܡܟ̣ܐܠܦܗ̈ ܐܠܢܐܡܘܣ܆ ܘܠܟ̇ܢܗ̇ ܥܙܐܗ ܒܟ̇ܠܐܡ ܪܘܚܐܢܝ65. ܘܨܐܪ ܒܝܢܗܡ ܒܥܕ ܕܠܟ ܡܘܕܗ66 ܟ̇ܬܝܪܗ67. ܘܩܒܠ ܡܢܗ ܐܠܡܥܡܘܕܝܗ68
As the Pope wondered at the sudden conversion of the man, he did not reproach him for the attempt, but rather comforted him with consoling words. From then on, so strong was the affection knit by God between the two, that (John) became the godfather of the children of the often-mentioned illustrious man.	And when the blessed man saw the sudden repentance that he had, he was amazed and did not reproach him for that illicit assault, but rather and even more comforted him with sweet words. And through this friendship they will be linked with each other, so that he will also receive the holy sacrament of baptism from the blessed man.	And when the saint saw the abundance of his compunction, he did not reproach him, but he received him and reassured him joyfully, and received his children from the baptism.	And when the Patriarch saw the fervent repentance that he had, and his warm faith, he marveled and did not reproach him for the action opposite to the law, but he began to console him with good speeches. And after this, a strong love occurred between them, and he received the baptism from him.

As we can see, the text of α is unsustainable: the patriarch John, instead of becoming σύντεκνος of the patrician Nicetas, baptizes him.69 We find the misunderstanding in the three witnesses of α (LPM) and in the Sinaitic fragment (S),70 while in π, it is clear that John became godfather of Nicetas’ children, who are explicitly mentioned (ܝܠܕܘܗ̈ܝ). The compiler of π, faced with an inconsistency of its antigraph (P), decided to turn to its superior quality source (Σ). The Arabic version here clearly follows the text of α, presenting the same error, almost ad verbum, of the manuscripts LPM. There are other passages in which a similar situation occurs, which cannot be examined here analytically.71

The close link between the Arabic version and α is confirmed by their comparison with β, as in the following passage.

Greek text = Festugière, 370.111-112	Syriac rec. α = Bedjan, 344.13-14	Syriac rec. β = Venturini (2020), 72.16-17	Arabic version = Vat. Syr. 202, f. 25v; DMMM 138, f. 113v
«Δάνεισαι καὶ ἀγόρασον αὐτόν∙ ὁ θεὸς γὰρ οἶδεν, πολύ ἐστιν καλοποδίνος καὶ εὐλογῆσαί σε ἔχει ὁ θεὸς δι’αὐτοῦ.»	ܐܬܦ̇ܪܣ ܘܙܒܢܝܗܝ. ܡܛܠ ܕܥܒܕܐ ܗܘ ܒܪܝܟܐ. ܘܡܬܒ̇ܪܟ ܐܢܬ ܡܛܠܬܗ.	ܫܐܠ ܘܙܒܢܝܗܝ. ܡܛܠ ܕܝ̇ܕܥ ܗܘ ܡܪܝܐ. ܕܪܓܠܐ ܛܒܬܐ ܐܝܬ ܠܗ. ܘܥܬܝܕ ܗܘ ܐܠܗܐ ܕܢܒ̇ܪܟܟ ܡܛܠܬܗ.	ܐܚܬ̇ܐܠ ܘܐܫܬܪܝܗ: ܦܐܢܗ̇ ܥܒܕ ܡܒܐܪܟ܇ ܘܣܘܦ ܝܒܐܪܟܟ ܐܠܠܗ ܡܢ ܐܔܠܗ̇܀
“Ask for a loan and buy him! God knows it, he is a bringer of happiness (καλοποδίνος) and God is going to bless you through him”.	“Devise a manner and buy him, because the servant is blessed, and you will be blessed through him”.	“Ask and buy him, because the Lord knows that he has a good foot, and God will bless you through him ”.	“Find a way and buy him, because the servant is blessed, and God will bless you through him”.

The literal translation ܕܪܓܠܐ ܛܒܬܐ (“with a good foot”) of the difficult term καλοποδίνος (“bringer of happiness”) is trivialized by α, whose compiler deduced the meaning of “blessed” from the following verb ܡܬܒܪܟ, which translates the Greek εὐλογῆσαι. The original translation had to seem unintelligible because of its extreme literality, to the scribe who introduced the variant. Here again, the Arabic ܡܒܐܪܟ reproduces the error of α. It is also noteworthy the precise translation of the Greek δάνεισαι (“ask for a loan”) in β, against the Arabic and α, which simplify the concept. Nevertheless, this passage reveals something else: α does not say that God bestows the blessing, although this is explicit in Greek, in β, and in Arabic.72 This attests to the better quality of the Arabic text if compared to α.

The Arabic version as a halfway witness between α and Σ

We can explain in two different ways the presence of the Greek’s original details in the Arabic version: a) the Arabic translator also consulted the Greek text to improve his translation; b) the Arabic version depends on an ancient form of the Syriac text, preserving an intermediate stage between α and Σ. The first hypothesis is challenging to demonstrate and refute because it involves a contamination, as the compiler mixed two different sources. To verify the second hypothesis, we need a passage that has undoubtedly been translated into Arabic from a Syriac model of better quality than α, without the possible help of a Greek source. I have found a sentence in the Arabic version that certainly reflects the Syriac original, while the Greek text expresses the concept using different words.

Greek text = Festugière, 366.23-26	Syriac rec. α = Bedjan, 336.10-12	Syriac rec. π = Venturini (2022), 20.30-32	Arabic version = Vat. Syr. 202, f. 21v; DMMM 138, f. 112r
πλὴν ἐκεῖνο ἀντιβολῶ τὴν τιμίαν σου κεφαλὴν μηδαμοῦ ἐντάξαι τὸ ὄνομα τῆς ἐμῆς ἀναξιότητος, ἀλλὰ μᾶλλον τοῦτο αἰτήσασθαι Χριστὸν ἐκεῖ με ἀπογράψασθαι ἔνθα ἀληθῶς ἡ ἀπογραφὴ μακάριστος.	ܘܡܦܝܣ ܐܢܐ ܠܟ: ܕܠܐ ܬܟܬܘܒ ܫܡܝ ܒܕܘܟܬܐ ܗ̇ܝ: ܕܡܬܝܗܒ ܛܘܒܐ ܠܐܝܠܝܢ ܕܡܬܟܬܒ ܫܡܗܘܢ ܬܡܢ.	ܘܬܘܒ ܒ̇ܥܐ ܐܢ̣ܐ ܡܢܟ ܕܠܐ ܬܟܬܘܒ ܫܡܝ ܣܟ ܒܡܕܡ ܕܒ̇ܢܐ ܐܢܬ ܐܠܐ ܨ̇ܠܐ ܥܠܝ ܕܢ̣ܬܟܬܒ ܫܡܝ ܒܣ̣ܦܪܐ ܕܚ̈ܝܐ	ܘܐܢܐ ܐܛܠܒ73 ܐܢ ܠܐ ܬܟܬ̇ܒ ܐܣܡܝ ܦܝ ܫܝ ܡܡܐ ܬܒܢܝܗ: ܘܠܟܢ ܨܠܝ ܥܠܝ ܐܢ ܝܟ̇ܬܒ ܐܣܡܝ ܦܝ ܐܠܡܘܜܥ ܐܠܕܝ ܝܥܛܐ ܐܠܛܘܒܐ ܠܟܠ ܡܢ74 ܝܟܬܒ75 ܐܣܡܗ̇ ܗܢܐܟ܀
Further, I pray your honorable head not to place the name of my indignity anywhere, but rather to ask Christ to inscribe me where the inscription is truly blessed.	I ask you not to write my name in that place, where beatitude is given to those whose names are written there.	And furthermore, I ask you not to write my name at all on what you build, but pray for me that my name will be written in the Book of Life.	And I ask your holiness not to write my name on what you build, but pray for me that my name will be written in the place where beatitude is given to everyone who asks for his name to be written there.

The Arabic here is almost identical to π, which is the only Syriac recension to preserve the correct text.76 It is indeed still immune from the saut du même au même produced in α by the confusion between the two identical terms ܫܡܝ ... ܫܡܝ (to which correspond the Arabic ܐܣܡܝ ... ܐܣܡܝ). The concordance of the Arabic text with π occurs this time against the Greek, which presents a different formulation of the sentence. In Greek (without any variant in the manuscripts) the conditions for a saut du même au même are totally missing, as there is not proximity between two identical terms. It is unlikely that the Arabic translator managed to achieve the same text of π independently, by consulting the different formulation of the sentence that we find in Greek. It is much more probable that the Syriac model of the Arabic translator preserved an earlier stage of the text than α, still immune from its oversight. The following stemma codicum summarizes the ecdotic relationship of the Arabic version with the Syriac manuscript tradition.

Albeit indirectly, the Arabic version is the only complete witness of such an ancient stage of the Syriac tradition, because a) the text of Σ is lost, and only partially reflected by the Story of Peter the Tax Collector (β); b) the recension π retains only sporadic elements of Σ, but the great part of its text depends on α; c) the Sinaitic fragment (S) preserves only two tiny excerpts of the work. The intermediate position of the Arabic version between Σ and α is enough to justify its knowledge of the Greek’s original details that disappeared in the Syriac manuscript tradition, without involving any awareness of the Greek original.

A simpler explanation for Greek-Arabic concordances against Syriac

As we have seen above, the earliest form of the Syriac version (Σ) is known only from indirect witnesses (β and π). To support the theory of a revision, all we can do is look for significative concordances between Arabic and Greek against β or π. In this paragraph, I will start with the examination of the latter case. The first passage where Arabic and Greek agree against π is found at the very beginning of the work.

Greek text = Festugière, 343.5-14	Syriac rec. α = Bedjan, 303.6-9	Syriac rec. π = Venturini (2022), 11.4-7	Arabic version = Vat. Syr. 202, f. 3v; DMMM 138, f. 106r
Ὁ μὲν σκοπὸς εἷς ἐστιν ἡμῶν τε καὶ τῶν πρὸ ἡμῶν φιλοπόνων καὶ ὁσίων ἀνδρῶν ὁ ἐπὶ τῇ παρούσῃ τοῦ ἀοιδίμου ἀνδρὸς τοῦ βίου διηγήσει, τουτέστιν τὸ πᾶσιν μὲν μίμησιν θεοφιλῆ καὶ ὠφέλειαν προσγενέσθα, δόξαν δὲ καὶ μεγαλοπρέπειαν τῇ ἁγίᾳ καὶ προσκυνητῇ τριάδι ἀναπέμψαι καὶ ἐν τούτῳ ὡς ἐν πᾶσιν τῇ πάντοτε κατὰ γενεὰν καὶ γενεὰν τοὺς οἰκείους φωστῆρας ἀναδεικνυούσῃ εἰς τὸ φωτίζειν τοὺς ἐν σκότει καὶ σκιᾷ θανάτου καθημένους τῆς ἁμαρτίας.	ܬܫܒܘܚܬܐ ܘܬܘܕܝܬܐ ܘܩܘܒܠ ܛܝܒܘܬܐ ܠܐܒܐ ܘܠܐܒܪܐ ܘܠܪܘܚܐ ܩܕܝܫܐ: ܗ̇ܘ ܕܡܕܟܐ ܘܡܙܗܐ ܠܩܕܝܫܘ̈ܗܝ: ܐܝܟܢܐ ܕܢܢܗܪܘܢ ܐܝܟ ܟܘ̈ܟܒܐ ܢܗܝܪ̈ܐ ܠܐܝܠܝܢ ܕܒܚܫܘܟܐ ܘܒܛܠܠܝ̈ ܡܘܬܐ ܝܬܒܝܢ.	ܐܬܚܙܝ ܠܝ ܐܚܝ̈ ܘܐܒܗ̈ܝ ܕܢܣܝ̣ܡ ܬܫܥܝܬܗ ܕܗܢܐ ܩܕܝܫܐ ܡܪܝ ܝܘܚܢܢ . ܕܢ̣ܗܘܐ ܝܘܬܪܢܐ ܠܐܝܠܝܢ ܕܩܪܝܢ ܘܫ̇ܡܥܝܢ . ܘܢ̇ܫ̣ܒܚܘܢ ܠܬܠܝܬܝܘܬܐ ܩܕܝܫܐ ܕܡܚܘܝܐ ܒܓܠܝܐ ܢܘܗܪܐ ܕܩܕܝܫ̈ܝܗ̇ . ܕܒܟܠ ܕܪ̈ܝܢ ܐܝܟ ܟܘܟ̈ܒܐ ܢܗ̈ܝܪܐ . ܠܐܝܠܝܢ ܕܝ̇ܬܒܝܢ ܒܚܫܘܟܐ ܘܒ̈ܛܠܠܝ ܡ̇ܘܬܐ .	ܪܐܝܢܐ78 ܐܠܐܒܗܐܬ79 ܐܠܕܝܢ ܣܒܩܘܐ ܘܟ̣ܒܪܘܐ ܣܝܪܗ̈ ܗܕܐ ܐܠܩܕܝܣ ܐܒܝܢܐ80 ܐܢ ܝܟ̇ܬܒ ܐܡܪ ܝܟ̇ܘܢ ܦܝܗ ܡܢܦܥܗ ܠܡܢ ܝܩܪܐܗ ܘܠܡܢ ܝܣܡܥܗ: ܘܝܣܒܚ ܐܠܬܐܠܘܬ ܐܠܩܕܘܣ ܐܠܕܝ ܦܝ ܟ̇ܠ ܚܝܢ81 ܝܜܗܪ ܩܕܝܣܝܗ: ܠܝܢܡܘܐ82 ܡܬ̣ܠ83 ܐܠܟ̇ܘܐܟ̇ܒ84 ܐܠܕܝܢ ܗܡ ܔܠܘܣ ܦܝ ܐܠܜܠܡܗ ܘܜܠܐܠ ܐܠܡܘܬ.
One alone is our objective and that of the zealous and pious men before us in the present narration of the biography of the celebrated man: that he becomes, on the one hand, for all an imitation loved by God and something profitable, on the other, to elevate glory and magnificence to the holy and worshipful Trinity, which always, from generation to generation, shows also in this, as in all things, its own luminaries to enlighten “those who are in darkness and shadow of death” of sin.	Glory and praise and thanksgiving to the Father, the Son, and the Holy Spirit, who purifies and makes his saints shine, so that they shine like bright stars for those who lie in the darkness and shadow of death.	It seemed suitable to me, my brothers and fathers, to present the story of this Saint John, so that those who read and hear may benefit and so that they glorify the Holy Trinity, which openly reveals the light of its saints, who illuminate in everything, like shining stars, those who lie in the darkness and shadow of death.	We have seen the fathers who came before us, describe the narration of this our holy father, something useful to those who will read and hear it, and we praise the Holy Trinity, which in every generation illuminates its saints, who like the stars (illuminate) those who are sitting in the dark and in the shadow of death.

In the Syriac recension α, the first lines of the prologue are missing, maybe because of a material damage in the antigraph. Here again, π reproduces better than α the original form of the prologue. As in the previous example, the Arabic version reflects the knowledge of an ancient form of the Syriac text, as reveals its concordance with π (against the Greek) in the mention of: the “fathers” (Ar.: ܐܠܐܒܗܐܬ; π: ܘܐܒܗ̈ܝ) instead of the “men” (Gr.: ἀνδρῶν); the benefit “of those who read and hear” (Ar.:ܝܩܪܐܗ ܘܠܡܢ ܝܣܡܥܗ ; π:ܠܐܝܠܝܢ ܕܩܪܝܢ ܘܫ̇ܡܥܝܢ) instead of “for all” (Gr.: πᾶσιν); “his saints” (Ar.: ܩܕܝܣܝܗ; π: ܕܩܕܝܫ̈ܝܗ̇) instead of “his stars” (Gr.: τοὺς οἰκείους φωστῆρας). Nevertheless, in this case the Arabic retains original details of the Greek which have been suppressed by π: the mention of the fathers “who preceded us” (Ar.: ܐܠܕܝܢ ܣܒܩܘܐ ; Gr.: τῶν πρὸ ἡμῶν), and “in every generation” (Ar.: ܟ̇ܠ ܚܝܢ ; Gr.: κατὰ γενεὰν καὶ γενεὰν).

This situation is the direct consequence of the metaphrastical nature of π. On the one hand, most of its text reproduces Paris BnF Syr. 235 (P), while its secondary source (Σ) was used by the compiler non-systematically, and mostly in case of necessity (for example, to fill lacunas). On the other hand, the compiler of π radically abridged its model, by cutting original details, sentences, and even entire episodes. Although sometimes π it the only witness preserving the Greek’s original details,85 it remains generally a lower quality source than the Arabic text. Depending on an ancestor of α, the Arabic version well preserves the author’s narrative scheme, which has been destructured by π. Therefore, it is totally to be expected that several passages of the Greek are better reproduced in Arabic than in Syriac. I will enumerate here only a few of them.

At chapter VI, the correct transliteration of ܒܪܛܐܢܝܐ (Vatican Syr. 202, f. 11v; cf. also the Latin Britania, Bohdziewicz-Warburg, p. 328, chap. 12, §18) = τῆς Βρεττανίας (Festugière, 354.43), against the Syriac ܕܟܪܛܢܝܐ (α) and ܕܩܪܝܛܘܢܝܐ (β);86 at chapter XI, the mention of the Nile ܐܠܢܝܠ (Vatican Syr. 202, f. 15v; cf. also the Latin Nili, Bohdziewicz-Warburg, p. 319, chap. 22, §3), which reflects the Greek τὸν ποταμὸν (Festugière, 358.11), omitted in Syriac; at chapter XII, the mention of the “universal prayer” ܐܠܩܐܬܘܠܝܩܝ, and of the “anaphora” ܐܠܐܢܐܦܘܪܐ (Vatican Syr. 202, f. 17r), which reflect the Greek τὴν καθολικὴν [...] καὶ τοῦ ἁγίου καταπετάσματος ὑψοῦσθαι μέλλοντος (Festugière, 360.29-30); at chapter XIV, the mention of the merchant ܐܗܠ ܐܠܣܘܩ (Vatican Syr. 202, f. 19r), which is the Greek μετά τινος τῶν τῆς πόλεως καπήλων (Festugière, 362.3), while the Syriac version has ܠܐܢܫ ܡܢ ܒܢܝ̈ ܡܕܝܢܬܐ (α) or ܠܐܢܫ̈ܝܢ (β).87 At chapter XXIX, the sentence ܐܢܢܝ ܠܡ ܐܣܦܟ̇ ܕܡܝ ܒܥܕ ܥܢܟ. ܟ̇ܡܐ ܐܡܪܢܝ ܐܠܡܣܝܚ, “I have not yet shed my blood for you, as Christ ordered me!” (Vatican Syr. 202, f. 35r), correctly reproduces the Greek οὔπω τὸ αἶμά μου ὑπέρ σου ἐξέχεα ὡς ἐνετείλατό μοι ὁ δεσπότης μου καὶ πάντων Χριστός (>Festugière, 381.12-13), which is not preserved in Syriac.88

The Story of Peter the Tax Collector: the Arabic version of the tale and its relationship with β, α, and the Greek text

The independent recension of the Story of Peter the Tax Collector (β), despite its fragmentary nature, represents the most direct descendant of the original Syriac translation (Σ) of the Life of John the Merciful.89 This excerpt reproduces the longest episode of the Life (it takes up approximately 10% of the work). The big size and the ecdotic relevance of the fragment allow a systematic comparison with the Arabic version (chapters 20-21 in Sinai Ar. 428), with the form of the tale included in the Syriac Life (α), and with the Greek text.90 It can be helpful to analyze the following questions.

• Many details of the Greek original are not (or are badly) reproduced by α: are they better reproduced by β or by the Arabic version?
• In the case of discordance between α and the Greek text, does the Arabic agree with α against the Greek and β, or does it agree with the Greek against α and β?
• It is possible to retrace the history of some variants to understand their progressive stages in the different stemmatic levels?

To answer to the previous questions, I will present a table comparing the four versions of the tale: Greek, Syriac 1 (β), Syriac 2 (α), Arabic. I give the English translation of each one, with the reference to the original text’s editions. The Arabic text, which is unedited, can be found in the Appendix (§8) at the end of this contribution. I examine here only the first half of the tale, going until the second vision of Peter (Festugière, 369.70), more an additional passage, coming from the conclusion (Festugière, 371.170-176).

Font styles’ legenda:

α, β, and Arabic vs Greek

Greek and β vs Arabic and α

Greek, Arabic, and β vs α

Greek, Arabic, and α vs β

Greek text = Festugière, 368.1-29	Syriac 1 (β) = Venturini (2020), 60-62, §1	Syriac 2 (α) = Bedjan, 340.5-341.5	Arabic version = Vat. Syr. 202, f. 23r; DMMM 138, f. 112va
When I was in Africa, something like this happened. I was administrator —he affirms— of a tax collector, extremely rich and merciless. One day, thus, in winter, while some poor people were sitting in the sun and were warming up, each one began to praise the houses of the merciful men and to pray for each one of them, and similarly to mock and revile the houses of the unmerciful men. So meanwhile came in the middle also the name of my master, the tax collector, and each one began to ask his neighbor: “In truth, brother, did you ever receive alms from that house?”. And —I swear it as in front of the Lord— all asking each other, nobody was found who has ever received anything from that house. One of them says then: “What do you give me, and today I will take alms from him?”. And, after making a deal with him, he goes and stands outside the door of the tax collector’s house, waiting for the moment when he comes back from the mass. Indeed, he, by determination of God, arrived at the same time as he, who was entering the door, and the beast of burden, loaded with loaves from the bakery, which was entering for his breakfast. So, after seeing the poor, he grabs out of anger, having not found a stone, a loaf from the basket of the beast and throws it at his face. That one then took it and left, swearing to his companions that — he says— “I received it from his own hands”.	My beloved, there was in Africa one of the noblemen, a tax collector, whose name was Peter, and he had absolutely no mercy towards anyone. And so one day it happened that, while the poor people were sitting in the winter days and were warming up in the sun, each one of them began to praise and bless the houses of the merciful men who gave them alms, praying for them. In the same way, therefore, they also blamed and cursed the houses of the unmerciful men, and among them they also put in the middle the name of Peter, who had no mercy, and each one began to ask his companion if he had ever received alms from his house. And after everyone had asked his companion and no one was found who said that he had received something from him sometime, one of them spoke and said: “What do you give me, and I receive alms from him today?”. And after making a deal with him on what they would have given him, he immediately went and stood in front of the door of that rich man, and waited for him until he returned. And then, by God's determination, were entering at the same time he and his beast of burden, which was carrying the loaves from the bakery for his meal. And having seen that poor, he got very angry, and wanting to throw at him a stone and not finding it, he grabbed a loaf from the basket that was on the beast, and threw it angrily at the poor's face, and hit him. That one then immediately bent down a little and took the loaf, and ran and went to his companions, swearing to them: “From his hands I have received it now!”.	There was one of the noblemen, and he was very rich, whose name was Peter, and he never had compassion towards anyone. And one day of the poor people, while they were sitting in the sun, remembered the merciful men who gave them alms and prayed for them. And then they started to mention those who had no mercy for them and to curse them, and they also remembered Peter, since he was not merciful. And they asked each other if he had ever given alms to someone of them, and none was found among them who had received anything from him. And one of them started talking and said, “What do you give me, and I will receive alms from him?”. And they promised to give him something. And he immediately went and stood at Peter's door, and waited for him until he came at his house. And by God's determination he arrived, and with him a donkey, which brought loaves for him to eat. And when the poor saw him, began to implore him, asking him for alms. Peter then, when he saw that poor, got very angry, and he looked for a stone to throw at him, but he did not find it. And he grabbed a loaf of those which his donkey was carrying, and threw it at the poor with great anger. And the poor took it, and returned to his companions, and began to swear that he had received it from Peter’s hand.	There was a noble man named Peter, and he was very rich and he had absolutely no mercy towards anyone. In some days, while the poor people were sitting in the sun, each one of them began to mention the merciful men and who gave them alms, and to pray for them, and to remember even those who had no mercy and to curse them. And then they also mentioned Peter, who had no mercy, and began to ask each other if he had given alms to any of them, and they did not find anyone of them who had received anything from him. So one of the poor men said: “What do you give me, if I go and receive alms from him?”. And they promised to give him something. And that poor went and sat at his door, waiting for him until he came. And with God's help he came, and with him there was an animal that was bringing a loaf from the bakery for the meal. And when he saw that poor, he got very angry and looked for a stone to throw at him, and he did not find it. And he took a loaf from those the animal was carrying, and threw it at his face with great anger, and hit him. And the poor took that loaf and brought it to his companions, and began to swear to them: “I received it from his hands”.

Some observations:

• Syriac 2 (α) and Arabic omit many important details of the Greek, which are regularly mentioned by Syriac 1 (β): the geographical setting of the tale “in Africa”; the profession of Peter, “a tax collector”; the temporal setting “in winter”; the poor people “warming up”; the blessing on “the houses” of the merciful men; the literal reproduction of the Greek εἰς μέσον, “in the middle”; the question whether a poor “had ever received” from Peter (Syriac 2 and Arabic: “whether Peter had ever given”); the deal of the poor with his companions; the basket carried by the beast of burden.
• When the Arabic version preserves an original detail of the Greek which is omitted by Syriac 2 (α), it is always found also in Syriac 1 (β): the specular reproduction of the Greek ἤρξατο ἕκαστος, “each one began”; the loaves coming “from the bakery”; the throw of the loaf “at the face” of the poor. There are no original details of the Greek, which are preserved exclusively by the Arabic version.
• Syriac 1 (β) does not specify that Peter was “very rich”. The Arabic version retains this original detail but, here again, is not the only one, as Syriac 2 (α) retains it too.91
• Syriac 2 (α) and Arabic say that the poor’s companions “promised to give him something”, while the Greek text says that they made “a deal with him”. It is noteworthy that Syriac 1 (β) presents a halfway form of the sentence: “after making a deal with him | on what they would have given him”. The concept, which in Greek is not so explicit, has been clarified by the Syriac translator, who added the second part of the sentence. The first part of the sentence has been later simplified into “they promised” by the common ancestor to Arabic and Syriac 2 (α).

Greek text = Festugière, 368.30-369.55	Syriac 1 (β) = Venturini (2020), 62-66, §2-3	Syriac 2 (α) = Bedjan, 341.6-342.9	Arabic version = Vat. Syr. 202, f. 23v; DMMM 138, f. 113ra
Indeed, after two days, the tax collector fell ill with a fatal disease, and sees in his sleep himself, called to account [for his actions], and all his actions weighed on a scale. In a scale pan, thus, some bad-looking blacks (μαῦροι) were gathering, the other belongs to others, clothed in white and fearsome in appearance, who, finding nothing good to counterbalance the wicked deeds gathered by the blacks in their scale pan, were very sad and gloomy and hesitating towards each other, and they said: “Don't we really have anything here?”. Then one of them says: “Really we have nothing but a single loaf that he gave two days ago to Christ, even this reluctantly”. And when they put the loaf, the scale pan was balanced. And those clothed in white who had appeared to him, tell him: “Come on, add [something] to the loaf, because these blacks will catch you”. When he woke up, thus, he acknowledged that what he had seen was not false, but true, since everything [experienced] from his youth, even what he had forgotten, he saw that those Ethiopians had gathered and carried on the scale. And he said: “Alas (Βαβαῖ)! If a single loaf, which I threw with anger, benefited me so, how many damages avoids he who gives his goods with simplicity to those who need them!”. And from then on, he got so wise —he says— that he became extremely merciful to the point that he did not even spare his own body.	And after two days, that rich man fell ill with a fatal disease. And he had a vision in his dream, as if he was standing in a court and was being judged for his actions. And he saw that some black and bad-looking ones came, and they gathered all his evil deeds and put them on a scale pan, on one side; on the other side of the scale, instead, there were other men dressed in white and good-looking, and they wanted to put something good on the scale pan which was next to them, against the evil things that had been placed by those blacks, and they found none. And then they were sad and very gloomy, and hesitating they were saying to each other: “So, what have we got? Don't we have anything to put here?”. So, one of them said: “We really have nothing but this loaf, which two days ago he gave to Christ unintentionally, for some reason”. And then they hurried, and they brought it and placed it, and immediately that scale pan matched with his counterpart. And immediately those dressed in white spoke and said to him: “Peter, go and add to this loaf, and know that otherwise those blacks will catch you, and we will not help you at all”. And when he woke up, he acknowledged that the things that had appeared to him were not a lie but reality, since all he had done or said from his youth until then, including what he had forgotten, he had seen that those blacks had collected and placed on that scale pan. And he said to himself: “Alas (ܒܒܝ)! If a loaf, that with anger I threw against that poor unintentionally, has benefited so much and was able to save me from the hands of those blacks, from how many damages is released he who gives his goods to the needy people with simplicity and good effort!”. And from then on, he became so honest that he was eminent and excellent in mercy towards the poor people, to the point that he would not even spare his own body.	And after two days it happened that the rich Peter fell ill. And he saw in his dream that he was standing in a courtroom, and he was judged for his evil deeds. And he saw some blacks, horrible looking, who came towards him and gathered his evil deeds and placed them on a scale pan. And near the other scale pan he saw some good-looking men, with white clothes, who were standing up and wanted to put some good deed against the evil ones which the blacks had placed. And they did not find it, and they were very sad, and said to each other: “Is there nothing here to put on this scale?”. And one of them answered and said: “Actually we have nothing here but a loaf, that two days ago he gave to Christ unintentionally, for some reason”. And they took the loaf and put it on the scale, and immediately he leaned over and was in balance with the other scale pan. And those dressed in white began to speak and said: “Oh Peter, go and add to this loaf! Otherwise, know that the blacks will catch you!”. And after they said these things, Peter awoke from his sleep, and acknowledged that the dream he had seen was true, because everything he had done and said from his youth until that day, he had seen that those blacks had brought it and placed it on the scale. And he said to himself: “If a loaf, which I gave to that poor unintentionally, has thereby helped and saved me from those blacks, how more he who gives to the poor people with a good heart is preserved from the evil!”. And from that day on, he gave alms to the poor people.	And that rich man fell ill, and he saw in his dream as if he was standing in the day of Judgment to be judged for his deeds. And he saw a group of black men, ugly face, coming towards him, after collecting all his evil deeds and putting them on a scale pan. And there was also another group of men, and they were standing on the other side of the scale, and they were good-looking, in white clothes, and they wanted to put something in the other scale pan to counterbalance, but they found no good deeds against the evil deeds of those blacks, and they were sad about that. And while they were in that situation, they said to each other: “So, don't we have anything here to put on the other scale pan?”. And one of them answered, saying: “We have nothing except a bread that he had given to Christ two days ago unintentionally”. And then they took it and put it on the scale pan, and it balanced it with the other scale pan. Then those who were clothed in white said: “Oh Peter, go and add something else good to this loaf! Otherwise, know that these blacks will catch you!”. And when Peter noticed that what he had seen was truth and not a lie —because he acknowledged that everything he had done, from his youth up to that time, he saw the blacks coming towards him and putting it on the scale pan, he said to himself, marveling: “If a loaf that I threw to that poor unintentionally, has benefited and has saved me from those blacks, from how many evils will be saved he who gives, with his good effort, what he owns to the poor people! And from that day on, he became merciful to the poor people.

• Syriac 2 (α) and Arabic do not mention the “fatal disease” of Peter; they do not say that the white men were “gloomy and hesitating”, but simply “sad”; they suppress the interjection ܒܒܝ, which in Syriac 1 (β) accurately translates the Greek βαβαῖ; they do not specify that the loaf was thrown “with anger”; they do not reproduce the Greek expression ἐν ἁπλότητι (“with simplicity”), as Syriac 1 (β) does; they suppress the whole final sentence, corresponding to the Greek consecutive clause ὥστε μήτε τοῦ οἰκείου σώματος φείσασθαι (Festugière, 369.55).
• Right in the last sentence, the Arabic version shows a halfway textual stage between Syriac 1 (β) and Syriac 2 (α): it omits, as Syriac 2 (α) does, the consecutive clause, but it well reflects the Greek ἐγένετο ἄκρος ἐλεήμων, “he became merciful” (albeit it does not reproduce ἄκρος), while Syriac 2 (α) trivializes the concept (“he gave alms”).
• The expression “with simplicity” (ἐν ἁπλότητι) is reproduced only by Syriac 1 (β), which moreover presents the addition: “and good effort”. The Arabic version retains only the latter addition, having lost the first member, which is more difficult to understand. Syriac 2 (α) goes on in the deterioration process of the original, changing the expression in “with a good heart”.

Greek text = Festugière, 369.58-70	Syriac 1 (β) = Venturini (2020), 66-68, §4	Syriac 2 (α) = Bedjan, 342.10-343.1	Arabic version = Vat. Syr. 202, f. 24v; DMMM 138, f. 113rb
Therefore, it happened once that he went after the night, as usual, to the customs office, and goes towards him a sailor (ναύτης), who came out of a shipwreck naked as he had been begotten, and fell at his feet, imploring him for a garment. Therefore [Peter], thinking that he was a poor, takes his own tunic off, which was expensive, and makes him swear with faith, saying: “By the Lord, brother, put it on and pray for me!”. After he left, ashamed to wear it, he gave it to a tailor to be sold. And coming back the tax collector, sees it hung and became very sad. And after he returned to his home, he could not eat anything, but after closing the door of his room he sat down, crying and pondering thus: “I was not worthy —he says— that the poor man had a memory of me”. While then he was discouraged, he fell asleep.	Therefore, it happened once that, as he was leaving the church to go to the customs office —since he was a tax collector—, came across him a sailor (ܢܘܛܐ), who had saved himself from the fury of the sea and who was naked as at birth. And he bowed before him, asking to give him a garment for his body. Therefore, after seeing him, and after thinking that he was a poor, he took off the expensive garment that was on him and gave it to the poor, making him swear a lot and saying: “By our Lord, wear it and pray for me!”. The poor man then, having gone away, was ashamed to wear that garment because of its elegance, and immediately gave it away to be sold. And the rich Peter, after waiting a while, came back and saw that garment hung in the market, and he became sad and regretted a lot. And after he went to his home, he could not eat anything, but closed the door of his room in front of him and sat down, weeping and pondering and saying: “I was not worthy to leave a memory to that poor”. And while he was regretting and moaning, he dozed off and fell asleep.	And one day, while he was leaving the church to go to the customs office —since he was the chief tax collector— came across him a sailor (ܡܠܚܐ), who had run away naked from the sea. And he fell to the ground before him, and implored him to give him a garment to wear. And when [Peter] saw him, he took off one of the sumptuous garments that he was wearing, and gave it to him and made him swear and said: “By our Lord, wear it and pray for me!”. And that poor, after walking away from him, was ashamed to wear it because of his elegance, and gave it to the market to be sold. And when Peter came back and saw that garment hung in the market to be sold, became very sad. And after going to his house, he could not eat because of his sadness. And he closed his door and sat down, and started to weep and moan and to say: “I did not deserve to be remembered by that poor”. And while he was crying and regretting, he dozed off and fell asleep.	And one day, while he was leaving the church to go to his court, —since he was the chief tax collector— and there he found a sailor (ܢܘܬܝ), who was saved from drowning in the sea and who was naked. And he fell in front of him, imploring and asking to dress him. And when Peter saw him, he felt compassion for him and took off a good [om. VD] garment and gave it to him [VD: made him wear it] and made him swear [VD: asked him], saying: “By God, accept this garment and pray for me!”. And when the sailor left and walked away, he was ashamed to wear it because it was elegant, and put it up for sale in the market. And the rich [om. VD] Peter, returning to his house, saw the garment hung [om. VD] in the market for sale, and he was very sad about it. [And when he arrived at his home, he could not eat anything for the great sadness, and closed his door and sat down, crying and sighing from the bottom of his heart, and saying: “I did not deserve to be remembered by that poor”. And while he was regretting and crying [om. VD], he dozed off and fell asleep.

• Syriac 1 (β) and Arabic still reflect some ancient features of the Syriac translation, like the calque ܢܘܛܐ, reproduced by the Arabic ܢܘܬܝ, and used to translate the Greek ναύτης. In the Syriac manuscript tradition, the calque has been trivialized not only by Syriac 2 (α), which substitutes it with the facilior ܡܠܚܐ, but also by one of the same manuscripts of β, which prefers the equivalent ܐܠܦܪܐ.92 The progressive corruption of the Syriac text did not affect the Syriac model of the Arabic version, which still preserved the reading ܢܘܛܐ.
• Nevertheless, the Syriac model of the Arabic version had already suffered some of the omissions that we find also in Syriac 2 (α): the loss of the main verb which introduces the episode (Gr.: Συνέβη, “it happened”); the sentence “after thinking that he was a poor”, which is essential for the comprehension of the whole passage (the sailor felt ashamed and refused the expensive garment, but Peter did not understand why, thinking that he was a poor asking for alms). Everything claims for a progressive corruption of the Syriac translation. The Arabic version reflects a precise stage of this corruption process, which is halfway between Syriac 1 (β) and Syriac 2 (α).

Greek text = Festugière, 371.170-176	Syriac 1 (β) = Venturini (2020), 80, §10	Syriac 2 (α) = Bedjan, 347.3-9	Arabic version = Vat. Syr. 202, f. 27r; DMMM 138, f. 115ra
He who was cooking went out running. But be careful that he is not fleeing, for he is a great servant of God. For when he came out, he said to me: “I say to you in the name of the Lord”, and immediately I saw a flame, which came out of his mouth and touched my ears: so immediately I heard, and I spoke.	The one who was serving in the kitchen came out running. But beware that he does not flee, because he is truly a servant of God! For when he went down, he said to me: “I tell you in the name of Christ: open!”. And immediately I saw that came out of his mouth like a flame of fire, and it came and reached my ears and my mouth, and immediately my ears heard, and my mouth spoke”.	“The man who was serving in the refectory went out in a rush. But beware that he does not flee, because he is a servant of God!”. And he told them: “When he came to me, he said: ‘In the name of our Lord Jesus Christ, open the door for me!’, I saw going out of his mouth like a tongue of fire, and it touched my mouth and my ears, and immediately I spoke with my mouth, and I heard with my ears.	And he said to them: “The one who managed the kitchen came out running, fleeing. And he said to me: 'I tell you in the name of Christ, open to me'. And I saw coming out of his mouth like a flame of fire, which reached my ears and my mouth”.

In this last passage, coming from the conclusion of the tale, the Syriac translator readapted the Greek text, by adding the detail of the mouth touched by the flame. Here the Arabic text and Syriac 1 (β) are almost identical, while Syriac 2 (α) inverts the terms (“my mouth and my ears”) and substitutes the substantive ܫܠܗܒܝܬܐ, “flame” (surely original, cf. Gr.: φλόγα), with ܠܫܢܐ, “tongue”. The perfect concordance of the Arabic with Syriac 1 (β), against Syriac 2 (α) on one hand, and against the Greek on the other, shows once more that the model of the Arabic translator was not the Greek original, but a Syriac manuscript belonging to a high branch of the tradition.

To answer to the previous questions:

• Syriac 1 (β) preserves original details of the Greek, which have disappeared in Syriac 2 (α), better than the Arabic version does. There is a first group of omissions and errors which is exclusive of Syriac 2 (α), whereas Arabic and Syriac 1 (β) faithfully reproduce the Greek. There is then a second group of variants, which are common to Syriac 2 (α) and Arabic, whereas only Syriac 1 (β) correctly reflects the Greek.
• I did not find any passage which is correctly reproduced only by the Arabic version, against Syriac 1 (β) and Syriac 2 (α). Quite the opposite, the Arabic version usually agrees with Syriac 2 (α), against Syriac 1 (β) and the Greek. Finally, it also happens that Arabic agrees only with Syriac 1 (β), against the Greek.
• It is sometimes possible to retrace the progressive development of the variants. The Greek text is often readapted by the Syriac translator, introducing some extensions and clarifications. Sometimes the Arabic version has lost the main features of the original, preserving only those extensions. Syriac 2 (α) often goes further on in this deterioration process.

None of these observations is compatible with a revision on the Greek original performed by the Arabic translator. It would be challenging to justify how the outcome of such a revision is often identical to Syriac 1 (β). If the Arabic translator had consulted a Greek source, the improvements of the Arabic text would be different from those of Syriac 1 (β). To support the theory of a “re-translation” from Greek, it is not enough to find details of the Greek original or calques from Greek words, which the Arabic may have derived from a Syriac source of high quality.93