Introduction

David Pingree

The Byzantine translations of astrological texts by Abū Maʿšar seem to have been executed, by unidentified scholars, in about 1000.[1] Among these translations is an abbreviated version of the Kitāb al-mudḫal al-kabīr, which forms book III of a collection entitled Ἀποτελεσματικὰ μυστήρια τῆς ἐπιστήμης. This edition of that text is based on three manuscripts.

E. Rome, Angelicus Graecus 29, 346 folia. The Μυστήρια were copied by Eleu­ther­ius ⟨Zebelenus⟩ of Elis[2] on Mitylene in 1388. Formerly the property of George, Count of Corinth,[3] the manuscript was described by F. Cumont and F. Boll in CCAG 5, 1; 4–57. Ff. 68–91 contain Μυστήρια III 1–14, E, and 15–68 (67–68 are not translated from Abū Maʿšar’s Kitāb al-mudḫal al-kabīr).

V. Vatican City, Vaticanus Graecus 1056, 244 folia. The main text (there are numerous scholia by later scribes) was copied in the fourteenth century, apparently from an archetype of the twelfth century.[4] The manuscript was described by I. Heeg in CCAG 5, 3; 7–64. Ff. 221^v–234 contain Μυστήρια III 1–6, 10–14, A–D, 16 (beginning only), 18–20, 23–24, 22, 25–36 (middle), and 45 (middle)–56; and ff. 24–27^v (V′) contain III 6–7 and 21.

W. Vatican City, Vaticanus Graecus 1058, 499 folia. The manuscript was copied in the middle of the fifteenth century. It was described by D. Pingree in The Astronomical Works of Gregory Chioniades, vol. 1, Amsterdam 1985, part 1, pp. 25–29. Ff. 454–457^v contain Μυστήρια III 21. Attention to this copy was drawn by O. Neugebauer, ‘Variants to the Greek Translation of Abū Maʿshar’s Version of the Paranatellonta of Varāhamihira and Teukros’, Bulletin de l’ Aca­démie royale de Belgique, Classe des Lettres et des Sciences Morales et Politiques, 5^e sér., 43, 1957, 133–140. I have used Neugebauer’s transcription of the text, which is preserved in the collections of the Department of the History of Mathematics at Brown University.

There are also extant eight manuscripts descended from E which I have not used as their scribes took many liberties in their transcriptions. These are

a. Milan, Ambrosianus B. 38. sup. Copied in the fifteenth century “from an old, worn-out manuscript written by an idiosyncratic scribe”. The manuscript was described by A. Martini and D. Bassi in CCAG 3; 6. Ff. 64^v–93 contain Μυστήρια III, with the omission of chapters 24–28. The apparatus in CCAG 11, 1; 87–93 gives variants of a to the chapter headings of s.

b. Berlin, graecus 173. Copied in the fifteenth century. This manuscript was described by F. Boll in CCAG 7; 48–63. Ff. 136–136^v contain Μυστήρια III 3.

f. Florence, Laurentianus 28, 33. Copied, as was v, in 1542. The manuscript was described by A. Olivieri in CCAG 1; 39–60. Ff. 138–192^v contain Μυστήρια III. This manuscript was used in part by F. Boll in his edition of III 21.

g. Paris, Bibliothèque nationale graecus 2419. Copied by George Midiates in the fifteenth century. Formerly properly of Nicolao Cardinal Ridolfi (died 1550) and of Catherine de Medici. The manuscript was described by M.A. Kugener and F. Cumont in CCAG 8, 1; 20–63. Ff. 71^v–72 contain Μυστήρια III 16–18, and ff. 137^v–138^v III 14, 19–20, and 12–13.

p. Paris, Bibliothèque nationale suppl. gr. 1148. Copied from b in the sixteenth century, before 1573. Described by P. Boudreaux in CCAG 8, 3; 81–87. F. 133^v contains Μυστήρια III 3.

r. Escorial I. R. 14. Copied in the fifteenth century. Described by O. Zuretti in CCAG 11, 1; 3–28. Ff. 143^v–146 contain Μυστήρια III 21.

s. Escorial I. Φ. 5. The relevant leaves copied by Petrus Carnabaca from Monemvasia, a scribe employed by Diego Hurtado de Mendoza (1503/04–1575)[5] in 1543. Described by O. Zuretti in CCAG 11, 1; 51–106. Ff. 286–316^v contain Μυστήρια III, with the omission, as in a, of chapters 24–28.

v. Vatican City, Vaticanus Graecus 1057. Copied, as was f, in 1542. Described by F. Cumont and F. Boll in CCAG 5, 1; 73. Ff. 73^v–101^v contain Μυστήρια III.[6]

The independence of V from E is proved by the fact that each manuscript preserves genuine parts of the translation that are omitted by the other. The value of W is demonstrated by the correctness of some of its readings in contrast to those of E and V. The headings were added sometime after the translation; they often reflect the beginning of the chapter rather than its actual contents, and chapter E in manuscript E was appended to chapter 14. For the most part, these chapter headings are missing from V as is the index of chapters prefaced to the text in E. Therefore, all manuscripts that preserve these headings are descended from E.

In the edition all the variants of manuscripts E, V, and W, are recorded, including some of the cases where the manuscript abbreviation is ambiguous. Emendations and additions (enclosed in ⟨ ⟩ if not obvious) are justified by the citation of the original Arabic text. The sections in that Arabic text on which the Greek is based are indicated in the margins of the Greek text. It remains unclear what principles if any guided the translator in his choice of passages to render into Greek.

The chapters of Μυστήρια III that have previously been edited are the following:

1, 3, and 5 by F. Cumont using E in CCAG 5, 1; 152–153.

12 and 13 by F. Cumont using E and g in CCAG 8, 1; 178–181.

14 by L. Parmentier using E in CCAG 5, 1; 154–155.

17 by W. Hübner using Eafgs in Hübner 1995,[7] v. I, 259–265.

21 by F. Boll using E and, from decan 2 of Libra to decan 3 of Capricorn, f, in CCAG 5, 1; 156–169.

22 by I. Heeg using E and V in CCAG 5, 3; 131–132.

32 by F. Cumont using E in CCAG 5, 1; 169–170.

49 by O. Zuretti using E, a, and s in CCAG 11, 1; 177–178; by I. Heeg using V in CCAG 5, 3; 132.

50–56 and 65–66 by O. Zuretti using E, a, and s in CCAG 11, 1; 178–202.

The excerptor of Abū Maʿšar’s text was not interested in the rational explanations of astrological doctrines, and misses out most of Parts I–V. He also considerably abbreviates the parts that he translates. It is not surprising that he should miss out crossreferences (e.g. in 45, [17]), but he also regularly omits references to authorities (e.g. Hermes and Valens in Part VIII). Occasionally, however, he will add a phrase: e.g. in 6, [6] he adds ‘Some people say that they have named these ⟨constellations⟩ from the shapes of the asterisms, but others say that it is from their effect’. Several of these additions take the form of glosses: in 45, [16] he adds the explanation that the coldness in Saturn is ‘unseparable’, whereas the dryness can depart from it, and that in Mars, the heat is unseparable, but the dryness can depart from it. Similarly, in 47, [13] he adds the explanation that ‘for Jupiter and Pisces are moist, and do not mix’ and that ‘for they are in the same house, but it (Mars) is hot and Aries is hot’. Similar additions, can be found in 47, [13]–[15].

He has a penchant for organisation. Thus in B [4]–[5] he gives a systematic list of powers that is not found in exactly these words in the Arabic: ‘The Sun is a greater benefic than all the ⟨other⟩ planets; after it is the Moon, after this is Jupiter and after this is Venus, and after this is Mars. Saturn is more of a malefic than Mars.’ He adds contrasts where there are none in Arabic: 26, [title] ‘filling and depleting’ (Arabic omits ‘and depleting’); 65, [41], ‘nobility and ignobility’ (Arabic: ‘nobility and class’); and 66, [6] ‘fat and emaciated’ where the Arabic has only ‘emaciated’.

In Part VII, the translator generally points out whether the configurations of the planets described are ‘strong’ or ‘weak’ (see also 66, [3]), and in Part VIII, he divides the significations of the lots into those when the lot is ‘made good’ or ‘in a good position’ (ἀγαθυνόμενος or καλῶς κείμενος) and those when it is ‘harmed’ or ‘in a bad position’ (κακούμενος, κακυνόμενος or κακῶς κείμενος): see 60, [2], 61, [2]–[3], 63, [2]–[3], 65, [148], [157], [160], [163], [241].

The translator obviously thinks about the meaning of the text. In 32 [2] he anticipates the last sentence of the paragraph ([8]) by drawing attention to the fact that the degrees he mentions are ‘according to the present time’. In 34, [3] he adds examples of animals; ‘that is to say, the horse, the ox and others like this’, but, when faced with two attributes in Arabic, he picks the wrong one, writing that ‘Leo indicates having claws (‘miḫlab’) like a dog’ when he should have chosen the word ‘ḏū nāb’ (‘with canine teeth’). Sometimes he spells out ‘the opposite’; e.g. in 65, [136] he adds ‘he will enjoy a peaceful death’, while the Arabic has only ‘the opposite’ (to the violent death mentioned in the first half of the sentence). In 45, [12] the translator brings together the example of the effect of mixture of water and wine with that of vinegar and honey, which is not mentioned in the Arabic. In 47 the translator translates the same passage twice ([4]–[5] is the same as [1]–[3]), but corrects a mistake that he had made in the first translation (writing ‘quartile aspect to the Sun’ instead of ‘trine’).

Correspondence between Arabic and Greek Texts

Passages marked with an asterisk have no equivalent in the Arabic text.