Description of the Arabic manuscripts and the relationship between them and the Latin Liber de orbe

Since analysis of the contents is required to collate the two Arabic manuscripts, I will first describe their details. In the course of the description, I will compare their contents with the Latin Liber de orbe to prove that they contain the Arabic text of this Latin work.

Description of MS B

Berlin, Staatsbibliothek zu Berlin, Ms. or. oct. 273 (=Ahlwardt no 5654).Footnote ¹⁰ Two initial flyleaves, sixty-six leaves, and one final flyleaf (folios i–ii, 1a–66b, and iii).

Beginning (folio 1b, lines 1–3):Footnote ¹¹

Ending (folio 66a, lines 13–14):

The copy is unnamed and undated. Ahlwardt estimates the date of its copying as around 600 H/AD 1203.Footnote ¹² On the front page (folio 1a), the copyist wrote the following note about the content of this codex:Footnote ¹³

Above this, another hand wrote the title and author as follows:

In the middle of the page, the third hand wrote the title and author as follows:

The copyist collated the copy with the exemplar. While he entered some notes in the margin (on folios 4a, 10b, 12b, 25b, 30a and 31b), he emended quite a few words and sentences by crossing out wrong words and writing the correction after them. The following are a few examples:Footnote ¹⁴

B: 3a, 9–10:

Cf. P: 21, 15:

B: 11b, 14–15:

Cf. P: 39, 3:

However, many miscopies and misdotted words are found in it. I list a few examples, as follows.

B: 3b, 1:

Cf. P: 21, 18–19:

B: 5b, 12–13:Footnote ¹⁵

Cf. P: 2, 13–14:

As for the contents, this copy contains an introduction and thirty-nine chapters. The introduction (folio 1b) only offers laudation to God in nine lines without any suggestion about the author. We must note that whereas LS omits it,Footnote ¹⁶ LL has a literal translation of it, but this part in LL is followed by a brief description of the aim and structure of the book, which cannot be found in the Arabic text.

In both MS B and MS P, each chapter has the heading without numbering. For the sake of convenience, I number and list them with English translation as well as with the corresponding Latin headings,Footnote ¹⁷ as follows.Footnote ¹⁸

Chapter 1: Discourse on the fact that the orb is managed and created (B: 1b, 10)

Chapter 2: Discourse on the four elements and how their Creator arranged them (B: 3b, 14; P: 22, 7–8)

Chapter 3: Discourse on motions (B: 4a, 8; P: 1, 1)

Chapter 4: Discourse on the natures and the motions (B: 5a, 15; P: 2, 5)

Chapter 5: Account of the fact that earth, water, air, fire, and the orb are spherical (B: 7a, 4; P: 27, 17)

Chapter 6: Account of the opposition of the four elements (B: 8a, 1; P: 28, 16)

Chapter 7: Discourse on the rotation of the Earth and the knowledge of night and day (B: 9a, 1; P: 36, 1)

Chapter 8: Discourse on the alteration of seas and rivers and the difference of the earth and mountains (B: 12b, 6–7; P: 46, 3)

Chapter 9: Discourse on the difference of water in the earth and [on] earthquakes (B: 14a, 3–4; P: 48, 1)

Chapter 10: Discourse on volcanoes in RomeFootnote ¹⁹ and Sicily (B: 16a, 1)

Chapter 11: Discourse on the flow and ebb of the sea by the Moon and its action on moisture (B: 16b, 1)

Chapter 12: Discourse on the food of the Nile in Egypt (B: 18b, 11–12)

Chapter 13: Discourse on the rotation of the orb and its motion and nature (B: 21b, 6; P: 29, 16)

Chapter 14: Account and discourse on the revolution of the orb and the Sun (B: 24b, 4)

Chapter 15: Discourse on the fact that any change does not reach the orb in its essence and in its motion (B: 29b, 4–5; P: 24, 5)

Chapter 16: Discourse on circles, chords and points (B: 32b, 7)

Chapter 17: Discourse on the difference of the Sun's rising and setting in [various] countries (B: 34b, 1–2; P: 40, 2)

Chapter 18: Account of the knowledge of the Sun's magnitude (B: 35b, 1–2)

Chapter 19: Discourse on the Moon's borrowing of the Sun's light (B: 38a, 1)

Chapter 20: Discourse on lunar eclipses (B: 40a, 1)

Chapter 21: Discourse on stars' borrowing of the Sun's light (B: 41a, 1–2)

Chapter 22: Discourse on the cause of solar eclipses (B: 42b, 1; P: 11, 1)

Chapter 23: Discourse on why the Moon becomes visible to some people and does not become visible to others and why it appears small or large (B: 44a, 5–6)

Chapter 24: Account on the difference of the Moon's light and stars' light in [various] cities (B: 44b, 10–11; P: 9, 5–6)

Chapter 25: Discourse on the number of the Moon's orbs (B: 46a, 1; P: 13, 1)

Chapter 26: Discourse on the two orbs of the Sun (B: 47a, 1; P: 14, 1)

Chapter 27: Discourse on the number of the orbs and their motions (B: 48a, 1; P: 17, 2)

Chapter 28: Discourse on the motion of the greatest orb (B: 50a, 3; P: 8, 3)

Chapter 29: Discourse on the orb of the zodiacal signs (B: 50a, 13; P: 8, 21)

Chapter 30: Discourse on the alteration of the natures of a season (B: 51b, 5; P: 4, 12)

Chapter 31: Discourse on the orbs of Saturn (B: 51b, 13; P: 4, 21)

Chapter 32: Discourse on Saturn's retrogradation and its returning into that from which it has left among the zodiacal signs (B: 53a, 11–12; P: 15, 5–6)

Chapter 33: Discourse on the orb of the fixed stars (B: 54b, 1; P: 19, 12)

Chapter 34: Discourse on the knowledge of how large the whole Earth is in miles (B: 54b, 7; P: 41, 1)

LL splits this chapter into Chapters 34 and 35:

Chapter 35: Discourse on the climates of the Earth (B: 55b, 13–14; P: 43, 5)

Chapter 36: Discourse on the two places where night or day continues during a whole day (B: 59a, 10–12)

Chapter 37: Discourse on winds (B: 60b, 7; P: 19, 18)

Chapter 38: Discourse on clouds, rain, thunder and lightning (B: 62b, 8–9; P: 31, 15–16)

Chapter 39: Discourse on plants (B: 64a, 7–8; P: 34, 4)

The above list shows that although Chapter 34 is divided into two, LL contains the entire chapters of MS B in the same sequence, and LS also follows the same chapter order, but omits several chapters. On the other hand, LL sometimes paraphrases the Arabic chapter headings, while they are literally translated in LS. For example, the chapter heading of Chapter 39 ‘on plants’ is translated accurately in LS, but LL renders it ‘on insensitive creatures’.Footnote ²⁰

The main texts of the LL and LS have the same characteristics as the headings. For example, at the middle of Chapter 3, the author introduces a figure to explain the motions of the four elements, as follows (B: 4b, 15; P: 1, 13–14):Footnote ²¹

This Arabic text is translated literally in LS without the reference to God, as follows:Footnote ²²

LL has a paraphrased version:Footnote ²³

Needless to say, detailed analysis of the relationship between the Arabic and the Latin should wait for a critical edition of the Latin Liber de orbe (LS and LL), but the above preliminary comparison obviously establishes that these two Arabic manuscripts contain the Arabic original of this Latin work, and that LS is a truncated version of it, while LL is paraphrased from it. Moreover, the variant readings in the above list of the chapter headings reveal that sometimes MS P has a different reading against MS B and the Latin. To clarify the characteristics of its variation, I will describe MS P in detail.

Description of MS P

MS P: Philadelphia, Pennsylvania University Library, MS LJS 439. One initial flyleaf, twenty-five leaves, and one final flyleaf (folio i, pp. 1–50, and folio ii).

MS P has page numbers written by a modern hand at the upper left corner. The chapters are not numbered. Comparison between MS B and MS P shows that MS P is misbound. The correct page sequence is as follows:

Thus the beginning is at page 21 and the ending is at page 50.

Beginning (page 21, line 1)

Ending (page 50, lines 12–13)

Then follows an Arabic verse written by another hand in six lines.

The copy is unnamed and undated. The library cataloguer estimates the date of its copying around the year 14 CE.Footnote ²⁴

Due to the lack of several leaves, including the first, it does not have a note on the title and author as well as the introduction, and, among thirty-nine chapters of MS B, it transmits twenty-five complete chapters (Chapters 2–8, 13, 17, 22, 24–35 and 37–39) and four incomplete chapters (Chapters 1, 9, 15 and 16).Footnote ²⁵ Most leaves have severe erased parts on the surface, so that quite a few sentences are unreadable. The copyist collated it with the exemplar. He entered several corrections in the margin (sometimes with the mark ‘corrected’).

As for the contents, it is remarkable that the chapter order in MS P is not the same as in MS B and in the Latin Liber de orbe, and some chapter headings are different from those in MS B and LS (for example Chapters 22, 26 and 28). In the main text, we find many additions and modifications against MS B and LS. For example, Chapter 39 begins as follows (B: 64a, 8–12; P: 34, 5–8):Footnote ²⁶

LS:

This part of Chapter 39 contains several examples, despite being a few lines:

Example 1

B:

P:

Example 2

B:

P:

Example 3

B:

P:

In the last example, ‘a place’, is rendered into the plural form ‘the places', in MS P; however, a redactor forgets to change ‘in it’, into ‘in them’, so that this sentence contains a grammatical mistake.

On the other hand, MS B has no additional reading against MS P and LS in this part, whereas it contains miscopies, such as instead of In fact, we find only a few examples of modification in MS B against MS P and LS throughout the whole text. One of them is found in Chapter 13:

B: 22a, 13–15

P: 30, 10–11:

LS:

This example shows that MS B has an additional explanation beginning with ‘I mean’; this phrasing suggests that MS B also contains a later addition, although we cannot completely deny the possibility that this is the text by the author.

Analysis of the textual relationship reveals that MS P transmits this work after a redaction by a later scholar, so that it has a different chapter order with contamination of quite a few additions and modifications. Thus, when editing the Arabic Liber de orbe, I will follow MS B, which contains the complete set of the chapters with minor modification, and emend its scribal errors by comparing it with MS P and LS.

Based on the above description, I begin to consider the authorship of the Arabic Liber de orbe (henceforce ArLO) and the date of its composition. First, I will outline its contents to obtain clues to the author and the date.

The contents of the Arabic Liber de orbe

As already described, the introduction of ArLO only has laudation to God in nine lines, so the author does not give any information about why and how he composed this work. After praising God, he suddenly begins the first chapter. In each chapter, he first states his argument, and proves it frequently with diagrams. He sometimes mentions several contrary opinions and disputes them rationally.

The argument of Chapter 1 is the regularity of the circular motion of the orb bestowed by God, and then in Chapters 2–4 and Chapter 6 he introduces the Aristotelian theory of the four elements constructing the sublunar world,Footnote ²⁸ where he explains the orb as a distinct entity because of its circular motion opposed to the rectilinear motions of the four elements. Thus it is clear that he suggests another element for the orb, the so-called fifth element, although he does not explicitly call it so.

Whereas he contrasts the four elements with the orb according to their motions, he declares in Chapter 5 that they all constitute concentric spheres. That is because, he says, the orb and earth are spherical in nature, and the middle elements, namely fire, air and water, follow the shape of these two. For example, he explains the reason for the sphericity of fire as follows (B: 7b, 9–12; P: 28, 10–12):Footnote ²⁹

[Fr. 1]

Then, in Chapters 7–12, he illustrates geological and meteorological phenomena, such as the transformation of oceans and mountains, the different tastes of water and the occurrence of earthquakes, volcanoes in Rome and Sicily, the relationship between the tidal ebb and flow and the Moon, and the flood of the river Nile, by using the theory of the four elements. For example, he explains in Chapter 9 the occurrence of earthquakes (B: 14a, 12–14b, 2; P: 48, 7–11):

[Fr. 2]

Here he explains that the qualitative opposition between heat and coldness makes the earth cracked, and wind from the earth shakes it; however, he does not elucidate why wind emerges from the earth under this condition. Aristotle, who is the main source for ArLO about natural philosophy, describes the process whereby moistened earth is heated and dissolves dry exhalation, which turns into wind.Footnote ³⁰ Thus the above quotation suggests that while focusing on the contradiction of the two qualities, the author does not give details about exhalation.

Following this is described (B: 14b, 5–7; P: 48, 12–14) another phenomenon caused by this wind:Footnote ³¹

[Fr. 3]

In a similar way to this example, he explains in six chapters a variety of geological occurrences rationally with recourse to the Aristotelian theory of the four elements.

After discussing phenomena in the sublunar world, he focuses on the nature of the orb in Chapters 13–15, where he confirms repeatedly that the orb consists of an element other than the four elements. Then he describes the structure of the superlunar world by using the orb in Chapters 16–34. First, in Chapter 16, he explains the main components of the orb; that is, three circles, seven points and three chords. Before introducing them, he states the aim of this chapter (B: 32b, 8–11):Footnote ³²

[Fr. 4]

Here he declares that these components are the basis of cosmology. Then he introduces the three circles (B: 32b, 11–33a, 5):

[Fr. 5]

Given that the east–west circle corresponds to the equator line, the author obviously explains these main components in the case of the equatorial region, so that the south–north circle is the meridian circle. Following the three circles, the three chords are described (B: 33a, 5–9):

[Fr. 6]

Then he lists the seven points; that is, the centre of the Earth (nuqṭat al-arḍ), the east point (nuqṭat al-mashriq), the midheaven (wasṭ al-samā’), the west point (nuqṭat al-maghrib),Footnote ³³ the point of the lowest cupola of the orb (nuqṭat asfal qubbat al-falak), the north point (nuqṭa fī al-shamāl, i.e. the northern equatorial pole) and the south point (nuqṭa fī al-janūb, i.e. the southern equatorial pole).

After setting the main components of the orb, he explains in Chapters 17–24 how lunar and solar eclipses occur, how all stars borrow solar light, and how to determine the Sun's magnitude. Then, in Chapters 25–33, he presents the cosmos as a combination of concentric orbs, and describes planetary motions by epicycle–eccentric models. What is remarkable about his cosmology is that the cosmos is thought to consist of ten orbs, not of the standard nine orbs (namely the greatest orb, the orb of the fixed stars, and the seven planets' orbs). That is because in ArLO the orb of the zodiacal signs is a separate starless orb (the ninth orb), which is divided into twelve zodiacal signs, and the fixed stars themselves are on the orb of the fixed stars (the eighth orb), which is under the ninth orb.Footnote ³⁴ In Chapter 29, he explains the motion of precession as follows (B: 50b, 9–51a, 5; P: 3, 9–20):Footnote ³⁵

[Fr. 7]

Here he notes that owing to this motion the eighth orb has no star in the place corresponding to that of a zodiacal sign in the ninth orb. This quotation impresses on us that he strives to examine the reality of zodiacal signs, one of the basic elements of astrology.

Further down, in Chapters 34–36, the author treats topics of geography, including how to determine the Earth's magnitude, the climate divisions and particular places where it stays bright all day without the arrival of night or vice versa. Chapters 37–38 are on meteorology: on wind, and on thunder, clouds, rain and lightning. Finally, the last chapter of the book, Chapter 39, covers topics on plants.

As is clear from the above overview, ArLO gives us a comprehensive account of phenomena in the sublunar world as well as in the superlunar world. To structure his arguments, the author describes the mechanism of sublunar phenomena with recourse to the theory of the four elements, and explains superlunar phenomena with geometrical explanation using plenty of diagrams. The reason for his thoroughly logical explanation of all phenomena lies in his ambition to show how rational the construction of the world is, which in turn proves that its creation was impossible except by the wisest God.

His intention is most apparent in Chapter 14. First, he proposes the argument that if God is the wisest, He will make the motion of the orb the perfect motion; that is, circular motion. Then he states the reason why, if the motion of the orb increased or decreased, daytime would also increase or decrease, which should affect tremendously the living beings on the earth. At the end of this discussion (B: 28b, 8–9), he quotes the following verse of Qur’ān [iii.191]:Footnote ³⁶

[Fr. 8]

This is the only quotation from the sacred texts in ArLO. This stanza makes it clear that his main objective is to prove the thesis of ‘the aimed creation of the world by the one God’.

We must remark that whereas the author shows the physical aspect of the cosmos in detail with a number of geometrical diagrams, he does not mention how to calculate astronomical quantities using them. The only exception is the calculation of the Earth's magnitude in Chapter 34, where he presents the process of determining the circumference of the Earth. First, he explains how to determine one degree of the Earth's circumference by using the astrolabe. Then he gives the result that one degree is 66⅔ miles, and calculates the circumference, namely 24,000 miles (=66⅔ ×360).

Excepting this example, the author never presents quantitative determination when describing astronomical phenomena. In fact, at the end of Chapter 20, he remarks the absence of quantitative description in this book, as follows (B: 40b, 2–4):

[Fr. 9]

This note shows that he concentrates on the qualitative features of the cosmos, intentionally skipping astronomical quantitative determination usually written in zījs, namely astronomical handbooks with tables.Footnote ³⁷ In the Islamic world, this genre of works specializing in the qualitative account of the cosmos came to be known as ‘books on ‘ilm al-hay'a [science of the configuration of the orbs]’ after the appearance of Ibn al-Haytham's (965–c.1040) On the Configuration [hay'a] of the World. Footnote ³⁸ArLO is distinctively characteristic as an ‘ilm al-hay'a work, because whereas books of this genre normally exclude subjects on physics and meteorology, it covers more comprehensive topics about the cosmos to show ‘the creation of the cosmos by the one God’.

This author's thesis is also evident from counterarguments he refers to. For example, in Chapter 1 he refutes the view of those who believe in the eternity of the world (that is, dahrīyasFootnote ³⁹), as follows (B: 3a, 1–5; P: 21, 10–12):

[Fr. 10]

Since the dahrīyas were one of the dominant opponents of this thesis, the denial of their doctrine results in better support for it.

What should be noted is that the dahrīyas in his version adopt one of the main principles of astrology, namely the effect of celestial entities upon the sublunar world. Although he does not give any view on astrology in this chapter, his attitude towards it is suggested in Chapter 29 (Fr. 7), where he examines the reality of zodiacal signs. In fact, his obvious refutation of astrology is found in Chapter 14.

In this chapter, the author proves the eternal circular motion of the orb and the Sun endowed by God with recourse to the theory of the four elements and the existence of another element constructing celestial entities, and then he quotes the mentioned Qur'anic verse (Fr. 8). At the end of this chapter, he criticizes astrological concepts (B: 28b, 9–29b, 2):Footnote ⁴⁰

[Fr. 11]

Here he is clearly addressing the astrological notions of ‘triplicities’ and ‘beneficence and maleficence of the planets’.Footnote ⁴¹ As already seen, he considers superlunar objects as consisting of the fifth element, and as having no quality existing in sublunar objects. Thus it is not surprising to find him refusing to assign any qualities to superlunar entities. In this quotation, he explains that many scholars use this concept as a metaphor to convince ordinary people, and he ends this section with a promise to explain astrological concepts in a later section; however, we cannot find a relevant section in ArLO.

Although the author refutes astrological accounts, Fr. 11 undoubtedly shows that he gives a privileged position to the Sun and the Moon. In Chapter 19, he plainly declares as follows (B: 38b, 2–3),

[Fr. 12]

Then he demonstrates with a diagram the Moon's reception of the Sun's light based on the lunar phases. In Chapter 11, he quotes the following statement by ‘the philosopher’ (B: 16b, 2–3):

[Fr. 13]

Then he explains the relationship between the tidal ebb and flow and the Moon's motion.

However, it is evidently problematic to support the two notions, of the specific position of the Sun and the Moon, and of the existence of the fifth element,Footnote ⁴² but in ArLO the author does not state the reason why the Sun and the Moon have these special qualities. Instead he seems to think it satisfactory to merely show that the statements about the nature of the Sun and the Moon, including the one by ‘the philosopher’, correspond to daily observations.

As in this example, the author brings up the opinions of predecessors several times, but his sources are not clear. For example, in Chapter 18 (B: 35b, 9–36a, 1) he clearly mentions his source as Ptolemy's Structure of the Orbs (Kitāb tarkīb al-aflāk) in determining the Sun's magnitude, and describes how Ptolemy demonstrated the Sun to be larger than the Earth by examining the shadow of the Earth. At the end of this chapter, he gives the datum of its magnitude, 166⅜, without the calculation behind it. This title, Structure of the Orbs, suggests that he might have referred to Ptolemy's Planetary Hypothesis, but we cannot find in this work a demonstration using the Earth's shadow. It is remarkable, however, that the datum 166⅜ in ArLO is more exact than the 166⅓ in the Planetary Hypothesis, and the 170 in the Almagest.Footnote ⁴³ Thus it seems reasonable to assume that the author uses something other than the Ptolemaic works, but attributes its source to Ptolemy intentionally or unintentionally.

Chapter 27 has another quotation from Ptolemy's work. When the author describes the proportionality of the rotation period of a planet and the size of its orb, he quotes Ptolemy's explanation utilizing layers of different pulleys (baraka). However, since Ptolemy never adopted such proportionality, this quotation cannot come from Ptolemy.Footnote ⁴⁴

In both of them, the author obviously does not use Ptolemy's works, but other Greek astronomical sources.Footnote ⁴⁵ This fact suggests that ArLO was written before Ptolemy's works were canonicalized in the Islamic world, although Ptolemy's authority was already recognized. That is to say, ArLO was written fairly early as a book on the ‘ilm al-hay'a.

Chapter 34 has information which suggests the date when this work was written. In the latter half of this chapter, the author explains the time difference of the beginnings of a solar or lunar eclipse at different places. To show this, he recommends recording the starting times of an eclipse in several places, and gives the following example of a solar eclipse (B: 55b, 9–12; P: 43, 3–4):Footnote ⁴⁶

[Fr. 14]

With a computer program, we can confirm the occurrence of this solar eclipse. Since the author uses this incident as an example to support and vivify his explanation, it must still be a fresh memory for his readers. Hence I would assume ArLO to have been composed not too long after this eclipse.

This date proves the early composition of ArLO, but then contradicts its attribution to Māshā'allāh.Footnote ⁴⁷ This attribution is doubtful also from the fact that the author of ArLO rejects the qualities of stars, whereas Māshā'allāh accepts them. Hence there emerges a question about its authorship, which is the main topic of the next section.

The authorship of the Arabic Liber de orbe

The above information about the solar eclipse also indicates that this book was written in or near the Andalusia region, since it was recorded in al-Andalus. Furthermore, the following features of ArLO are suggestive of the place of its composition:

• 1 In Chapter 17 the author gives an example of comparison between the Ifrīqīyya (i.e. Tunisia), Egypt and Iraq.

• 2 The entirety of Chapter 12 is devoted to the flood of the river Nile.

• 3 Chapter 10 is on volcanoes in Rome and Sicily.

These clues lead us to a tentative conclusion that it was written around AD 940 in the Andalusia–Maghreb area, especially a region close to Egypt and Sicily; that is, in the Ifrīqīyya.

At that time, the Ifrīqīyya was the capital of the Fatimid dynasty, a scholastic centre of the Maghreb.Footnote ⁴⁸ One of the most prominent scholars was Isaac Israeli (c.855–c.955), a Jewish physician and philosopher in the Fatimid court.Footnote ⁴⁹ What is remarkable is that he had a disciple well versed in astronomy: Dūnash ibn Tamīm (d. after 955).Footnote ⁵⁰

Biographical information on Dūnash is very scarce, but two of his works have descended to us, namely Treatise on the Armillary Sphere, and Commentary on the Sefer Yezira. The Treatise on the Armillary Sphere is extant only in Istanbul, as Ayasofya MS 4861 (henceforth MS A).Footnote ⁵¹ In this work, he describes in great detail how to construct and operate the armillary sphere, an astronomical instrument consisting of rings. Thus its contents confirm his specialism in astronomy. We are also informed from the Commentary on the Sefer Yezira that he wrote other astronomical works.

The Sefer Yezira, or ‘Book on Creation’, is a Hebrew esoteric book on cosmogony, and several scholars wrote commentaries on it.Footnote ⁵² Among them, the earliest extant is the Arabic commentary by Saadia Gaon (884–942), a contemporary of Isaac Israeli.Footnote ⁵³

Dūnash wrote the next-earliest commentary in Arabic. However, we do not have a complete Arabic manuscript of it. Fortunately, about 75 per cent of the text was salvaged by George Vajda and Paul Fenton from Judaeo-Arabic documents, and the whole text is transmitted in several Hebrew versions.Footnote ⁵⁴ Because it was written fairly early as a commentary on the Sefer Yezira, it is an important source to explore the early stage of the reception of the Sefer Yezira.

In this work, Dūnash comments on the passages of the Sefer Yezira one by one. In the course of the commentary, he mentions and criticizes Saadia's comments several times. We must remark that he gives some information about himself, especially in the commentary on 1.5,Footnote ⁵⁵ which is only extant in the Hebrew translations.Footnote ⁵⁶

At the beginning of this part, Dūnash comments on the passage ‘the limit of beginning, the limit of end’. He interprets it as ‘the limit of the creation and termination of the world’, and explains that this limit can be perceived by human beings through their rational faculty as well as through traditional stories. As for the rational perception, he notes that they can realize the existence of the Rotator, namely God, by observing the eternal rotation of the orb. On the other hand, he gives Genesis as an example of the traditional accounts of God's creation of the world. When describing the details of Genesis, he mentions his current date: 344 H (=955–956).Footnote ⁵⁷ This date was almost at the time when ArLO was being written.

Next, Dūnash moves to the passage ‘the limit of good and the limit of bad’, and remarks that good and bad are not substances but attributes, and he declares the oneness of God. Then he refutes the dualists who believe the existence of the good and the bad as the light and the dark, by using the example of lunar phases which show that light is produced by the Sun and dark occurs when something blocks the light. This example is concluded by the following note:Footnote ⁵⁸

In this note, he explicitly mentions one of his astronomical works containing a section on the ‘ilm al-hay'a, a section on astronomical calculation, and a section on astrology.

Afterwards, Dūnash comments on the passage ‘the limit of east and the limit of west’. Here he focuses on the unreality of the east and the west, because there are unlimited terrestrial points, each of which has east and west. Then he adds the following remark:Footnote ⁵⁹

In this note, he tells us that he explained the unreality of the determination of constellations (i.e. zodiacal signs) in the treatise on the weakness of the principles of astrology, which is the second chapter of Book on the Configuration of the Orb dedicated to the third Fatimid Caliph, al-Manṣūr (r. 945–952). As is evident from its title, this book is also a work on the ‘ilm al-hay'a.

These quotations show that around AD 940 Dūnash wrote at least two books containing the ‘ilm al-hay'a. Moreover, what is remarkable is that the contents of his comments on 1.5 indicate Dūnash's sharing of some distinctive opinions with the author of ArLO:

• 1 Dūnash defends ‘the creation of the world by the one God’ and rejects opinions contrary to his thesis, such as dualism; this is also the main thesis of ArLO.

• 2 He has explained lunar phases with a diagram in one of his books on the ‘ilm al-hay'a; ArLO also has a section (Chapter 19) on lunar phases with a diagram.

• 3 He is sceptical about astrology, and his remark on ‘the unreality of constellations’ is comparable to Fr. 7 of ArLO, where the author describes that the places of the zodiacal signs in the orb of the zodiacal signs are not actual.

As is clear from the above comparison, Dūnash is a promising candidate for the author of ArLO. To confirm this possibility, I will further analyse the contents of Dūnash's two works.

Dūnash's Treatise on the Armillary Sphere

Before beginning to examine the contents of the Treatise on the Armillary Sphere, I will describe MS A itself.Footnote ⁶⁰

Description of MS A

Istanbul, Ayasofya MS 4861, folios 1a–60a. Two initial flyleaves and sixty leaves (folios i–ii, and 1a–60a).

Beginning (folio 1b, lines 1–5):

Ending (folio 60a, lines 13–15):

Colophon (folio 60a, lines 15–19):

The colophon shows that this copy was completed on 613 H/1216 by ‘Abd al-Qawī ibn ‘Abd al-Mu‘ṭī for personal use. The title and the author are given at the beginning of the treatise, as already described. They are also noted by the copyist on the front page (folio 1a, lines 5–7):

The copy has been collated with the exemplar. The copyist entered corrections in the margins (sometimes with the mark ‘correct’ – or ).

The Contents of the Treatise of the Armillary Sphere

The Treatise of the Armillary Sphere consists of an introduction and two Parts In the introduction (folios 1b–3b), Dūnash explains the reason why he composed this treatise.Footnote ⁶¹ Then Part 1 contains the following three topics preliminary to the operation of the instrument:

Part 2 in forty-three chapters shows how to operate the instruments, as well as how to show celestial phenomena by using it (Chapters 23, 28, 29, 39, 40 and 41).Footnote ⁶² Although this treatise is technical on the operation of the armillary sphere, it contains three examples to be compared with ArLO.

Example 1

At the beginning of Part 1, Dūnash introduces the main components of the orb, as follows (folio 3b, lines 12–16):Footnote ⁶³

Then he enumerates the following components as the main parts:

Dūnash clearly lists the same components as in ArLO, Chapter 16 (=Fr. 5 and Fr. 6). Furthermore, we must remark that as does the author of ArLO (Fr. 4), Dūnash presents them as the basis on which to understand the orb.

Example 2

At the end of Part1, Dūnash explains the sphericity of the orb, the Earth and the three elements (folio 9b, line 16–10a, line 10):

Here Dūnash describes the cosmos as concentric orbs. According to his explanation, the reason for the sphericity of water and air is that they are between the two spheres; that is, the Earth and the orb. His cosmology and the reasoning of the sphericity of the elements are strikingly similar to ArLO, Chapter 5 (Fr. 1).

Example 3

In Part2, Chapter 28, ‘on the knowledge of the climates and their notions’, he calculates the circumference of the Earth (folio 34a, lines 1–13). He uses the magnitude of one degree taken by the predecessors – that is, 66⅔ miles – and he calculates the circumference as 24,000 miles (=66⅔ × 360). This calculation is the same in ArLO, Chapter 34, which I have mentioned.

The above three examples illustrate that Dūnash and the author of ArLO share specific topics and opinions. Next, I will examine the contents of the Commentary on the Sefer Yezira.

The Arabic original of the Commentary on the Sefer Yezira

In Dūnash's commentary on the Sefer Yezira 1.5, we have found several striking similarities to ArLO. However, this part has the significant discrepancy that Dūnash calls the greatest orb the ninth orb, whereas in ArLO it is the tenth orb, but we should consider the fact that the comment on 1.5 is left only in its Hebrew translations. As Vajda elucidates,Footnote ⁶⁴ the Hebrew versions sometimes deviate from the Arabic original; for example, some Hebrew translators make Dūnash's criticism against Saadia Gaon milder. Thus it is possible that in some of the Hebrew translations ‘tenth’ has been changed to ‘ninth’ to accommodate Dūnash's cosmology into the nine-orbs system, a more generally accepted model. Therefore the part of the commentary extant in the Judaeo-Arabic documents, namely the portion after the commentary on 1.6, is important to comprehend Dūnash's thought directly.

Throughout the text, we notice that Dūnash tries to explain the Sefer Yezira's three-elements (fire, air and water) theory by employing the four elements.Footnote ⁶⁵ In the comment on 2.5b, ‘He [God] finds that all the creatures are produced in one name’,Footnote ⁶⁶ he explains how the four elements were produced:Footnote ⁶⁷

In this comment, he presents the four elements as concentric orbs, just as the author of ArLO does in Chapter 2. We must note that although the process of their generation is never mentioned in ArLO, it is explained here. In the comment on the Sefer Yezira 1.11 – ‘Three: Water from Air. He engraved and hewed out through it Emptiness and Void, Mud and Mire. He engraved it like a kind of garden bed; He raised it like a kind of wall, He surrounded it like a kind of ceiling’Footnote ⁶⁸ – Dūnash has already elucidated the transformation of the elements with respect to God's creation:Footnote ⁶⁹

What is remarkable about this quotation is that he mentions geological phenomena caused by huge wind, such as earthquakes and eclipses, namely the Sun's darkening. This description clearly matches ArLO, Chapter 9 (Fr. 2 and Fr. 3).

The above examples of Dūnash's views on the sublunar world and the four elements bear a strong resemblance to the corresponding contents of ArLO. As for Dūnash's attitude toward the superlunar world, it is worth noting that in the comment on the Sefer Yezira 6.4 –

– he discusses the natures of the zodiacal signs:Footnote ⁷²

Here he refers to the notion of ‘triplicities’, which is used in Chapter 14 of ArLO, and then explains that the author of the Sefer Yezira utilized this notion towards ordinary people. The example and reasoning Dūnash presents is very similar to that in ArLO.

Since the above comparison between ArLO and Dūnash's two works gives us enough evidence about authorship, we can fairly conclude that the author of ArLO was Dūnash. Then, by using some clues, we will identify ArLO among his works.

Conclusion: the identification of the Arabic Liber de orbe

Coming back to the account on astrology in Chapter 14 of ArLO (Fr. 11), we are reminded that this book does not have the section on astrological notions which was promised by its author in this chapter. This is unusual for ArLO, because the author explains all the topics he promises to describe afterward. This anomaly leads us to think that ArLO is a part of a large work, lacking a section on astrology. This possibility is congruent with its imperfect introduction, only having the laudation to God; for the original introduction could have become inappropriate after the extraction of ArLO from the larger work. Finally, given that the only sacred phrase in ArLO quoted to verify the main thesis is from the Qur’ān (Fr. 8), this work was probably written for Muslims. Consequently, the above analysis gives us promising candidates for the author and the title of ArLO – Dūnash ibn Tamīm, Book on the Configuration of the Orb (dedicated to al-Manṣūr) – although it lacks the astrological section ‘on the weakness of the principles of astrology’.

The identification of ArLO reveals its importance in the history of astronomy as well as in the history of philosophy. If it is correct, this work is definitely one of the earliest works on the ‘ilm al-hay'a, which provides us with valuable materials for investigating the formation of the ‘ilm al-hay'a tradition in the Islamic world before the time of Ibn al-Haytham. As a philosophical work, it is also a precious textual source disclosing the development of Isaac Israeli's philosophical system, which was immensely influential on the Latin Aristotelianism. It is to be hoped that my critical edition of ArLO will profit both fields.