Early Science and Medicine 15 (2010) 237-265 www.brill.

nl/esm

Al-Fārābī on the Method of Astronomy

Damien Janos
Ruhr-Universität Bochum*

Abstract
is article analyzes al-Fārābī’s (d. 950) conception of the astronomical method by
examining rarely studied texts such as the K. al-mūsīqā and K. al-burhān and by addres-
sing key issues such as the subject matter of astronomy, the techniques used to derive
the first principles of this science, the relation between astrology, astronomy, physics,
and metaphysics, and the place of al-Fārābī in the Arabic astronomical tradition. e
analysis indicates that al-Fārābī’s theories combine material from the Greek astro-
nomical tradition, especially Geminus, as well as from the logical works of Aristotle,
particularly the Posterior Analytics. Moreover, it enables us to view al-Fārābī as a link
between the Greek astronomers on the one hand and Ibn Sīnā and Naṣīr al-Dīn al-Ṭūsī
on the other.

Keywords
al-Fārābī, Arabic astronomy, Arabic philosophy, scientific method

Introduction
The need to analyze the early falāsifah’s conception and practice of
astronomy has already been underlined by some scholars, but to this
day very little research has been devoted to this subject.1 This neglect

* Ruhr-University, Bochum, SH 1/193, Universitätsstraße 150, D-44780 Bochum,

Germany (dtjanos@yahoo.com). I warmly thank Professors A.I. Sabra, F.J. Ragep,
and R. Morrison for their insightful remarks on previous drafts of this article. Any
remaining shortcomings are entirely my own.
1)
A.I. Sabra, “Configuring the Universe: Aporetic, Problem Solving, and Kinematic
Modeling as emes of Arabic Astronomy,” Perspectives on Science 6 (1998), 288-330,
316-317, stresses the need to study the astronomical works of the falāsifah. G. Saliba,
Islamic Science and the Making of the European Renaissance (Cambridge, Mass., 2007),

© Koninklijke Brill NV, Leiden, 2010 DOI: 10.1163/157338210X493941

238 D. Janos / Early Science and Medicine 15 (2010) 237-265

is perhaps partly due to the fact that the Arabic astronomical tradition
must be carefully studied before any attempt is made to understand the
contribution of the falāsifah to this field. Yet it seems that there exists
sufficient information on ʿilm al-hayʾa today to begin this enterprise
and to provide a sketch of the role that astronomy plays in the works
of al-Kindī, al-Fārābī, and Ibn Sīnā.2 In the following paragraphs, I will
examine al-Fārābī’s conception of the method of astronomy and its
connection with the Greek and Arabic philosophical and scientific trad-
itions. After clarifying the relation between astronomy and astrology, I
discuss al-Fārābī’s conception of the subject matter, method, and first
principles of astronomy, as well as of the place of mathematics in the
scientific methodology.
There are no extant works in which al-Fārābī provides a systematic
exposition of the astronomical method. His views must be reconstructed
from a variety of sources, the most important of which is the K.
al-mūsīqā al-kabīr (henceforth K. al-mūsīqā).3 Although primarily

171-185, and R. Morrison, Islam and Science: e Intellectual Career of Nīẓām al-Dīn
al-Nīsābūrī (Routledge, 2007), 13-16, 30-35, briefly discuss the place of hay’a in Arabic
philosophy. To my knowledge, the only attempt to study the place of astronomy in
al-Fārābī’s works was made by the Russian historian A. Kubesov, who devoted several
studies to al-Fārābī’s cosmology in the 1970s and 1980s. Unfortunately, they were
based on a manuscript purporting to be al-Fārābī’s Sharḥ al-majisṭī, but which later
turned out to be a copy of a text originally composed by Ibn Sīnā (see B. Goldstein,
“Book Review of F. Sezgin’s Geschichte des arabischen Schrifttums, vol. VI, Astronomie
bis ca. 430 H.,” Isis 71/2 (1980), 341-342). Kubesov must nevertheless be given credit
for being one of the only historians to have examined the cosmology and astronomy
of early Arabic philosophers. For an assessment of Ibn Sīnā’s astronomical output, see
F.J. Ragep and S.P. Ragep, “e Astronomical and Cosmological Works of Ibn Sīnā:
Some Preliminary Remarks,” in Sciences, techniques et instruments dans le monde iranien
(Xe-XIXe siècle), edited by N. Pourjavady and Ž. Vesel (Téhéran, 2004), 3-17; see also
S. Ragep, “Ibn Sīnā,” in e Biographical Encyclopedia of Astronomers, Springer Refer-
ence, edited by T. Hockey et al. (New York, 2007), 570-572.
2)
is is especially true given Saliba’s studies on the early phase of hayʾa: G. Saliba
“Early Arabic Critique of Ptolemaic Cosmology: A Ninth-Century Text on the Motion
of the Celestial Spheres,” Journal for the History of Astronomy 25 (1994), 115-41; and
“Islamic Astronomy in Context: Attacks on Astrology and the Rise of the Hayʾa Tra-
dition,” Bulletin of the Institute for Inter-Faith Studies 4 (2002), 73-96.
3)
I have used the following edition: Kitāb al-mūsīqā al-kabīr, edited by Ghaṭṭās ʿAbd
al-Malik Khashabah (Cairo, 1960).
 D. Janos / Early Science and Medicine 15 (2010) 237-265 239

concerned with musicology, this work nevertheless contains a wealth

of information on philosophical methodology and the theory and prac-
tice of the sciences, as well as insight into al-Fārābī’s interpretation of
the Posterior Analytics. When read in conjunction with other works such
as the Iḥṣāʾ al-ʿulūm, his two treatises on astrology, and his metaphysical-
political treatises (namely, the Mabādiʾ ārāʾ ahl al-madīnah al-fāḍilah
and the Al-siyāsah al-madaniyyah),4 the K. al-mūsīqā enables us to recon-
struct a relatively satisfactory picture of al-Fārābī’s approach to astron-
omy.
It should be noted that al-Fārābī is also credited by the bio-bibliog-
raphers with a commentary (sharḥ) on the Almagest, and both Brock-
elmann and Sezgin mention two manuscripts, one in the British Library
and the other in the Majlis Library in Tehran, which purport to be this
sharḥ.5 Unfortunately, the manuscript in the British Library is in fact a
commentary by Ibn Sīnā that was mistakenly attributed to al-Fārābī.6
As for the Tehran manuscript, either it has been lost or the reference
rests on another misattribution.7 The conclusion, then, is that as far as
we know, no copy of al-Fārābī’s sharḥ has come down to us.

4)
Iḥṣāʾ al-ʿulūm lil-Fārābī, 2nd ed., edited by O. Amine (Cairo, 1949); the two trea-
tises on astrology have been edited: Maqālah fīmā yaṣiḥḥu wa mā lā yaṣiḥḥu min
aḥkām al-nujūm, in Al-aʿmāl al-falsafiyyah, edited by Jaʿfar Āl Yāsīn (Beirut, 1992),
279-313; Maqālah li-Abī Naṣr al-Fārābī fī al-jihah allatī yaṣiḥḥu ʿalayhā al-qawl fī
aḥkām al-nujūm, in Nuṣūṣ falsafiyya muhdāh ilā al-duktūr Ibrāhīm Madkūr, edited by
M. Mahdi (Cairo, 1976), 69-74; Al-Farabi on the Perfect State: Abū Naṣr al-Fārābī’s
Mabādiʾ ārāʾ ahl al-madīna al-fāḍila: A Revised Text with Introduction, Translation,
and Commentary by Richard Walzer (Oxford; New York, 1985); Kitāb al-siyāsah al-
madaniyyah al-mulaqqab bi-mabādiʾ al-mawjūdāt, edited by M. Najjār (Beirut, 1964).
5)
See al-Qifṭī, Tarīkh al-ḥukamāʾ, edited by J. Lippert (Leipzig, 1903), 279, and
Ibn Abī Uṣaybiʿah, ʿUyūn al-anbāʾ fī ṭabaqāt al-aṭibbāʾ, edited by N. Riḍā (Beirut,
1965), 608; F. Sezgin, GAS, 1967-, Band 6, 195, which also includes the references
to Brockelmann.
6)
B. Goldstein was the first to notice this in his “Book Review of F. Sezgin’s Geschichte,”
342. I myself was able to consult a copy of this manuscript and can confirm that
although al-Fārābī’s name appears twice on the first folios, a comparison with Ibn Sīnā’s
commentary on the Almagest as it appears in the “Mathematics” of the K. al-shifāʾ
reveals that it should be ascribed to the latter.
7)
In spite of several attempts, I could not obtain a copy of this manuscript from the
Majlis Library, and its staff was apparently unable to identify it.
240 D. Janos / Early Science and Medicine 15 (2010) 237-265

e Relation between Astronomy and Astrology

T.-A. Druart has already shown that al-Fārābī wrote several treatises in
which he attempted to distinguish not only between astronomy and
astrology, but also between the valid and invalid parts of astrology, i.e.,
between those parts that rely on a sound scientific method and those
that do not.8 An example of a sound astrological pursuit according to
al-Fārābī consists in observing the effects that the heat emitted by the
celestial bodies has on sublunary beings. In contrast, he considers that
the attempt to predict the deaths of kings and other dignitaries from
the positions of the stars and planets is devoid of any scientific validity
and rests on flawed analogical reasoning.9 In fact, al-Fārābī rejects any
kind of astral prognostication. I will limit myself here to a few additional
comments on al-Fārābī’s perception of astrology and its relation to
mathematical astronomy.
First, it is important to realize that al-Fārābī subsumes astronomy
and astrology under a single science, which he calls “science of the stars”
(ʿilm al-nujūm). As al-Fārābī explains in the Iḥṣāʾ, ʿilm al-nujūm is div-
ided into two main parts. The first one, astrology, or rather judicial
astrology (ʿilm aḥkām al-nujūm), focuses on the planetary signs and
indications that can reveal the existence of past, present, and future
events. The second part, mathematical astronomy (ʿilm al-nujūm
al-taʿlīmī), investigates the exterior properties of the heavenly bodies,
especially those that pertain to numbers and measurements, such as the
sizes, distances, and motions of the planets.10
Al-Fārābī’s classification in the Iḥṣāʾ shows that he did not succeed
in separating scientific astronomy completely from other cosmological
disciplines in the way that Ibn Sīnā, al-Ṭūsī, and others would several
centuries later. According to Ibn Sīnā, astronomy, or the ‘science of the

8)
T.-A. Druart, “Astronomie et astrologie selon Fārābī,” Bulletin de philosophie médié-
vale 20 (1978), 43-47; see also “Le second traité de Fārābī sur la validité des affirma-
tions basées sur la position des étoiles,” Bulletin de philosophie médiévale 21 (1979),
47-51; and “Al-Fārābī’s Causation of the Heavenly Bodies,” in Islamic Philosophy and
Mysticism, edited by P. Morewedge (New York, 1981), 35-47. See also Y. Michot,
Réfutation de l’astrologie (Beyrout, 2006), 55-60.
9)
Al-Fārābī, Maqālah fī al-jihah; and Maqālah fī mā yaṣiḥḥu, sections 4 ff., 17, 20,
and 23-24; Druart, “Astronomie,” 46-47.
10)
Al-Fārābī, Iḥṣāʾ, 84.
 D. Janos / Early Science and Medicine 15 (2010) 237-265 241

configuration [of the world]’ (ʿilm al-hayʾa), is an independent math-

ematical science (ʿilm riyāḍī) that is exclusively concerned with the
motion, arrangement, and external properties of the celestial bodies
and superlunary phenomena. Astrology (aḥkām al-nujūm), on the other
hand, deals with the effects of the planets on the sublunary world and
is classified as a physical science (ʿilm ṭabīʿī).11 As far as we know, Ibn
Sīnā was the first faylasūf to clearly achieve this separation between
mathematical astronomy and physical astrology. In contrast, in apply-
ing the term ʿilm al-nujūm to both astronomy and astrology, al-Fārābī
is following other contemporary classifications of the sciences, such as
those of the Ikhwān al-Ṣafāʾ and al-Khwārizmī, which placed mathema-
tical astronomy and astrology under a single overarching discipline.12
From a historical perspective, al-Fārābī’s classificatory scheme is
closer to that of ancient Greek thinkers than to that of Ibn Sīnā and
the hayʾa practitioners of the post-eleventh century, chiefly on account
of a common terminological ambiguity. Like ʿilm al-nujūm, ἀστρο-
νομία can be used as a generic term to refer to both astronomy and
astrology. In fact, as can be seen from the various entries in Liddell and

11)
See for example the Risālah fī aqsām al-ʿulūm al-ʿaqliyyah, edited by A. Hundīh,
reprinted in Islamic Philosophy, vol. 42, Abū ʿAlī al-Ḥusayn ibn ʿAbdallāh ibn Sīnā,
Philosophical Treatises, Institute for the History of Arabic-Islamic Science at the Johann
Wolfgang Goethe University (Frankfurt am Main, 1999), 120-121.
12)
Al-Khwārizmī, Kitāb mafātīḥ al-ʿulūm, edited by G. van Volten (Lugduni-Bata-
vorum, 1968), 210-232; Ikhwān al-Ṣafāʾ, Rasāʾil Ikhwān al-Ṣafāʾ wa-khillān al-wafāʾ,
edited by Khayr al-Dīn al-Ziriklī (Miṣr, 1928), vol. 1, 73; F.J. Ragep, Naṣīr al-Dīn
al-Ṭūsī’s Memoir on Astronomy (al-Tadhkira fī ʿilm al-hayʾa), 2 vols. (New York, 1993),
vol. 1, 34-35. To my knowledge, al-Fārābī uses the term ʿilm al-hayʾa only once in his
works (in the short treatise Qawl al-Fārābī fī al-tanāsub wa al-taʾlīf, in Al-manṭiqiyyāt
lil-Fārābī, edited by Muḥammad Taqī Dānishʾpazhūh (Qum, 1987), vol. 1, 504-506,
505) and usually refers to astronomy as ʿilm al-nujūm. In the Qawl, however, al-Fārābī
clearly has mathematical astronomy in mind, since he mentions the Almagest and the
“demonstrative proofs” of this science. is raises the question of how much import-
ance should be attributed to the account in the Iḥṣāʾ and its use of the generic term
ʿilm al-nujūm, especially since Gutas has shown that this treatise may be a reworking
of a late-antique Alexandrian classification of the sciences; see D. Gutas, “Fārābī,” IV.
Fārābī and Greek Philosophy, in Encyclopaedia iranica, edited by E. Yarshater (Lon-
don, 1982-), 219-220; idem, “Paul the Persian on the Classification of the Parts of
Aristotle’s Philosophy: A Milestone between Alexandria and Baġdād,” Der Islam 60
(1983), 231-67, 225-260.
242 D. Janos / Early Science and Medicine 15 (2010) 237-265

Scott’s Greek-English Lexicon and in Brill’s New Pauly, Greek authors of

the classical period often used the terms ἀστρονομία and ἀστρολογία
very loosely and without distinguishing clearly between the two disci-
plines. It appears that it is only towards the end of antiquity that a more
specific terminology was developed to refer to the mathematical aspects
of astronomy.13
In spite of the foregoing, al-Fārābī establishes an important termin-
ological and conceptual distinction between astronomy and astrology.
Terminological, in that scientific astronomy (ʿilm al-nujūm al-taʿlīmī)
is contrasted to astrology (ʿilm aḥkām al-nujūm); and conceptual, in
that the subject matter and epistemological criteria that correspond to
these two disciplines are different. Whereas astrology focuses on the
effects of the celestial bodies here on earth, which it studies either
empirically and through direct observation of the physical world, or
through analogical reasoning, astronomy relies on demonstration to
establish proofs concerning the motions of the celestial bodies and their
spatial relation with regard to the earth. Moreover, whereas the subject
matter of astrology consists of things that occur sometimes or for the
most part, astronomy studies what occurs necessarily and has for subject
matter the eternal and unchanging celestial beings.14
In brief, it may be said that al-Fārābī anticipated in some ways the
distinction between astronomy and astrology that one finds in the works
of Ibn Sīnā, al-Bīrūnī, and the hayʾa practitioners of the thirteenth and
fourteenth centuries, although he did not carry out this separation to
its full extent. On the one hand, al-Fārābī used different Arabic words
to distinguish mathematical astronomy from astrology and was very
suspicious of certain aspects of astrology, which he criticized in treatises
he wrote specifically to this effect. In his critical assessment of astrology,
al-Fārābī departs markedly from Ptolemy, who fully endorsed it, as can
be seen in the Tetrabiblos, a voluminous work devoted to this subject.15

13)
See the articles ‘astronomy’ by F. Krafft and ‘astrology’ by W. Hübner and
H. Hunger in the Brill’s New Pauly 2002; I. Mueller, “Physics and Astronomy:
Aristotle’s Physics II.2.193b22-194a12,” Arabic Sciences and Philosophy 16/2 (2006),
175-207, 178, note 10.
14)
Al-Fārābī, Iḥṣāʾ, 84; Druart, “Astronomie et astrologie selon Fārābī,” 44.
15)
For the influence of Ptolemy’s Tetrabiblos on the formation of early Arabic astrology
and its demarcation from astronomy, see C. Burnett, “e Certitude of Astrology: e
 D. Janos / Early Science and Medicine 15 (2010) 237-265 243

On the other hand and in spite of this, al-Fārābī subsumed astronomy

and astrology under a single overarching discipline which he called ʿilm
al-nujūm and did not go as far as Ibn Sīnā in isolating mathematical
astronomy completely from the other cosmological disciplines.

e Subject Matter of Astronomy

The Iḥṣāʾ, which contains al-Fārābī’s most systematic description of the
classification of the sciences, provides much information on astronomy.
In this treatise, astronomy is presented as a mathematical science,
together with arithmetic, geometry, music, and optics.16 Al-Fārābī then
divides the subject matter of mathematical astronomy into three parts.
The first one deals with the exterior aspects of the heavenly bodies, such
as their shapes, positions, and sizes. It also includes an examination of
the earth and asserts its stationary position in the world. The second
part deals with celestial motion, both the general motion shared by all
the celestial bodies and the particular motions of the planets. The third
part focuses on the earth and related geographical, climatological, and
demographical questions. I here provide a new translation of the entire
passage, because it sheds light on al-Fārābī’s understanding of astron-
omy:

Mathematical astronomy [ʿilm al-nujūm al-taʿlīmī] examines three aspects of the

celestial bodies and the earth:
First, [it examines] their shapes, the positions of some of them vis-à-vis others,
and their order in the world, as well as the sizes of their bodies, the relations that
exist between them, and the measures of the distances between them. [It also
shows] that the earth as a whole does not move from its center or in its center.
Second, [it examines] how many motions the celestial bodies have and the fact
that all of their motions are circular. [It studies] those [motions] that are common
to the planets and other non-planetary bodies, and those [motions] common to
all the planets as well as those that are specific to each. [It also examines] the num-
ber of each kind of these motions, the directions toward which they move, and

Scientific Methodology of al-Qabīṣī and Abū Maʿshar,” Early Science and Medicine
7/3 (2002), 198-213.
16)
Al-Fārābī, Iḥṣāʾ, 43; see also Naṣṣ al-tawṭiʾah aw al-risālah allatī ṣuddira bi-hā
al-manṭiq, edited by R. Al-ʿAjam (Beirut, 1985), vol. 1, 55-63, 59, which also includes
astronomy in the mathematical sciences.
244 D. Janos / Early Science and Medicine 15 (2010) 237-265

from which direction each one of these motions originates. It also makes known
the means to establish the place of each star one by one in the parts of the zodiac
at each moment and with the totality of its kinds of movements.
It investigates also into everything that is concomitant with the celestial bod-
ies and each one of their motions in the zodiac and what pertains to the relation
between them due to their conjunction, separation, and the diversity of their posi-
tions.
In brief, [it examines] everything that pertains to their motions insofar as it
relates to the earth, like the eclipse of the sun. [And it investigates] everything that
occurs to them on account of the place of the earth among them in the world,
such as the eclipse of the moon. [It looks into] the number of these occurrences,
in what state and at what time and how often they appear, like the rising and set-
ting of the sun and other such things.
ird, it studies the inhabited and uninhabited regions of the earth. It estab-
lishes how many parts are inhabited and how many are its major regions which
are the climes, and it classifies the places that happen to be inhabited at a partic-
ular time, as well as the place of each inhabited region and its organization in the
world. Moreover, it studies what necessarily affects each one of the climes and
inhabited zones due to the common revolution of the world in the universe [ʿan
dawrat al-ʿālam al-mushtarakah lil-kull], which is the cycle of day and night, on
account of the position of the earth: like the rising and setting of the sun, the
length of days and nights, and other similar things. All of this is comprised by this
science.17

Al-Fārābī’s general description of the subject matter of mathematical

astronomy is historically significant, for it anticipates several features
of the classification later found in hayʾa works. The inclusion of what
today belongs to the disciplines of geography and climatology into
astronomy is one of the marking features of this classification. The
Tadhkirah of al-Ṭūsī, for instance, is divided into hayʾat al-arḍ and
hayʾat al-samāʾ, the former corresponding broadly to the third part of
al-Fārābī’s account in the Iḥṣāʾ, the latter to the two first parts.18 More-
over, according to the third paragraph of the Iḥṣāʾ account, astronomy
is concerned not only with the planets, but also with the non-planetary
celestial bodies, that is, with the celestial spheres, which are conceived
of as concrete existents. Like later hayʾa authors, al-Fārābī believes that
the motion and arrangement of the spheres represent important aspects

17)
Al-Fārābī, Iḥṣāʾ, 84-86.
18)
Ragep, Naṣīr al-Dīn al-Ṭūsī’s Memoir on Astronomy, vol. 1, 36.
 D. Janos / Early Science and Medicine 15 (2010) 237-265 245

of the astronomical project. Finally, al-Fārābī’s outline of the subject

matter of astronomy clearly reflects his belief that astronomy is limited
to studying the exterior aspects and properties of the celestial bodies
and that it is not suited to examine their substance. In fact, he says
nothing here about the inner nature of the spheres or about aether, the
simple element of the heavens, which he discusses at length in other
treatises.19 This omission is due to the fact that, as a mathematical sci-
ence, ʿilm al-nujūm al-taʿlīmī is devoted to the study of those properties
of bodies that can be abstracted from matter, while it is the task of
physics to study the celestial substance.
That the investigation of the celestial substance is reserved for phys-
ics, not astronomy, is confirmed by another section of the Iḥṣāʾ, where
al-Fārābī explains that it is a part of physics (the second in his classifica-
tion) that must investigate “whether simple bodies exist, and if they do,
what kind of bodies they are, and how many they are?”20 By “simple
bodies,” al-Fārābī means not only the four sublunary elements (fire, air,
water, and earth), but also aether, the Aristotelian first body or fifth
element, which is conceived of as a simple, incorruptible substance. In
fact, Aristotle’s discussion of aether in De caelo I.2-4 is explicitly men-
tioned in this passage.
Hence, for al-Fārābī, astronomy and physics are separate sciences,
which can nevertheless study two different aspects of the same subject
matter, i.e., the celestial bodies. Whereas astronomy studies the exterior
aspects of the planets, physics inquires into their substance, what they
are made of, and how this substance relates to the sublunary elements.
While distinct in their methods, physics and astronomy share a com-
mon subject. This leads to certain important implications when it
comes to these sciences’ search for first principles; but I shall return to
this point later on.
It is notable that al-Fārābī’s conception of how astronomy and phys-
ics are related in terms of their common subject matter agrees in its

19)
See, for instance, his refutation of Philoponus’ criticism of aether in M. Mahdi,
“Alfarabi against Philoponus,” Journal of Near Eastern Studies 26/4 (1967), 233-260.
20)
Al-Fārābī, Iḥṣāʾ, 96. In this passage, al-Fārābī divides physics into eight parts, the
second of which inquires into the bodies that are simple, as opposed to the fifth part
that inquires into composite bodies.
246 D. Janos / Early Science and Medicine 15 (2010) 237-265

main lines with the views of some ancient Greek thinkers, such as
Simplicius. Simplicius made a distinction between the physical study
of the cosmos (embodied in the De caelo tradition) and the mathemat-
ical-astronomical approach, which studies those aspects of the celestial
bodies that can be abstracted from matter.21 This basic methodological
position, which al-Fārābī inherited from the late-antique commenta-
torial tradition, also appears in a modified form several centuries later
in the works of the hayʾa authors, such as Naṣīr al-Dīn al-Ṭūsī. As F.J.
Ragep writes, for al-Ṭūsī “...it was for ʿilm al-hayʾa to examine the
outward manifestations of simple bodies, whereas it was for al-samāʾ
wa-’l-ʿālam to investigate their essential nature.”22
More must be said about al-Fārābī’s conception of astronomy as a
mathematical science, which follows a well-established Greek tradition.
Plato in Republic VII, Aristotle in Metaphysics XII.8.1073b and possibly
in Physics II.2,23 Ptolemy in the Almagest, and Simplicius in his com-
mentary on the Physics,24 had all classified astronomy as a mathematical
science. This is not to say, however, that these thinkers conceived of the
place of mathematics in astronomy in an identical way.
There is in fact a great diversity in their approaches, which is due
among other things to the status of mathematical objects in their phil-
osophy. In the case of Plato, mathematics is inextricably linked to his
theory of the forms and of an ideal world beyond the realm of sense
perception. Mathematical objects have a privileged status due to their
immateriality and their ontological proximity to this purely intelligible
dimension, although Plato did not go as far as Speusippus in making
them the primary entities of his metaphysical doctrine. This explains
why Plato in the Republic argues that astronomy should be studied
“by means of problems, as we do geometry.”25 For Plato, astronomy is

21)
See Simplicius’ commentary on this passage in On Aristotle Physics 2, translated by
Barrie Fleet (London, 1997), 290,1-293,15.
22)
Ragep, Naṣīr al-Dīn al-Ṭūsī’s Memoir on Astronomy, vol. 1, 39.
23)
Mueller, “Physics and Astronomy.”
24)
See again Mueller, “Physics and Astronomy,” 179-184, and Simplicius, On Aristotle
Physics 2, 290,1-293,15.
25)
Republic, VII.530b, translated by G. M.A. Grube, revised by C.D.C. Reeve, in
Plato: Complete Works, edited by J.M. Cooper and D.S. Hutchinson (Indianapolis,
1997).
 D. Janos / Early Science and Medicine 15 (2010) 237-265 247

non-physical and does not deal primarily with bodies; rather, it deals
with “true numbers” and “geometrical figures.”26 If this mathematical
approach to astronomy is adopted, it can lead us closer to the divine
world, and, moreover, it should play an important role in the curricu-
lum of the guardians.
In the case of Aristotle, his intention in defining astronomy as a
mathematical science is grounded in methodological issues rather than
metaphysical ones. Aristotle believes that astronomy is primarily inter-
ested in the exterior aspects of the celestial bodies, which it studies
regardless of their inner nature and composition, albeit in connection
with motion. In contrast to the physicist, it is the privilege of the math-
ematician to be able to conceive of objects by abstracting them from
their matter, although these objects have no real, independent existence
outside of matter. Nevertheless, Aristotle also recognizes a physical
aspect to astronomy in that it focuses on the properties of real, moving
bodies, an ambiguity that led several scholars to the belief that astron-
omy is a physical science.27
As for Ptolemy, he had a completely different conception of the
mathematical dimension of astronomy. He may in many ways have
been influenced by the Platonic and Neoplatonic tendency to treat
mathematics as a special discipline that bears a close relation to the ideal
world of nous.28 But Ptolemy’s interest in mathematics is overwhelm-
ingly tied to the concepts of observation and scientific accuracy. Ptol-
emy undermines physics and metaphysics as cosmological disciplines
on the grounds that they produce speculative and unverifiable theories
about the universe. Mathematics, on the other hand, is able to formu-
late proofs that are logically compelling and demonstrative in essence.
This accounts for Ptolemy’s interest in observation and in the accumu-
lated data of past astronomical endeavours. Ptolemy’s interest in math-
ematics is therefore methodological and epistemological, and he regards
this science as the foundation of sound astronomical practice.

26)
Republic, VII.529d, translated by G.M.A. Grube.
27)
Mueller, “Physics and Astronomy,” argues, against Ross, that Aristotle classified
astronomy as a mathematical, not a physical, science, in spite of its having real bodies
as subject. At any rate, this is how Simplicius and other late-antique authors inter-
preted Aristotle.
28)
See L. Taub, Ptolemy’s Universe (Chicago, 1993), 135-155.
248 D. Janos / Early Science and Medicine 15 (2010) 237-265

Al-Fārābī’s conception of the relation between mathematics and

astronomy is likely to have been inherited from the Alexandrian school
tradition, and it may best be described as a hybrid between the Aristo-
telian and Ptolemaic positions on this issue. On the one hand and at a
methodological level, al-Fārābī broadly follows Aristotle’s classification
of the sciences as it was interpreted by the later commentatorial trad-
ition and Simplicius in particular, and more specifically the idea that
astronomy is a mathematical science that focuses on the exterior aspects
of the celestial bodies. He explains in the Aims of Aristotle’s Metaphysics
and the K. al-burhān that it is the task of mathematicians to treat figures
in abstraction from matter, although these have no real existence outside
of matter.29 As for the inner nature of the heavens, it is addressed by
physics, not astronomy, as the De caelo and later commentaries had
made clear. On the other hand, as we shall see shortly, al-Fārābī empha-
sizes the observational and experience-based aspects of astronomy, and
in that sense he is close to, and perhaps directly indebted to, Ptolemy.
Moreover, al-Fārābī’s various treatises on algebra and geometry and
especially his commentary on the Almagest indicate that he was genu-
inely interested in the various branches of mathematics, and, as a corol-
lary, in the mathematical dimension of astronomy.30

e Astronomical Method
One of the main questions addressed by al-Fārābī in the first introduc-
tory section of the K. al-mūsīqā concerns the epistemological founda-
tions of the particular sciences and especially music. In order to
strengthen his arguments, al-Fārābī compares music to other sciences,

29)
e Aims of Aristotle’s Metaphysics, in Classical Arabic Philosophy: An Anthology of
Sources, translated by J. McGinnis and D.C. Reisman (Indianapolis, 2007), 79; and
al-Fārābī’s discussion of Aristotle’s and Plato’s views on mathematical objects in his K.
al-burhān, in Al-manṭiq ʿinda al-Fārābī, edited by R. al-ʿAjam and M. Fakhry (Beirut,
1985), vol. 4, 68-69.
30)
See G. Freudenthal, “La philosophie de la géométrie d’al-Fārābī,” Jerusalem Studies
in Arabic and Islam 11 (1988), 104-220; and “Al-Fārābī on the Foundations of Geom-
etry,” in Knowledge and the Sciences in Medieval Philosophy, Proceedings of the Eighth
International Congress of Medieval Philosophy (S.I.E.P.M.), Helsinki 24-29 August 1987,
edited by M. Asztalos, J.E. Murdoch, and I. Niiniluoto (Helsinki, 1990), vol. 1, 54-61.
 D. Janos / Early Science and Medicine 15 (2010) 237-265 249

such as astronomy and medicine, and by so doing provides insight into

their method as well.
Al-Fārābī begins by classifying these disciplines into various categor-
ies depending on the method required to establish their first principles.
There are sciences, he tells us, whose first principles are acquired intui-
tively and from a very young age. There are other sciences some of
whose first principles are acquired in this manner, while others are
derived from separate sciences. Finally, there are sciences that rely on
both methods and in addition establish first principles through experi-
ence (tajribah).31 Al-Fārābī does not provide specific examples to illus-
trate the categories he posits. But it becomes clear shortly afterwards
that astronomy belongs to the third category, which means that some
of its principles are innate to humans, some are derived from other
sciences, while still others are reached as a result of experience. In the
following paragraphs, I discuss the latter two methods, namely, experi-
ence, and the transfer of principles from one science to another.

Experience and Induction

Al-Fārābī’s K. al-mūsīqā, together with his K. al-burhān, contains his
most systematic and detailed exposition of the importance of experi-
ence, observation, and induction in philosophy. Experience (tajribah)
in particular is discussed at some length in the K. al-mūsīqā.32 In this
work al-Fārābī explains that experience consists in the repeated sen-
sation (iḥsās) of facts, which enables the intellect (ʿaql) “to perform
an act” (yaf ʿalu), that is, to make a universal judgement on the basis
of these facts. As a result, experience can produce certain knowl-
edge and provides us with some of the first principles necessary for
demonstration. As al-Fārābī puts it, “the things acquired through

31)
Al-Fārābī, Mūsīqā, 96-97.
32)
See pages 92-96. Tajribah corresponds to Aristotle’s ἐμπειρία, which is described in
several of his works, for instance in Prior Analytics I.30.46a18-22 and Posterior Analytics
II.19.100a1-9. See J. McGinnis, “Scientific Methodologies in Medieval Islam,” Journal
of the History of Philosophy 41/3 (2003), 307-327; idem, “Avicenna’s Naturalized Epis-
temology and Scientific Method,” in e Unity of Science in the Arabic Tradition, edited
by S. Rahman, T. Street, and H. Tahiri (Dordrecht, 2008), 129-153; and J. Janssens,
“Experience (tajriba) in Classical Arabic Philosophy,” Quaestio 4 (2004), 45-62, for an
analysis of Ibn Sīnā’s and al-Fārābī’s interpretations of Aristotle’s theory of experience.
250 D. Janos / Early Science and Medicine 15 (2010) 237-265

experience [tajribah] become first principles [mabādiʾa ūlā] in demon-

strations [barāhīn].”33
Al-Fārābī provides more specific information on the role of experi-
ence in astronomy. At one point he writes that “the situation of this
science [the musical science] is like that of other sciences in which many
of the principles [mabādiʾ] are acquired through the experience of sens-
ibles [tajribat al-maḥsūsāt], as in astronomy [ʿilm al-nujūm] and most
of optics and medicine...”34 And shortly afterward, he adds that “many
of the principles of astronomy are acquired by the observer as a result
of sensation through observations by means of instruments [kathīr min
mabādiʾ ʿilm al-nujūm taḥṣulu lil-nāẓir fīhi ʿan al-iḥsās bi-al-arṣād bi-
al-ālāt].”35 According to al-Fārābī, then, the astronomer may derive
astronomical principles from the “experience of sensibles” and from
“observations by means of instruments.” A similar point appears in the
K. al-burhān, where al-Fārābī describes experience (tajribah) as a source
of knowledge for mathematical astronomy (ʿilm al-nujūm al-taʿlīmī).36
Now Janssens argues in his article “Experience in Classical Arabic
Philosophy” that experience and observation are close but different
concepts for al-Fārābī. “It is clear,” he writes, “that experience is not
opposed to observation, but is closely linked with it: they both pay
special attention to things, or events. But experience transcends ob-
servation, in that, contrary to the latter, it does not simply notice
particulars, but in addition tries to establish a kind of universality out

33)
Al-Fārābī, Mūsīqā, 96, my translation.
34)
Ibid., 100.
35)
Ibid., 101.
36)
Al-Fārābī, K. al-burhān, vol. 4, 71. It is in light of this emphasis on observation that
one should understand al-Fārābī’s distinction in the K. al-mūsīqā, 98-101, between
the theoretical and practical sides of astronomy. e theoretical astronomer, al-Fārābī
tells us, need not know how to use astronomical instruments, as long as he can rely
on someone else to perform the observations for him. He is like the musicologist
who does not need to know how to play an instrument if he can rely on someone
else to play the notes and tunes for him. As al-Fārābī explains, if, for some reason, an
astronomer is not able to benefit from the help of an observer or does not possess the
required technology to carry out the observations himself, then he is dependent on
the findings of his predecessors. e K. al-burhān, 75, also contains some comments
on the relation between the practical and theoretical dimensions of the sciences and
mentions astronomy as an example.
 D. Janos / Early Science and Medicine 15 (2010) 237-265 251

of a number of particulars.”37 Janssens is undoubtedly right in attribut-

ing to experience a claim to universal knowledge, and he adduces a
number of convincing passages from al-Fārābī’s works to buttress his
point. On the other hand, Janssens does not define observation in the
context of al-Fārābī’s philosophy, nor does he give the Arabic term that
would correspond to this concept. In fact, in the works studied by
Janssens and in the K. al-mūsīqā al-Fārābī does not use a special word
to express the concept of observation. True, in the latter work, he refers
to the “observations by means of instruments” (al-arṣād bi-al-ālāt) of
the astronomers, but it is unlikely, I think, that arṣād here refers to a
full-blown theory of observation in the way that tajribah refers to a
theory of experience.
On the other hand, the K. al-mūsīqā distinguishes clearly between
experience (tajribah) and induction (istiqrāʾ).38 What, then, is the dif-
ference between these two concepts? According to al-Fārābī, experience
involves a judgement of the intellect, which can extract universal know-
ledge from specific events or facts and lead to certainty. “Experience,”
al-Fārābī says in the K. al-burhān, “is what produces certitude of know-
ledge through a universal judgement.”39 This universal judgment is
reached through intellect, as the K. al-mūsīqā makes clear: tajribah is
defined as “the determination of the sensation of various things a
repeated number of times in order that the intellect … may act with a
special act [fiʿl khāṣṣ] and reach certainty …”40 This passage closely
mirrors the one found in al-Fārābī’s Talkhīṣ jawāmiʿ kitāb al-nawāmīs
li-Aflāṭūn: “the meaning of experience is the attentive consideration of
the particulars of a thing, more precisely forming a judgment about the
thing’s universality inasmuch as experience finds that universality in
these particulars.”41 The “special act” of the intellect in experience, then,

37)
Janssens, “Experience,” 50.
38)
Al-Fārābī, Mūsīqā, 94-96; this passage may have been inspired by Posterior Ana-
lytics I.31, where Aristotle explains that sense perception per se cannot lead to certain
knowledge and demonstration. See also Aristotle’s discussion of induction in Prior
Analytics II.23.
39)
Al-Fārābī, K. al-burhān, vol. 4, 24-25, translated in Janssens, “Experience,” 52.
40)
Al-Fārābī, Mūsīqā, 95, my translation.
41)
Al-Fārābī, Talkhīṣ jawāmiʿ kitāb al-nawāmīs li-Aflāṭūn, edited by .-A. Druart in
“Le sommaire du livre des lois de Platon,” Bulletin d’études orientales 50 (1998), 109-
155, 124. e translation is taken from Janssens, “Experience,” 50.
252 D. Janos / Early Science and Medicine 15 (2010) 237-265

is to form a universal judgment on the basis of the particular sensations

and extract a universal meaning out of these particulars. In contrast to
experience, induction does not involve this special intervention of the
intellect and simply consists in the gathering of data. Moreover, it does
not produce universal, certain knowledge in itself, although it represents
a starting point for experience to take place.42
The foregoing shows that al-Fārābī regarded induction and especially
experience as important features of the philosophical method in general
and of the astronomical method in particular. Al-Fārābī’s statements in
the K. al-mūsīqā reveal the thoroughly inductive quality of his approach

42)
It is noteworthy that the emphasis on experience and observation found in the
K. al-mūsīqā also appears in other works by al-Fārābī. In one of the two treatises on
astrology already mentioned, al-Fārābī explains that experience is necessary to under-
stand the effects of the celestial bodies on sublunary existents, especially the manner
in which the celestial bodies transmit heat to plants and other organisms (see the
Maqālah li-Abī Naṣr al-Fārābī fī al-jihah, sections 3-4, translated in Druart, “Le second
traité de Fārābī,” 48-50). In this context, experience enables valid astrological inquir-
ies to take place, i.e., those that study the things that occur with regular frequency.
e epistemological importance of experience is also highlighted in the K. al-millah,
where it is presented as a source of knowledge for the practicing physician (see Alfar-
abi, the Political Writings: Selected Aphorisms and Other Texts, translated and annotated
by Charles E. Butterworth (Ithaca, N.Y., and London, 2001), 105: “Clearly, he [the
physician] could not have acquired this determination [how to cure a particular per-
son] from the books of medicine he studied and was trained on, nor from his ability
to be cognizant of the universals and general things set down in medical books, but
through another faculty developing from his pursuit of medical practices with respect
to the body of one individual after another, from his lengthy observation of the states
of sick persons, from the experience acquired by being occupied with curing over a long
period of time, and from ministering to each individual.” is passage may be based
on Aristotle, who also uses the medical art as an example in Metaphysics I.1.981a1 ff.
Finally, the Risālah fī al-khalāʾ shows that al-Fārābī did not hesitate to carry out prac-
tical experiments to solve physical questions such as the existence of the void (see his
“Risāla fī’l-khalāʾ: Fārābī’s Article on Vacuum,” edited and translated into English by
N. Lugal and A. Sayılı (Ankara, Türk Tarih Kurumu Basimevi, 1951), 1-16, 21-36).
It is interesting that Ibn Sīnā in his K. al-burhān also provides an elaborate discussion
of induction and experience in connection with the acquisition of first principles. In
many respects, such as their mutual endorsement of experience as a valid method of
investigation (one that is in fact more valid than induction), al-Fārābī’s and Ibn Sīnā’s
accounts share many parallels. See the insightful articles by McGinnis, “Scientific
Methodologies” and “Avicenna’s Naturalized Epistemology,” which also discuss Ibn
Sīnā’s criticism of these two concepts in Aristotle.
 D. Janos / Early Science and Medicine 15 (2010) 237-265 253

to astronomy and the other sciences, such as music and medicine, and
especially the important role that experience plays in producing know-
ledge. This attitude can be explained by the influence of the Posterior
Analytics, a work quoted several times in the K. al-mūsīqā. The import-
ance that this text had in shaping al-Fārābī’s methodology appears
clearly when he writes that “the first principles of absolute demonstra-
tions in every science only reach the soul through the sensation [iḥsās]
of individual and particular things, as has been shown in the Posterior
Analytics [anālūṭīqā al-akhīrah].”43 Together with the K. al-burhān, then,
the K. al-mūsīqā testifies to the profound impact that the Posterior
Analytics had on the second master, and it is probably al-Fārābī’s thor-
ough acquaintance with this text that can best explain his interest in
the methodology of particular sciences such as music and astronomy.
It leads him to emphasize the value of experience and induction, and
it probably played a decisive role in the conceptual and epistemological
distinction he made between astronomy and astrology.
Al-Fārābī’s treatment of experience and to a lesser degree of observa-
tion and induction in connection with astronomy has several precedents
in Greek and Arabic philosophy. Aristotle alludes to the importance
of cosmological observations in De caelo II.13 and Book Lambda
8.1073b1-20, two passages which point to his belief that astronomy
undergoes periodic progress thanks to the gradual accumulation of
new data. An even more striking precedent occurs in Prior Analytics
I.30.46a19-22, where Aristotle states that “astronomical experience sup-
plies the principles of astronomical science.”44 In Ptolemy’s Almagest,
observation is described as one of the methodological pillars of astron-
omy, and one on which mathematical and astronomical theories rely.45
With regard to the Arabic world, P. Adamson and H. Wiesner have

43)
Al-Fārābī, Mūsīqā, 92, my translation.
44)
Translated by A.J. Jenkinson, in e Basic Works of Aristotle, edited by Richard
McKeon (New York, 2001).
45)
Ptolemy’s Almagest, translated and annotated by G.J. Toomer (London, 1984), I.1
H8, alludes to the work of previous scientists and the importance of empiricism; I.2
H9 refers more directly to the role of observation: “We shall try to provide proofs
in all of these topics by using as starting points and foundations, as it were, for our
search the obvious phenomena, and those observations made by the ancients and in
our own times which are available”; see also IV.1 H266, and B. Goldstein, “Saving the
254 D. Janos / Early Science and Medicine 15 (2010) 237-265

emphasized the important role that sense perception and empirical

observation play in al-Kindī’s approach to physics and cosmology.46
Furthermore, al-Fārābī’s position should be contextualized in terms
of the various astronomical observations carried out in Baghdad and
other cities during the ninth and tenth centuries by astronomers such
as al-Battānī, al-Farghānī, and Thābit ibn Qurra.47 This concern for
precise observations later grew in importance and culminated in the
works of the hayʾa practitioners. In this respect, recent scholarship has
shown that Arabic astronomers systematically tested and considerably
expanded the corpus of observational data they inherited from the
Greeks.48
In light of these facts, it is not surprising that al-Fārābī makes expe-
rience and observation one of the methodological foundations of as-
tronomy. Al-Fārābī considers astronomy to be, at least partially, an
empirical science that relies on the accumulation of data. Particularly
noteworthy is his mention of “instruments” (ālāt), which betrays a keen
interest in the quantitative dimension and practical side of astronomy.
Furthermore, it is noteworthy that al-Fārābī’s ideas on the method of

Phenomena: e Background to Ptolemy’s Planetary eory,” Journal for the History

of Astronomy 28/1 (1997), 1-2.
46)
P. Adamson, “Al-Kindī and the Reception of Greek Philosophy,” in e Cambridge
Companion to Arabic Philosophy, edited by P. Adamson and R.C. Taylor (Cambridge,
2005), 32-52, 42-43; H.S. Wiesner, “e Cosmology of al-Kindī,” PhD dissertation,
Harvard University, 1993, 35-38.
47)
See A. Sayılı, e Observatory in Islam and its Place in the General History of the
Observatory (Ankara, 1960); and D. King, “Astronomy in the Islamic World,” in
Encyclopaedia of the History of Science, Technology, and Medicine in Non-Western Cul-
tures, edited by H. Selin (Dordrecht; Boston, 1997), 125-134, who provide an over-
view of the various astronomical observations carried out during ʿAbbāsid times.
48)
A.I. Sabra, “e Astronomical Origin of Ibn al-Haytham’s Concept of Experi-
ment,” Actes du XIIe Congrès International d’Histoire des Sciences, T.III A, Paris 1968
(Paris, 1971), reprinted in A.I. Sabra, Optics, Astronomy, and Logic in Arabic Science
and Philosophy (Aldershot, 1994); Bernard Goldstein, “eory and Observation in
Medieval Astronomy,” Isis 63 (1972), 39-47; R. Morelon, “ābit b. Qurra and Arab
Astronomy in the 9th Century,” Arabic Sciences and Philosophy 4/1 (1994), 111-141;
A.I. Sabra, “Configuring the Universe,” 290 ff.; G. Saliba, Islamic Science, especially
Chapter 3; F.J. Ragep, “Islamic Reactions to Ptolemy’s Imprecisions,” in Ptolemy in
Perspective: Use and Criticism of His Work from Antiquity to the Nineteenth Century,
edited by A. Jones (Springer-Verlag, 2009), 121-134.
 D. Janos / Early Science and Medicine 15 (2010) 237-265 255

astronomy are inscribed in his general conception of the origin and

development of philosophy as expressed in the Fī ẓuhūr al-falsafah and
K. al-ḥurūf.49 According to these accounts, it is the accumulation and
classification of knowledge over an extended period of time, as well as
the development of specialized terminologies, which enabled the vari-
ous branches of philosophy to flourish.

Astronomy, Physics, and Metaphysics

Apart from experience, al-Fārābī mentions in the K. al-mūsīqā another
source from which the principles of astronomy are derived, namely,
natural philosophy. In an important passage, he explains that in order
to account for certain celestial phenomena, such as the planetary mo-
tions and their causes, astronomers must borrow physical principles
from natural philosophy. He writes:

Moreover, the case [in music] when we are unable to perceive the individual
entities is like the case of many of the sciences whose first principles [mabādiʾuhā
al-uwal] are proven in other sciences, and the practitioner of this science takes an
accepted principle which has been established in these [other] sciences. When he
is asked to prove it, he refers to the specialists of these sciences. is is what the
astronomer [munajjim] does when he wants to explain the causes [asbāb] of the
various motions of the celestial bodies that appear through observation [arṣād].
He can only explain these causes, such as the eccentrics and epicycles, when it is
posited that these planetary motions are in themselves regular [mustawiyyah]. He
cannot prove this at all in astronomy, but only by borrowing accepted [premises]
[musallamatan] from the natural scientists...50

This passage explicitly and vividly describes the dependence of astron-

omy on physics. It does not on the other hand describe in any detail
the nature of the premises or principles that are borrowed from natural
philosophy. But al-Fārābī provides a hint when he mentions that the
celestial motions are regular (mustawiyyah). Although it is not spelled
out, the assumption is that the heavens are made of a simple element
that possesses a unique motion, namely, circular motion, which, in its

49)
Al-Fārābī, Fī ẓuhūr al-falsafah, in Ibn Abī Uṣaybiʿah, ʿUyūn, 604-605; al-Fārābī,
Kitāb al-ḥurūf, 2nd edition, edited by M. Mahdi (Beirut, 1990), Part II.
50)
Al-Fārābī, Mūsīqā, 102, my translation.
256 D. Janos / Early Science and Medicine 15 (2010) 237-265

constancy and harmony befits the divine or exalted nature of the heav-
ens. The theory of a unique celestial substance possessing a regular
circular motion was elaborated by Aristotle in the De caelo I.2-4 and
subsequently became one of the fundamental physical assumptions
underlying the works of many ancient and medieval astronomers. In
spite of the fact that Aristotle’s discussion of aether in the De caelo is
one that falls plainly within the purview of physics (since it connects
the various existing motions (rectilinear and circular) with different
types of elements and concludes that aether is a unique celestial sub-
stance), it was often used as a premise or principle in astronomy. An
illustration of this is Ptolemy’s mention of aether in Book One of the
Almagest in order to reinforce his claim that the heavenly bodies are
spherical.
It is likely that al-Fārābī had the De caelo in mind when he wrote this
passage of the K. al-mūsīqā and that he accepted the Aristotelian cor-
relation made between the simplicity of aether and the regularity of
circular motion, at least on behalf of the astronomers. In light of the
foregoing, al-Fārābī’s argument may be reconstructed as follows: astron-
omy is unable to account for the heavenly bodies’ regular and uniform
circular motion and thus can devise models of planetary motions only
if it borrows this principle from physics. Physics can explain the cause
of the regular, circular, heavenly motion through a discussion of simple
bodies and the principles of motion and rest.
What this means is that astronomy is dependent on another science
for some of its principles and is hence not a completely self-contained
discipline. Experience and observation, as well as the mathematical
theories built on their data, are insufficient for one to acquire a com-
prehensive knowledge of the heavenly phenomena, if one does not use
in addition certain fundamental physical principles. Stated otherwise,
although some problems, such as the sizes and distances of the planets,
can be solved by astronomers using a combination of observation and
mathematical calculations, knowledge of the causes and devices respon-
sible for celestial motion requires that one transfer physical principles
to astronomy, where they are used as first principles (mabādiʾ uwal,
which corresponds to the Greek ἀρχαί).51

51)
In upholding astronomy’s dependence on physics, al-Fārābī anticipates one of the
defining features of the later hayʾa tradition, which is illustrated for instance in al-Ṭūsī’s
 D. Janos / Early Science and Medicine 15 (2010) 237-265 257

The previous passage from the K. al-mūsīqā, and al-Fārābī’s claim

that astronomy is dependent on physics in particular, point to his
implicit belief that an astronomical account should be in harmony with
physical reality. This can be vindicated by the fact that al-Fārābī clearly
considered the celestial orbs and the various devices responsible for
carrying the planets, such as the eccentrics and epicycles, to be real,
corporeal entities. This view is reflected in his general account of the
structure of the heavens in the Ārāʾ where he describes the spheres as
“circular bodies” (sing. jism kurī) and “celestial bodies” (sing. jism
samāʾī).52 In the same spirit, al-Fārābī also refers to the minor spheres
(probably epicycles) contained by the main spheres as “corporeal circles”
(sing. dāʾirah mujassamah).53 Finally, it should be noted that in the
passage from the Iḥṣāʾ quoted above, al-Fārābī also refers to the “non-
planetary bodies” (ajsām…ghayr al-kawākib), by which he clearly means
the transparent spheres of the heavens.54

Tadhkirah. However, it should be noted that some of the hayʾa authors who flourished
after al-Ṭūsī, such as the fifteenth-century astronomer al-Qūshjī, attempted to “free”
astronomy entirely from physical and metaphysical borrowings, chiefly for religious
reasons; see F.J. Ragep, “Freeing Astronomy from Philosophy: An Aspect of Islamic
Influence on Science,” Osiris 16 (2001), 49-64, 66-71.
52)
Al-Fārābī, On the Perfect State, 118-119; 128-129.
53)
Al-Fārābī, On the Perfect State, 128-129.
54)
ese passages indicate that al-Fārābī, at least in his philosophical treatises, ascribed
real existence to the eccentrics and epicycles devised by Ptolemy. Moreover, there is
no reason to think that al-Fārābī would have considered the physical-cosmological
model outlined in his emanationist treatises to be incompatible with his mathematical
treatment of astronomy in the Sharḥ al-majisṭī, even though the disappearance of the
Sharḥ makes it impossible to confirm this hypothesis. I do not wish to enter here the
debate between instrumentalism and realism, which, on my view, belongs to the past
scholarship on the history of astronomy. As early as the 1970s, scholars undermined
Duhem’s dichotomy and argued that most ancient astronomers sought to achieve a
balance between mathematical and physical theories (see G.E.R. Lloyd, “Saving the
Appearances,” e Classical Quarterly 28 (1978), 202-223; and L. Wright, e Astron-
omy of Eudoxus: Geometry or Physics? Studies in History and Philosophy of Science 4
(1973), 165-172). Recent studies have provided a much more accurate assessment of
ancient astronomy and cosmology without even referring to Duhem’s framework. e
priority at this point is to try to understand how the various mathematical, physical,
and metaphysical theories interact in the works of ancient authors. As an illustration
of this new approach, see James Evans and J. Lennart Berggren, Geminos’ Introduc-
tion to the Phenomena: A Translation and Study of a Hellenistic Survey of Astronomy
258 D. Janos / Early Science and Medicine 15 (2010) 237-265

Al-Fārābī’s understanding of the relation between astronomy and

physics is indebted to a long tradition in ancient philosophy, a starting
point of which is Aristotle’s Physics II.2, which attempts to clarify the
link between astronomy, physics, and mathematics. According to the
traditional interpretation illustrated in Ross’ commentary on this pas-
sage, Aristotle concludes that astronomy is more physical than math-
ematical, but this view has recently been challenged.55 Regardless of
whether Aristotle ultimately defines astronomy as a mathematical or a
physical science, he acknowledges the close relation between astronomy
and physics as a result of their studying the same subject matter, i.e.,
the celestial bodies, albeit from different angles and using a different
method.
As for Ptolemy, it is well known that although he claimed to ground
his method entirely in mathematics, he at times relied strongly on phys-
ics to elaborate his proofs and explanations. This can be seen, for
instance, in his treatment of the heavens’ sphericity in Almagest I.3. It
is nevertheless in the work of the Greek astronomer Geminus that
al-Fārābī’s view finds a much closer parallel. Al-Fārābī’s general position
concerning the importance of natural philosophy in astronomy is mir-
rored in Geminus’ Introduction to the Phainomena and especially in his
shorter treatise entitled Concise Exposition of the Meteorology of Poseido-
nios.56 In these treatises, Geminus, or Poseidonios as reported by
Geminus in the case of the latter treatise, makes the following points:
astronomy and physics focus on different aspects of the celestial bodies,
the former on their motion and exterior features, the latter on their
substance; both sciences use a different method to prove the same thing,
but because the astronomer cannot know the true nature of superlun-
ary things, he must take his first principles from physics; finally, there
is an emphasis on the importance of knowing causes.57 The dependence

(Princeton; Oxford, 2006); Taub, Ptolemy’s Universe; Ragep, Naṣīr al-Dīn al-Ṭūsī; and
R. Morrison, Islam and Science.
55)
For an account and criticism of Ross’ view and a new interpretation of the rela-
tion between astronomy and physics in Aristotle’s Physics II.2, see Mueller, “Physics
and Astronomy.”
56)
Both texts have been translated into English and analyzed by Evans and Berggren in
Geminos; see also A.C. Bowen, “e Demarcation of Physical eory and Astronomy
by Geminus and Ptolemy,” Perspectives on Science 15/3 (2007), 327-58, esp. 331 ff.
57)
Evans and Berggren, Geminos, 53-58, 252-255; see also Bowen, “Simplicius.”
 D. Janos / Early Science and Medicine 15 (2010) 237-265 259

of astronomy on physics is clearly conveyed when Geminus writes “that

he [the astronomer] must take from the physicist the first principles,
that the motions of the stars are simple, uniform, and orderly...”58
As we have seen, all of these points may be found in al-Fārābī’s writ-
ings. Al-Fārābī considers that astronomy and physics study different
aspects of the same subject matter. He also presents physics as one of
the sources from which astronomical principles are derived. Finally, he
cautions that astronomers can acquire only partial knowledge of the
causes of celestial phenomena if they do not rely on natural science.
These glaring parallels suggest that al-Fārābī is indebted to a particular
trend of Greek astronomical theory, which is perhaps most clearly
embodied in Geminus. Did al-Fārābī read part of the Arabic translations
of Geminus’ works? This hypothesis seems reinforced by the fact that,
as Evans and Berggren write, Geminus’ “remarks constitute the clearest
statement of this relationship [i.e., between astronomy and physics] we
find in any of the Greek astronomical writers.”59
Al-Fārābī follows Geminus on another notable point. Geminus, as
reported by Simplicius, had argued that physics and astronomy can
both reach the same conclusion on the sphericity of the celestial bodies
by mustering different proofs and by identifying different causes.60
Al-Fārābī provides a similar discussion and example in his K. al-burhān.
According to al-Fārābī, physics studies bodies that are enmattered and
insofar as they are in motion, while mathematics studies bodies
abstracted from their matter. As a consequence, different sciences can
study the same object by elaborating different proofs and by identifying
different causes. For al-Fārābī, there is therefore no contradiction and
opposition between the methods of physics and astronomy. Rather,
these sciences can collaborate together toward a common end. He
writes: “For this reason, it is possible for mathematics and physics to
cooperate in [the examination of ] a single thing, and for the former to
provide a cause and the latter to provide another cause. In this fashion

58)
Evans and Berggren, Geminos, 254-255. is statement strikingly resembles
al-Fārābī’s previously quoted assertion in the Mūsīqā on the relation between astron-
omy and physics.
59)
Evans and Berggren, Geminos, 252.
60)
See Simplicius, On Aristotle Physics 2, 290,25 ff.
260 D. Janos / Early Science and Medicine 15 (2010) 237-265

the sphericity of the world and the sun and the moon is examined by
both mathematics and physics.”61
In view of all these parallels, coupled with the fact that at least parts
of Geminus’ Introduction are known to have been translated into Arabic
and Hebrew,62 we may hypothesize that al-Fārābī may have known
Geminus’ theories, perhaps through Simplicius’ writings, and that they
may have stood as a model for al-Fārābī’s conception of the astronomical
method. Regardless of the impact that the Almagest and other Ptolem-
aic works had on al-Fārābī, in this particular instance the second mas-
ter is closer to Geminus than to Ptolemy.
In addition to physics, it appears that metaphysics also plays a crucial
role in al-Fārābī’s conception of the astronomical method, although the
second master is much more laconic on this subject. He does not say
anything to this effect in the K. al-mūsīqā, but this view is reflected in
his other works on methodology, such as the K. al-burhān. There it is
argued that the particular sciences are all dependent upon first philoso-
phy (i.e., metaphysics) for their principles, since metaphysics is the
universal science that studies being qua being. Al-Fārābī writes:

the particular sciences (e.g., physics, mathematics) are all below the First Philos-
ophy, participating in it in so far as all their subjects are below the Absolute Exis-
tent. is science will employ universal premises which all the particular sciences
employ in the way we have described, while the particular sciences employ prem-
ises which are demonstrated in that science [First Philosophy].63

This idea is echoed in the Aims of Aristotle’s Metaphysics, where al-Fārābī

explains that it is the task of first philosophy to provide the principles
and subjects of the particular sciences.64
The dependence of the particular sciences on the first universal
science is further elaborated in another passage of the K. al-burhān,
where al-Fārābī follows Aristotle’s distinction in Posterior Analytics I.13

61)
Al-Fārābī, K. al-burhān, vol. 4, 68, my translation.
62)
R.B. Todd, ‘Geminos,’ in Dictionnaire des philosophes antiques (Paris, 1989), vol. 3,
473. Todd notes that the Hebrew translations were based on Arabic versions.
63)
Al-Fārābī, K. al-burhān, 65, translated in G. Endress, “Mathematics and Philosophy
in Medieval Islam,” in e Enterprise of Science in Islam: New Perspectives, edited by J.P.
Hogendijk and A.I. Sabra (Cambridge, Mass.; London, 2003), 139.
64)
See McGinnis’ and Reisman’s translation in Classical Arabic Philosophy, 80.
 D. Janos / Early Science and Medicine 15 (2010) 237-265 261

between knowledge of the fact and knowledge of the reasoned fact, or

inna and lima (or innī and limmī) proofs in Arabic. On the basis of this
distinction, al-Fārābī argues that knowledge of the cause together with
knowledge of the existent is always preferable to knowledge of the exist-
ent alone.65 Now metaphysics is the science which par excellence is able
to provide such knowledge, since it examines the primary and most
fundamental causes of existence. This explains why the premises that
the particular sciences such as physics and astronomy borrow from
metaphysics can account for the causes (asbāb) of things. As al-Fārābī
writes, “the prior sciences provide the posterior sciences with knowledge
of the causes or of the causes and existence together…”66 This means
that in theory some of the principles of astronomy will be taken from
metaphysics, the first, universal science, although it should be said that
al-Fārābī does not provide a concrete example to illustrate this idea. At
any rate, astronomy and physics are dependent on metaphysics and
occupy a subordinate position in al-Fārābī’s classification of the sci-
ences.67
It should be noted, however, that the relation between astronomy,
physics, and metaphysics is not unilateral but rather reciprocal. If
astronomy depends on metaphysics and physics for some of its prin-
ciples, it appears that astronomy can also benefit the latter sciences by
providing them with knowledge concerning certain existents, i.e., the
celestial bodies, with which physics and metaphysics also deal. In the
K. al-burhān, al-Fārābī writes that “things that are proven in astronomy
are used as first premises in metaphysics and physics.”68 Al-Fārābī elab-
orates this point shortly afterwards by adding that

65)
Al-Fārābī, K. al-burhān, vol. 4, 26 ff.
66)
Ibid., vol. 4, 66, my translation.
67)
Al-Fārābī’s subordination of astronomy to other sciences such as physics and meta-
physics is likely to have been inspired by passages of Aristotle’s Posterior Analytics,
especially 74a38-75b20, 78b35-79a15 and 87a32-38. Aristotle in general maintains
the autonomy of the sciences, but in some passages he also hints at their interconnec-
tion and subordinates some sciences to others (e.g., optics to geometry). While he may
have been influenced by this work, al-Fārābī nevertheless goes beyond it and provides
a different classification of the sciences according to which astronomy is subordinated
to both physics and metaphysics, but collaborates with them in the broader cosmo-
logical enterprise.
68)
Al-Fārābī, K. al-burhān, 66, my translation. e Arabic reads: “fa-li-dhālika
262 D. Janos / Early Science and Medicine 15 (2010) 237-265

e prior sciences provide the posterior sciences with knowledge [maʿrifah] of the
causes or knowledge of the causes and existence (wujūd) together, whereas the pos-
terior sciences provide the prior sciences with knowledge of the existents alone.
For example, the art of astronomy provides physics and metaphysics with [the
knowledge of ] many aspects of the existents that are comprised by them...69

According to this passage, which makes a distinction between various

types of knowledge (maʿrifah), the posterior sciences can provide the
prior sciences with knowledge about the existents that compose their
subject matter. This idea should be interpreted in conjunction with
what was said previously about the various types of proofs in the theor-
etical sciences. Al-Fārābī follows the Aristotelian distinction between
proof of the fact and proof of the reasoned fact, and he argues that the
premises one science borrows from another can be used either as proofs
of existence or proofs of existence and causes. As al-Fārābī writes, “the
premises [muqaddamāt] used as principles [mabādiʾ] in a certain science
and demonstrated in another can be used either as causes [asbāb] or as
indications [or signs] [dalāʾil].”70 Hence, particular and posterior sci-
ences, in this case astronomy, can benefit metaphysics by providing it
with knowledge of certain existents and by formulating demonstrations
of the fact, or innī proofs, while metaphysics can provide knowledge
of the causes of beings and formulate demonstrations of the reasoned
fact, or limmī proofs.71
It is noteworthy that al-Fārābī’s belief that physics and metaphysics
are a source of astronomical principles was widespread in the later
Arabic astronomical tradition. For example, it is found several centur-
ies later in the works of two of the most important hayʾa practitioners,
Naṣīr al-Dīn al-Ṭūsī and Muʾayyad al-Dīn al-ʿUrḍī.72 At the beginning

tustaʿmalu ashyāʾ tabarhanat fī ʿilm al-nujūm muqaddamāt uwal fī al-falsafah al-ūlā

wa fī al-ʿilm al-ṭabīʿī.”
69)
Ibid.
70)
Ibid.
71)
Very little has been written about the innī/limmī proofs in Arabic philosophy,
and even less in the astronomical context. See Paul Lettinck, Aristotle’s Physics and its
Reception in the Arabic World: With an Edition of the Unpublished Parts of Ibn Bājja’s
Commentary on the Physics (Leiden, 1994), 105-113; Ragep, Naṣīr al-Dīn al-Ṭūsī,
vol. 2, 386-388.
72)
Sabra, “Configuring the Universe,” 307-308, 313; and Ragep, Naṣīr al-Dīn al-Ṭūsī’s
Memoir on Astronomy, vol. 1, 38-46.
 D. Janos / Early Science and Medicine 15 (2010) 237-265 263

of his Tadhkirah, al-Ṭūsī explains that “those of its [astronomy] prin-

ciples that need proof are demonstrated in three sciences: metaphysics,
geometry, and natural philosophy.”73 Although al-Ṭūsī does not elabor-
ate on the metaphysical principles he has in mind, he nevertheless
establishes the subordination of astronomy to metaphysics in a fashion
that recalls al-Fārābī’s classification of these sciences. Al-Ṭūsī, like
al-Fārābī, integrates astronomy in a hierarchy of the sciences that cul-
minates with metaphysics, which is seen as the first and primary science
upon which the other disciplines depend. Moreover, al-Ṭūsī devotes a
whole section of his treatise to the physical principles that are necessary
for the study of astronomy.74 In spite of major differences between these
two thinkers, one may conclude that al-Fārābī anticipated some of the
features that later became hallmarks of the hayʾa tradition in Islam. But
the exact relation between the early falāsifah and the hayʾa thinkers
deserves another and more detailed study.

Conclusion
In spite of the brevity of al-Fārābī’s remarks on the method of astron-
omy, it is nonetheless possible to reach certain conclusions on the basis
of the previous analysis. To begin with, it should encourage us to re-
examine some aspects of al-Fārābī’s affiliation to the Aristotelian, Pla-
tonic, and Neoplatonic traditions. In his description of the astronomical
method, al-Fārābī appears as a quite thorough Aristotelian, and more
specifically, as a careful reader of the Posterior Analytics, although he
also uses material derived from the works of ancient astronomers, espe-
cially Geminus and Ptolemy. He is one of the earliest thinkers in Islam
to emphasize the importance of observation, induction, and especially
experience in astronomy and music and to have applied some of Aris-
totle’s methodological directives to these particular sciences. His K.
al-mūsīqā indicates that he perceived the significance of induction and
experience in the search for first principles and reflected deeply on the
relation between scientific practice and theory.

73)
Translated in Ragep, Naṣīr al-Dīn al-Ṭūsī’s Memoir on Astronomy, vol. 1, 90.
74)
Ragep, Naṣīr al-Dīn al-Ṭūsī’s Memoir on Astronomy, Book I, Chapter 2, 98-102.
264 D. Janos / Early Science and Medicine 15 (2010) 237-265

Al-Fārābī here departs markedly from the ancient Platonists and

Neoplatonists according to whom all true knowledge consists of
glimpses into the intelligible and immaterial world and who in general
discarded the realm of sense perception. This attitude is attested with
regard to astronomy by the opening pages of Proclus’ Hypotyposis.75 In
contrast to these thinkers, al-Fārābī condones the study of the physical,
perceptible world and shows a genuine, albeit cursive, interest in the
empirical and practical aspects of astronomy.
In addition, his use of mathematics within astronomy seems more
informed by the scientific method of Ptolemy and the methodological
considerations of Aristotle than the metaphysics of the Platonists and
Neoplatonists, who had often elevated mathematical numbers or objects
to a divine or quasi-divine status. Al-Fārābī conceived of mathematics
as a useful tool to investigate the world around us, but there is no sug-
gestion, whether in his mathematical or philosophical treatises, that he
regarded this science as a special means of accessing metaphysical know-
ledge. In fact, the Aims of Aristotle’s Metaphysics and the K. al-burhān
indicate that al-Fārābī sided with Aristotle on the issue of the nature of
mathematical objects and that he defended the view that these objects
may be abstracted from matter in the mind only, but not in reality.76
On this point, al-Fārābī departs not only from the Greek Neoplatonists,
but also from al-Kindī, who was apparently influenced by Proclus’ com-
mentary on Euclid’s Elements, and according to whom mathematical
objects enjoy an intermediary position between physical objects and
intelligible beings.77 In brief, al-Fārābī’s approach to the astronomical
method is more indebted to the astronomical and Aristotelian traditions

75)
See Lloyd, “Saving the Appearances,” 207, for a translation and discussion of the
relevant excerpt.
76)
Al-Fārābī, K. al-burhān, vol. 4, 68-69. As G. Freudenthal has shown, al-Fārābī
endeavoured to develop an analytical approach to mathematics for didactic reasons (in
addition to the synthetic, deductive method inherited from Euclid), and he strongly
believed in the pedagogical virtue of proceeding gradually from concrete physical bod-
ies to more abstract geometrical entities in the teaching of this discipline; see Freuden-
thal, “La philosophie de la géométrie d’al-Fārābī,” and “Al-Fārābī on the Foundations
of Geometry.”
77)
See D. Gutas, “Geometry and the Rebirth of Philosophy in Arabic with al-Kindī,”
in Words, Texts, and Concepts Cruising the Mediterranean Sea, edited by R. Arnzen
and J. ielmann (Leuven; Dudley, Mass., 2004), 195-211, especially 204-205, 208.
 D. Janos / Early Science and Medicine 15 (2010) 237-265 265

than to the Neoplatonist one, an intriguing realization in view of the

otherwise many Neoplatonic features of his cosmology.78
The above conclusions agree with and strengthen the views developed
by J. McGinnis and F.J. Ragep in recent articles, which emphasized the
importance of induction, experience, and observation in the works of
Ibn Sīnā and Arabic astronomers respectively.79 Their articles stress the
discernment with which the Muslims received, criticized, and tested
the knowledge of the Greeks. Ragep’s paper in particular enables us to
contextualize al-Fārābī’s position vis-à-vis Greek Neoplatonic attitudes
toward astronomy and to better assess the differences between their
approaches.
In terms of the history of Arabic astronomy, the early falāsifah, and
foremost among them al-Fārābī, may be seen as a link in the long chain
that connects ancient Greek astronomers such as Geminus with the
medieval Arabic exponents of a new form of astronomy in the Islamic
world, which culminated with the achievements of the Marāgha school.
At the methodological and doctrinal levels, al-Fārābī’s ideas on astron-
omy often stand mid-way between those of his Greek predecessors and
those of his Arabic successors. As a result, it is not unreasonable to
surmise that he may have influenced later thinkers such as Ibn Sīnā and
that some of his views anticipated certain developments in ʿilm al-hay’a;
but this question requires more extensive research.

78)
By the same token, this invalidates M. Mahdi’s repeated claim in his Alfarabi and
the Foundation of Islamic Political Philosophy (Chicago; London, 2001), 121-122,
124, to the effect that al-Fārābī’s cosmology as exposed in his philosophical works
is a “political cosmology” devoid of scientific foundations. As this study has shown,
al-Fārābī was genuinely interested in the scientific foundations of astronomy, which he
discussed in his writings on scientific method, and he believed that astronomy, physics,
and metaphysics are interrelated sciences that each play a vital role in the cosmological
project. is explains why the cosmology of his metaphysical treatises is informed by
the available scientific knowledge of his time.
79)
McGinnis, “Scientific Methodologies” and “Avicenna’s Naturalized Epistemology,”
and Ragep, “Islamic Reactions to Ptolemy’s Imprecisions.” Ragep provides an inter-
esting discussion of the theological and social reasons that may be responsible for the
gap between the Greek Neoplatonists and the early Arabic thinkers in their approach
to the physical world.
Copyright of Early Science & Medicine is the property of Brill Academic Publishers and its content may not be
copied or emailed to multiple sites or posted to a listserv without the copyright holder's express written
permission. However, users may print, download, or email articles for individual use.