§23. Empiricism in Islam

In the eighth century, Muslim officials and scholars chiefly in Baghdad began to commission Arabic translations of the major works of Greek sci­ence and philosophy. They soon possessed often very good translations of Euclid, Ptolemy, Hippocrates, Galen, and nearly all of Aristotle and his an­cient commentators.

Alongside the translation movement, associated research flourished on the new problems these works posed. That included the cultivation of philos­ophy, especially devoted to Aristotle. The falsafah of Islam is a development of a translation project that begins in the ninth century when al-Kindi reads Metaphysics. In the tenth century, Aristotelian teachers from the schools of Alexandria were brought to Baghdad to teach, ensuring that the Arabic in­troduction to Greek culture was through the declining school traditions of late antiquity. A scholar observes that “what was notably different about the Aristotelian tradition when it finally enters Arab hands is that the system bears all the signs of having been taught, and with a thoroughness and ritual that immediately suggests ‘syllabus’ and ‘curriculum.’ ”¹⁸⁵

Instruction in this secular literature, notably philosophy, was private.

To consider empiricism under Islam, let us begin with their experience. From the eighth to the fourteen century, Arabic scholars commanded the Greek sources and developed the most advanced science of the age in prac­tically every field. For reasons that remain debated, their scientific tradi­tion went into decline after the thirteenth century. Their most impressive accomplishments of an “empirical” tenor are in observational sciences— optics, astronomy, and medicine.

(1) The eleventh-century Kitab al-Mana, or Book of Optics, by Ibn al- Haytham (Alhazen to Latins), is the most penetrating contribution to this sci­ence between Ptolemy and the seventeenth century. The work incorporates new explanations, including the camera obscura, which the author used to demonstrate the rectilinear propagation of light. He experimented with plane and curved mirrors, improved Ptolemy's instrument for measuring refraction, and studied the bending of light at plane and spherical surfaces with air-water, air-glass, and water-glass interfaces. His analysis of vision transformed Aristotle's intromission account (forms enter the eyes without their material) into a mathematical theory as satisfactory as that achieved for the extramission, geometrical theories (rays flow out of the eyes) by Euclid and Ptolemy.

His theory of visual perception initiates a psychological approach charac­teristic of European theories from George Berkeley to Hermann Helmholtz (§68). He did not know what Kepler knew about the lens-property of the eye's crystalline humor, but he has much of the rest of Kepler's account. Between the center of vision and the seen object the flattened spherical surface of the eye's crystalline humor displays a configuration of illuminated points of color corresponding to their arrangement in the field of vision. Perception and all its visual properties (size, shape, and distance) consist in a mental reading of this color mosaic, which is first sensed on the crystalline surface. Since this mosaic is two-dimensional, al-Haytham draws the conclusion that depth perception is not optical but psychological. Features of the retinal image function as signs for mental operations; size on the crystalline surface is a cue to size, brilliance a cue to distance.¹⁸⁸

(2) Astronomical observatories were established in the early ninth cen­tury in Baghdad and Damascus. They employed the best instruments of the time to study planetary motion, the obliquity of the ecliptic, the procession of equinoxes, and other phenomena. They accumulated substantially more data than Greek and Babylonian records had, and introduced new observations, such as planetary brightness and transits. They improved the measurement of constants, composed new star catalogs, and introduced trigonometry. It was these scholars who first distinguished between astronomy and astrology, that is, horoscopes.¹⁸⁹

As soon as they mastered Ptolemy, the Baghdad scholars began to find problems, both internally, and in reconciling Ptolemy and Aristotle, from whom Ptolemy made some troubling departures. He criticizes Aristotle’s explanation of the movement of the spheres, which has them transmitting motion to each other like rollers, something Ptolemy regards as an invalid extrapolation from terrestrial observation.

Ptolemy’s theory also has technical problems, which these Arab readers soon discovered. For instance, Ptolemy insists that he is describing physical realities, spheres and bodies moving with the motions he assigns to them. Yet these readers discovered that it is not possible to physically realize uni­form circular motion with Ptolemaic machinery. Many of the astronomical parameters of Ptolemy’s Almagest proved defective, and a program of obser­vation was initiated to correct them. Obviously these astronomers were not merely preserving Greek astronomy, but were overcoming its problems and creating a new science. Alert scholars in Europe were aware of this, and went to the East for the best scientific work, especially in the more observational, empirical sciences of astronomy and medicine. Copernicus benefited from innovations in mathematical astronomy introduced in the thirteenth cen­tury by Tusi and Urdi at the observatory at Maragha, in Mongol Iran. Theirs are the only theorems Copernicus uses that are not in Euclid or Ptolemy, prompting the suggestion that we see Copernicus as the last Maragha astron­omer rather than forging a new astronomy on his own.¹⁹¹

(3) Although the Qur'an makes no reference to medical practice, some Hadith (preserved sayings of Mohammad) record the Prophet's medical advice, and one name of the Qur'an is al-Shifa, Restorer of Health. Islamic religious tradition recognizes medicine, its goal being in harmony with the Quranic vision of human well-being. The Islamic hakim, wise man, is invari­ably a physician. Islamic medicine was basically Galen, systematized and ex­panded; for instance, by the description of diseases unknown to the Greeks, including measles and smallpox.

Islamic hospitals are in operation from the eighth century, beginning in Iraq. By the tenth century there were many, and they are integrated with medical teaching. A tenth-century Baghdad hospital employed twenty-five physicians, including occultists, surgeons, and bonesetters, working in sep­arate halls for different types of patients, and with an outpatient dispensary. These hospitals were secular institutions, the physicians an indiscriminate mix of Muslim, Christian, Jew, even pagan. They were also places of intense and prestigious theory and practice, where Islam's best physicians produced their most advanced research. Ibn al-Nafis, a thirteenth-century physician at Cairo, gave the first account of the minor circulation of blood between heart and lungs. Another physician exhorts colleagues to make the best use of the opportunities afforded them by service in the hospital. “[The physician] should frequently enter and serve in hospitals; he can study rare diseases which he encounters [there]. One often witnesses in such places diseases which are unheard of and which one does not see [discussed] in the written [medical literature].... If he sees any such rare condition, he should record it in his notebook and thus preserve it so that he and others can benefit from it.”¹⁹³

The research and practice surveyed evince vibrant traditions of empirical observation, description, measurement, and recording, including mastery of the most advanced mathematics of the age. So much for their experience with experience as a means of knowledge. What of their empiricism and phi­losophy of knowledge?

“Knowledge” translates the Arabic word ‘ilm, which began as a Bedouin term for “way sign.” Orientalist Franz Rosenthal writes forcefully of the sin­gular value of knowledge for medieval Islam.

All its intellectual efforts and indeed all its educational efforts constantly stress the importance of the term “knowledge” and the terms “teaching” and “learning,” which, among the languages of Muslim peoples, the lin­guistic extension of Arabic at least makes almost one and the same thing as knowledge.

It is hard to conceive of any place, no matter how remote, where there might have existed anyone, no matter how humble and uneducated, who was not filled with awe when he heard the word “knowledge.” And it was gener­ally recognized that any genuine betterment of an individual’s standing in society depended on his share of knowledge, in preference to uncertain avenues of birth, power, and wealth.¹⁹⁴

Rosenthal summarizes the idea of knowledge in the Qur’an, where, he says, the root and cognates are prominent. “The student of the Qur’an finds him­self confronted with the thought presented forcefully and inescapably that all human knowledge that has any real value and truly deserves to be called ‘knowledge’ is religious knowledge... specifically identical with the content of the divine message transmitted by the Prophet.” The Scripture commends knowledge as “the goal of all worthwhile aspirations of mankind, the true synonym of religion.” The Qur’an also recognizes a mundane, human know­ledge, acquired by eye, ear, and heart, and acknowledges the specialties, disciplines, and sciences, which are understood to be interdependent and ul­timately serving the same purpose of salvation.¹⁹⁵

Topical Hadith collections usually include a chapter on “knowledge,” which sometimes serves to introduce the whole collection. Knowledge is the result of the process of learning—“Knowledge is attained only through study.” The view is attributed to the Prophet that the disappearance of know­ledge and scholars will herald the end of the world. These prestigious sources underscore the urgency of knowledge, the necessity of instruction, and the importance of avoiding error, which accounts for the singular value in this tradition of certainty (yaqin). This is distinctive in their theory of knowledge, and the value of certainty in European philosophy comes from this source.¹⁹⁶

Certainty is described as a light in the heart, the source of faith and piety. In the Qur’an it is a synonym for truth and essential insight, and above all syn­onymous with knowledge, though generally a superior knowledge, purified of error and fallibility. Knowledge is the firm belief that a thing is; certainty is one’s satisfaction with and assuredness of what one knows, eliminating all doubt. “A little bit of certainty removes all doubt from the heart, and a little bit of doubt removes all the certainty from the heart.” Doubt connotes a trou­bled relation to religion. The principal weapon of heretics and unbelievers in their opposition to Islam is the planting of doubts. Rosenthal claims it was difficult for these thinkers to distinguish between doubt and outright unbe­lief. Doubt, in whatever way indicated, “became the true pariah and outcast of Muslim civilization. It stands for all that is shunned like the plague. No worse fate can befall a man than being tossed into the sea of doubts and left there to flounder and possibly to drown, deprived of the certainty that reli­gion is intended and equipped to give him”¹⁹⁷

Aristotle did not mention certainty as a quality of episteme-science, and it was not much considered in Latin philosophy until Albert the Great noticed it in his Arabic sources, where it had silently replaced the traditional identifi­cation of the end of demonstration as episteme or scientia. Al-Farabi, preem­inent among the Baghdad Aristotelians of the ninth century, wrote an entire treatise on the subject, Conditions of Certitude. “Certainty means that we are convinced, with respect to what we assent to, that it cannot possibly be dif­ferent from our conviction. Moreover, we are convinced that this conviction about it also cannot be otherwise, to the point that when one reaches a given conviction concerning his initial conviction, he maintains that it, too, cannot be otherwise, and so on indefinitely”¹⁹⁸

This sounds Stoic, like the idea Zeno demonstrated with the closed fist (knowledge) and clenching the fist with the other hand (certainty), though there is apparently no evidence that the philosophers of medieval Islam were acquainted with Stoic ideas. To Albert it seemed like an overlooked scientific value, and by the time of Scotus and Ockham the defense of certainty is ex­pected in natural philosophy, especially in the proemial parts that began to address questions of method and evidence, forerunners of epistemology.

Scholars draw attention to two ideas in Arabic thought on the value of knowledge. One is to esteem it, as we have seen. The other encourages the hu­mility to remember the imperfection of current knowledge and seek better knowledge wherever it may be. According to a proverb, “A man remains knowing as long as he searches for knowledge and continues to study. When he thinks he knows, he has become ignorant.” This questing spirit was not a theme of Greek philosophy, apart from the medical tradition. Drawing this spirit deeper into elite culture gives a different texture to Islamic rationalism, which is less chauvinistic than Greek philosophy, more interested in learning from others, which is a kind of empiricism in its own right.¹⁹⁹

The same spirit that piqued them to seek out Greek science persuaded the scholars of Islam that, however excellent, it cannot be the last word. It is a body of fascinating claims that need to be carefully checked. An entire “doubts” (shukuk) literature sprung up. The astronomers discovered doubts about Ptolemy, the doctors discovered doubts about Galen. Remarkably, they found little to doubt in Aristotle, except for the radical theologians who doubted everything the incoherent philosophers tried to advance. A deci­sive fact for the Islamic experience with Aristotelianism is that these scholars learned their Aristotle from Alexandrians drenched in Neoplatonism. It seemed normal. Ibn Sina could read a treatise of Plotinus convinced that it was by Aristotle. This singular blend of Alexandrian Platonism and Aristotle was Islam’s only philosophy—it was falsafah, and not just one school in a philosophical marketplace, as the ancient Greek cities were.²⁰⁰

Religious opposition to falsafah developed almost instantly and never stopped finding support. But the philosophers did not stray from their syn­cretic Neoplatonism. “The world was ruled by the First Being, the First Cause, which was eternal and perfect without matter and without form, the absolute One without any other specification or qualification.... Centuries of un­questioned philosophical tradition had given to a highly controversial and hypothetical postulate like this the appearance of self-evidence.” Another scholar describes Islamic intellectual life as reason in the service of a non- rational revealed code of conduct, namely, the sharia law, which is to Islam as theology is to Christianity. Sharia is given; the reasons for the commands are inaccessible to the human mind.²⁰¹

Not fertile ground, one might think, for philosophical empiricism despite the profound empiricism of Islam’s scientific practice. Al-Birundi, a Persian thinker of the eleventh century, seems to have had the most exalted view of the senses, which are the basis of all knowledge and are rightly trusted be­cause perception (tasawwur) is possible only of something that has complete or partial correspondence in existence (William Ockham calls this “intui­tive cognition” [§32]). Al-Birundi is notable for his study and translation of Hindu thought, and says he learned his idea of sense perception from the Indians. This perception is made for knowing, and the pleasure of knowledge stimulates people to seek it.²⁰²

The physicians of Islam were urged to value observation and experience over words. “Observation is always much stronger than words,” a thirteenth­century Baghdad physician says. “Observation is even more valid than Galen. Despite his rank among the scientists in investigation, and his meticulousness in all that he said or practiced, observation is still more true than him.” Yet these doctors would probably be offended by association with “empiricism” animadversions against which are a theme in their commentaries on the first Hippocratic Aphorism. Here it is in a modern English translation. “Life is short, art long, opportunity elusive, experience treacherous (peira sphalere), judgment difficult” Islamic commentary is notably interested in the words “experience treacherous” or (translated from Arabic) “empiricism is risky” What is empiricism, and what is the risk?²⁰³

There were no self-described Empiricists among these physicians or in the works they collected from Alexandrian medicine, and it is evident that they were not well informed on the ideas of a school they nevertheless deplore. The oldest complete commentary on the Hippocratic Aphorism is from the mid-eleventh century. The explanation begins by stipulating that “empir­icism (tajribah) is of two kinds. It may be the consideration of the known general rules with reference to observable details. Everybody must do this, and it entails no risk. Or it may be testing something without any theoretical speculation leading to it, or anything resulting from a principle or rule. It is obvious that this kind of empiricism is untrustworthy and, therefore, risky” The physician should know, understood with Aristotle as knowing the cause. Medicine is too lofty, and the human body too noble, to be a plaything for experimenters.²⁰⁴

The argument is elaborated in a commentary ofthe latter twelfth century by a Cairo physician and contemporary of Maimonides. He says that the apho­rism is a warning against empiricism. “Through this aphorism [Hippocrates] intended to inform us about the faults and defects that affect the craft [as practiced by] the Empiricists, and about the protection and freedom from such faults that come to it [as practiced by] the Dogmatists” What is wrong with empiricism? “If, with regard to the entire medical knowledge, we relied upon mastering it through empiricism, we would be forced to study empiri­cally every detail of every sickness or healthy state and of every medicine or food, be it simple or compound or used in combination, under every possible condition” In effect, condemning medicine to futility.²⁰⁵

This author's idea of empiricism seems to come from Aristotle and the idea of a complete induction, which was not Hippocratic thinking nor that of the Alexandrian Empiricists. For them a medical education was a matter of accumulating sapient experience and trusting qualified analogy. Their un­derstanding of medical techne was never supposed to mesh with the princi­ples of Aristotelian episteme, to which it is here held accountable. Certainly, if you want to be a empiricist, you will have to check your Aristotelianism, and that is why neither the doctors nor the philosophers of Islam want to be empiricists. Nothing interests them in this third way between Plato and Aristotle.

Another argument from this commentary is that “we can rely on em­piricism only when the same action is repeated in many things that agree with each other as to one and the same condition in every respect or in most respects. "1 his is difficult to find.” Perhaps that is why the art is long, but not why experience is treacherous. Coming to that he says, empiricism is risky because “a person proceeds with an action without insight and surety................................................................................................

'1 he human body is something noble, and proceeding to treat it empirically without being sure of good and healthy results is risky.” A physician should not be experimenting and investigating. He should already know all that need be known, and bring certainty to the bedside. “'1 he physician must pos­sess insight into all he must do. He must understand every change.”²⁰⁶

Like the metaphysician and theologian, the doctor should be certain, and not be a reckless experimenter. “The physician treats the patient in order to cure, and not in order to experiment.... '1 he experimenter is a learner. If we are always experimenting, when, I would like to know, can we cure the sick?” Obviously this presupposes that medical experiments are risks not worth taking. Experimentation is fine for Ptolemy or Ibn al-Haytham, with their mirrors and tubes, but not for physicians. '1 he risk is unprofessional.²⁰⁷

Islamic philosophy and medicine reached their summit together in the work of the Persian Ibn Sina, or Avicenna to the Latins, writing in the elev­enth century. On the question of experience and empiricism in science his position is Aristotelian, though with interesting criticism and modification. Like Aristotle, the empiricism is theorematic rather than problematic, an empiricism of sensory evidence for syllogisms rather than experimental dis­covery. Yet in a departure from Aristotle, he criticizes and rejects the place of epagoge (istiqra, induction) in science, and proposes in its place a qualified, methodic experience, his tajriba, which is suitable to discover principles of science.

'1 he experience he has in mind is not just desultory looking around. It has to be a rigorous, methodical experience, a thesis that resembles Epicurean ca­nonic (§19) and Galen's idea of qualified experience (§6). Methodical experi­ence is the experience or memory of regularly observed concurrence without falsifying contraries. Such regularity cannot be an accident. Reason requires a causal connection, which must be investigated. Avicenna's example recurs endlessly in later Latin literature on experimental reasoning: scammony, with its supposed power to purge bile. The drug is administered to the same purgative result in many observed cases. The experience leads the physician to disregard the possibility that it is an accident, judging instead that it is an intrinsic characteristic. “Since chance is not always or for the most part... it is known that [purging] is something which the scammony necessitates by [its] nature, since it does not turn out that there is anything arbitrary about it ”208

Considerations of falsification distinguish this experience from a mere induction or history. The scientific use of experience requires the observer to note variables and background conditions and take these into account in assessing regularity and the absence of falsification. With all things duly considered, a qualified conclusion can be stated. “It is necessary that the ex­perimental judgment we possess is that the scammony commonplace to us and perceived [before us], either from its essence or from the nature in it, purges bile, unless it is opposed by an obstacle.” The range and number of observations and the absence of falsification reasonably provoke the judg­ment that the quality belongs to the nature of the stuff. Methodical experi­ence cannot identify the cause but assures us that a cause is operating.²⁰⁹

I say more about Avicenna later, specifically, his idea of the inner senses (§26). Avicenna and his entire tradition consistently distinguish sense from intellect and tolerate no admix. Yet the most original development in this philosophy comes from the search for something to mediate these. What they found is the maani, or in Latin, intentio, intentionality.²¹⁰

Aristotelianism is empiricism only with many qualifications, and is not much like the authentically empirical philosophy of Democritus. The Islamic philosophers were as Neoplatonic as they were Aristotelian, and have even less appetite for experience. They acknowledge the usual senses, assign them the place Aristotle did in Posterior Analytics, and, with rare and fascinating exceptions like al-Birundi, that is the last we hear of them. They have more to say about intuition and intellection as higher faculties of knowledge. It seems that nothing worth knowing could be known principally from experi­ence. The philosophers are free to ignore experience because nothing made it problematic for them. Opposition to their rationalism comes not from any consideration of experience, but from Islamic tradition, their opponents being theological literalists.

A late contribution to this tradition identifies precisely this evaluation of experience as the fatal flaw in Islamic philosophy. These philosophers made the catastrophic mistake of setting a Greek evaluation of experience above the Qur’an, which has a different teaching and should have been better con­sidered by the philosophers in their estimation of experience. Writing in 1930, Muhammed Iqbal criticizes Islam’s philosophical tradition for being much too impressed by Greek thought, and not recognizing the Qur’an’s empirical outlook. “The spirit of the Qur’an is essentially anti-classical.” The Scripture “constantly calls upon the reader to observe the perpetual change of the winds, the alternation of day and night, the clouds, the starry heavens, and the planets,” and is said to “regard hearing and sight as the most valu­able divine gifts and declare them to be accountable to God for their activity in this world.” That is what “the earlier Muslim students of the Qur’an com­pletely missed under the spell of classical speculation. They read the Qur’an in the light of Greek thought.” Specifically, Greek rationalism.²¹¹

In this tradition I see an empiricism of scientific practice that never be­came a philosophical principle, and on this point Islam bears an analogy to traditional China. Their scientific practitioners were able experimenters, but the philosophers were uninterested in their practice. There are adventures in Islamic philosophy, but empiricism is not one of them.