Justified True Belief

(●) The first self-described philosopher of the Arabic tradition, who argued that the past cannot be endless and that a world with a beginning must have a Maker. + A hundred years after he died, a Baghdad bookseller named Ibn al-Nadim sat down to catalog everything for sale in his city and worked his way through the writings of one man for page after page. Books on metaphysics and logic, on arithmetic and the spheres, on music, astronomy, geometry, medicine, politics, the weather, the making of glass and perfume and swords. By the end he had listed something over two hundred and forty titles under a single name: Abu Yusuf Ya’qub ibn Ishaq al-Kindi. Of that shelf, only a few dozen works survive, and most of the philosophy that survives does so for one reason: a single battered codex, one handwritten book with its sheets sewn along the spine, came to rest in a library in Istanbul. He is usually called the first philosopher of the Arabs, and the title is not flattery. Before him, philosophy in the Islamic world meant Greek books read by Christians who spoke Syriac, a form of Aramaic that the eastern churches used for worship and for scholarship, and argued over in translation. Al-Kindi was the first man of Arab descent to take up that inheritance in Arabic, give the language a philosophical vocabulary it did not yet have, and write as a philosopher rather than as a translator or a commentator. He did all of it from inside the household of the caliph, the ruler of the whole Muslim empire and Muhammad’s successor at the head of it, tutoring one of the caliph’s sons and dedicating his hardest book to the boy’s father. For Christian readers, one particular argument is the reason to know him. Aristotle had taught that the world always existed, and al-Kindi set out to prove that it could not have, because an endless stretch of past time is a thing that could never have been finished. That argument did not begin with him and it did not end with him, and tracing where it came from and where it went afterward is most of the reason to read him at all. One name causes trouble. The Arabic Apology of al-Kindi, a Christian defense of the faith written at roughly the same period, is the work of a different man with a similar name, and the two have been mixed up since the Middle Ages.

(●) c. 801 AD: Born at Kufa, the son of an Abbasid governor, into a family that traced itself to the kings of Kinda. + Kinda was an old South Arabian royal house, and al-Kindi’s people claimed descent from its chieftains. That mattered in ninth-century Baghdad more than a modern reader might guess. The men doing philosophy in the caliphate were overwhelmingly Persians, Syrians and Christians of the eastern churches, and a philosopher of pure Arab stock was unusual enough that the phrase "philosopher of the Arabs" stuck to him for good. His father, Ishaq ibn al-Sabbah, held a governorship under the early caliphs of the Abbasid house, the dynasty that had taken the caliphate in 750 and built Baghdad a dozen years later to rule the empire from, and he died while his son was still a boy. After that the sources disagree about the details, in the way sources usually do at this distance. Most place the birth at Kufa, a garrison city on the edge of the Iraqi desert; some place it at Basra, the great port at the head of the Gulf; and the governorship is handed now to the one city and now to the other. Nearly all of them agree that the boy did his first schooling at Basra. Nobody at all recorded the year he was born, so the c. 801 is a convention, reckoned backwards from later reports of how old he was, and the death date rests on the same sort of reckoning.

(●) c. 820 AD: He came up to Baghdad while the caliphs were paying to have Greek philosophy turned into Arabic. + Imagine a government deciding that the most useful thing it could buy was books in a language almost nobody at court could read, and then buying them for the better part of two centuries. That is roughly what happened. Under the caliph al-Ma’mun, who ruled until 833, Baghdad ran what came to be called the Bayt al-Hikma, the House of Wisdom: a library and translation office attached to the caliph’s household, where Greek medicine, mathematics, astronomy and philosophy were rendered into Arabic, often by way of Syriac. Al-Kindi walked into the middle of all this as a young man, and he stayed in the city for the rest of his life. He came to oversee one of the two great translation circles of the century. The other belonged to Hunayn ibn Ishaq, a physician of the Church of the East, which is the ancient church of Mesopotamia and Persia that lived and worshipped outside the Roman empire altogether, and he was the finest translator of the age. Al-Kindi himself seems to have had little Greek or none. What he did instead was commission the work, go over the finished Arabic and correct it, and then explain to his readers what it meant. The single most consequential book his circle produced was the so-called Theology of Aristotle, made by Ibn Na’ima al-Himsi, a Syrian Christian from Homs, and corrected by al-Kindi for the young prince Ahmad. Aristotle did not write it. It is a paraphrase of the Enneads of Plotinus, the third-century pagan mystic whose thought had already shaped Augustine, and the mistaken title meant that for centuries afterward Muslim and then Christian readers took Plotinus for Aristotle’s own last word on God.

(●) c. 833 AD: The caliph al-Mu’tasim handed him a son to educate, and got On First Philosophy in return. + Al-Mu’tasim came to the throne in 833 and reigned until 842, and under him al-Kindi reached as high as a scholar could reach. He became tutor to the caliph’s son Ahmad, and a good many of his surviving treatises are letters written to that pupil, patient pieces that begin by defining their terms because the reader is a boy. The book he addressed to the caliph himself is the one that survives best and matters most. Its title, On First Philosophy, is Aristotle’s own name for the study of what it is for anything to exist at all, and of the first causes behind everything else; later readers called the same subject metaphysics. Al-Kindi opens it by defining philosophy as knowledge of things as they truly are, and by arguing that the highest of the philosophical disciplines is knowledge of the First Truth, which is the cause of every other truth. Then he spends the body of the work proving that the universe is finite and had a beginning, and closes by describing the One who began it. Four sections came down to us. The rest of the book is lost.

(●) c. 840s AD: He put numbers on things other men left to judgment: the strength of a medicine, the intervals of a song, the letters of a cipher. + The philosophy is only a corner of the catalog. Ibn al-Nadim sorted al-Kindi’s output into seventeen departments, and the mathematical and scientific titles outnumber the philosophical ones several times over. Three of them are worth meeting, because the same habit of mind is at work in all three: take something everyone treats as a matter of taste or judgment, and ask what it would look like counted. - Medicine. Physicians already graded a drug by degrees, first through fourth, the way a cook speaks of a mild or a fierce chili. Al-Kindi asked what happens when you mix them, and proposed that the strength of a compound rises in geometric steps as the ingredients are combined. Gerard of Cremona later translated the result into Latin as De gradibus, and medieval European doctors argued about it for centuries. - Music. He wrote the earliest Arabic treatises on music theory that survive, worked out the mathematics of intervals and rhythm, and is credited with adding a fifth string to the oud, the short-necked lute that is the ancestor of the European lute and, at some distance, of the guitar. - Codes. In a short epistle on deciphering encrypted messages, he described what is now called frequency analysis: count how often each letter appears in the enciphered text, compare it to how often each letter appears in ordinary writing of that language, and the commonest symbol is almost certainly standing in for the commonest letter. It is the earliest known description of the technique, and it sat unread in an Ottoman archive in Istanbul until 1987.

(●) After 847 AD: Two rival scholars turned a caliph against him, and he was flogged in public while his library was carried away. + The story comes from Ibn Abi Usaybi’a, a physician who wrote his lives of the doctors four centuries later, so it is a late report rather than an eyewitness account. It runs like this. Muhammad and Ahmad, two of the three sons of Musa ibn Shakir who were known together as the Banu Musa, were brilliant engineers and mathematicians with the ear of the caliph al-Mutawakkil, and by this account they had a habit of destroying anyone whose learning threatened their standing at court. They worked on the caliph until he had al-Kindi beaten, fifty lashes in front of a crowd, when the old man was about sixty. His books were confiscated and shelved in a repository that Baghdad nicknamed the Kindiyya, after the man they had been taken from. What brought them back is the sort of detail that sounds invented and probably is not. The caliph set the Banu Musa to cut a canal, and the brothers handed the survey to al-Farghani the astronomer, who set the mouth of the channel deeper than the rest of its bed, so that water enough to fill the opening would never rise high enough to run down the length of it. A blunder on a caliph’s canal could have cost the brothers their heads. The man who vouched for their figures and saved them was Sanad ibn Ali, an astronomer they had earlier turned the caliph against and had packed off to Baghdad, and the report has him refusing to say a word for them until the library went back to its owner. The episode is easy to mistake for something larger than it was. Al-Mutawakkil did reverse his predecessors’ religious policy and came down hard on the rationalist theologians, the party that made reason the arbiter in disputed questions of doctrine, so it would be easy to read al-Kindi’s fall as a philosopher persecuted for his opinions. The report itself says otherwise. This was a quarrel between scholars over standing at court, and the weapon nearest to hand happened to be the caliph.

(●) c. 873 AD: He died in Baghdad, and within two centuries most of what he wrote was gone. + He died in Baghdad in the early 870s, an old man out of fashion. Nothing about the funeral is recorded, and the death date is inferred rather than attested. The eclipse came fast. Al-Farabi, born about the year al-Kindi died, and Avicenna after him built a far more thoroughly Aristotelian philosophy in Arabic, and beside their systems al-Kindi’s treatises looked like a first draft. Arabic readers stopped copying him, which is why so little survived: most of the philosophical works that remain do so in a single manuscript in Istanbul, and one or two are known only in Latin. That Latin is the second half of the story. From the twelfth century, translators working in Toledo turned his optics, his pharmacology and his little treatise on the intellect into Latin, and the Christian West read him under the name Alkindus. Roger Bacon leaned on his optics. The schoolmen, which is what the teaching theologians of the new universities were called, argued for generations over his account of the intellect. The man who had written to teach a caliph’s son ended up, without ever knowing it, on the reading lists of Paris and Oxford.

What He Taught

(T1) Truth belongs to whoever found it first, and a foreign name on a book is no reason to refuse it. + Everyone has met the reflex. A good idea arrives with the wrong label on it, from the wrong country or the wrong sort of person, and the first instinct is to look for something wrong with it. Al-Kindi lived with a strong version of that reflex. He was recommending pagan Greeks to Muslim readers in a city where a serious body of opinion held that everything worth knowing had come down with the Qur’an and that the rest was foreign meddling, and he knew perfectly well what he was up against. His answer was not that the Greeks were harmless. It was that truth does not belong to anybody. If a man says something true, the truth was not his to begin with, and his nationality has no vote in the matter. Al-Kindi goes further and says we owe a debt of gratitude to the ones who got there first, since they are the reason we have it at all, and he borrows the point from Aristotle, which is itself the argument in miniature. Christians had said the same thing three centuries earlier under a different sky. Justin Martyr held that whatever any man had said truly belonged to Christians, because the divine Word who became flesh in Christ had scattered seeds of truth among the nations. Augustine put it as a rule of thumb: when the philosophers have said something true, take it from them as from unjust possessors and put it to better use. Al-Kindi arrived at the same working principle from inside another faith, and he needed it to do anything at all.

(Q) "We ought not to be ashamed of appreciating the truth and of acquiring it wherever it comes from, even if it comes from races distant and nations different from us. For the seeker of truth nothing takes precedence over the truth, and there is no disparagement of the truth, nor belittling either of him who speaks it or of him who conveys it." + Source: al-Kindi, On First Philosophy I (Ivry translation). The sentence just before this one quotes Aristotle on being grateful to those who handed down any part of the truth, and al-Kindi says how beautiful the remark is before making it his own. The passage is the closest thing the Arabic philosophical tradition has to a manifesto, and it opens the book that a caliph was meant to read.

(T2) The world cannot have existed forever, because an endless stretch of past time could never have been finished. + A man works down a list of errands in order, and at any hour of the morning you can tell how far he has got by what is crossed off. Now hand him a list with no first line on it. Not a long list: a list that has no beginning at all, so that every errand on it has another errand above it that has to be done first. He will never cross off a single thing, and not because he is slow. There is nowhere for him to start. That is al-Kindi’s picture of the past, with days in place of errands. Before today could arrive, yesterday had to arrive first, and before yesterday the day before that, and so on with no first day anywhere in the series. If the series truly has no beginning, then the arrival of this perfectly ordinary morning required an infinite number of days to have gone by one after another and been completed. His claim is that this is not merely hard to imagine but impossible, because a series with no beginning is a series that was never started, and what was never started was never finished either. Yet here is today, finished and arrived. So the days must have started somewhere. Behind the picture stands a distinction he inherited from Aristotle and then put to a use Aristotle would have hated. An infinite in potentiality is a series you can always add one more to: whatever number you have counted to, you can count one higher, and nothing will ever stop you. An actual infinite is a different animal. It is an infinite quantity that is all there at once and complete, the whole endless collection delivered to the door like a set of encyclopedias with no last volume. Aristotle allowed the first kind and denied the second, and on that much al-Kindi agreed with him. Then he pressed the point past where Aristotle would go. Time, he says, is a quantity, measured out by the motion of bodies, and the body of the universe is finite in size, so the time belonging to that body cannot be actually infinite either. He backs it up with a piece of arithmetic worth following slowly. Take a finite stretch away from an infinite one, say a single century lifted out of a beginningless past, and ask what is left. If what is left is finite, then putting the century back gives you something finite, and the past you began with was never infinite in the first place. If what is left is still infinite, then you have one infinite quantity sitting inside another and smaller than it, which is not what infinite means. Either answer is absurd, and for al-Kindi that closes the question. The daring in this is easy to miss now. Aristotle was the master, and Aristotle taught that the cosmos had no beginning. Al-Kindi was willing to take Aristotle’s own principles and turn them against Aristotle’s conclusion, partly because he thought the Qur’an was right that the world began, and partly because he thought the point could be shown by argument instead of merely asserted. From a beginning to a Beginner is then a short step: what comes to be does not bring itself about, so the cause of the whole series has to stand outside it.

(Q) "Furthermore, if a definite time cannot be reached until a time before it is reached, nor that before it until a time before it is reached, and so to infinity; and the infinite can neither be traversed nor brought to an end; then the temporally infinite can never be traversed so as to reach a definite time." + Source: al-Kindi, On First Philosophy II (Ivry translation). To traverse is simply to cross, to get all the way from one end to the other. The sentence that follows in the text draws the conclusion: since a definite time plainly has been reached, time is finite, and so the being of a body is finite, and it is impossible for body to be eternal.

(T3) God is one in a way that nothing else is one: no parts, no kind, no description. + Call a married couple one, call a choir one, call a country one, and everybody understands you perfectly well. Everybody also understands that the choir is eighty people in matching shirts. Al-Kindi’s point is that every oneness we ever meet is that kind of oneness. A body is one and has halves. A species is one and has members. Even a soul, which has no halves that anything could cut, is one thing with several powers, since it remembers and wants and reasons. In every case the unity is something that has come to a thing which in itself is many. A unity that has come to something was given to it, and whatever is given has a giver. So he goes through the kinds of oneness one at a time, showing of each that its unity is borrowed rather than its own, and then he argues that the borrowing cannot go on forever. At the end of the chain stands what he calls the True One. It is not the largest member of the series and not the simplest member either, but the source from which every member has its unity at all. Because it is the source, the ordinary ways of describing a thing all fail on it. It has no matter and no form. It has no quantity and no quality. It has no genus, a genus being the general kind a thing belongs to, the way a dog belongs to the kind animal; and it has no specific difference either, a specific difference being the feature that marks off one sort within that kind, the way the power to reason marks a human being off from the rest of the animals. Strictly speaking the True One cannot be described at all, because every description sets a thing beside the other things of its type, and here there is no type. Christian readers will recognize most of this and should. It is close to what Western theology calls divine simplicity, the doctrine that God is not assembled out of anything and has no parts to come apart, held by Augustine and Aquinas and defended by Protestant and Catholic writers since. Where the paths separate is over how far the silence goes. Al-Kindi presses the denial until nothing at all may be said of God in his own right. Scripture presses in the other direction, saying that God is love and light and a Father, and that the Word who was with God and was God became flesh and was seen (John 1:1, 14). Being past our measuring is not the same thing as being past our knowing, and the Christian claim has always been that the One who cannot be measured has told us his name.

(Q) "The True One, therefore, has neither matter, form, quantity, quality, or relation, is not described by any of the remaining intelligible things, and has neither genus, specific difference, individual, property, common accident or movement; and it is not described by any of the things which are denied to be one in truth. It is, accordingly, pure and simple unity, i.e., (having) nothing other than unity, while every other one is multiple." + Source: al-Kindi, On First Philosophy IV (Ivry translation). The parentheses and the "i.e." are the translator’s, supplying in English what the Arabic carries in a word. A property, in the old logic he is using, is a feature that belongs to every member of a kind without being what that kind is, and a common accident is a feature that merely happens to be true of a thing. This is the summary sentence at the end of a long march through every kind of oneness al-Kindi can name, each one shown to be borrowed (T3).

(T4) To create is to bring something out of nothing, and only God has ever done it. + A potter at the wheel is not, strictly speaking, making anything. The clay was already there, dug out of a bank somewhere, and all the potter has done is rearrange it; take the clay away and he is a man waving his hands over a spinning disc. Al-Kindi noticed that every cause we are able to watch is like the potter in this respect. Fire heats water that already exists. Parents have a child out of material already to hand. Nothing in ordinary experience produces its own stuff. For the other kind of making he kept a separate Arabic word, ibda’, usually rendered into English as origination, and what it means is bringing a thing to be where there was no thing, no material, and no earlier state of affairs to work on. Origination of that sort takes no time, on his account, because there is no process to run through; and it has no subject, because there is nothing lying underneath waiting to be changed into something else. That, he says, is what God does, and it is what nothing else does. Everything else is an agent, meaning a thing that genuinely acts rather than merely gets acted upon, only by a sort of courtesy. Everything else had to be pushed before it could push, and what it hands on is the push it received. Against the Greeks this was a real break, and it is the point where a Muslim philosopher and the Christian tradition stand on the same side. Plato’s craftsman in the Timaeus shapes material that was always there. Aristotle’s god moves the world without ever having made it. Genesis begins somewhere else: in the beginning, God created the heavens and the earth (Genesis 1:1), and the letter to the Hebrews puts the same point as plainly as it can be put, that what is seen was not made out of things that are visible (Hebrews 11:3). Al-Kindi reached for the Greek toolkit and used it to argue for the doctrine the Greeks did not have.

(Q) "We say that the true, first act is the bringing-to-be of beings from non-being. It is clear that this act is proper to God, the exalted, who is the end of every cause." + Source: al-Kindi, On the True, First, Complete Agent and the Deficient Agent that is Metaphorically an Agent (Peter Adamson’s translation in the Stanford Encyclopedia of Philosophy, from the Arabic of Abu Rida 1950, pages 182 to 184). The treatise is a short one, possibly a surviving piece of something longer, and it exists to draw one distinction: between the agent who acts without being acted on and the agents who only pass along what was done to them.

(Q) "The True One is therefore the First, the Creator who holds everything He has created, and whatever is freed from His hold and power reverts and perishes." + Source: al-Kindi, On First Philosophy IV (Ivry translation). The last lines of the surviving book. Having argued the world to a beginning and the beginning to a single source, al-Kindi ends not with a principle but with a hand: creation is held in being moment by moment, and let go of, it stops.

What Christian Thinkers Made of Him

(†) His argument against an eternal world came from a Christian, and Christians took it back. + Three hundred years before al-Kindi, a Christian grammarian in Alexandria named John Philoponus had done something almost nobody in the ancient world dared to do: he attacked Aristotle head on. A grammarian in that world was not a man fussing over commas but a professional teacher of literature and language, which in Alexandria meant a man who had read everything. His Against Aristotle on the Eternity of the World and his reply to the pagan Proclus argued that a beginningless past would be an actual infinite, and that an actual infinite cannot be gone through or added to. Philoponus was read in Arabic, and al-Kindi’s finitude arguments run along his tracks. From al-Kindi the argument passed into Muslim theology, and two centuries later al-Ghazali gave it the compact shape that carried it down the years: whatever comes to be has a cause, the world came to be, therefore the world has a cause. Then Christians met it coming back the other way, and they did not agree about it. Bonaventure argued in thirteenth-century Paris that an infinite past could never have been traversed, which is exactly the move al-Kindi had made (T2). Thomas Aquinas, in the same city and the same generation, took the opposite view, holding that Scripture teaches a beginning but that reason on its own cannot demonstrate one. The question was live among Christians, in other words, and not settled by authority. The modern form is the one a reader is most likely to have met. Christian philosophers call it the kalam cosmological argument, after kalam, the Arabic word for reasoned theological argument, and William Lane Craig revived it in the 1970s by tracing it back through al-Ghazali to al-Kindi and to Philoponus behind him. So the chain runs from a Christian in Alexandria to a Muslim in Baghdad to a Muslim in Persia and back to Christians in Paris, in Oxford, and in our own century, each of them handing on an argument they did not invent. Al-Kindi would have thought that entirely as it should be (T1).

(†) The Greek books reached him through Christian hands, and reached the Christian West again through his. + The translation movement al-Kindi worked inside was staffed, to a striking degree, by Christians. Hunayn ibn Ishaq of the Church of the East ran the rival circle and set the standard for the whole enterprise. Qusta ibn Luqa, a Melkite from Baalbek, meaning a Christian of the churches that held to the faith of the emperor at Constantinople, translated Greek mathematics and wrote philosophy of his own. Ibn Na’ima al-Himsi, the Syrian Christian who produced the Arabic Theology of Aristotle for al-Kindi’s circle, had learned his Greek in a tradition that had been reading Plato and Aristotle in monasteries and schools for centuries. The Syriac churches were the pipe through which Athens reached Baghdad. Then the direction reversed. From the middle of the twelfth century, Toledo in Christian Spain became a workshop where Arabic books were turned into Latin, and several hands there put al-Kindi into the language of the European schools. His treatise on the intellect crossed over twice: Gerard of Cremona made one version, and the one that circulated widely as De intellectu came out of the workshop of Dominicus Gundissalinus. It was among the first works of Arabic philosophy the schoolmen had in front of them, and it helped set the terms of a long Latin argument about how a human mind comes to understand anything at all. Gerard also gave the West the optics, De aspectibus, which taught Europe to treat seeing as a matter of rays traveling in straight lines, and which stands behind the work of Roger Bacon and the other thirteenth-century students of light. There is a lesson in the traffic. Christians and Muslims in these centuries argued with each other hard and often, and the same men also copied, corrected and carried each other’s books. Both things are true, and a history that keeps only one of them is not the history.

(†) Where the gospel parts company with him: he wrote against the Trinity, and a Christian philosopher of Baghdad answered him. + Al-Kindi did not leave this to inference. He wrote a short treatise arguing that the Christian doctrine of the Trinity is incoherent, and the argument is built out of the same materials as everything else he wrote. The tools came from the Isagoge of Porphyry, a third-century introduction to logic that every student in Baghdad worked through before he was let near anything harder. Porphyry sets out the ways a term can be said of a thing: as its genus, the broad kind it falls under; as its species, the narrower kind inside that; as its difference, the feature marking the narrow kind off from the rest of the broad one; and then as a property or as an accident, the two looser ways of belonging. Those are the same categories al-Kindi had used to strip every description away from the True One (T3), and here they come back as weapons. The case runs like this. If Father, Son and Spirit each have something in common with the others and something of their own, then each of them is a composite, meaning a thing put together out of parts, in this case out of the shared part and the private part. Whatever is put together must have had something put it together, and whatever has a composer is caused. So each person of the Trinity would be a caused thing, and a caused thing is not God. Set that beside the True One of (T3), of whom no genus and no difference may be spoken at all, and the two accounts of God cannot both be true. A century later Yahya ibn ’Adi, a Christian philosopher of Baghdad and one of the best logicians of his day, wrote a reply that survives. His answer was that al-Kindi had smuggled in an assumption about what "common" means. Something can be common to several things without being divided up among them like a cake, and the attributes the three persons share are eternal and uncaused, so the composition al-Kindi needs never gets going; nor does a relation that has always held require one of its terms to have come first in time. Christians had been here before, and Ibn ’Adi knew it. The Cappadocian fathers, the fourth-century bishops of central Asia Minor who did more than anyone to fix the church’s language for the Trinity, had answered an argument of the same shape from Eunomius, who reasoned that God’s whole essence is to be unbegotten, so that the Son, being begotten, could not be God in the full sense. Ibn ’Adi was drawing on that inheritance. The disagreement is genuine and it is not small. Al-Kindi held that God’s oneness rules out any distinction within him whatever. The Christian claim is that the one God is Father, Son and Holy Spirit, that the Son is the radiance of the glory of God and the exact imprint of his nature (Hebrews 1:3), and that disciples are baptized into the one name of the three (Matthew 28:19). Both sides say God is one; they mean different things by saying it, and no amount of goodwill dissolves that. What can be said is that al-Kindi argued the point rather than shouting it, and that a Christian in his own city answered him in the same register, which is more than most centuries have managed.