By the account of his youth that he gave late in life to his pupil and secretary Abū ʿUbayd al-Jūzjānī, Avicenna read Aristotle's Metaphysics forty times as a young man in Bukhara, learned it by heart, and still could not see what the book was for. A short treatise by al-Fārābī, the tenth-century philosopher who had worked in Baghdad, explained Aristotle's aims to him at last. The same account says he mastered logic, mathematics and physics largely on his own, was let into the palace library of the Sāmānid ruler of Bukhara after treating him as a physician, and had finished his studies by eighteen. Dimitri Gutas, the historian who has studied that account most closely, reads it as a lesson as much as a memoir. It illustrates Avicenna's own theory of knowledge: a strong mind can reach the truth without a teacher, because it sees at once the middle term that makes an argument work. He called that gift ḥads, intuition.
What he built from there was a single system meant to explain everything, religion included. Its centre is a proof of God that starts not from motion, as Aristotle's did, but from the difference between what must exist and what merely can. Whatever merely can exist, Avicenna argued, owes its existence to something else, and the chain of causes ends in one being that owes it to nothing: the Necessary Existent (wājib al-wujūd), which is God. In everything but God, what a thing is does not include the fact that it is. His encyclopedia al-Shifāʾ (The Healing) closes on this metaphysics and ends with prophecy and religious law, giving Islam itself a philosophical account.
The theologians answered. Al-Ghazālī, the jurist and theologian, wrote his Incoherence of the Philosophers around 1095 and named three of Avicenna's teachings as unbelief: that the world has no beginning in time, that God knows particular things only in a universal way, and that souls do not return to bodies after death. Even so, later thought in the Islamic East was largely built in Avicenna's terms. Fakhr al-Dīn al-Rāzī, a theologian, pressed hard against Remarks and Admonitions, Avicenna's last summary of philosophy, in a commentary on it, and Naṣīr al-Dīn al-Ṭūsī, an astronomer and philosopher, wrote another to defend it. In the seventeenth century the Persian philosopher Mullā Ṣadrā was still writing glosses on the metaphysics of al-Shifāʾ.
His account of the soul starts from a thought experiment. In Kitāb al-Nafs, the book on the soul in al-Shifāʾ, Avicenna asks the reader to imagine a man created in empty air, his eyes veiled, touching nothing. That man would still know that he exists without knowing that he has a body, so the self he knows is not the body. The rational soul, the book argues, thinks without any bodily organ and survives death. Translated into Latin in the twelfth century, the books on the soul and on metaphysics in al-Shifāʾ shaped the scholastics, Thomas Aquinas among them. In Cairo, Maimonides, the Jewish philosopher writing in Arabic, read him too, though he ranked al-Fārābī above him.
He wrote all this while serving rulers who kept changing. In 999 the Turkic Qarakhanids took Bukhara, and Avicenna, an official of the fallen Sāmānid state, left; in his own words, necessity led him to go. He served one court after another as physician and administrator, and in Hamadan became vizier to the Buyid emir Shams al-Dawla. After the emir's death the court suspected him of dealing with ʿAlāʾ al-Dawla, the ruler of Isfahan, and held him for four months in the castle of Fardajān. He spent his last years in ʿAlāʾ al-Dawla's service in Isfahan, and wrote for him a summary of philosophy in Persian; most of his books are in Arabic, the language of learning across the Islamic world. He began al-Shifāʾ in Hamadan at his students' request and finished it in Isfahan by about 1027. His Canon of Medicine (al-Qānūn fī al-ṭibb) taught medicine in European universities until the seventeenth century.
He died in 1037 of colic at Hamadan, on one of his patron's campaigns, and was buried there. Popular stories in the Persian- and Turkish-speaking world later made him a saint and a mystic, or a magician. Most reference works give 980 as the year of his birth. Gutas, who wrote the Stanford Encyclopedia of Philosophy's entry on him, holds that this date is mistaken and that he was born some years earlier, in the 970s.