The Canon of Medicine Is Completed

Ibn Sina's writing career unfolded amid the political fragmentation of the eastern Islamic world. He served rulers as physician and administrator, moved between courts, and at Hamadan was a vizier, went into hiding, and was imprisoned for a time. The neat date attached to the finished book should not conceal that unsettled setting or the long work of later transmitters (Gutas).

1. Ninth–tenth centuries: translation into Syriac and Arabic and the compendia of earlier Arabic-writing physicians created the textual and technical foundations on which Ibn Sina worked (Musallam).
2. 980: Ibn Sina is born at Afshana, near Bukhara, and is educated in the Samanid capital.
3. 1012–1014: after travelling west, he settles at Jurjan and begins the Canon. His disciple and biographer Abu ʿUbayd al-Juzjani joins him there (Bertolacci).
4. 1014–1015: Ibn Sina works at Rayy. Book 1 was composed before 1015, showing that the work already had substantial form before his longer Hamadan period (Musallam).
5. 1015–1024: while serving at Hamadan, Ibn Sina completes the five-book work. “c. 1025” is the conventional rounded date used for this timeline entry (Bertolacci).
6. Second half of the twelfth century: Gerard of Cremona translates the Canon from Arabic into Latin at Toledo and adds a glossary (Weisser).
7. Thirteenth–fourteenth centuries: Arabic commentaries critically investigate the text; in western Europe, medical scholars use it in the thirteenth century and university courses adopt selected sections in the fourteenth. A Montpellier syllabus of 1309 lists it as optional; by 1340 it was established in several programmes (Fancy; Weisser).
8. 1472–1473: a partial Latin edition of Book 3 appears in 1472; the earliest dated complete Latin edition is printed at Milan in 1473. Printing multiplied particular Latin versions rather than recovering a pristine text (Weisser; Qatar National Library).
9. 1527–1593: Andrea Alpago's corrections to the Latin text, based on Arabic manuscripts and accompanied by a glossary, appear posthumously at Venice in 1527. The Arabic text is printed at Rome in 1593, chiefly for export to eastern markets (Weisser).

This sequence offered readers several routes into the work: principles for teaching, an alphabetical drug book for consultation, an organ-by-organ therapeutic survey, and a formulary. Surviving copies show that readers sometimes extracted only the anatomy, while Latin university lecturers concentrated on selected sections rather than teaching all five books at equal depth (NLM; Weisser).

In Book 2, Ibn Sina says that a drug's powers could be known through reasoning or through tajriba, experience. He sets seven conditions for making experiential claims: the substance should not have an acquired quality caused by preparation or storage; the illness should be “simple” rather than composite; essential and accidental effects should be distinguished; the strength and timing of the effect should be considered; the result should occur consistently in most cases; and a drug intended for people should ultimately be assessed in the human body (Karimullah).

These are important medieval reflections on how observation can support a causal inference. They are not a protocol for randomisation, blinding, control groups, statistical analysis, or informed consent. The rules also presuppose a humoral account of diseases and medicinal “powers.” Kamran Karimullah traces them to Ibn Sina's reworking of Galen's qualified experience within Aristotelian debates about knowledge. Describing the passage as the invention of the modern clinical trial turns resemblance into identity and obscures its actual intellectual setting (Karimullah).

The distinction matters medically. The Canon documents what a learned eleventh-century physician considered plausible and well reasoned; it does not establish that its diagnoses map onto modern diseases or that its remedies are safe and effective by current evidential standards.

More than two dozen Arabic commentaries on the Canon were written between the twelfth and fourteenth centuries. Nahyan Fancy's study of Fakhr al-Din al-Razi (d. 1210) and Ibn al-Nafis (d. 1288) rejects the old picture of commentary as uncritical repetition. Their method of taḥqīq—verification or critical investigation—could retain much of the shared medical system while disputing particular arguments (Fancy).

One surviving NLM manuscript makes this layered reception visible: its margins preserve extensive extracts from Ibn al-Nafis's commentary on anatomy. Ibn al-Nafis faulted Ibn Sina for dispersing anatomical discussion through organ-specific sections. In the same commentary he rejected a Galenic passage between the heart's ventricles and described blood reaching the left ventricle by way of the lungs. That later argument was produced through engagement with the Canon, not simply contained in Ibn Sina's original text (NLM).

This commentary tradition also corrects a Eurocentric transmission story. The work had long Arabic afterlives in scholarly centres across the Islamic world, alongside abridgements, marginal notes, and Persian and Hebrew engagements. Its Latin career was one branch of a wider history, not the destination that gave the Arabic book its meaning.

Gerard of Cremona's twelfth-century Latin translation answered the needs of a developing scholastic medicine. The Canon's ordered framework suited teaching in which medical theory was connected to Aristotelian natural philosophy. Yet adoption was gradual: the strongest evidence puts its use by western medical scholars in the later thirteenth century and in university courses from the fourteenth, not immediately after translation (Weisser).

Curricula selected. Lectures frequently concentrated on the physiology of Book 1 and the theory of fevers in Book 4; Book 3's vast organ-by-organ survey was often excerpted or used for examinations. Professors such as Dino del Garbo, Gentile da Foligno, and Jacques Despars wrote commentaries that mediated the text for students and sometimes added observations of their own (Weisser).

Print extended but did not freeze this tradition. Fifteenth- and sixteenth-century editions combined Gerard's Latin with glosses, corrections, indexes, and commentaries. Humanist physicians then attacked Arabic-Latin terminology and particular prescriptions while preferring newly recovered Greek texts; Paracelsus's famous burning of the Canon at Basel in 1527 was a dramatic but unrepresentative rejection of textual authority in general. Decline was therefore contested and uneven rather than a single scientific overthrow (Weisser).

The completion chronology is reconstructed from Ibn Sina's movements and the history of his writings. The core narrative of his life depends heavily on an autobiography recorded by al-Juzjani and the continuation that the disciple supplied. Dimitri Gutas cautions that this combined account is the source behind later narrative biographies, not one witness among many independent lives (Gutas). The rounded date 1025 should consequently be read as a historical marker, not false precision.

Manuscripts answer different questions. NLM MS A 53 is a complete, unsigned and undated Arabic copy probably made in an early fifteenth-century Timurid workshop in Iran. Its five illuminated openings, cranial-suture diagrams, emendations, Persian notes, repairs, and rebordering preserve evidence of production and later ownership (NLM). A complete 1597 copy now at Yale names its scribe, ʿAbd al-Karim al-Qutbi al-Hanafi, in a colophon and carries later Arabic and Persian notes (Library of Congress). Neither manuscript is an eyewitness record of composition six centuries earlier.

The Canon itself is a prescriptive work of elite learned medicine, not a hospital register or a collection of patient narratives. It is strong evidence for the categories and therapies Ibn Sina wished to systematise, but much weaker evidence for how widely any remedy was used, who could obtain it, or what patients and non-textual healers thought of treatment. Its social impact must be reconstructed from manuscripts, curricula, commentaries, and other records rather than inferred from the author's instructions alone.