The Western Shelf · 300–1600

About a hundred and sixty entries, not a library. The test for inclusion was consequence: was it read, copied, imitated, translated, argued with or banned by those who came after? Thirteen centuries of Europe produced far more good books than any shelf can hold, and many fine ones fail that test and are not here.

The shelf is built around authors rather than titles, so that a life stands behind each book, and anonymous or collective works are gathered under the tradition that produced them. That is why the Icelandic family sagas, the chansons de geste, the Frankish annals, the Scottish makars and the mystery cycles each occupy a single spine rather than eight. The list this grew from ran to four hundred and seventy-two separate titles; grouping them this way loses nothing and makes the shape of the thing visible.

Latin runs the whole length of it and is still the majority language in 1550. The vernaculars arrive at wildly different moments — English and Irish early, French and German in the twelfth century, Italian in the thirteenth, Castilian and Portuguese later still — and the language filter, laid against the period filter, is the quickest way to watch that happen.

Byzantium is included. A canon of the West that stops at the Adriatic is a canon that cannot explain where the West got its Plato, its Aristotle, its Greek grammars or its idea of what an empire looks like. Procopius, Psellos and Anna Komnene belong on this shelf as surely as Gregory of Tours does.

Where the shelf ends

Nominally at 1600; in practice at about 1616, which is where Shakespeare and Cervantes both stop. Cutting at the round number would have severed King Lear, Macbeth, The Tempest and the second part of Don Quixote from everything that produced them, which no reader would thank us for. Works that clearly belong to the next age have been left for the next shelf: Grotius, Harvey, Descartes, Hobbes, Burton's Anatomy of Melancholy, Browne's Religio Medici, Marino's Adone, and Galileo's Dialogue of 1632 — though his Sidereus Nuncius of 1610 is here, since it belongs with Copernicus and Kepler rather than with what came after.

Corrections made along the way

The source list carried a number of duplicates, now merged: Cosmas of Prague's Chronica Boemorum appeared twice under two titles, as did Dante's De Monarchia, Vesalius's Fabrica, Münster's Cosmographia, Cellini's Autobiography, Einhard's Life of Charlemagne, Ariosto's Orlando Furioso, Rojas's Celestina, Alemán's Guzmán and Oviedo's General and Natural History. Vincent of Beauvais's Speculum Historiale is a part of his Speculum Maius, not a separate work, and is shown as such.

Andrzej Frycz Modrzewski's major book is De Republica emendanda; the Satyr credited to him in the source is by Jan Kochanowski. Piero della Francesca's perspective treatise is De prospectiva pingendi. Cennino Cennini's book is Il Libro dell'Arte, in Italian, not a Latin De pictura — that title belongs to Alberti. Ermoldus Nigellus wrote in honour of Louis the Pious, not a verse Vita Karoli. The Book of the Dean of Lismore is a Gaelic anthology, not a Scots one.

Absences that were harder to defend

The source list arrived without Gregory of Tours, Ammianus Marcellinus, Bernard of Clairvaux, the troubadours, the Golden Legend, Meister Eckhart, Ramon Llull, Ficino, Sannazaro, Leonardo, Tyndale, Las Casas, Kochanowski, Tirant lo Blanch, Don Juan Manuel, Thomas Middleton and the English mystery cycles. All are here, along with Macrobius, Martianus Capella, Pseudo-Dionysius, Eriugena, Hrotsvitha, the Carmina Burana, the Romancero, Catherine of Siena, Ausiàs March and the Corpus Iuris Civilis — which is arguably the single most consequential Latin text on the shelf and had been left off it entirely.