web analytics
26 Jun

Correcting Lucretius: Dúngal and Carolingian Cosmology

King and Court Astronomer
Share on FacebookTweet about this on Twittershare on TumblrShare on LinkedInGoogle+Print this pageEmail to someone

In January 1417 a man called Poggio Bracciolini pulled a book from a shelf in a German monastic library. The text was De Rerum Natura, a long poem written by the Roman Lucretius in the 1st century BC outlining the tenets of Epicureanism. Much has been written about that event and its effect on the Renaissance, even suggesting it was central in creating modernity1. The story of the text before it was found is less well known.

De Rerum Natura was a poem written in the 1st century BC outlining the tenets of Epicureanism, a philosophical school founded by Epicurus (c. 341-c. 271 BC)2. Epicureanism grew to be one of the major philosophical schools, declining in popularity from the 2nd century on. The oldest manuscripts of the poem that survive are held in the library of Universiteit Leiden: Voss. Lat. F. 30, from the early 9th century nicknamed O and Voss. Lat. Q. 94, nicknamed Q3.

O was copied from an exemplar now lost in a Carolingian scriptorium in the early ninth century (c. 800), probably in north-west Germany or north-east France. Substantial corrections were made soon after in an insular hand (ie the style of writing used in Britain and Ireland). Further corrections were made in the mid-ninth century by a Carolingian hand, and glosses added in the late tenth or early eleventh century4. These corrections (some 20% of which were inaccurate, which allows us to trace them) appear in the Poggio manuscript 5, meaning that the book Poggio found was a descendant of O.

Page from "O" showing corrections in darker ink (c) Leiden University Library, ms. VLF 30, fol. 22r

Page from “O” showing corrections in darker ink. Leiden University Library, ms. VLF 30, fol. 22r

Correction was not a mechanical process: judgement was required to correct Latin usage, to decide what an unclear word should be based on context or, in the case of poetry, what word fitted the metre. Interest also affected corrections: the second corrector made most changes to Book I and Book VI, perhaps reflecting fluctuating enthusiasm. The first, on the other hand, made many more corrections, mainly to Books VI and V which had the most cosmological material 6.

That first corrector has been identified by Bernard Bischoff as the Irish scholar Dúngal (active c. 800-c. 827), based on comparisons with existing samples of his writing 7. If this identification is correct, it explains the interest of the corrector in the cosmological books. Dúngal’s best-known work explained the structure of the universe to Charlemagne.

Dúngal is first mentioned in a letter written to Irish monks by Alcuin, the most senior scholar in the Carolingian court, around the 790s. That letter describes Dúngal as “a venerable brother, a doctor of your learning…a most devout person leading a regular life”. Charlemagne consulted Dúngal in 801 regarding the substance of nothing, a topic of debate at the time8. In 811, Charlemagne wrote to the abbot of St Denis requesting that Dúngal explain how two solar eclipses could occur in the same year (a total eclipse was seen in Western Europe on 30 November 810, but a bishop of Constantinople had informed the Carolingian Court of an eclipse seen earlier in 810)9.

Dúngal’s detailed answer outlined the well-known medieval picture of a cosmos with a central earth orbited by the sun, moon and planets. It draws mainly on Macrobius’s early fifth century commentary on Cicero’s Dream of Scipio, a commentary known to the Irish. Data on solar and lunar latitudes came from Pliny the Elder (from memory, since Dúngal noted that he did not have the manuscript to hand), along with information about predicting eclipses. Dúngal closed with an attempt to date the first eclipse using techniques rooted in computus and looked forward to the possibility of accurate prediction of celestial events10.

Dúngal’s answer required choosing between different accounts of the cosmos. Most classical authorities agreed that the moon was the closest body to the earth and the outer planets were Mars, Jupiter and Saturn, but there were different orders suggested for the bodies in between. Dúngal’s letter includes Macrobius’ outline of three of these. There was the Chaldean system (Moon – Mercury – Venus – Sun) which Pliny favoured, the Egyptian system (attributed to Plato: Moon- Sun – Mercury – Venus) which Macrobius preferred and the Later Platonic (Moon – Sun – Venus – Mercury)11.

Dungal’s cosmos,
“Dungalus, Epistola ad Carolum Magnum de duplici solis eclipsi anno 810” (Dungal’s letter to Charlemagne)
Royal MS 13 A XI f. 127.r
British Library

Dúngal’s letter not only gives preference to the Chaldean system, following Pliny, but gives the impression that Macrobius is in agreement. The most plausible explanation is that this was not a mistake, but deliberate. Dúngal accurately quotes Macrobius, but does so selectively, leaving out the stronger of the two defences of the Platonic system and Macrobius’ explicit statement of which he prefers 12. Monks such as Dungal were not mechanically repeating the ancients but attempting to create a consistent account of the universe. “The story of Dungal’s reinterpretation of Macrobius is a nice example of Carolingian concern to rationalise disagreements in doctrine about the natural world and to establish a unified authoritative teaching”13.

The letter was widely copied and circulated in the 9th century. The earliest extant manuscript includes additional extracts from Macrobius, Isidore’s Etymologies and Bede’s The Nature of Things. It also contains diagrams, including one which corresponds to Dúngal’s preferred planetary order. This diagram is also found in later manuscripts of Macrobius. Eastwood suggests the letter achieved high status and was copied for educational use in the 9th century, with the diagram then making its way into later Macrobius copies14,

This image from a 15th century Macrobius manuscript shows the Dungal order of planets (MS Typ 7, Houghton Library, Harvard University)

This image from a 15th century Macrobius manuscript shows the Dungal order of planets (MS Typ 7, Houghton Library, Harvard University)

The letter makes no use of Lucretius, but that is not surprising. Lucretius’ cosmology suggests a flat earth: he is not convinced that solar eclipses are caused by the intervention of the moon and he explicitly mocks the idea of a spherical earth 15 His cosmology cannot be aligned with Macrobius or Pliny.

Dúngal’s career did not end with the letter to Charlemagne. In 825 his school in the Carolingian capital at Pavia is recommended as the primary centre for those seeking instruction in Milan, Lombardy, and Piedmont. Verses attributed to him survive, along with a treatise written in 827 arguing against the iconoclastic ideas of Bishop Claudius of Turin, a protégé of the emperor Louis the Pious. Dúngal outlines the history of Christian use of images, assembles quotes from the church fathers supporting their use and argues for their utility in education16.

Dúngal retired to the monastery at Bobbio bringing 28 treatises with him, six of which survive at Bobbio. A copy of Lucretius recorded at Bobbio in the tenth century (now lost) may also have been brought there by him. He died and was buried there in the early 830s17.

Carolingian cosmology did not end with Dúngal. The structure of the cosmos continued to be debated and different classical authorities used. For example, the 5th century Martinius Capella’s De nuptiis Philologiae et Mercurii (“On the Marriage of Philology and Mercury”) suggested that Mercury and Venus orbited the sun, which in turn orbited the earth. Three separate commentaries written in the ninth century survive for De nuptiis, including one by John Scottus Eriugena. Eriugena went even further than Martinius, suggesting that Mars and Jupiter also orbited the sun18. However all existing models were to be superseded by Ptolemy’s Almagest, due to that work’s greater accuracy in predicting planetary motions. Translated into Arabic in the 9th century, the Almagest was translated from Arabic into Latin in the 12th century.

Featured image: King and court astronomer, from a 15th century Macrobius manuscript: MS Typ 7, Houghton Library, Harvard University

Further Reading

Irene O’Daly (2014) “The Beauty of Mistakes” on MedievalFragments – examines the corrections made to “O”.

Danièle Cybulskie (2014) “Copycat: The Life of a Medieval Scribe” on Medievalists.net.

The Medieval manuscripts page, Universitet Leiden Library.

Jeremy Norman (2012) “The Earliest Surviving Text of Lucretius’s De rerum natura (Circa 825)” on History of Information

Jeremy Norman (2015) “Poggio Rediscovers Lucretius’s De rerum natura (1417 – 1473)” on History of Information

NASA: Catalog of Solar Eclipses: 0801 to 0900

Information on Irish scholars involved in the 9th century Carolingian Renaissance, including links to Peter Adamson’s “History of Philosophy without any gaps” podcasts on the period.

Richard Sharpe (2010) “Books from Ireland, fifth to ninth centuries”, Peritia 21, pp. 1—55 (online pdf). Explores Irish book culture in the early medieval period.

David Butterfield (2013) The Early Textual History of Lucretius’ De rerum natura Cambridge University Press.

Bruce Eastwood (2007) Ordering the Heavens: Roman Astronomy and Cosmology in the Carolingian Renaissance , Leiden, NL:Brill


  1. See for example, “The Swerve: How the World Became Modern” by Stephen Greenblatt (2011, W. M. Norton & Co.)
  2. For more on Epicurus and Epicureanism, see Tim O’Keefe “Epicurus (341—271 B.C.E.)”, The Internet Encyclopedia of Philosophy, ISSN 2161-0002, URL = http://www.iep.utm.edu/epicur/ , 01 June 2016 and Konstan, David, “Epicurus” in Edward N. Zalta (ed.) The Stanford Encyclopedia of Philosophy (Summer 2014 Edition), URL = http://plato.stanford.edu/entries/epicurus/
  3. Images of pages from Voss. Lat. F. 30 and Voss. Lat. Q. 94 are available from the website of library of Universiteit Leiden. There was also a 9th century copy in Germany, of which only fragments survive. Note that these are only surviving copies: many more may have been in existence and since lost.
  4. David Butterfield (2013) The Early Textual History of Lucretius’ De Rerum Natura, Cambridge University Press, pp. 6-7, 204-5, 228-9, 246.
  5. Butterfield (2013), pp. 22, 28-9
  6. Butterfield (2013), pp. 204-220 228-233
  7. Bischoff’s identification is based on glosses in books annotated by Dúngal and given by him to the library at Bobbio, see Butterfield (2013) p. 218. Not all accept this identification, see Michael Reeve (2007) “Lucretius in the middle ages” in The Cambridge Companion to Lucretius, Cambridge University Press, pp. 205-213.
  8. David Ganz (2004) “Dúngal (fl. c.800–827)”, Oxford Dictionary of National Biography, Oxford University Press, doi:10.1093/ref:odnb/8263.
  9. Bruce Eastwood (2007) Ordering the Heavens: Roman Astronomy and Cosmology in the Carolingian Renaissance , Leiden, NL:Brill, p. 46
  10. Ganz (2004). Eastwood (2007) pp. 175-7
  11. Eastwood (2007), pp. 37-43 goes through the Macrobius text in detail and outlines how it was misread by later readers
  12. Eastwood (2007), pp. 46-52
  13. Eastwood (2007), p. 52
  14. Eastwood (2007),pp. 50-52.
  15. See 1.951–1117 (Book I) and 5.416–770 (Book 5) of De Rerum Natura. The verses describing the structure of the cosmos are from here on. Lucretius on eclipses is here. The lines explicitly saying the idea of a spherical earth is mistake is here.
  16. Ganz (2004)
  17. Butterfield (2013), p. 31. Ganz (2004)
  18. Bruce Eastwood and Gerd Grasshoff (2004) “Planetary Diagrams for Roman Astronomy in Medieval Europe, ca. 800-1500” Transactions of the American Philosophical Society, Vol. 94, No. 3, pp. i-iii, v-xiv, 1-21, 23-71, 73-115, 117-147, 149-158 – see pp. 7-8. Dermot Moran (2008) “John Scottus Eriugena” in Edward N. Zalta (ed.), The Stanford Encyclopedia of Philosophy (Fall 2008 Edition), URL = http://plato.stanford.edu/archives/fall2008/entries/scottus-eriugena/

Leave a Reply

Your email address will not be published. Required fields are marked *