Bryn Mawr Classical Review

BMCR 2019.06.18 on the BMCR blog

Bryn Mawr Classical Review 2019.06.18

Marcus Deufert, Kritischer Kommentar zu Lukrezens 'De rerum natura'. Texte und Kommentare, Band 56.   Berlin; Boston:  De Gruyter, 2018.  Pp. x, 516.  ISBN 9783110414714.  €149,95.  

Reviewed by Nicoletta Bruno, Thesaurus Linguae Latinae, München​ (

Table of Contents

The Kritischer Kommentar zu Lukrezens 'De rerum natura' by Marcus Deufert is the result of tireless work on the Lucretian text and is partly a tribute to Karl Lachmann—as Deufert himself acknowledges (p. V)—and to the practice of writing commentaries that exclusively address questions of textual criticism. It may seem old-fashioned to publish a critical commentary together with the critical text, but there is nothing more useful than a work like this to justify textual choices and to guide the reader step by step in explaining the text. The praefatio of a critical edition can only provide a general survey of the method and stemmatic relationships, but cannot go into detail on specific choices. Together with the Prolegomena, published in 2017, and the Teubner critical text, published in April 2019, this provides a very comprehensive outline of the Lucretian text, which represents over twenty years of Deufert’s Lucretian scholarship.

The De rerum natura (DRN) is famous for its abundant repetitions, which have been interpreted in various ways: some argue that they are a mnemonic tool to help readers memorize Epicurean precepts; others think that they are a sign of Lucretius’s poetic immaturity; and still others take them as strong evidence of interpolation. In an earlier study, Deufert argued for the deletion of about 370 verses as later interpolations.1 In the Kommentar, Deufert is more cautious but he still firmly believes that most of those suspicious verses are spurious. By contrast, David Butterfield, editor of the future OCT Lucretius, believes that the DRN is incomplete: for him the work was unrevised. In this respect, Butterfield agrees with H. Diels, R. Heinze, C. Bailey and E. J. Kenney, who are all reluctant to find interpolations among the textual inconsistencies, and who focus on other difficulties in the text such as incomplete verses and inconsistences of content and style. Deufert is not inclined to accept radical interventions, such as conjectures and transpositions, which he often relegates, even the most convincing ones, to the apparatus. In this respect his approach is generally conservative.

Deufert proposes the removal of about 220 verses across 60 passages, which he considers spurious. Moreover, he makes frequent use of cruces desperationis. The critical question is this: is the DRN an incomplete and imperfect work or a complete and interpolated one? According to Deufert, the text is complete and was interpolated in the early stages of its circulation.2 Since the supposed interpolated lines are already in the Lucretian text as it is known to authors of the first century AD, such as Seneca and Quintilian, Deufert believes that the interpolations were very ancient, within approximately fifty years of Lucretius’s death. The basis for his belief is found in the Anecdoton Parisinum (Paris BN lat. 7530, ff. 23r-29r), which reports that the grammarian Probus (late 1st century AD) added critical notes to various poetic texts, including that of Lucretius. This report would have come from a lost work by Suetonius. Furthermore, Jerome states that he had commentaries on Lucretius in Adversus Rufinum 1, 16. These are commentaries that could perhaps be related to Probus’ work.

The spurious passages identified by Deufert are certainly problematic from a stylistic and syntactic point of view and inconsistent in their content. In what follows, I will briefly show how Deufert faces three complex Lucretian passages: the preface of Book 4, verses 1-25 and verses 1211-1217 and 1341-1349 of the fifth book.

One of the most discussed problems concerns the proem of Book 4, lines 1-25, which is almost the same as 1, 926-950. Deufert refers to the long demonstration carried out in his book of 1996 (pages 81-96). There were four possibilities to explain this long repetition in the text: 1) Lucretius is the author of only the vv. 926-50, written in Book 1 ; 2) an interpolator moved to the first part of Book 4 the verses written in Book 1; 3) Lucretius first wrote the verses of the proem of Book 4 and then he moved them to Book 1; 4) an interpolator added the verses of Book 4 to Book 1. Deufert’s idea is that the author of the verses of Book 1 was Lucretius, and therefore the verses are authentic, while those of Book 4 were written by an interpolator. Lines 921-50 are consistent in themselves, since they resume the controversy against Heraclitus and the end of the first preface and introduce the end of the first book. The interpolator would have written only nam to 4, 11 instead of sed of 1, 936. Deufert (p. 201) also criticizes Butterfield:

“Bei der Übertragung der Verse 926-960 an den Beginn des vierten Buchs dürfte die glattere und leichtere kausale Konjunktion nam—sei es bewusst, sei es unbewusst—die anspruchsvollere, aber untadelige adversative Konjunktion sed verdrängt haben, die Butterfield (2016 p. 27-29)3 gewiss zu Unrecht in 1, 936 durch nam ersetzen möchte: Butterfield übergeht in seiner Paraphrase des Zusammenhangs, dass Lukrez den Vers 935 negativ formuliert hat (non … uidetur), was die Voraussetzung dafür ist, dass er mit adversativem sed fortfahren kann” (p. 201).

According to Butterfield, Lucretius himself did write the verses that stand in the manuscripts as both 1, 926-50 and 4, 1-25. Moreover, Deufert, in the Kommentar believes that Book 4 of the DRN originally did not have its own proem, unlike the other books. To support this, he quotes a recent study by Michael Erler, who cites as a parallel the proem-less beginning of Book 2 of Apollonius Rhodius’s Argonautica.4 In general, if Lucretius has neither revised nor published an edition of his poem himself, it is unlikely that infelicities in the text can all be explained as errors in transmission. Butterfield comes to this cautious conclusion: “the Lucretian editor should certainly not improve the ordering of the poem for him by rearranging paragraphs and arguments, regardless of what Lucretius might have intended, unless the transposition could be correcting a genuine error of transmission”.5 On the other hand, if we accept Deufert’s hypothesis, that the poem had been completed by Lucretius, the deletion of 4, 1-25, and many other passages that Deufert removes, would be consistent and acceptable.

There are several sections that Deufert prefers to print in his new text compared to his 1996 book. One is 5, 1211-1217, which he formerly considered spurious since it broke a sequence in style and content. These lines arise from a doubt among men regarding the movement of the stars: should they believe the Epicurean doctrine about the beginning and the end of the world (as indicated in 1212-4) or believe that the gods will guarantee an eternal duration of the universe (1215-7)? The authenticity of the verses has been questioned not only by Deufert in 1996 (pages 299-301) but also by Gerhard Müller,6 who removed the vv. 1215-7. In this regard, Deufert wrote: “1211-7 sind durch die Konstruktion dubiam mentem (1211)… ecquaenam (1212) … et simul ecquaean (1215) … eindeutig als eine einheitliche Periode konstruiert” (Deufert 1996, p. 301). Deufert in the Kommentar (p. 349), with mature and concise analysis, changes his previous idea and states that he no longer shares Müller’s arguments for the excision of three lines, because: “gewiss ist zu labentia (genauso wie zuvor in 1215 zu aeterna donata salute) grammatisch moenia mundi aus 1213 zu ergänzen, aber die moenia mundi stehen hier als pars pro toto für den mundus selbst: Genau wie diesen denkt sich die mens dubia auch dessen Grenzmauern unsterblich und in ruhig-gleitender Bewegung befindlich”. He concludes his demonstration to preserve vv. 1211-1217 by providing some examples taken from the ThLL that show the presence in the same sentence of labi and mundus.

Another controversial passage is 5, 1341-1349, where the logic is hard to follow but it is clear that Lucretius, or the interpolator, tries to explain why men could have done such incredible actions in war. The peculiarity of 1341-9 for Deufert lies in the fact that Lucretius allows doubts to remain. This is contrary to what he elsewhere sought to achieve: resolving doubts, removing fears, finding answers as close as possible to the truth (vera ratio). Although today it may seem bizarre, Housman (CR 42, 1928, 122-3) hypothesized that lines 1341-3 and 1347-9 were a sarcastic comment by Cicero, whom he accepted as Lucretius’s first editor. Bailey’s remarks on these verses are extreme: “this paragraph ... makes me wonder whether Jerome was not right, and Lucretius’ mind was from time to time deranged”. According to Deufert, the purpose of Lucretius’ work is to explain the nature of things, based on empirical and analogical demonstrations, without moralistic and catastrophic judgments or predictions about the future: “Lukrez ist weder geisteskrank noch der Nostradamus der Antike, sondern ein klar denkender Anhänger der epikureischen Philosophie”.7

The Kritischer Kommentar is a mine of information. The enormous merit of this text is its precise discussion—by words or by sections of the text—of almost all the verses of the entire DRN: this characteristic makes it an essential text for consultation. The commentary, detailed and rich in information, but never verbose, offers a lucid and effectively critical reading of the text and the history of its interpretation.

To conclude, the volume of Deufert leaves the impression, more disconcerting than one might expect, that there are numerous passages of the DRN that still require further treatment or are in a desperate state. He offers many solutions, but leaves almost as many questions. We may have to wait for a careful reading of his critical edition.


1.   Deufert, M. Pseudo-Lukrezisches im Lukrez: Die unechten Verse in Lukrezens „De rerum natura“ (Berlin; New York 1996),
2.   There are no contemporary sources on the biography of Lucretius or composition of the DRN apart from the witnesses of Cicero and Jerome.
3.   Butterfield, D. J. “Some Problems in the Text and Transmission of Lucretius” in R. Hunter—S. P. Oakley (edd.), Latin Literature and its Transmission. Papers in Honour of Michael Reeve. Cambridge 2016, 22–53.
4.   Erler, M.. “Lukrez und Apollonios Rhodios. Zur Frage des Proömiums zu Buch IV” in Kulte, Priester, Rituale—Beiträge zu Kult und Kultkritik im Alten Testament und Alten Orient: Festschrift für Theodor Seidl zum 65. Geburtstag, edited by S. Ernst—M. Häusl. St. Ottilien 2010, 473-482.
5.   Butterfield, D. J. The Early Textual History of Lucretius’ De rerum natura. Cambridge 2013, 273.
6.   Müller, G. “Die Problematik des Lukreztextes seit Lachmann” in Philologus 103, 1959, 53-86.
7.   Deufert, M. Pseudo-Lukrezisches im Lukrez: Die unechten Verse in Lukrezens „De rerum natura“. Berlin; New York 1996, 274. It is worth retracing the editorial choices that have divided Lucretian scholars over the last two hundred years. Lachmann reads sic instead of si in 1341 (conjectured by Marullus). Furthermore, he inverts the order of 1342-3 to 1343-2, which certainly brings more consistency to the content. However, Lachmann deletes 1344-6 on the grounds that they were written by a lector philosophus or an interpolator (interpolator irrisor). Munro deletes verses 1341-46 as a whole, agreeing with Lachmann about lines 1344-46. Giussani believes that these lines are a note written by Lucretius himself but at a later time. According to Ernout, the conclusion of this section of verses is quite disconcerting. Ernout follows Lachmann in expunging only 1344-6. Diels, like Lachmann, expunges 1344-6 as certainly interpolated, and believes that there is a gap before verse 1341, which he fills with the words sic miseri sero cognorunt damna ferarum. Martin transposes 1347-9 after 1340, but leaves 1341-6.

Read comments on this review or add a comment on the BMCR blog

Read Latest
About BMCR
Review for BMCR
Support BMCR

BMCR, Bryn Mawr College, 101 N. Merion Ave., Bryn Mawr, PA 19010