The study of logic was foundational to philosophical education in the late antique and early Byzantine periods and focused on Aristotle’s Categories, On Interpretation, and Prior Analytics 1.1-7. The place of Aristotle’s Sophistical Refutations within the study of logic is more difficult to determine. An interest in the Sophistical Refutations seems to have developed in the second and early third centuries AD (Galen and Alexander of Aphrodisias) and reached its zenith in the Alexandrian Neoplatonic commentators of the late fifth and sixth centuries (Ammonius, John Philoponus, and Elias), probably due to the increased attention given to paralogisms (fallacious arguments) in Neoplatonic education during this period.[1]
There is little evidence that a continuous commentary on the whole of the Sophistical Refutations was produced during the late antique and early Byzantine periods. By the middle Byzantine period, however, a large number of scholia relating to the interpretation of the Sophistical Refutations had been collected from various sources. In the first half of the twelfth century, these scholia were used by Michael of Ephesus to produce a commentary on the Sophistical Refutations.[2] Michael’s work became the principal source for all later Byzantine commentaries on the Sophistical Refutations.[3]
Gysembergh’s book reports the discovery of previously unknown fragments relating to the interpretation of the Sophistical Refutations. This includes thirty-one fragments of Herminus, thirty fragments of Alexander of Aphrodisias, and references to comments by Aspasius, Syrianus, and John Philoponus.[4] These exegetical remarks would date from the first half of the second century AD (Aspasius and Herminus), the early third century (Alexander), the first half of the fifth century (Syrianus), and the sixth century (Philoponus).
The fragments are given in Latin translation by Agostino Nifo (ca. 1470-1538) in his commentary on the Sophistical Refutations.[5] In the preface to that commentary, Nifo remarked that the book containing Herminus’s comments “that has come down to us has been cut up into many fragments” (qui pervenit ad nos, multis fragmentis sit mutilatus). Gysembergh assumes that Nifo is referring to a manuscript containing Herminus’ commentary that was mutilated and had become difficult to read, leading Nifo to repeatedly qualify his reporting with si recte intellexi or si recte percepi.[6] It seems more likely that Nifo’s manuscript contained scholia excerpted from earlier works. Nifo’s difficulty may have arisen from trying to understand excerpts that did not report the whole argument but had been cropped or cut short (common meanings of mutilatus).
As Gysembergh notes, most of the comments that Nifo attributed to Greek expositors are ascribed to Alexander and Herminus. Nifo often presents these two authors as commenting on the same passages in the Sophistical Refutations and agreeing in 14 cases, whereas in 10 cases Alexander criticizes Herminus and distinguishes his own position from that of his predecessor.[7] This raises the possibility that Nifo’s manuscript contained excerpts from a commentary of Alexander that reported and interacted with interpretations previously advanced by Herminus.[8] In this case, Herminus’ views may not have been available directly, but only through the mediation of Alexander.[9]
Gysembergh’s translation of Nifo’s Latin text is clear and accurate, as are the footnotes accompanying the translation. A few minor corrections might be suggested. On p. 16, prosequamur may not mean “attack,” but rather “pursue,” which provides a more appropriate contrast with the phrase sed ut fugiamus that follows. On p. 25 n.3, the reference to Michael of Ephesus is given twice. On p. 46, after “if it were put forth,” some text is missing, and one should supply “whether every dog barks and the answerer granted it”; and.” On p. 58, after “for no one will,” one should read “reprove” (redarguet) in place of “prove.” On p. 68 n.1, for “passion” read “passive.”
The fragments attributed to Alexander and Herminus closely follow the text of the Sophistical Refutations. Their aim is to explain the meaning of the text to students who may have studied rhetoric but had only an elementary knowledge of philosophy. As Gysembergh notes, more attention is given to the rules of dialectical disputation than to apodeictic reasoning.[10] As a result, about one-fifth of the fragments focus on Sophistical Refutations 17, which discusses the role of the answerer in dialectical disputation, and connect this with Aristotle’s remarks about dialectical disputation in Topics 8.
Gysembergh demonstrates that the fragments transmitted by Nifo were also known to Michael of Ephesus, who used them selectively in writing his commentary on the Sophistical Refutations. It is therefore useful to compare Nifo’s Latin translation of the fragments with the Greek text of Michael’s commentary. In cases where Michael simply reproduced the scholium with little adaptation, much of the Greek text of Nifo’s source can be plausibly reconstructed.[11] In other cases, Michael is not much help in reconstructing the Greek text, since he is inconsistent in his use of sources or even misunderstands them.[12]
Another challenge in recovering the original text is that Nifo tends to paraphrase the Greek authorities he cites.[13] Furthermore, in at least one case, Nifo’s rendering of his Greek source into Latin is shaped by Boethius’s Latin translation of the Sophistical Refutations.[14] Nifo is thus not directly rendering the Greek text of the scholium but there is an additional layer of mediation involved between the Greek text and Nifo’s Latin version. In some fragments, it is also difficult to determine which words belongs to Nifo’s source and which represent Nifo’s own interventions in the text.[15]
Gysembergh’s study concludes with a helpful summary of the content and philosophical significance of each fragment. He also notes that in discussing the text of the Sophistical Refutations, the fragments mention several interesting variant readings.[16] Two of the variant readings discussed by Alexander are found in a twelfth century Greek manuscript, Basileensis 54 (F II 21; designated u in the editions of Waitz and Ross).[17] Herminus and Alexander also offer some conjectural emendations to improve the sense of the text.[18]
Future research on the new fragments published by Gysembergh might consider two issues. First, it would be helpful to look for other attestations of these fragments in Greek and Latin texts. Nifo is a very late witness. Although Gysembergh’s argument that Michael of Ephesus made use of these fragments is likely correct, it would strengthen the case for authenticity if other evidence for the transmission and use of the fragments could be found.
Second, although the instruction in these new fragments is not directed toward advanced students, the fragments do shed some light on the specialized terminology used in late antique logic, and this might be further explored in subsequent research. For example, Herminus fr. 16 considers the fallacious arguments that can arise when a word has several meanings.[19] “Dog,” for instance, can denote a terrestrial animal, a fish (dogfish), and a star (the Dog Star).[20] Where disambiguation has not occurred, the question “Does every dog emit heat?” does not admit of being affirmed or denied because “dog” does not yet signify anything definite.[21] Only a conditional answer that makes a distinction can be given: “If the question is asked about the heavenly dog, it is so that every dog emits heat; if it is about the earthly one or the aquatic one, it is not so.” Herminus describes this answer as “conditional by disjunction,” a term apparently drawn from Stoic logic that is also attested once in Philoponus.[22]
Gysembergh’s book makes an important contribution to our understanding of the reception of the Sophistical Refutations in late antique philosophy and will be of interest to anyone studying that topic.
Notes
[1] Byard Bennett, “The Afterlife of Manichaeism in Neoplatonic Education,” in Manichaeism and Early Christianity: Selected Papers from the 2019 Pretoria Congress and Consultation, ed. Johannes van Oort, Nag Hammadi and Manichaean Studies 99 (Leiden: Brill, 2022), 401-431, at 427. The two sophisms published by Sten Ebbesen may also date from the sixth or seventh centuries; see Ebbesen, “The Un-Byzantine Byzantine on Two Sophisms,” in Aristotle and His Commentators: Studies in Memory of Paraskevi Kotzia, ed. Pantelis Golitsis and Katerina Ierodiakonou, Commentaria in Aristotelem Graeca et Byzantina 7 (Berlin: De Gruyter, 2019), 195-206.
[2] On Michael, see Pantelis Golitsis, “Michel d’Éphèse,” in Dictionnaire des philosophes antiques, ed. Richard Goulet, vol. 7 (Paris: CNRS Éditions, 2018), 609-616.
[3] Sten Ebbesen. Commentators and Commentaries on Aristotle’s Sophistici Elenchi: A Study of Post-Aristotelian Ancient and Medieval Writings on Fallacies. Vol.1: The Greek Tradition (Leiden: E.J. Brill, 1981).
[4] The comments of Syrianus are attributed to “Philoxenus.” Marinus (Life of Proclus 11) refers to Syrianus as the son of Philoxenus. Boethius also seems to know this tradition, indicating that the cognomen of Syrianus was Philoxenus (Commentary on Aristotle’s On Interpretation 2 [87.30-31]). Presumably the ascription to Philoxenus originated from a misunderstanding of the attribution ΣΥΡΙΑΝΟΥ ΤΟΥ ΦΙΛΟΞΕΝΟΥ that preceded the excerpt.
[5] Agostino Nifo, Expositiones Magni Augustini Niphi Medices Philosophi Suessani in Libros de Sophisticis Elenchis Aristotelis (Venice, 1534).
[6] Gysembergh, 32 (Herminus fr.8); 45 (Herminus fr. 15=Alexander fr. 15); 51 (Herminus fr. 18); 59 (Herminus fr. 21) with Gysembergh’s comments in vii; 13; 27 n.1; 42 n.5; 46 n.4.
[7] Gysembergh, 80.
[8] A similar question arises in Boethius Commentary on Aristotle’s On Interpretation 2.26, There, as John Magee has noted, “Boethius or his source sees an implicit repudiation of Herminus in Alexander’s view of the matter, for it is only after having introduced the latter that Boethius makes mention of the need to correct Herminus” (Boethius on Signification and Mind [Leiden: Brill, 1989] , 94; emphasis his).
[9] Gysembergh seeks to adopt a nuanced view, arguing initially that Nifo’s manuscript contained a commentary of Herminus (vii, 1), but afterwards conceding that at least some of Herminus’s views are known only through the mediation of Alexander: “In at least one case, Nifo explicitly states that Alexander quoted Herminos (fr. 14); this implies that while much of the Herminos material may have been independent, at least some of it was embedded within fragments from Alexander” (3).
[10] Gysembergh, 81.
[11] Gysembergh, 15 (Herminus fr. 1); 20 (Herminus fr. 2=Philoponus fr. 1); 33 (Alexander fr. 8); 39 (Alexander fr. 12).
[12] As Gysembergh (60 n.1) notes in discussing Alexander fr. 21, Michael sometimes artlessly joins together brief scholia that offer incompatible interpretations of a passage, without noticing or resolving the disagreements in the sources he uses. In other cases, Michael misunderstands the argument in his source and leaves out an important argument or conclusion; see, e.g., Gysembergh (66 n. 2) on Alexander fr. 22.
[13] Gysembergh, 95-96.
[14] See Gysembergh, 67 (Alexander fr. 23), where the Greek terms τέμνεται-καίεται-αἰσθάνεται are given the Latin equivalents secatur-uritur-operatur. The use of operatur to translate αἰσθάνεται is initially puzzling until one recognizes that Nifo is dependent on Boethius’s Latin translation of Sophistical Refutations 22, 178a12-13, and that by posita est ab interprete alia dictio, Nifo is referring not to himself but to Boethius.
[15] For example, the words Per quae…accepit in Alexander fr. 4=Herminus fr. 4 may belong to Nifo and not to Alexander or Herminus as Gysembergh (25 n.3) suggests. The discussion of εὐθύς in Herminus fr. 9=Alexander fr. 9 (Gysembergh, 34) may also be an intervention by Nifo.
[16] Gysembergh, 81.
[17] Gysembergh, 28 (Alexander fr. 6a); 78 (Alexander fr. 28).
[18] Gysembergh, 38 n.1; 58 n.4; 66 n.3.
[19] Gysembergh, 47 (Alexander fr. 16=Herminus fr. 16).
[20] Compare Clement of Alexandria, Stromateis 8.4.12.5; Dexippus, On Aristotle’s Categories 19.17-30.
[21] Compare Galen On Fallacies 2 (6.12-13).
[22] Gysembergh (47 n.2) notes that the corresponding Greek phrase κατὰ διάζευξιν ὑποθετικός is found in Philoponus Commentary on the Prior Analytics 245.4; compare Ammonius Commentary on Aristotle’s On Interpretation 67.5-19. Gysembergh observes that the phrase is also found in a paraphrase of the Prior Analytics ascribed to Themistius (CAG 23.3); that work, however, is now attributed to the Byzantine scholar Sophonias and reproduces the passage from Philoponus.