Mayo ————— Greek, Roman, and Byzantine Studies 62 (2022) 417–437 Article copyright held by the author(s) and made available under the Creative Commons Attribution License CC-BY https://creativecommons.org/licenses/by/4.0/ Herophilus and Common Opinions Andrew C. Mayo HOUGH THE THIRD-CENTURY BCE Alexandrian phy- sician Herophilus of Chalcedon is a pivotal figure in the history of medicine, what we know of him comes from the scattered testimony of later authors.1 Of this evidence one piece that has proved to be of great interest is the title of one of Herophilus’ lost works, namely Πρὸς τὰς κοινὰς δόξας, which we hear about from the second-century CE author Soranus. The title is usually translated as Against Common Opinions. This translation, however, is not required by the Greek. Though there are a number of titles with πρός followed by an accusative noun where the sense is clearly ‘against’, these cases involve individuals or groups of people. With abstract or otherwise im- personal nouns it almost always indicates a more general kind of specification. In this paper, I maintain that the same holds for Herophilus’ work and that the title does not mean “against common opinions,” but rather “regarding common opinions.”2 1 I cite Herophilus by the number of the fragment or testimonium in H. von Staden, Herophilus: The Art of Medicine in Early Alexandria (Cambridge 1989). All translations are my own. 2 I acknowledge that the point is not altogether conclusive. The parallels I adduce below are a mixed bag: none is quite equivalent to the title Πρὸς τὰς κοινὰς δόξας and they would seem at first to pull in different directions. The best parallels among other titles—titles for which a polemical context makes much better sense than for the present case—do tend in the direction of a hostile connotation; but the best parallels when our chief consideration is what “common opinions” ought to mean (above all in Aristotle) suggest rather a neutral connotation. A strictly philological analysis can only go so far, and so it is ultimately on the basis of what we know about Herophilus that “regarding common opinions” emerges as the better interpretation. T 418 HEROPHILUS AND COMMON OPINIONS ————— Greek, Roman, and Byzantine Studies 62 (2022) 417–437 After I have argued the case, I draw some interesting impli- cations of this revised title for our understanding of the scope and aims of this treatise. For, as I will show, the title “against common opinions” is not in line with what we know of this work and Herophilus’ method in general. The conventional translation “against common opinions” is generally taken to suggest that Herophilus denied the value of established opinion in scientific inquiry;3 however, the place of established opinion (δόξα) in science is a commonplace in Aristotelian scientific method, by which Herophilus seems to have been influenced.4 My reading of the title, on the others hand, suggests that Herophilus advocated a provisional acceptance of the claims of established opinion in science. 1. κοιναὶ δόξαι and the authenticity of the title Our primary evidence for the title of Herophilus’ work comes from a passage in Soranus of Ephesus’ Gynecology (Gyn. 1.27.2, 29.1 = T203–204 von Staden): Now, as Herophilus too mentioned in his work Regarding Common Opinions, of previous physicians some say that menstruation has a beneficial effect for both health and procreation. Themison, however, and most of our school say it is beneficial only for begetting children, while of the more distinguished physicians some say that is beneficial neither for health nor for producing children. But following different approaches, Herophilus and Mnaseas say that for some women menstruation is helpful for health, but for others it is harmful … But Herophilus says that 3 An anonymous referee points out that Πρὸς τὰς κοινὰς δόξας with the plural can refer just as well to some specific set of common opinions as to common opinions in general. I will explain later why it seems to me likely that Herophilus’ work addressed the problem of ἔνδοξα in Aristotelian science and dialectic, which has to do with the theoretical standing of common opinions in general. If I am right that “πρὸς X” flags a zētēma or problēma (see below), then it must have concerned itself with common opinions in general, though in the context of medicine. 4 For Aristotelian scholarship on this issue see n.45 below. ANDREW C. MAYO 419 ————— Greek, Roman, and Byzantine Studies 62 (2022) 417–437 menstruation is sometimes harmful for some women, since (he says) some are healthy without impediment when they are not menstruating; and often, on the contrary, they grow paler and weaker when menstruating and undergo the onset of disease; while other times in some cases it is helpful, so that, whereas previously pale and malnourished, later, after menstruation, they have a hearty color and are well-nourished.5 From Soranus’ testimony we learn that in this work Herophilus discussed prior medical opinions in relation to his own, for example whether menstruation is ever harmful to women.6 The fact that he speaks of the medical opinions of his predeces- sors has led most scholars to translate the work “Against Common Opinions.”7 In so doing, they tacitly assume that titles with ‘πρός 5 ἔνιοι µὲν οὖν τῶν ἔµπροσθεν, ὡς καὶ Ἡρόφιλος ἐν τῷ Πρὸς τὰς κοινὰς δόξας ἐµνηµόνευσεν, ἐπὶ συµφέροντι λέγουσιν γεγονέναι τὴν κάθαρσιν καὶ πρὸς ὑγείαν καὶ πρὸς παιδοποιίαν, Θεµίσων δὲ καὶ οἱ πλεῖστοι τῶν ἡµετέρων πρὸς µόνην παιδογονίαν, τινὲς δὲ τῶν ἐπισηµοτέρων οὔτε πρὸς ὑγείαν οὔτε πρὸς παιδοποιίαν. Ἡρόφιλος δὲ καὶ Μνασέας κατὰ διαφόρους ἐπιβολὰς τισὶ µὲν τῶν γυναικῶν πρὸς ὑγείαν ὠφέλιµον λέγουσιν εἶναι τὴν κάθαρσιν, τισὶ δὲ βλαβεράν … Ἡρόφιλος δὲ ποτὲ µὲν καί τισιν τῶν γυναικῶν βλαβεράν φησιν εἶναι τὴν κάθαρσιν, καὶ γὰρ ἀνεµποδίστως τινὰς ὑγιαίνειν µὴ καθαιροµένας καὶ πολλάκις τοὐναντίον καθαιροµένας ὠχροτέρας γίνεσθαι καὶ ἰσχνοτέρας καὶ παθῶν λαµβάνειν ἀφορµάς, ποτὲ δὲ καὶ ἐπί τινων ὠφέλιµον, ὥστε πρότερον ἀχροούσας καὶ ἀτροφούσας ὕστερον καὶ µετὰ τὴν κάθαρσιν εὐχροῆσαί τε καὶ εὐτροφῆσαι. 6 Cf. H. von Staden, “Rupture and Continuity: Hellenistic Reflections on the History of Medicine,” in P. J. van der Eijk (ed.), Ancient Histories of Medicine: Essays in Medical Doxography and Historiography in Classical Antiquity (Leiden 1999) 143–187, at 145. 7 F. Kudlien, “Herophilos und der Beginn der medizinischen Skepsis,” Gesnerus 21 (1964) 1–13, at 3; von Staden, Herophilus 373; T. Tieleman, “Dialectic and Science: Galen, Herophilus and Aristotle on Phenomena,” in P. H. van der Eijk et al. (eds.), Ancient Medicine in its Socio-Cultural Context II (Amsterdam 1995) 487–495, at 494. This may also be motivated by the observation at Ps.-Gal. Opt.sect. I 109 K. (= T54 von Staden) that Herophilus did not trust foolish opinions (µὴ δόξαις ἠλιθίαις ἀποπιστεῦσαι). Tieleman (494) takes this passage to mean that Herophilus distinguished between “observable phenomena, which he examined, and opinions (doxai), in which 420 HEROPHILUS AND COMMON OPINIONS ————— Greek, Roman, and Byzantine Studies 62 (2022) 417–437 + accusative’ tend to be works directed ‘against’ opponents.8 Plutarch’s On Common Conceptions against the Stoics gives us a very good example of this usage. Here πρός clearly has a hostile connotation, and the group towards which Plutarch is hostile, the Stoics, is marked by the accusative. While this sense of πρός is common when referring to individuals or groups, it is almost never found when referring to abstract or non-personal nouns. Plutarch himself illustrates this well: at On Common Conceptions against the Stoics 1063D10, when discussing Chrysippus, he men- tions the idea that “people philosophize with reference to common opinions” (πρὸς τὰς κοινὰς ἐννοίας φιλοσοφοῦσι).9 This certainly does not seem to mean “against common con- ceptions” but “with reference to common conceptions.”10 This seems to hold generally for constructions of this kind. ___ he did not put his trust.” Yet the language here is consistent with a range of interpretations. Doubtless Kudlien’s influential paper is partially responsible for the standard translation “against common opinions”: the hostile sense of πρός best suits his characterization of Herophilus as a skeptic critical of established beliefs (so with Sextus Emp. Pyr. 2.48.9, Math. 9.44.1). 8 An anonymous referee notes that such hostile uses of πρός with the accusative are often influenced by the language of the lawcourt (LSJ s.v. C.I.4, with titles of judicial speeches being especially relevant, though the dominant expression takes the form κατά τινος). This is part of the reason why πρός in a hostile sense naturally takes a personal object, πρός τινα or πρός τινας (note, however, that even with individuals the sense need not be hostile; cf. LSJ s.v. C.I.5). Titles like Serapion’s πρὸς τὰς αἱρέσεις (discussed further below) are best understood this way as well: it is as though Serapion is facing the sects and addressing them. The other chief context in which πρός naturally takes on a hostile connotation involves fighting (cf. Cambridge Greek Lexicon s.v. B.4). Although in English “to fight against poverty,” e.g., is idiomatic, abstract objects are much less natural in Greek (on this point cf. J. D. Denniston, Greek Prose Style [Oxford 1952] 23–40) and would probably resort to personification. 9 I thank an anonymous referee for this illuminating parallel. 10 Cherniss translates the phrase as “with a view to the common con- ceptions”: H. Cherniss, Plutarch’s Moralia XIII.2 (Cambridge [Mass.] 1976) 695; cf. De Stoic. repug. 1042D–E. ANDREW C. MAYO 421 ————— Greek, Roman, and Byzantine Studies 62 (2022) 417–437 Herophilus’ work therefore is more naturally construed as Regarding Common Opinions. What then does it suggest concern- ing its contents? κοιναὶ δόξαι suggests his concern with method, with the things that are presupposed in medical inquiry. Hein- rich von Staden11 brings up in this connection the notorious ἔνδοξα of Aristotle’s Topics (1.1 100b21–23)12 as well as the fact that Aristotle sometimes uses the same expression we meet with in the title, κοιναὶ δόξαι, ostensibly with reference to the same class of ἔνδοξα held to be true by all or most experts in a given domain. Soranus’ testimony, according to which Herophilus evaluates the established views of medical experts, is consistent with Herophilus having had something like Aristotle’s κοιναὶ δόξαι in mind. Even if the book title may not have been assigned by Herophilus himself, an issue raised by von Staden,13 it must at least have been a reasonable description of either a work or part of a work written by Herophilus. I believe, however, that the title probably goes back to Herophilus: it is common 11 In Ancient Histories of Medicine 146–147. 12 On the controversy surrounding Aristotelian ἔνδοξα see D. Frede, “The Endoxon Mystique: What Endoxa Are and What They Are Not,” OSAPh 43 (2012) 185–215. 13 In Ancient Histories of Medicine 162 n.47. In support of the view that the title is not by the author one might draw attention to πρὸς τὰς Κνιδίας γνώµας, which Galen associates most closely with a portion (the proem) of the Hippocratic On Regimen in Acute Diseases (In Hipp. de vict. acut. XV 452.14 K.). Yet it is clear that πρὸς τας Κνιδίας γνώµας was used of the work as a whole (a point discussed below). Since the most plausible scenario is that Herophilus, a well-known Alexandrian intellectual, published the work in his own lifetime, and since HVA is not a particularly doxographical text, the fact that some assigned On Regimen a title structurally similar to Πρὸς τὰς κοινὰς δόξας does not particularly bolster the alternative hypothesis that Herophilus’ work was entitled by assimilation to the doxographical tra- dition. If the title was assigned by later Herophileans, it was surely chosen as a suitable description of the contents of the work. 422 HEROPHILUS AND COMMON OPINIONS ————— Greek, Roman, and Byzantine Studies 62 (2022) 417–437 neither in doxographical nor in medical writing;14 in fact, both in its use of πρός with an abstract object and in its reference to κοιναὶ δόξαι, it is unusual enough to render unlikely the hypothesis that the title was conceived and added to the work later. The Peripatetic character of a work dealing with κοιναὶ δόξαι15 points rather in the direction of Herophilus himself.16 I will say more about this shortly. Herophilus was a prominent figure in Alexandrian intellectual culture17 and, by this time, prominent intellectuals who wanted their works to be read probably did, as a rule, have some concern for the titles of their works, even if this did not guarantee the success and survival of their designation.18 14 I have not been able to find any titles of the form πρὸς … δόξας other than the Hippocratic work discussed in n.13 and below. Other kinds, which do seem to have been in use in medical doxography, include the πρὸς τὰς αἱρέσεις of Serapion (discussed below), the ἀρέσκοντα of the Herophilean Alexander Philalethes (De diff. puls. VIII 726.11 K. = T24 von Staden), and the περὶ τῆς ἐµπειρικῆς αἱρέσεως of Heraclides (De libr. propr. XIX 38.19–20). 15 Noted by von Staden (Herophilus 395). 16 As often with ancient works, the title may have been taken from its opening words. Cf. E. Nachmanson, Der griechische Buchtitel (Gothenburg 1941) 7–8. So von Staden, in Ancient Histories of Medicine 162 n.47. 17 Aside from his enormous influence in medical thought, there is his ap- parent association with Diodorus Cronus in an anecdote (Sext. Emp. Pyr. 2.245 = T15 von Staden) that reflects his contemporary prominence. Note also the reference to Herophilus’ discoveries in Callimachus’ Hymn to Artemis (see H. Oppermann, “Herophilos bei Kallimachos,” Hermes 60 [1925] 14– 32) and perhaps his Hymn to Delos (see G. Most, “Callimachus and Hero- philus,” Hermes 109 [1981] 188–196). Consider also Herophilus’ apparent connections with the Ptolemaic court (Cels. Med. pref. 23 = T63a von Staden), on which see F. Schironi, “Enlightened King or Pragmatic Rulers? Ptolemaic Patronage of Scholarship and Sciences in Context,” in P. Bosman (ed.), Intellectual and Empire in Greco-Roman Antiquity (New York 2019) 1–29, at 9. 18 Nachmanson argues that earlier authors (like those of the older Hippocratic texts) did not, in general, assign their works titles, a practice that begins with the Sophists (Buchtitel 31–34; cf. 9–13 on the titles of the ANDREW C. MAYO 423 ————— Greek, Roman, and Byzantine Studies 62 (2022) 417–437 If the title was not Herophilus’ own, one plausible explana- tion is that it arose from the assimilation of his work into the doxographical tradition; alternatively, a portion of one of his works may have taken on this title in transmission (both alter- natives are simultaneously possible). That there is not much evidence for other titles in common use with πρὸς and δόξας makes these hypotheses less appealing. I have found only one title of similar form, which designates a work on the opinions of the Knidian physicians attributed by Julius Pollux to Hip- pocrates (Ἱπποκράτης δὲ ἐν τῷ πρὸς τὰς Κνιδίας δόξας, Onom. 10.87.5). Because its usual form, however, seems to have been Πρὸς τὰς Κνιδίας γνώµας,19 the parallel to Herophilus’ title is not as close as it seems. Another important difference between them is that it involves a group of people (the Knidian phy- sicians), precisely the type of object with which πρός more naturally bears a hostile connotation. Galen makes clear that this was an alternative title for On Regimen in Acute Diseases (ἡ δίαιτα ἐπὶ τῶν ὀξέων νοσηµάτων), the one he himself ordinarily uses.20 It must be noted that on the whole On Regimen is hardly ___ works of Protagoras and Plato). It is otherwise with the surviving treatises of Aristotle, as Nachmanson (at 13) observes, given his students’ role in editing and disseminating his treatises. From the examples adduced by Nach- manson one may infer behind this shift a greater concern for the author’s public persona. 19 Galen, who thinks that Hippocrates’ authorship was plausible (and mentions competing attributions to Euryphon, Thessalus, and Polybus), refers to it with Πρὸς τὰς Κνιδίας γνώµας (Diffic. resp. VII 891.6–10 K.; cf. PHP IX 6.5.5; In Hipp. de fract. XVIIIb.467.6; In Hipp. de vict. acut. XV 452.14). This variant is not surprising, since δόξα and γνώµη must have been deemed interchangeable by later writers. Πρὸς τὰς Κνιδίας δόξας must have been derived from Πρὸς τὰς Κνιδίας γνώµας (rather than the reverse), not least because the title of the Knidian Sentences to which it ostensibly responds has γνώµη (e.g. Gal. In Hipp. epid. XVIIa.886.4–5 K.). 20 Multiple, different titles should not surprise us if we accept that the works of the Hippocratic corpus were assigned titles in the Hellenistic period, probably by Alexandrian scholars, and for the most part lacked 424 HEROPHILUS AND COMMON OPINIONS ————— Greek, Roman, and Byzantine Studies 62 (2022) 417–437 a polemic against the Knidian Sentences, even if its proem is hostile. It is not devoted to taking down the Knidian Sentences in the same way, e.g., that Plutarch’s Πρὸς Κωλώτην is a polemic against Colotes. Perhaps it was given this title because it re- sponds to the Knidian Sentences, even though it is not concerned throughout with the Knidian Sentences.21 Even if ex hypothesi the translation Against the Knidian Sentences best captured the sense of the Greek (the title may have been narrowly motivated by the proem),22 the salient fact remains that the “Knidian sentences” are those of Knidian physicians. Therefore, if πρός can have a hostile sense here, it is because talking about Knidian opinions entails talking about Knidian physicians.23 All things con- sidered, this parallel hardly settles the interpretation of Hero- philus’ title.24 ___ them previously (so Nachmanson, Buchtitel 18). 21 It is of interest that the Knidian physician Euryphon was considered a possible author of On Regimen (that is, of Πρὸς τὰς Κνιδίας γνώµας). Galen mentions elsewhere (In Hipp. epid. XVIIa.886.4–5 K.) that some attributed a Knidian Sentences to Euryphon (κἀν ταῖς Κνιδίαις γνώµαις). The Knidian Sentences seem to have reported views of the Knidian physicians (cf. W. D. Smith, “Galen on Coans versus Cnidians,” Bulletin of the History of Medicine 47 [1973] 574–580). That the Knidian physician Euryphon could be placed on either side of this suggests viewing On Regimen as participating in a pro- ductive dispute. On this view, Πρὸς τὰς Κνιδίας γνώµας would portray it as an inquiry into the subject matter of the Knidian Sentences. 22 An anonymous referee rightly points out that this was a common prac- tice, including with Hippocratic texts such as On Humors and On the Nature of Bones. 23 Accordingly, though Galen at one point (In Hipp. de vict. acut. XV 452.14 K.) understands the title as having hostile connotations with reference to the proem of the work, he explains this as involving criticism “against the Knidian physicians” (ἐδείχθη γὰρ ἐν τῷ δευτέρῳ µηκέτι πρὸς τοὺς Κνιδίους ἰατροὺς τὸν λόγον ποιούµενος, “for he was shown as no longer arguing against the Knidian physicians in the second book,” 452.13–15, 453.1). 24 Perhaps Herophilus conceived of doctors as a group in writing on “common opinions,” but the connection with a distinct group of people is ANDREW C. MAYO 425 ————— Greek, Roman, and Byzantine Studies 62 (2022) 417–437 When it comes to interpreting the scientific significance of κοιναὶ δόξαι, a passage from Aristotle’s Physics is especially sug- gestive (Phys. 4.6, 213a19–22): In the same way, just as with place, we should also consider it to be the natural scientist’s task to investigate void, whether it exists or not, and how it exists, and what it is; because it involves nearly equal belief and disbelief because of what is held about it. For those who discuss it posit the void as a sort of place and vessel, and it is thought to be full whenever it contains the kind of mass that it is capable of receiving, but void whenever it is deprived, on the supposition that the same thing is void, full, and place, but their being is not the same. We should begin the inquiry by grasping the things that are said by those who claim there is void, and in turn what things are said by those who deny there is one, and thirdly the common opinions about them.25 The approach Aristotle takes in this passage to investigating the existence and character of void begins, as often, by surveying the contrasting claims experts have previously made, and then goes on to consider the “common notions” about them (Ari- stotle has earlier in this book taken a similar approach to the investigation of place, 4.1, 208a27–209a30).26 To judge from the Soranus passage on menstruation, Herophilus seems to ___ surely much looser in this case, all the more so if “common opinions” has something like its Aristotelian sense. 25 τὸν αὐτὸν δὲ τρόπον ὑποληπτέον εἶναι τοῦ φυσικοῦ θεωρῆσαι καὶ περὶ κενοῦ, εἰ ἔστιν ἢ µή, καὶ πῶς ἔστι, καὶ τί ἐστιν, ὥσπερ καὶ περὶ τόπου· καὶ γὰρ παραπλησίαν ἔχει τήν τε ἀπιστίαν καὶ τὴν πίστιν διὰ τῶν ὑπολαµβανοµένων· οἷον γὰρ τόπον τινὰ καὶ ἀγγεῖον τὸ κενὸν τιθέασιν οἱ λέγοντες, δοκεῖ δὲ πλῆρες µὲν εἶναι, ὅταν ἔχῃ τὸν ὄγκον οὗ δεκτικόν ἐστιν, ὅταν δὲ στερηθῇ, κενόν, ὡς τὸ αὐτὸ µὲν ὂν κενὸν καὶ πλῆρες καὶ τόπον, τὸ δ’ εἶναι αὐτοῖς οὐ ταὐτὸ ὄν. ἄρξασθαι δὲ δεῖ τῆς σκέψεως λαβοῦσιν ἅ τε λέγουσιν οἱ φάσκοντες εἶναι καὶ πάλιν ἃ λέγουσιν οἱ µὴ φάσκοντες, καὶ τρίτον τὰς κοινὰς περὶ αὐτῶν δόξας. 26 On this passage see G. E. L. Owen, “Tithenai ta phainomena,” in S. Man- sion (ed.), Aristote et les problèmes de méthode (Louvain 1961) 83–103, repr. G. E. L. Owen, Logic, Science and Dialectic: Collected Papers in Greek Philosophy (Ithaca 1986) 239–251, at 241–242. 426 HEROPHILUS AND COMMON OPINIONS ————— Greek, Roman, and Byzantine Studies 62 (2022) 417–437 have done something similar, if not necessarily the same: he mentioned previous physicians who held menstruation to be beneficial for health and procreation, while he himself held a more nuanced position. All this very much suggests that Herophilus began by surveying the opinions held by earlier authorities and then tried to synthesize them, or at least pre- sented his own as accounting for what was true in the opinions of his predecessors. Soranus mentions later physicians who held that men- struation was not beneficial for health, but affirmed that it was for producing children (most of the Methodist school of medicine); and he adds that a smaller number of Methodists deemed menstruation useful neither for health nor for pro- ducing children. It is not clear whether Herophilus was aware of these positions specifically (the Methodists may be picking up on an older tradition), or if he only surveyed the views of those who claimed menstruation to be beneficial on both fronts. If Herophilus was only aware of those who held it to be bene- ficial,27 then in fact what we find here is Herophilus preserving what he took to be the “common opinion”: he agrees with his predecessors in taking menstruation to be beneficial for producing children, and also agrees that in many cases it is beneficial for health as well. In this case, however, he would have qualified this view by observing cases where menstruation is harmful to health; yet he could have done so without disputing that his predecessors had observed it to be beneficial. Herophilus emerges much more in line with the common opinion he was considering than the later Methodists. If cor- 27 It is of course also possible that Herophilus did mention other views of menstruation as harmful or as neutral; it is possible that Herophilus in some way combined prior observation of menstruation as beneficial with prior observation of menstruation as harmful. But the point stands that the prior opinion that Soranus tells us Herophilus discussed is mostly in line with the opinion that Soranus tells us Herophilus held himself. ANDREW C. MAYO 427 ————— Greek, Roman, and Byzantine Studies 62 (2022) 417–437 rect, this interpretation of the evidence weighs against taking πρός as ‘against’. Instead it suggests a nuanced engagement with this common opinion that in the end preserves the core of his predecessors’ consensus. This is the sort of approach we find in Physics 4.6, one that tries to find and hold on to those elements in previous views that are well-founded. The affinity in method and terminology (the shared recourse to κοιναὶ δόξαι) between Herophilus and Aristotle lends additional sup- port to considering the title authentic. There are no other testimonia or fragments of Herophilus explicitly connected with this work. But there is one additional piece of relevant evidence in what Herophilus is reported to have said about the healing of wounds of different shapes (Cassius Iatrosophist Problemata 1 1.144 Ideler = T236 von Staden): Why do round wounds heal with more difficulty than others? The Herophileans assign the cause of this by making use of geo- metrical demonstration: they say that the circular shapes of these wounds appear to be small in circumference, but are not in fact like this. Rather, they have surfaces larger in area than they seem. And that which is larger needs more time for scarring over. Therefore, it is to be expected that wounds of this kind seem to heal with more difficulty, even if they also seem small: in truth they are not so, but they are larger. Asclepiades evaded this entirely. If when there is a round wound present someone made a cross-incision of the bodies lying around it, in such way that the shape of the wound becomes more elongated from the cross-incision, scarring over would take place more quickly: and this is the opposite of what Herophilus thinks. For if the size of the wound, as they say, is the cause of the difficulty in healing it, since the same size is present and another is added from the cross-incision these wounds should rather become more difficult to heal.28 28 διὰ τί τὰ στρογγύλα ἕλκη δυσαλθέστερα καθέστηκε τῶν ἄλλων; οἱ µὲν οὖν Ἡροφίλειοι τὴν αἰτίαν ἀποδιδόασι γεωµετρικῇ χρώµενοι ἀποδείξει· φασὶ γὰρ 428 HEROPHILUS AND COMMON OPINIONS ————— Greek, Roman, and Byzantine Studies 62 (2022) 417–437 From other sources we learn that the slower healing of round wounds drew attention not only after Herophilus (since Asclepiades is mentioned) but also before him. Aristotle mentions it in the Posterior Analytics (1.13, 79a10–16), when he contrasts the physician’s knowledge of the fact with the geometer’s knoweldge of the reason. The Hippocratic corpus is aware of the importance of the skin drawing together for the healing of wounds (Of Places in Man 38, VI 329 Littré),29 which could supply another explanation that is broadly akin to the one Cassius attributes to Asclepiades. Herophilus’ reasoning seems rather to have been that circular wounds have the smallest ratio of periphery to surface area, something the geometer can attest. While the point is speculative, this might be an instance of the sort of common opinion that Herophilus addressed in his lost work: Aristotle gives the impression that the consensus of physicians was that round wounds heal more slowly; Herophilus seems to have accepted this common opin- ion and proposed his own explanation of it. This lends further support to the view that in his lost work on common opinions Herophilus engaged κοιναὶ δόξαι in medicine in a more con- structive than destructive fashion, even if nothing positively ties this testimony with that work. ___ ὅτι τὰ κυκλικὰ σχήµατα τῶν ἑλκῶν µικρὰ µὲν φαίνεται τῇ περιοχῇ, οὐ τοιαῦτα δ’ ἐστίν, ἀλλ’ ἔχει τῇ δυνάµει µείζονα τὰ ἐµβαδὰ ἤπερ φαίνεται. τὸ µεῖζον δὲ πλείονος χρόνου δεῖται πρὸς τὴν ἐπούλωσιν· ὥστε εἰκότως τὰ τοιαῦτα ἕλκη φαίνεται δυσαλθῆ, εἴ γε καὶ µικρὰ φαίνεται· κατὰ δὲ τὸ ἀληθες οὐχ οὕτως ἔχει, ἀλλ’ ἐστὶ µείζονα. τοῦτο δὲ περικειµένως διεκρούσατο Ἀσκληπιάδης· εἴ τις στρογγύλου ἕλκους ὑποκειµένου ἐπιδιέλῃ τὰ παρακείµενα σώµατα, ὥστε ἐκ τῆς ἐπιδιαιρέσεως γενέσθαι ἐπιµηκέστερον τὸ σχῆµα τοῦ ἕλκους, θᾶττον ἂν γένοιτο ἡ ἐπούλωσις· τοῦτο δ’ ἐναντίον τῷ τοῦ Ἡροφίλου ἀρέσκοντι. εἰ γὰρ τὸ µέγεθος τοῦ ἕλκους, ὡς αὐτοί φασιν, αἴτιον γίνεται τῆς δυσθεραπευσίας, ἐχρῆν τοῦ αὐτοῦ ὑποκειµένου µεγέθους καὶ ἑτέρου προσγινοµένου ἐκ τῆς ἐπιδιαιρέσεως, µᾶλλον γίνεσθαι δυσιατότερα ταῦτα τὰ ἕλκη. 29 Cf. G. Majno, The Healing Hand (Cambridge [Mass.] 1975) 154. ANDREW C. MAYO 429 ————— Greek, Roman, and Byzantine Studies 62 (2022) 417–437 2. The meaning of πρός When it does not go with a named individual, the usual meaning of πρός with the accusative in scholarly30 and intel- lectual contexts (e.g. Pl. Plt. 306A10 or Arist. Cael. 307a8) is the neutral, non-polemical one.31 This meaning suits Herophilus’ aim of relating his own views to those of his predecessors (as discussed below). In scholarly usage πρός with accusative is regularly used to indicate an object of inquiry, a zētēma.32 Accordingly, it is certainly possible to render the title of his work “with reference to common opinions” or “in response to common opinions.” Its content would have elucidated where the author stood in relation to common (medical) opinions. Several reasons make this preferable to “against common opin- ions.” First, as noted above, when not used with a personal 30 Cf. R. Nünlist, “Topos didaskalikos and anaphora—Two Interrelated Principles in Aristarchus’ Commentaries,” in M. R. Niehoff (ed.), Homer and the Bible in the Eyes of Ancient Interpreters (Leiden 2012) 113–126, at 121–124; F. Schironi, The Best of the Grammarians: Aristarchus of Samothrace on the Iliad (Ann Arbor 2018) 703 n.249. 31 An anonymous referee points out that in a therapeutic context (especially in pharmacology) πρός with a malady can often be translated ‘against’ (although with a treatment in mind it may also mean ‘with refer- ence to’ or ‘directed at’ the malady). Πρὸς δυσωδίαν µυκτήρων, Πρὸς τὰς ἐκ τόκου ῥαγάδας, and Πρὸς τὰ ἐν τῇ κεφαλῇ ἐκβράσµατα are three examples from the list of remedies in Crito’s lost work on cosmetics (Galen On Compound Drugs According to Places 1.3, XII 446.14–449.7 K.). Yet even here some instances support the neutral sense of πρός, either with the condition sought or addressed by a given treatment or with the person or bodily part the treatment is to act on (bodily part in view at XII 387.2 and 432.14–15; a person at XIII 27.15). Galen also uses πρός for what is done with a view to prognosis and treatment (πρὸς τὴν τῶν ἀποβησοµένων πρόγνωσιν, ἴασίν τε τῶν παρόντων, “with regard to the prognosis of things that will ensue, and the healing of things already present,” XII 442.2). Thus, in a medical context πρός with the accusative flags the target or goal of a course of action, whether an outcome, a malady, a part of the body, or a type of patient. 32 So Schironi, The Best of the Grammarians 703 n.249. 430 HEROPHILUS AND COMMON OPINIONS ————— Greek, Roman, and Byzantine Studies 62 (2022) 417–437 name πρός with accusative is not typically polemical; and second, when ‘against’ is the better choice but the preposition does not govern a personal name the wording makes this clear (e.g. by its use with ἀντιλέγω vel sim., as in Gal. De usu part. III 316.5 K.). We do find Sextus Empiricus using πρὸς δόξας polemically, but this seems rather a contextual connotation than a denotation: a Pyrrhonian skeptic who discusses his posi- tion vis-à-vis established opinions should be expected to attack them (Pyr. 2.48.9, Math. 9.44.1). The normal meaning of πρός with δόξας is ‘in relation to’ (see examples below). Perhaps the Empiricist Serapion of Alexandria was similar to Sextus.33 Galen wrote two outlines of Empiricism on Serapion’s work πρὸς τὰς αἱρέσεις (De libr. propr. XIX 38.13–14 K.). An Empiricist, just like a Pyrrhonian skeptic, might be expected to attack established doctrines while articulating his position in relation to them. I view this as the most compelling evidence against my interpretation of the title of Herophilus’ work. Here it is quite plausible that in a title πρός with an impersonal object bears a polemical sense. Yet it is also possible that the sense of πρὸς τὰς αἱρέσεις is “with reference to the sects,” so that Serapion wrote treatises in response to (not against) the sects. Furthermore, to judge from Galen Serapion seems to have outlined his own Empiricist approach in these books. This sug- gests a combination of constructive and polemical engagement. Most significant for my argument, however, is that specific groups of people are in view under the term αἱρέσεις. If ‘against’ is preferred, given that Serapion must have gone through specific sects, once again the choice fits nicely under the well-defined category of hostile πρός with people. It is true that any abstract idea may be set in relation to the people who hold it—there is perhaps not a great difference between speak- 33 On Serapion see G. Cambiano, “Philosophy, Science and Medicine,” in K. Algra et al. (eds.), The Cambridge History of Hellenistic Philosophy (Cam- bridge 1999) 585–613, at 605. ANDREW C. MAYO 431 ————— Greek, Roman, and Byzantine Studies 62 (2022) 417–437 ing against an idea and speaking against those who hold it. But I submit that “sects” much more concretely point to groups of people than “opinions.” This is the more so with “common opinions.” Although in its Aristotelian sense experts are in view, Aristotle’s usage strongly suggests a neutral rather than a polemical sense.34 A review of other titles with the form πρός plus accusative shows that they are plausibly neutral or else involve individuals.35 At the fundamental level of denotation, πρός with accusative arguably bears the sense ‘with reference to’, but in certain contexts it may connote ‘against’. Only with distinct people does this connotation cross into the denotative and take an un- ambiguously polemical acceptation.36 Consider, in closing, Herophilus’ remarks on Hippocrates’ Prognostic, twice mentioned by Galen: διασκέψασθαι περὶ τῶν ὑπὸ Ἡροφίλου πρὸς τὸ Προγνωστικὸν Ἱπποκράτους ἀντειρηµένων, “to examine the things said polemically by Herophilus against (πρός) the Prognostic of Hippocrates,” and τὰ κακῶς ὑπὸ Ἡροφίλου γεγραµµένα πρὸς τὰς Ἱπποκράτους προγνώσεις, “the things written 34 Cf. in particular Cael. 2.13, 293a17, and 3.7, 306a5–9 (discussed below). To put the point simply, the phrase is Aristotelian but to write “against common opinions” is un-Aristotelian. 35 Titles of numerous works prior to Galen are conveniently listed in his On My Own Books (De libr. propr.). 36 This is clear in Homeric scholia. To refer to the questions surrounding Pylaemenes, e.g., we find πρὸς τὰ περὶ Πυλαιµένους, πρὸς τὸν Πυλαιµένεα, or πρὸς τὰ Πυλαιµένους (schol. Il. 1.320a, 2.517c, 2.837–838, 4.295, and 15.515a). ἡ ἀναφορά with πρός is found pointing to an object of inquiry (e.g. schol. Il. 12.17a) and with the names of individual scholars where it should mean ‘against’ (e.g. ἡ δὲ ἀναφορὰ πρὸς Ζηνόδοτον, schol. Il. 3.339). I owe these examples to an unpublished paper by Francesca Schironi on the use of πρός in Homeric scholia. Cf. further C. M. Schroeder, “A New Monograph by Aristarchus?” JHS 127 (2007) 138–141; Nünlist, in Homer and the Bible 121–124; Schironi, The Best of the Grammarians 703 n.249. Such usages are flagged by distinct wording suggestive of the specification of πρός (as ἀντιλέγω before). 432 HEROPHILUS AND COMMON OPINIONS ————— Greek, Roman, and Byzantine Studies 62 (2022) 417–437 by Herophilus in a hostile spirit against the prognoses of Hip- pocrates” (Gal. In Hipp. Progn. 1.4, XVIIIb 16.5 K. = T32–33 von Staden). From Caelius Aurelianus’ translation of Soranus’ work on chronic diseases we learn that Herophilus wrote a book (liber) on this subject (quem ad Hippocratis Prognosticum scripsit, Tardae passiones 4.8.113 = T31 von Staden). It seems likely, therefore, that there was a Herophilean work known as πρὸς τὰς Ἱπποκράτους προγνώσεις. Two points are relevant here. First, that the object of πρός, though not a personal name, occurs in close connection with a personal name. Second, that the work in fact need not have been written in a uniformly hostile spirit: Galen gives more weight to Hippocrates’ than to Herophilus’ authority; even if he does not always yield to Hippocratic authority, he may have taken a failure to show suitable deference for hostility.37 The evidence here (see T261–275) does suggest that Herophilus’ polemical stance towards Hip- pocrates may be largely Galen’s misreading of his intellectual independence. From the puzzling Hippocratic use of both πρόγνωσις and πρόρρησις, Herophilus seems to have derived an epistemological distinction between types of prognosis, namely that πρόγνωσις refers to reliably accurate prognosis while πρόρρησις refers to a more approximate, less reliably accurate kind of prognosis.38 37 W. D. Smith, The Hippocratic Tradition (Ithaca 1979) 191–193, argues that there was no such work of Herophilus πρὸς τὸ Προγνωστικὸν Ἱππο- κρατηους and that Galen was inflating some of Herophilus’ remarks on common opinions into a full-blown polemic. The attribution in Caelius’ translation of Soranus would simply be a doxographical distortion. But von Staden (Herophilus 442) is probably right in viewing such denials as special pleading. Smith’s suspicion that the polemic itself might be Galen’s distor- tion, however, strikes me as well-founded. Von Staden (429–432) concedes that Herophilus’ work may have been not so much “general Streitschrift” as “partially polemical exegesis” (431). 38 Gal. In Hipp. prorrh. 5.9.2 CMG = T262 and T264 von Staden. Cf. von Staden, Herophilus 431. ANDREW C. MAYO 433 ————— Greek, Roman, and Byzantine Studies 62 (2022) 417–437 Since there are not, as I have shown, any good parallels to lend support to “against common opinions” as the translation of πρὸς τὰς κοινὰς δόξας, and there is no connection to distinct individuals in this expression, the default English rendering of the title should be “with reference to common opinions.” 3. The significance of the title In this section I turn to a discussion of what was involved in Herophilus’ speaking “in reference to” common opinions. A well-known passage on the relationship between phenomena and preexisting belief in Aristotle’s On the Heavens provides an important parallel. Aristotle attributes failure to make claims that fit the phenomena to explanations of things with reference to already determined opinions (πρός τινας δόξας ὡρισµένας) rather than beginning from first principles.39 In an earlier passage Aristotle complains of those who inquire “not with reference to the phenomena” (οὐ πρὸς τὰ φαινόµενα τοὺς λόγους καὶ τὰς αἰτίας ζητοῦντες, Cael. 2.13, 293a17) but rather they pull the phenomena in line with their beliefs (ἀλλὰ πρός τινας λόγους καὶ δόξας αὑτῶν τὰ φαινόµενα προσέλκοντες, 293a26–27). It seems likely that Herophilus used the expression πρὸς τὰς κοινὰς δόξας, “with reference to common opinions,” with an eye to scientific method: for one thing, as already noted, κοιναὶ δόξαι are integral to Aristotle’s method in dialectic and science. Furthermore, Herophilus seems to have discussed scientific method in Aristotelian terms. This is most evident in what he said about beginning from the phenomena: λεγέσθω δὲ τὰ φαινόµενα π[ρ]ῶτα, καὶ εἰ µὴ ἔστιν πρῶτα, “let the phenomena be 39 συµβαίνει δὲ περὶ τῶν φαινοµένων λέγουσι µὴ ὁµολογούµενα λέγειν τοῖς φαινοµένοις. τούτου δ’ αἴτιον τὸ µὴ καλῶς λαβεῖν τὰς πρώτας ἀρχάς, ἀλλὰ πάντα βούλεσθαι πρός τινας δόξας ὡρισµένας ἀνάγειν, “It happens that when they talk about the phenomena what they say does not accord with the phe- nomena. The cause of this is their not properly taking account of first prin- ciples, but rather wanting to refer everything to some already-determined body of opinions” (Cael. 3.7, 306a5–9). 434 HEROPHILUS AND COMMON OPINIONS ————— Greek, Roman, and Byzantine Studies 62 (2022) 417–437 said to be first, even if they are not first” (Anon.Londin. 21.22–23 = T50a von Staden; cf. X 107 K. = T50b von Staden). Since Herophilus was expressly concerned with scientific method in Aristotelian terms and κοιναὶ δόξαι were a point of interest in Aristotelian science, Herophilus’ work on κοιναὶ δόξαι probably dealt at least in part with method. A crucial question in Aristotle’s natural science is how one should proceed with reference to opinions established prior to scientific investi- gation. This is the issue Herophilus meant to address in this work, as opposed to mounting a broad polemic against his predecessors. The question of how to handle common opinions (in medicine, at least) is the philosophical zētēma or problēma to which the work would have addressed itself. Of course, to say that Aristotle thought that in the ideal case science does not proceed merely from established belief does not mean that he thought doing so illegitimate in all cases. In fact, even in On the Heavens (2.12, 292a7–9) Aristotle seems to proceed from the claims of the ancient Babylonian and Egyptian astronomers, though these claims have very spe- cifically to do with what prior astronomers have observed as the phenomena of planetary motion themselves, rather than with more abstruse doctrines.40 Another noteworthy instance is found in the Generation of Animals, though here with important qualifications (Gen.An. 3.11, 760b27–33): From a theoretical account and from what is thought to be the case (συµβαίνειν) concerning them, the matters touching the generation of bees seem have this character. In truth the attendant facts (τὰ συµβαίνοντα) have not been sufficiently determined, but if ever they are determined then we ought to put more credence in perception than in theoretical accounts, and in theoretical accounts only if what they show agrees with 40 I thank an anonymous referee for drawing my attention to this and the following passage from the Generation of Animals. ANDREW C. MAYO 435 ————— Greek, Roman, and Byzantine Studies 62 (2022) 417–437 the phenomena.41 Here the contrast is between theoretical accounts (οἱ λόγοι) and what is thought to be the case (τὰ συµβαίνειν δοκοῦντα) on the one hand, and what is observed to be the case (τὰ συµβαίνοντα, ἡ αἴσθησις, τὰ φαινόµενα) on the other. Aristotle makes clear that preexisting opinion (especially when it comes to the testi- mony of experts regarding the phenomena, as in Cael. 2.12, 292a79) can provide valuable starting points for inquiry, but that direct observation is much to be preferred. If Herophilus was operating within an Aristotelian scientific framework, then it would make good sense for him to investigate established opinions and, where the phenomena were not at hand, to make use of them as Aristotle does. This point about the title of the work is not mere quibbling about translation: if it were right to render it “against common opinions,” we could take Herophilus to have argued that well- established views should be given no credence simply by virtue of their being well-established.42 Yet, according to Galen, even though Herophilus denied that knowledge of causes is possible, he nevertheless accepted them on a hypothetical basis; his justi- fication was that “it appears so to many people” (utens eis dicendo multis hominibus sic videri, CP 197 = T59a von Staden).43 Argu- ably, then, Herophilus held that appeal to established opinion is legitimate, but also that anything claimed on this basis should 41 ἐκ µὲν οὖν τοῦ λόγου τὰ περὶ τὴν γένεσιν τῶν µελιττῶν τοῦτον ἔχειν φαίνεται τὸν τρόπον καὶ ἐκ τῶν συµβαίνειν δοκούντων περὶ αὐτάς· οὐ µὴν εἴληπταί γε τὰ συµβαίνοντα ἱκανῶς, ἀλλ’ ἐάν ποτε ληφθῇ τότε τῇ αἰσθήσει µᾶλλον τῶν λόγων πιστευτέον, καὶ τοῖς λόγοις ἐὰν ὁµολογούµενα δεικνύωσι τοῖς φαινοµένοις. 42 So Tieleman, in Ancient Medicine in its Socio-Cultural Context II 494. 43 Though J. Allen has convincingly argued that Galen misunderstands Erasistratus in this work (“Galen as (Mis)informant about the Views of his Predecessors,” AGPh 83 [2001] 81–89), I think we can trust Galen here, both because he is not so hostile to Herophilus as he is to Erasistratus and because there is nothing especially complicated for Galen to misunderstand. 436 HEROPHILUS AND COMMON OPINIONS ————— Greek, Roman, and Byzantine Studies 62 (2022) 417–437 be open to revision. This suits Herophilus’ position on the usefulness of menstruation as Soranus relates it: the prior view that menstruation is always beneficial is not, for Herophilus, entirely wrong, but too categorical, since menstruation is sometimes beneficial, sometimes harmful. Perhaps Herophilus considered this doctrine the result of over-hasty generalization. 4. Conclusion I have argued that Greek usage does not lead us to expect a title of the form πρὸς τὰς κοινὰς δόξας to bear a hostile con- notation, even if it does not entirely preclude it. What else we know about Herophilus and his work makes it less rather than more likely that he should have written a general polemic against common opinions. From these two points, taken to- gether, I conclude that the title of the work should be rendered “in reference to common opinions” or “in response to common opinions,” and not “against common opinions.” Herophilus has been variously seen as a skeptic in his approach to natural science.44 Yet from the present case his approach to the role of established opinion in natural science should seem constructive when compared to Aristotle’s in its more exacting expressions (e.g. at An.post. 2.19, 99b32–100b17):45 while the evidence of what is generally believed to be the case should not supply premises for Aristotelian scientific investigation, Herophilus 44 Kudlien, Gesnerus 21 (1964) 1–13, and R. J. Hankinson, “Saying the Phenomena,” Phronesis 35 (1990) 194–215, at 207–212. Cf. von Staden, Herophilus 117–124 and 244–247; D. Leith, “Causing Doubts: Diodorus Cronus and Herophilus of Chalcedon on Causality,” CQ 64 (2014) 592– 608; M. Vegetti, “L’epistemologia della medicina ellenistica,” Lexicon Philo- sophicum (2018) 264–278, at 264–268. 45 Admittedly, things are more complicated when one looks at the way Aristotle actually approaches scientific investigation in the extant treatises. On the issues surrounding ἔνδοξα, dialectical argument, and Aristotelian science see Owen, in Logic 239–251; Tieleman, in Ancient Medicine in its Socio- Cultural Context II 487–495; H. Baltussen, Theophrastus against the Presocratics and Plato (Leiden 2000) 31–70; Frede, OSAPh 43 (2012) 185–215. ANDREW C. MAYO 437 ————— Greek, Roman, and Byzantine Studies 62 (2022) 417–437 seems to have considered addressing established opinion not only as a crucial part of the scientific method but also as the grounds, at least in some cases (as in CP 197 = T59a von Staden), for making qualified claims about the world.46 August, 2022 The University of Michigan acmayo@umich.edu 46 I owe many thanks to Francesca Schironi, who read through multiple drafts of this article and whose advice has proved invaluable. I also thank the anonymous referees and José M. González for their suggestions, which greatly improved the article.