Galen ad multos annos

Gden ad multos annos This paper focuses on demonstrating the weakness of the traditional date for Galen's death. It shows that there are good grounds fkom both within the Galenic Corpus and outside it for thinking that he lived at least until he was eighty. Information from Byzantine and Arab scholars from the sixth century onwards suggests that he had died in the reign of Caracalla, perhaps in 216. BIBLID [O21 1-9536(1995) 15; 25-39] Fecha de aceptación: 14 de marzo de 1995 When did Galen die? The standard answer has been to repeat the conclusions of the entry in the Suda lexicon, written about 1000 (1): Galen, the most distinguished physician, a Pergamene, lived in the reigns of the emperors Marcus, Cornmodus and Pertinax in Rome. The son of a landsurveyor and architect, Nicon, he wrote much on medicine and philosophy, besides grammar and rhetoric. Because they are universally famous, 1 thought it inappropriate to draw up a list of them here. He died aged seventy. On first examination, this biographical account has much in its favour. There are no obvious errors, and, although Nicon is never specifically called a land-surveyor, geometres, in our present Galenic Corpus, this could easily have been deduced from the references to his father's ski11 and education in architecture and geometry, V 42, VI 755 R, or preserved in 1 (*) The Wellcome Institute for the History of Medicine. 183 Euston Road, London NWI 2BE (United Kingdom) . (1) Suidae kxicon, ed. A ADLER, vol. 1, Leipzig, Teubner, 1928, S.V. Galenos. H. FLACH, Biographigraeci qui ab Hesychio pendent, Berlin, S. Calvary, 1883, no. 160. DkRTAMlS. Acta Hisp. Med. Su. Hist iüus., 15,1995, 25-39. a treatise now lost to us (2). But two features are worth noting. The first, and for the purposes of this essay, the most interesting, is that this short biography must derive from an earlier account, and one that was deliberately abbreviated: the personal «I thought it inappropnate» is a rare, if not unique, comment in this lexicon, and irnplies that a whole section was lifted bodily from an earlier source (3). Ada Adler, the Suda's most recent editor, suggested in her preface that this was the Lives by Hesychius of Miletus, who compi1e.d a series of short biographies of learned and famous men at the end of the sixth century, a suggestion that has been widely accepted. Secondly, apart from the references to Nicon and to Galen's age at death, al1 the information presented here is taken from Galen's De libris propriis, XIX 848 K. His books are described in the order in which they are recorded there, and the final cornrnent on his writings on grammar and rhetonc derives from the heading to its final section, XIX 48 K. The absence of Septimius Severus fi-om the list of emperors under whom Galen lived is signficant, for in this treatise the only living emperors mentioned are Marcus Aurelius, Cornrnodus, and Pertinax (4). The date of death must come fi-om another source, obviously not from Galen himself directly, and appears here almost as if tacked on to the end of an earlier entry. This entry in the Suda lexicon is thus not without its merits: it is, at least in part, based on the evidence of Galen himself, and, in al1 likelihood, it derives from a much earlier biography, perhaps one written at the end of the sixth century. Hesychius may also have been the source for the entry in the Chronography of George Syncellus, which was written about 810: (2) References to Galen are by the volurne and page in the edition of C. G. Kühn, Claudii Galeni Opera ornnia, Leipzig, C. G. Knobloch, 1821-1832, repr. Hildesheim, G. Olms, 1964, = K. (3) Rather as large sections from Isidore's Life of Darnascius, ed. R ASMUS, Das Leben des Phiímphen lsidoros von Damaskios aw Damaskos. Leipig, Teubner, 1911, are incorpomted verbatim in the Suda's entries on medicine and philosophy in the later fifth cenhuy. The comment about the need to abbreviate was omitted in one of the oldest manuscnpts, V. (4) The cede and death of Lucius Verus in 169 is mentioned at XIX 18 K., but the compiler of this entry could well have singled out only those emperors whom he thought Galen had served. DkRrAMlS. Acta Hisp. Med. Su. Hist Ziius., 15,1995,25-39. Galen ad multos annos 27 Galen, by birth a Pergamene, a most excellent doctor, flourished under Marnis Aurelius, also called Verus (5). But there were other stones circulating in Byzantium that gave different dates for Galen's life. Mchael Glykas in the twelfth century had to warn his readers not to believe in the popular story that Galen had disnissed with Mary Magdalene in Rome Christ's healing of the man born blind, John, chapter 9, and had explained to her that Jesus was weil acquainted with the healing properties of mineral earths. Leamed chronographers, like Tzetzes (d. lOS(11085) and Joel (active after 1204), declared that Galen had lived into the reign of Caracalla, i.e. 211/2-217, a date that is incompatible with that in the Suda. Their source, although not necessarily their only or, indeed, their oldest one, was the Chronicle of George Hamartolus (d. ca. 870), whose entry reads as follows: After Antoninus there reigned Antoninus Caracalla for 26 years, and he was killed in warfare. In his time there was also Galen the doctor (7). George's account is a mess. Caracalla did not reign for 26 years, and in his previous entry on Antoninus (whom he seems to see as different from Caracalla), George reports that this emperor reigned with his brother Geta for a few months before killing him, ruled for a further six years, and was then murdered by his own household. This discrepancy between the years could well be illusory, the result of a simple copyist's error (S), but it is also clear that George was himself conhsed over the narnes of the emperors and was also trying to reconcile two different sources for the same event, the murder of Caracalla in 216 by his own officers while on campaign against the Persians. George's claim that Galen lived into the (5) George Syncellus, Ecloga chronographica, ed. A. A. MOSSHAMMER, Leipzig, Teubner, 1984, p. 431. (6) Michael Glycas, Annales, ed. 1. BEKKER Bonn, 1836, p. 430; Tzetzes, Chiliades, ed. P.A.M. LEONE, Naples, Librena scientifica editrice, 1968, XI 99&XII 28; Epistulae, ed. P.A.M. LEONE, Leipzig, Teubner, p. 121; Joel, Chronographia, ed. F. IADEVAIN, Messina, Edas, 1979, 707-709. (7) George Hamartolus (= George the Monk), Chronicle, ed. P. WIRTH, Stuttgart, Teubner, 1978, p. 480. (8) ETH Z becomes ETH HZ, which becomes ETH KZ. But if this was a scnbal error, it had certainly entered the tradition by the time of Joel, who also gives the number of years as twenty-six. 1 DWAMlS. Acta Hisp. Med. Sci. Hist Illw., 15, 1995, 25-39. reign of Caracalla could be the result of a similar misunderstanding, especially because Commodus, Caracalla, and Elagabalus as emperors were all styled M. Aurelius Antoninus. But in the Greek tradition as a whole, and in the Byantine chronographers in particular, the first two are distinguished as Commodus and Caracalla. Whatever his uncertainty over Antoninus, George Hamartolus was clear in his own rnind on the difference between Commodus and Caracalla: whether his sources were so precise is another question. Nonetheless, the possibility must remain that the transfer of Galen to the reign of Caracalla was not simply the result of George's own misunderstandings. The change of date is all the more intnguing in the light of the alrnost universal agreement among our oriental sources that Galen died in his eighties, perhaps as old as 87, a date that, in this tradition, goes back to the early tenth century, to the Lives of the physicians by Ishaq ibn Hunain (9). Ishaq, no mean scholar himself and a prominent translator of Galen, clairned to have taken his information from a Chronology ofthe physicians by John the Grammarian of Aiexandria, an enigmatic authority, who perhaps lived in the sixth century and whom some have sought to identify with the we11-known John Philoponus (10). Ishaq's own chronology is fa, from faultless, and there are signs of evident confusion over narnes and dates, although some of this may have been the result of a process of transrnission that has seen most of his history preserved only as an epitome. Nonetheless, he is ostensibly sure of the dates of Galen's life, with which John had ended his own chronology. If the epitome is to be trusted, and there is no sound reason not to do so, the final section of John's book concluded with a reference to Alexander of Aphrodisias and his relationship with Galen. Just when John's book was composed is more controversial. Fritz Zimmermann drew attention to features which, he thought, indicated that it was a (9) Variants for this year, e.g. 88, can be explained by rnistaking an age for the year; even the incorrect supposition of Sliwa, De pam'archis nestorianorum, ed. H. GISMONDI, Rome, 1897, 1. p. 3, that he died in year five of the reign of Commodus, if changed to year five of Caracaila, i.e. 216, would then also give 87 for Galen's age at death. (10) ROSENTHAL, E. (1954). Ishaq b. Hunayn's Ta'rikh al-atibba. Oriens, 7, 55-80 (reproduced, with identicai pagination, as chapter 11 in his Science and medicine in Islam, Aldershot, Varionim, 1990). For the date, see p. 76, and for Ishaq's source, p. 75. DMVAMIS. Acta Esp. Med. Sci. Est Illus., 15, 1995,25-39. Galen ad multos anms 29 pseudonymous production by a Nestorian scholar writing about 800, but he also adrnitted that both later Byzantine and Arabic scholars were aware of some of the problems revealed in Ishaq's chronology, notably his synchronism of Galen and Christ, and that this pointed instead to a joint 1ateCreek source, most likely one composed at Alexandria, or at least dependent on Alexandrian sourca (11). There is a more serious objection to Zimmermann's theory. Much of this History of the physicians goes back directly to comments, quotations, and ideas found in the Galenic Corpus, and, as Rosenthal hirnself noted, there are parallels between the account in Ishaq and the abundant antiinfohon contained in a Commentary on the I-iippcnfic Oath, ascribed, 1 believe correctly, to Galen. This work was not translated into Syriac or Arabic until the rnid ninth century, by Hunayn, and into Arabic shortly after that, by Hubaysh, and it is unlikely that it had been earlier available to the largely Syriac speaking Nestorians (12). John must have taken his inforrnation direct from the Greek, a strong pointer to an Alexandrian rather than a N. Syrian or Persian source. This contention would be further strengthened if it could be proved that the Byzantine chronological tradition first represented by George Harnartolus, with its reference to Caracalla, also went back to John (13). So far we have identified two different dates for the death of Galen in Byzantiurn, one first appearing in the eighth century, the other in the tenth, but both possibly to be traced back to scholars working in the sixth century. On a pnori grounds there is little or nothing to choose between them. Neither John the Gramrnarian nor Hesychius of Miletus was incapable of error and confusion, yet each was a man of erudition, energy, and ability. How, then, can the historian decide between these two dates ? In (11) ZIMMERMANN, F. (1975). The chronology of Ishaq ibn Hunain's Ta'rih al'atibba. Jix Hunap ibn Ishaq: Collection d'artides publiées 2 l'occas1'on du onzieme centénaire de sa mort, Leiden, Brill, pp. 324330. (12) ROSENTHAL, E. (1954), op. cit. (h. lo), p. 56; cf his An ancient comrnentary on the Hippocratic Oath'. Bull. Hist. Med., 30, 1956, 52-87, at p. 55 (reprinted, with identical pagination, as chapter 111 in Science and medicine in Islam). (13) GERO, S. (1990). Galen on the Christians. Orientalia Chnstianapenodica, 56,395439, discusses the Arabic evidente, adding further references to Galen's death, and doubts the equivalente of John the Gramrnarian and John Philoponus. The confusion between the various rulers narned Antoninus is found in Arabic sources as well as in George Hamartolus, although this need not point to a joint source. DYNAMIS. Acta Hisp. Med. Su. Hist IIlus., 15, 1995, 25-39. the rest of this paper 1 sW employ three separate arguments, each of which, by itself, offers considerable supp01-t for the later date, and which, taken together, create an insuperable obstacle to believing that Galen died at the age of seventy. In 1984, 1 argued that a passage in an Arabic author, &distani, who died around 983, incorporated a quotation from Alexander of Aphrodisias to the effect that Galen had wasted eighty years of his life before coming to the conclusion that he knew nothmg (14). 1 suggested there that Alexander was alluding to Galen's philosophical autobiography, De proprüs placitis, in which, in several places, and especiaily in chapters 2 and 3, Galen emphasised his agnosticism about many of the standard philosophical topics of the day, which were often phrased in the form of 'either/or' questions, inviting assent or repudiation (15). Galen, for good reasons, rejected this simplistic approach to knowledge, repeating again and again that he saw no way of deciding without solid proof between what he saw merely as assertions. A hostile opponent could easily take this agnosticism for a confession of ignorance, an opportunity unlikely to have been rnissed by Alexander the Aristotelian, especially as he is known to have opposed Galen on a variety of philosophical issues. But has as-Sijistani preserved a genuine fragment of Alexander, or is it, as Hilary Wiesner and Silvia Fazzo have suggested, merely a piece of biographical fiction intended to link two contemporaries together in a neat opposition, philosopher versus physician, Platonist versus Aristotelian, and so on? (16) It is true that the Greek Alexander rarely refers to Galen by name, and equally true that in the treatises against Galen's views on motion and on possibility, parts of which are preserved only in Arabic, Alexander's (14) mON, V. (1984). Galen in the eyes of his contemporaries. Bull. Hist. Med., 58, 315324 (reprinted, with identical pagination, as ch. III. Zn: From Democedes to Harvey. London, Variorum, 1988). (15) MANSFELD J. (1990). Doxography and dialectic. Zn: W. Haase ( ed.). Aufstieg und Niedergang der romischen Welt, Teil U, hcipat, 4. Teilband, Philosophie, Berlin and New York, W. De Gruyter, pp. 3.057-3.229. (16) FAZZO, S.; WIESNER, H. Alexandre contre Galien, an unpublished paper delivered at a Conference Perspectives médiévales, held in 1993 at the Institut du Moyen Age, Paris. 1 arn grateful to the authors for letting me see a draft of their paper, which corrects some of my earlier views. DWAMTS. Acta tasp. Med. Sci. Hst Illus., 15, 1995, 25-39. Galen ad multos amos 31 opponent is almost aIways anonymous. It could indeed be, as Mesner and Fazzo argue, that the titles of these treatises were added by later scribes already familiar with the story of a quarrel between the two men; bibliogmphy responding to fiction. But not every reference to Galen is an editorial assertion, a translator's attempt at clarifícation, or a scribal rnistake (l?), and the closer in time the earliest Arabic descriptions of a debate between the two men are moved to the age in which their works kst became available in Arabic, the more likely it is that these descriptions have a basis in texts rather than in sheer invention. Ishaq's source, John the Grammarian, had also spoken of Alexander in connection with Galen, which would take knowledge of this link back probably to the si& century (18). Besides, the date implied by asSijistani, i.e. that Galen lived at least to 209/210, is unusual in that it corresponds to neither of the two dates of death under discussion here. Its singularity is a point in its favour. But, for the moment, let it be conceded that this quotation forms part of a late antique attempt to depict the doctrinal differences between Alexander and Galen in terms of a personal quarrel, and that it may have been one of the sources, if not the source, that suggested to a later biographer that Galen had lived beyond the canonical three score years and ten. My second argument relates to the chronology of Galen's writings, which, as far as the Severan period is concerned, has not been studied carefully for over half a century. Although Galen was a prolific author, there are physical lirnits to what a man can write or dictate, and the more treatises that have to be placed in the Severan period, the more likely it is that the traditional death date has to be moved later and later. In his fundamental survey of Galenic chronology, Ilberg listed seventeen tides that he thought were written then, including al1 of De locis affectis, the two tracts De compositione medicamentorum secundum genera and secundum locos, and the later books of the Methodus medendi and De temperamentis et facultatibus simplicium medicamentorum (19). Bardong, in his revi- (17) RESCHER, N.; MARMURA, M. E. (1965). Alexander against Galen on motion, Islamabad, Islamic Research Institute: for the mention of Gaien's name in the text of the manuscnpt, see fols. 64a20, 67a1, and the quotation from Gaien on possibility, p. 69. (18) ROSENTHAL, E. (1954), op. cit. (h. lo), p. 79. (19) ILBERG, J. (1896) Über die Schrdbteiierei des Maudios Galenos, m, Rheinisches Museum, DW'S. Acta Hisp. Med. Sci. Hist Illus., 15,1995,25-39. sion of Ilberg, added to them the last books of Admúlistrationes matomicae (20). Certainly f?om this period are also De propxiis placitis (Galen's last work), De publice dictis sub Pertinace, De moribus, which mentions the conspiracy of Perennis against Comrnodus in 185, and De animi peccatonrm dignotione et cura, which cites the latter work (21). More questie nable are the dates of the commentary on De natura hominis, that on De aeris, aquis, locis, and AdversusJulimum, al1 of which are likely to postdate 189 (22). In my opinion, space has to be found also for De thenaca, ad Pisonem, and the commentary on the Hippocratic Oath, which is not mentioned in De libris propriis but which was certainly written by a learned man of the time who was very familiar with Pergamum (23). Besides, we cannot a priori exclude from this period at least some of the treatises which are, as yet, undatable or which are now lost to us. Although Galen was a prolific author, this list of works written in a busy (and, in political terms, confused) time of life is one that can be squeezed only with the utmost of difñculty into the years from 193 to 200. Some of these works have an even tighter fit: De antidotis, so Simon Swain argues in his forthcoming book, is unlikely to have been written before 197 (24), and the way in which Galen refers to Severus and to the fact that he was forced to use stocks of material going back to Trajan or Hadrian, XIV 65 R, implies that some time has elapsed since then. The later De antidotis is placed, the later one must put De libris propriis and the later still goes De propriis placitis. Galenic chronology is elastic, but not so elastic as to hold al1 these works in the same package, amounting in total to a minimum of 3300 fidl pages in n. F., 51, p. 196 (repnnted, Darmstadt, Wissenschaftliche Buchgesellschaft, 1974, p. 90). (20) BARDONG, K. (1941). Beitrage zur Hippokrates und Galenforschung. Nachnchten von der Akademie der Mssenschaften zu Gottingen, phii.-hist Ki., p. 640. (21) WALZER, R (1962). Greek into Arabic, Oxford, B. Cassirer, p. 144, n. 7, although his supposition that the tract was written between 185 and 192 is unlikely, since Galen would have been extremely foolish to remark favourably then on the steadfast behaviour of the slaves of the chief conspirator. (22) See BARDONG, K. (1941), op. cit. (h. 20), p. 639. (23) STROHMAIER, G. (1970). Asklepios und das Ei. In: R Stiehl; H. E. Stier (eds.). Beitrage zur Aten Geschichte und deren Nachieben, Festschrií?fiFranzAtheim, Band ll, Berlin, W. de Gruyter, pp. 143153. (24) Hellenism and Empire, Oxford, Clarendon Press, 1996.1 am gateful to Dr. Swain for letting me see a copy of the relevant pages while in draft form. DMVAMIS. Ac~cta Hisp. Med. Sci. Hist iiius., 15, 1995, 25-39. Galen ad multos amos 33 Kiihn, all to be written within at most half a dozen years of a busy Roman practice (25). But the key to resohng the &ole problem of date rests on an investigation of the authenticity of the tract De theriaca ad Pisonem. Here is a text securely dated to the second half of the reign of Severus, which, if genuinely by Galen, would resolve once and for al1 the the choice between the two traditional dates. As Anton von Premerstein proved alrnost a century ago, the account of the accident to the son of Piso during a performance of the ceremonial Lusus Troiae, XIV 212 R, can refer only to the secular games of 204 (26). If this work was written by Galen, he must have lived for at least five years beyond 199. At first sight, those who defend the authenticity of this treatise face ovenvhelming odds. The modern editor of the Arabic version, Lutz Richter- Bernburg, boldly stated in his title that this work was not by Galen, and Konrad Schubring, in his bibliographical introduction to volume XX of the reprint of Kühn, bluntly cailed it «unecht» (27). The learned Ackermann, in the biographical and bibliographical introduction to Kiihn's edition, confessed that he could not easily see how and when Galen had written this book under Severus, pointing to what he considered stylistic anomalies - an absence of references to others and to De antidotis (or if that tract was the later one, any mention therein of Ad Pisonem), exuberant flattery, lack of organisation, and a style more like that of a young man (28). Those modern writers who defended its authenticity, either, like Cotuni, argued that it was a juvenile work by Galen - a theory that went back to the sixteenth century and had long been rightly discarded by serious scholars - or, like Watson and Bowersock, were apparently unaware of the consequen- (25) Calculations are made from the lists in Ilberg and Bardong, with the additions mentioned above: 1 have added up the pages in R, and allowed 65 pages for each book not included in that edition. 1 have not included figures for such lost tracts as the genuine De remediis parabilibus, which is also to be-dated to this period. (26) PREMERSTEIN, A. von (1898). Das Troiaspiel und die Tnbuni celerum. Festschrift Otto Benndorf; Vienna, pp. 261-266. See also FUCHS, H. (1990). Lusus Troiae. Diss., University of Cologne, pp. 5%64. (27) RI(3ITER-BERNBURG, L. (1969). Eine arabische version derpseudogalenischen Schrift De theriaca ad Pisonem. Diss., Gottingen; K. Schubring, XX p. XLVII K. (28) ACKERMANN, J. C. G. In: I, p. XXXVII-XXXVIII K. DkRAMlS. Acta Hisp. Med. Sci. Hist Iilus., 15, 1995, 25-39.