Bryn Mawr Classical Review 95.12.08

Tryggve Göransson, Albinus, Alcinous, Arius Didymus. Göteborg: Ekblad & Co., Västervik, 1995. Pp. 257. ISBN 91-7346-282-9.

Reviewed by Brad Inwood, University of Toronto (

It was not so long ago that doxography was, in the eyes of most classicists, a moderately stable field. Certain broad facts about the ancient tradition of recording the philosophical and scientific opinions of earlier authors were largely unquestioned, and those not directly engaged in the business could comfortably rely on the theories produced by Hermann Diels and other (usually German) authorities of the nineteenth and early twentieth centuries. Very little was certain, of course, but few doubted that the results of early doxographical research had produced the best results possible given the available evidence. We all knew about the contributions of Theophrastus, Aëtius, and Arius Didymus.

That has changed in recent years. One result of the explosion of interest in philosophical texts from the Hellenistic and Imperial periods has been a critical reexamination of many cherished views, and there is now relatively little that can be taken for granted by the researcher. On many topics, scholarly respectability now requires that one critically re-examine the results of earlier research before building on them. To some extent Theophrastus' Opinions of the Natural Philosophers is still special: its character as a funnel through which a high proportion of our information about the Presocratics passed down to the world of later antiquity makes Diels' theories about the Presocratics more secure (or at least less exposed to reasonable challenge). But for much of the rest of ancient philosophy outside the fourth century B.C., a revolution in our understanding of the surviving sources is under way.

Tryggve Göransson has produced a trenchant, often dryly witty book which forces a reexamination of several traditional source-critical and doxographical theories touching on middle Platonism and on the doxographical accounts found in book 2 of Stobaeus' Anthology. Writing with the kind of radical scepticism demanded by the field and with a keen eye for the logical structure of historical argumentation, G. returns to the year 1879, when two important works appeared. Jacob Freudenthal published Der Platoniker Albinos und der falsche Alkinoos and Diels' Doxographic Graeci established a new standard for and orthodoxy in doxographical studies.

G. traces the evolution of the "school of Gaius" theory from Freudenthal's claim that the Didaskalikos was actually written by the known Albinus (a student of the known Platonist Gaius) rather than by the otherwise unknown Alcinous to whom it is attributed by the manuscripts. The arguments of Giusta and Whittaker against Freudenthal's identification are now accepted almost universally, and G. builds on that conservative foundation in his re-examination of many middle Platonic texts and doctrines. One key feature of the traditional view had been the further belief that the Didaskalikos, whoever wrote it, was crucially dependent on the work of Arius Didymus, the doxographer and friend of Augustus. G.'s critique of the development of middle Platonic school doctrine also casts this proposition into doubt. This leads him to a more general reconsideration of the traditional portrait of Arius Didymus as a doxographer and to much more radical arguments which (if successful) will force a serious reassessment of the date and character of the major doxographies of Stoic and Peripatetic ethics preserved in book 2 of Stobaeus. The traditional portrait of Arius Didymus has not yet lost its standing as a "fact" about the history of ancient philosophy, and had recently been assumed uncritically by the present reviewer in the article "Arius Didyme" in R. Goulet's Dictionnaire des philosophes antiques vol. 1 (1989) and defended at length by David Hahm in "The Ethical Doxography of Arius Didmyus", ANRW II.36.4 (1990).

There are, thus, two main lines of argument in G.'s book: the critical re-examination of middle Platonism, which is the logical result of carefully thinking through the results of Whittaker's conclusions about Alcinous, and the radical reassessment of Arius Didymus. In addition to methodological care, G.'s book is also distinguished by a number of common-sensical assumptions about the history of philosophy. For instance, G. declines to believe that Platonist teachers and philosophers could be so substantially dependent on doxographical sources for their knowledge of Plato's doctrines -- whose works were, after all, widely available (see p. 22). Moreover, G. puts us on more solid ground with his sensible assumption of pluralism in middle Platonism, the contrary preference for Einzelquellen having been a product of source-critical methodology (see pp. 183-4).

A short review cannot do justice to this tight and closely reasoned book. The reader will perhaps be best served by a rapid summary of the results claimed for the understanding of middle Platonism and a quick sketch of the argument about Arius Didymus. The former claims seem destined to achieve a significant measure of agreement; the latter argument is more contentious but will have to be confronted by anyone working on Stoic and later Peripatetic ethics.

Chapters 2 to 5 deal with Gaius and Albinus. The modest results are:

  • Albinus was active in the Greek east (we know he once lectured in Smyrna) in the middle of the second century B.C., Gaius some years earlier (possibly based in Pergamum or Athens).
  • Albinus' Prologos is probably a part of a now lost set of notes by Albinus based on introductory lectures by Gaius on Plato, the whole work having been called Hupotuposeis.
  • As to commentaries on Platonic dialogues, there is "no clear evidence" for any by Gaius, while Albinus probably did publish commentaries on Phaedo and Timaeus.
  • Negative conclusions include the rejection of Lucian's character Nigrinus as a pseudonym for Albinus and the denial that Albinus' lost book on incorporeals with the pseudo-Galenic De Qualitatibus Incorporeis.
  • Chapter 5 deals with Albinus' classification of the Platonic dialogues and its relation to that found in Diogenes Laertius. This is slippery ground, but the proposals G. presents are both ingeniously simple and plausible.

    Chapters 6 to 8 analyse Alcinous' Didaskalikos and another work once claimed for the capacious "school of Gaius", the De Platone of Apuleius. Of course, with Alcinous distinguished from Albinus there was no longer a case for treating the De Platone as a product of the "school". But this still leaves open the question of the relationship between Alcinous' and Apuleius' texts. Both works turn out to rely on a variety of sources, one of which was indeed common: a summary of Platonic ethics. But there turns out to be no particular reason to identify this source with the first-century B.C. Platonist Eudorus. G. modestly and wisely offers no candidate for the identity of the common source. Only a prosopographical horror vacui could drive us to guesswork; an Anonymus Ignotus suffices. But G. does point out that once Alcinous is distinguished from Albinus, there is no longer any external evidence for Alcinous' date. The possible range established by internal criteria is from the first century B.C. to the rise of Neoplatonism (p. 133).

    One result of this loosening of the chronological range for Alcinous is to make possible a more realistic assessment of the relationship between Arius Didymus fragment 1 (preserved first by Eusebius) and chapter 12 of the Didaskalikos. The traditional view is that the Didaskalikos here adapts the text of Arius Didymus (a claim on which much of the speculation about the school of Gaius was ultimately based). But as G. shows in a careful comparison of the two texts, it is not very convincing that the influence ran in that direction; the main reason for accepting that view was a set of prior beliefs about their relative chronology. But if Alcinous could in principle be earlier than has been thought, and if Arius Didymus is not necessarily to be identified with the Augustan philosopher (as G. goes on to argue -- see below), then it becomes possible to accept that Alcinous or his source for ch. 12 was a source for Arius Didymus. No definite chronological conclusions follow from this, but it is a striking illustration of how much one needs to reconsider once one begins to examine the history of middle Platonism in a more critical manner.

    The figure of Arius Didymus used to do a lot of work in accounts of middle Platonism, and if G. is right his utility there is effectively eliminated. Just as the school of Gaius has been broken up by a more critical assessment, we turn out to have more sources, many of them unknown. This is undoubtedly a more realistic picture of the development of Platonism in the early Imperial period. But chapters 9 to 11 cut even more deeply and produce results which, if accepted, might force changes in our picture of the history of Stoicism and Aristotelianism as well.

    The nub of the issue is the material preserved by Stobaeus in book 2, chapter 7 of his massive anthology. Prima facie there are three relevant blocks of material here, conveniently labelled by Hahm (1990, 2945) as Doxography A (37.18-57.12), Doxography B (57.13-116.18), and Doxography C (116.19-152.25). B and C are attributed to the Stoics and Peripatetics respectively, and each is evidently an internal unity. A is less obviously a unity and lacks an explicit source-attribution in our manuscripts of Stobaeus. Because a paragraph from C is quoted elsewhere in Stobaeus (4.39.28) as being from the Epitome of Didymus, it is reasonable to assume that all of C, the Peripatetic ethical doxography, derives from the Epitome of Didymus. But the similarities between B and C are such that it is reasonable (though far from certain) to assume that B came from the same source. Hahm (1990, 2979-3012) argues that A, B, and C are all from the same source; but even a sympathetic reader must concede that the argument for similarity to C is considerably weaker for A than for B.

    G. accepts Didymus' authorship of C (p. 220), though he is more tentative about assigning B to the same author (pp. 220-1). When it comes to A, however, his argument is sharp and sound; his conclusion is negative (pp. 221-226). A is so unlike B and C, despite the serpentine arguments of Hahm, that the common authorship of A and C must be rejected even if one still inclines to accept the common authorship of B and C.

    The crucial question, then, is about the date and authorship of B and C, the Stoic and Aristotelian doxographies of ethics. If the Didymus who wrote C (and probably B) is the Arius Didymus of Diels, then we know just who compiled the doxographies and when he did so. For Diels' Arius Didymus is the friend and court philosopher of Augustus. But in chapter 10 G. examines the history of this hypothesis and the surviving evidence which might support it.

    There are three important stages in the history of this identification. (1) Meineke proposes it in 1860; (2) Diels claims to prove it (1879); (3) Hahm claims to prove it again (1990). The evidence cannot be reviewed here, but G. establishes that both Diels and Hahm effectively beg the question which they purport to prove. That is to say, they show that one can consistently believe that the doxographer Arius Didymus is the Augustan philosopher, but they do not show that one must. G., unlike Diels and Hahm, is scrupulous about the distinction between an argument which permits a certain belief and one which compels it. G. is quite right to conclude (p. 218) "that there have never been put forward any reasons for regarding Arius Didymus the doxographer as the same person as Arius, Augustus' court philosopher". And once that identification is doubted, the range of dates possible for the common author of B and C is now very wide indeed, anywhere "between the middle of the first century B.C. and the end of the second century A.D., perhaps as late as the third century A.D." (p.216).

    Should we then abandon the identification of the doxographer with the friend of Augustus? Not necessarily. For all that G. has established is that the hypothesis of Meineke and Diels is just that, a mere hypothesis. Diels and Hahm have argued for the possibility of the hypothesis while mistakenly thinking that they were arguing for its necessity or its probability. But it still remains a possibility, does it not?

    G. seems to think not, arguing (pp. 217-8) that "there are considerable reasons telling against such an identification". But what are these reasons? G. purports to offer two. The first is the "double-name argument" of Heine (see pp. 211-2), that there is a consistent pattern of reference: Augustus' friend is always called "Arius" and the doxographer is always called "Didymus" or "Arius Didymus". But G. concedes (p. 217) that Diels and Hahm have answered this objection to the identification. The second, then, is the only positive argument against identification. It consists simply in the belief that since Augustus' friend was a Stoic, he could not have had, qua Stoic, any interest in compiling non-polemical doxographicalaccounts of other schools' doctrines.

    Here G.'s customary sagacity deserts him with a breathtaking suddenness. For how can one know this about a Stoic of the period? By the same process of common-sensical guesswork as drove Diels? Nothing else is available. But worse yet, G. has forgotten the evidence which he adduced just a few pages earlier (pp. 209-211). As G. puts it, "in none of the passages referring to Arius' relations to Augustus is it mentioned to which philosophical school he belonged". The claim that Arius was a Stoic rests on a speculative identification with a name found in an index of philosophers treated in Diogenes Laertius book 7; this speculation is supported by arguments which (like those of Diels and Hahm) are no more than permissive (p. 210), certainly not compelling. G. follows this suggestion with a rather lame discussion of two passages of Tertullian which actually distinguish the Stoic view on the fate of souls after death from that of a certain Arius, passages which Diels not unreasonably took to count in favour of his identification of the doxographer with Augustus' friend.

    G. is ruthless in discounting Diels' and Hahm's hypothesis and in distinguishing merely permissive arguments in support of it from considerations which might offer it positive support or which could count as proof. It is a shame that the same standards are not applied to his own argument. When they are so applied, then one must conclude that the hypothesis of Meineke still has much to be said for it. Diels and Hahm do not go as far as one would like in supporting it; in fact, there is little to support it except the conventional doxographers' liking for economy -- why posit two philosophers with the same name when one will do?

    For students of Stoicism and the Peripatetic tradition doxographies B and C are of central importance. It matters most of all that they should be taken to be reliable reporters of school doctrine down to the first century B.C. This assumption is of course easier to make if they are datable to the Augustan period than if they were written 150 years later. G. has shown that this dating is insecure to a degree that few scholars and philosophers in the last hundred years have appreciated; he has not, however, offered any positive reason to reject it.

    Fortunately, for both texts there are other good reasons for treating these doxographies as reliable, beyond the hypothetical identification of Arius Didymus with Arius the friend of Augustus. The content itself and its relationship to other accounts of the doctrines in question remains important. But in the aftermath of G.'s sceptical reconsideration of the evidence considerably more caution will be called for. In this respect this book bears comparison with Sandbach's Aristotle and the Stoics.

    Albinus, Alcinous, Arius Didymus is written in a clear and economical English style, only occasionally unidiomatic. It includes an Index Locorum and separate Bibliographies of Editions and of Works Cited.