Anonymus Seguerianus - HEATH, M. (2005)

Embed Size (px)

Citation preview

  • 8/10/2019 Anonymus Seguerianus - HEATH, M. (2005)

    1/30

  • 8/10/2019 Anonymus Seguerianus - HEATH, M. (2005)

    2/30

    MALCOLM HEATH , N OTES ON THE A NONYMUS SEGUERIANUS

    ground-breaking edition 4 that the author was called Cornutus, a suggestion that isstill able to trap the occasional unwary non-specialist into a confusion with thefirst-century Stoic, who is known to have written on rhetoric. 5 But that cannot be

    right: AS is essentially a compilation of material drawn from named authorities(most extensively, Neocles, Alexander son of Numenius, and Harpocration), atleast some of whom date to the second century AD. Graeven, of course, was notguilty of an elementary chronological blunder. His idea was rather that there musthave been a third-century rhetorician of that name. The argument, which cruciallydepends on the fact that the definition of klon at AS 242 ( klon m n on stidianoaj mroj partzon prj teron klon parakemenon , with thefollowing example) is identical to that attributed to Cornutus in a fragment of thefifth-century sophist Lachares ( RG 7.931.1-14 Walz), is not compelling. There is asimpler explanation for the parallel. Unwary non-specialists are right to think ofthe first-century Stoic; but he was the author, not of the whole text, but of theshort definition of klon that an anonymous third-century rhetorician inheritedfrom one of his proximate sources. 6

    Dilts and Kennedy, in the introduction to their edition of AS, 7 report thatGraevens theory that Anonymous [ sic ] Seguerianus is a shortened version of awork on rhetoric by a rhetorician of the third century named Cornutus... has beendiscredited (xi). That, however, is misleading. Though the attribution to ahypothetical third-century Cornutus has not convinced subsequent researchers, itremains a virtual certainty that the text which Sguier discovered is a shortenedversion of a work on rhetoric by a rhetorician of the third century. We can beconfident that AS in the form that has been preserved is an abridgement because

    parts of a more extensive version have been preserved in the indirect tradition.How widely traces of the unabridged version are dispersed is, as we shall see, stilla matter for debate. But the minimum is beyond dispute: an anonymouscommentary on [Hermogenes] On Invention contains passages which cite thesame range of authorities as AS in the same manner; these passages partiallyoverlap with AS, but also present material missing from the version preserved inthe direct tradition. 8

    Previous editors of AS have not though it their business to edit the materialfrom the anonymous commentary (some do not even advertise its existence withany energy), and getting at it has hitherto posed a formidable challenge. Parts ofthe commentary were printed in volume 7 of Walzs Rhetores Graeci ( RG 7.697-860); the parts which Walz omitted were ones which had already been printed in

    RG 5 as parts of another, later commentary ( RG 5.363-436, attributed to

    4 Graeven 1891.5 E.g. Boys-Stone 2003, 210 n.31.6 I assess Graevens arguments briefly in Heath 2003a, 152.7 Dilts and Kennedy 1997. Patillon refers to another recent edition, which I have not seen: Vottero2004, including text, Italian translation, and commentary.8 Kennedy 2003, 300: One controversy that I hope may be regarded as settled relates to the

    Anonymous Seguerianus ... I hope to have demonstrated (Dilts and Kennedy 1997, xi-xv) that Anonymous Seguerianus is not an epitome of a single text but is an abstract of the views of certain

    second-century authorities. The latter has never been in doubt; the question is whether AS is anepitome of such an abstract, and Kennedy has provided no argument to the contrary.

    2

  • 8/10/2019 Anonymus Seguerianus - HEATH, M. (2005)

    3/30

    MALCOLM HEATH , N OTES ON THE A NONYMUS SEGUERIANUS

    Planudes). However, the text in RG 5 was that of the derivative commentary; thesuperior text of the RG 7 commentary had to be constructed out of the variantsreported in Walzs (not always unambiguous) apparatus. RG is not a user-friendly

    work at the best of times, and only the most obsessive devotees of fragmentaryrhetoricians would contemplate the task of reassembling the lost sections of ASfrom Walzs diaspora. It is therefore cause for rejoicing that we have the relevant

    parts of the anonymous commentary (henceforth AC), properly edited fromsuperior manuscript evidence, in Patillons accessible Bud. Though I shall arguein 4 that the relationship between AS and the material preserved in AC is not asinscrutable as P. supposes (xxiii), the decision to place this material separately inan Annexe rather than contaminating the two versions is methodologically correct.In what follows I cite AC using Patillons section numbers. 9

    Since I am obsessive enough to have made some tentative investigations into

    the material which Walz scattered between RG 5 and RG 7, I am particularlyaware of how much of a benefit Patillon has conferred on us. I should, perhaps, bemore relieved than embarrassed to discover how unequal my obsessive tendencieswere to the challenge that Walz had posed: the list of references I compiled, 10 happily now completely superseded, is glaringly incomplete. But obsessions arehard things to shake off, and Patillons work has driven me back to the questionwhether there may be further, unidentified material from the unabridged AS in thecommentary and other sources. It could well be argued that an edition is not the

    place to explore such inevitably speculative and elusive territory; and thesometimes unconvincing arguments of Graevens introduction (he, too, wiselyrefrained from contaminating his text) are a discouraging precedent. If I have a

    complaint, therefore, it is only that Patillon has done less than he could to alertreaders to the possibilities. I shall return to this point in 5. But first I shouldcomment on what Patillon has done, focusing on his text (2), and on the analysisof the sources of AS in his introduction (3).

    2. Textual problemsAS 1 : politikj toi dikanikj lgoj ej tssara mrhdiairetai t prokemena: crzomen gr n at prooimwn m nprj t prosecestrouj poisai toj kroatj, dihgsewj d prjt didxai t prgma, tn d pstewn prj t kataskeusai

    naskeusai t prokemenon: toj d pilgouj pgomen prj tpirrsai tn koonta ej tn p r mn yfon.

    The opening section illustrates both the conservative and the interventionisttendencies in Patillons editorial practice. He retains the words toi dikanikj which other editors have plausibly suspected as an interpolation; he is similarly

    protective of AS 18.8 toi to ntidkou . In AS 1 suspicion, at least, isencouraged by the omission of these words in John of Sardis ( PS 358.6 Rabe =

    9 Patillon indicates that his forthcoming edition of [Hermogenes] On Invention will include theanonymous commentary in its entirety. The sections included with AS are AC 1-2, 78-87, 170-198,

    225-227.10 Heath 2004, 268 n.29.

    3

  • 8/10/2019 Anonymus Seguerianus - HEATH, M. (2005)

    4/30

    MALCOLM HEATH , N OTES ON THE A NONYMUS SEGUERIANUS

    RG 6.511.5 Walz). Patillon mentions and discusses this evidence in a note, butdoes not acknowledge it in his apparatus. That is not the only case in which P.sapparatus is less informative than it should be: an editions users are entitled to

    expect that the apparatus will supply them with all the relevant evidence, andshould not have to hunt about in notes.

    A few lines later, however, Patillon adopts Aujacs dihgsewn in place of thetransmitted dihgsewj . This achieves consistency with the adjacent plural

    proems, proofs and epilogues. But anyone who compares, for example, 2.1 ( perprooimwn) with 40.1 ( per d tj dihgsewj ), or considers the promiscuousmix of singular and plural proem(s) and epilogue(s) in 27-28, is likely to concludethat this intervention is unnecessarily fussy. Similarly, in technographic prose(especially in technographic prose which we know has been subjected toabridgement) it seems misguided to assume that every ti-clause needs to be

    furnished with an explicit lgein (40.6) or epen (43.3, following Kennedy).Editorial intervention is, of course, necessary when one is dealing with an

    abbreviated technical text preserved in a single manuscript, and the judgementsthat need to be made are inevitably delicate and uncertain. So it is not surprisingthat Patillons proposals do not all command assent. For example, I find thedeletions at AS 43.2 and 72.2 persuasive, and the lacuna marked at 107.3; in 195.5Patillon is right to adopt Kaysers excellent metaceirizmenoj ; at 225.4diafqorj is attractive, as is the treatment of 241.1. On the other hand, thesupplement j at 150.4 seems to me unnecessary, and the supplement ath at208.7 implausible; at 172.2 Wilamowitzs ka ej strikes me as preferable;Kennedys diagnosis of the problem at 183.3 is more convincing; at 248.6 I wouldlike to see a convincing parallel for this use of sugkatstasij (for Finckhs snkatastsei compare Syrianus 2.127.11f. Rabe).

    I select here a few illustrative passages for comment. Points in some other passages will arise in the course of discussion in subsequent sections.

    AS 14 : ka sumboulesaj, sper par' `OmrJ Nstwr esgeiautn toj per Kaina ka 'Exdion sumboulesanta, ka tipeiqnioi san at, otwj ka prteron dhlseij ka atj, jpeisqntej m n katrqwsan, m peisqntej d plonto.

    This passage has been variously tinkered with by editors. Patillons proposal is to

    transpose sumboulesaj to follow prteron , which (he says) makes the syntaxof ka prteron more natural. But the corruption is not readily explicable, andthe position of sumboulesaj in the transmitted text is not in itself problematic;indeed, it seems positively useful to have the focus on symbouleutic speechestablished at the start of the sentence. If the syntax is felt to be unnatural, wouldit not be more sensible to transpose ka prteron ? If we were to read otwjdhlseij ka atj, j ka prteron peisqntej m n katrqwsan , then thesimilarity between otwj and j would give at least the glimmer of anexplanation for the confusion. But I am not sure that any change is needed: the

    prominent position of ka prteron achieves the desired strong emphasis on the past occasions on which the speakers advice proved beneficial.

    4

  • 8/10/2019 Anonymus Seguerianus - HEATH, M. (2005)

    5/30

    MALCOLM HEATH , N OTES ON THE A NONYMUS SEGUERIANUS

    AS 20 : o mnon d tn prooimwn es tinej lai, aj o crmeqan toj pilgoij, ll ka tn pilgwn, n ok stin n tojprooimoij crea. poisomen t proomion, e tn kefalawn tn

    nagkawn n totoij tj polyeij lhymeqa, n d t pilgJngkh psa n ti lambnein atn ej prrwsin parathsin.

    Patillon prints his own supplement, poisomen t proomion . Heclaims support from the parallel in RG 4.428.27f. (about which I shall have moreto say in 5.4), which offers lkopoisomen gr t proomion . Patillondescribes that variant as difficult, but the corruption from kakn... poisomen tolkopoisomen... is barely credible. Though the expression is surprising, theimage is appropriate and compelling: you should address your opponentsarguments 11 only where you need (and have the opportunity) to counter them; togive them free publicity in the proem is to turn the proem into (as we might say) aself-inflicted wound. 12

    AS 74 : ka n toj tropikoj noij kurwj cro, sperDhmosqnhj t necatise: di mij gr lxewj lon plrwsenhma.

    This is from the discussion of lexical brevity. Patillon supplements the text asfollows: ka n toj tropikoj . Thesupplement, which assumes a saut du mme au mme , is recovered from a parallel

    passage in John of Sardis commentary on Aphthonius (22.18-20 Rabe). But thecontext in John is constructed out of alternating extracts from AS and Theon, 13 and these words are found in Theon (84.10f. Spengel). The most rationalassumption is therefore that John has inserted a note from Theon into a passagefrom AS. Patillon counters (82 n.3): mais on la lit aussi chez J. Doxapatrs, quine dpend pas de J. de Sardes. But that is simply false. There is no doubt thatJohn of Sardis was one of Doxapatres sources in general; 14 in this particular

    passage he presents a conflation of AS and Theon similar to that in John ofSardis, 15 and one striking agreement proves his direct dependence:sunupakoetai gr t file (John Doxapatres 2.229.6 = John of Sardis 22.24,inspired by Theon 84.12) in place of the original lepei gr t file (AS75.2f.). Patillons supplement is therefore to be rejected decisively.

    AS 149 : d p to prgmatoj pstij gnetai kat trpouj trej,kat t ekj, kat tekmrion, kat pardeigma.

    This is from the account of Neocles classification of artificial proofs based onfact. Although three categories are specified, the following discussion confusingly

    11 For this technical use of nagkaon see Heath 2002a, 662-6.12 The verb is an echo of Aeschines 3.208; cf. sch. ad loc . (446ab Dilts) and Lexicon Seguerianums.v. lkopoisai .13 John of Sardis 21.5-18 Rabe = AS; 21.18-22.3 = Theon; 22.3-18 = AS; 22.18-20 = Theon;22.20-23.2 = AS; 23.2-4 = Theon.14 Rabe 1928, xi-xii.15

    John Doxapatres 2.228.15-19 Walz = John of Sardis 21.16-20 Rabe; 228.19-229.4 = 22.5-19;229.4-8 = 22.22-23.2.

    5

  • 8/10/2019 Anonymus Seguerianus - HEATH, M. (2005)

    6/30

    MALCOLM HEATH , N OTES ON THE A NONYMUS SEGUERIANUS

    also refers to shmeon . Patillon eliminates the confusion by reading tttaraj , andinserting after ekj. This is an initially attractive solution to a

    passage that I have always found difficult. But on reflection, I think the problem is

    with Neocles exposition rather than the text. shmeon does not have a sufficientlystable independent existence in the following discussion to merit this place in theclassification: in AS 152 shmeon is defined as a kind of tekmrion , and asequivalent in customary usage to ekj. Moreover, now that Patillon has made ACso much more accessible it is easier to see the structure of Neocles theory of

    proof in its entirety. I shall return to this in more detail in 4, where I argue that anunderstanding of the structure of Neocles theory helps to solve the puzzle abouthow the AC material is to be integrated into the material transmitted in AS. For

    present purposes it is enough to note that when Neocles maps his classification of proofs onto a classification of epicheiremes, he is working with the three-partclassification without any reference to shmeon (AC 81-82).

    AS 170 : tn d tpwn nioi mn, j Neoklj fhsi, koinj tinakat pasn tn stsewn erkasin, o d dwj ksthj stsewj,'Aristotlhj d ka koinoj ka douj toj m n plestouj eaden,per d tn dwn dialgetai sumfwnn ka atj EdmJ tkadhmak.

    This is the transmitted text. Most editors see a distinction between theorists,reading koinoj tinaj... erkasin... douj (Spengel): among topics, some havefound ones that are common across all the issues, others ones that are special toindividual issues. Yet the opening genitive creates the expectation of a distinction

    between kinds of topic, as (for example) at AS 145. Patillon meets thatexpectation by preserving koinj tina... erkasin and reading dioi for dwj:some topics have spoken certain things in common across all the issues, whileothers are special to individual issues. 16 Elsewhere in the treatise, this kind ofsubdivision is usually done with o m n... o d , while nioi is used to introducethe opinions of unnamed persons (AS 3, 132, 158, 242, AC 190). However, AC186 provides a parallel: tn d gnwmn niai m n cwrj podexewjlgontai... a d met podexewj . Yet Patillons text burdens Neocles with astrange way of making his point: can koinj tina... erkasin really betranslated sont des noncs appliqus communment? The passive lgontai inAC 186 is far more straightforward. Moreover, the more common approach givesthe section a continuity lost if Patillons proposal is adopted: some have foundtopics that are common across all the issues, others ones that are special toindividual issues; but Aristotle found ( eren Spengel, erhke Volkmann) bothcommon and special topics. This, however, raises an issue to which I shall returnin 3 below.

    16 Dilts and Kennedy adopt Spengels text but translate it as dividing topics: Some topics... inventsomething in common to all stases, but others are specific to each stasis. The first part of thistranslation leaves the masculine koinoj tinaj unexplained; the second would work better with

    Patillons dioi. The English translation in this edition is often unhelpful, and sometimes positivelymisleading.

    6

  • 8/10/2019 Anonymus Seguerianus - HEATH, M. (2005)

    7/30

    MALCOLM HEATH , N OTES ON THE A NONYMUS SEGUERIANUS

    AS 243 : d tn pilgwn frsij paqhtik sti ka tolmhrotratoj nomasi ka taj deinsesi ka toj scetliasmojperipaqestra.

    Patillons nmasi is neater than Wilamowitzs scmasi , and gives the passagemore point: it comments on both vocabulary and figuration, rather than makingfirst a general and then a more specific comment on figures. But the editors whoretain nomasi are, I think, correct. Despite the theoretical distinction betweenthought and expression, ancient rhetoricians realised that frsij and nhma areinseparable (like body and soul: PS 337.25-338.1). The style must fit the contentand vice versa ; so one cannot define the style appropriate to the epilogueindependently of the appropriate thoughts to express there. Compare AS 100,which combines advice on frsij , nomata , paradigms and lxij ; and since theadvice on frsij at AS 240 includes spermatikj cous tina tn

    pragmtwn , it is clear that frsij cannot be a matter simply of vocabulary andfiguration. (I return to this passage in 5.5.)

    3. Identifying the sourcesI have suggested (2) that Patillons conjecture in AS 170 breaks the

    continuity of the some-others-Aristotle sequence. But Patillon would not see thatas a disadvantage. If the sequence is maintained, then Neocles is the source for theAristotle citation as well as for the initial some-others contrast. If the sequence is

    broken, it becomes possible to suppose that the reference to Aristotle was inserted by AS himself. And that is what Patillon suggests (xxvii-xxviii). He speaks of AShaving undertaken une vaste enqute chez les thoriciens anterieurs, and adds:Aristote en particulier a t lu pour cette occasion (lxxxix). Yet ASs knowledgeof Theodorus and Apollodorus appears to be mediated by Alexander (see, forexample, AS 49-51, where their definitions of narrative are reported, along withAlexanders criticisms, as a prelude to Alexanders own definition), and I havelittle doubt that his knowledge of older tradition was also indirect. At AS 207-8the cluster of references to Plato, Chrysippus and Aristotle is suggestive of adoxographic source; Neocles refers to the Stoics at AC 181; and the reference toAristotle at AC 191 surely derives from Alexander.

    Patillons willingness to credit AS with direct use of Aristotle is characteristicof his distinctively high estimation of the author, to whom he attributes about 57%of the text (the passages attributed to AS are listedand miscountedon p.xxx).On Patillons view, the author moves fluidly between reporting the views ofnamed authorities and providing his own summary of common doctrine. Bycontrast, Dilts and Kennedy see him as a pure compiler, probably not himself ateacher of rhetoric, who never advances an opinion of his own (xi). I am notunsympathetic to Patillons position in principle: it is too easy to think of thecomposition of technical works as a process of mindless compilation fromsources. Yet AS is overtly compilatory in its approach. And we know that theabsence of a name does not prove the absence of a source: the definition of pqoj at AS 6 is word-for-word the same as that at AS 223, but only the latter passagenames the source (Neocles).

    7

  • 8/10/2019 Anonymus Seguerianus - HEATH, M. (2005)

    8/30

    MALCOLM HEATH , N OTES ON THE A NONYMUS SEGUERIANUS

    Assigning material to a source is not always easy. Dilts and Kennedy produce(with due acknowledgement of its inherent uncertainty) a distribution that ismanifestly flawed (xiii-xiv). The mechanical application of a principle that we is

    characteristic of Alexander and you of Neocles splits the systematicallyconstructed analysis of the telos of the proem in AS 9b-18 between Alexander and

    Neocles. Similarly, they attribute AS 215-218 to Alexander, though these sectionsunpack the scheme summarised in 214, where Neocles is the named source. 17

    Even where we have a source clearly identified, it is wise to be cautious.Compare the list of definitions of epicheireme with the list of definitions ofenthymeme:

    AC 78-80 ( RG 7.752.5-9 + RG5.395.13-15)

    AS 157-159 (= AC 170-172 ) + AC 173 ( RG 7.762.18-763.10)

    picerhm stin, nqmhma d stin,j m n `Arpokratwn , qsijnomatoj ej ti mrouj zthma, prjtn kaqlou ka genikn zthsincon tn naforn.

    j m n Neoklj , lgojproeirhmnwn tinn per tozhtoumnou, ka per tokaqhgoumnou ato, ka tinasunchsin cntwn tn kroatnt ndon kefalaiwdj kasuneilhmmnwj prostiqej.

    j d Neoklj picerhm stisullogismj met tj okeaj tnmern podexewj.

    j d nioi, nqmhm sti toprohgoumnou piceirmatojsumprasma prosagmenon tzhtmati n mi peridJ.

    lloi d otwj: picerhm stilgoj xwqen lambanmenoj prjpdeixin tn pokeimnwn.

    j d `Arpokratwn , nqmhm stilgoj prj pdeixin lambanmenojtn pokeimnwn. tin j d otwj: nqmhm stin sugkataskeuzei t prokemenonkeflaion. Neoklj d ka otwj rzetai:nqmhm sti sunestrammnojlogismj telj kaq' n xwma,oc plon moion oc pljkfermenon.

    Two things in this tabulation awaken suspicion. 18 First, the definition ofenthymeme attributed to Harpocration is parallel in form to the definition ofepicheireme attributed to others, and quite different in form from the definitionof epicheireme attributed to Harpocration. Second, the definition of epicheiremeattributed to Harpocration appears again at AC 176 ( RG 7.763.16-20); but there it

    17 Unless a name has dropped out, the fhs in 219 indicates that we are still dealing with Neocles,the last named source; the next explicit change of source is at 221, to Alexander.18 The fact that Neocles makes two appearances in the list of definitions of enthymeme should notexcite suspicion: as Graeven (xii n.1) and Patillon (127 n.2) note, the first formula is concernedwith the content, the second with the form, of the enthymeme. In the introduction to the second

    formula the ka (absent from the manuscripts used by Walz, so not known to Graeven) shows thatthe duplication is not accidental.

    8

  • 8/10/2019 Anonymus Seguerianus - HEATH, M. (2005)

    9/30

    MALCOLM HEATH , N OTES ON THE A NONYMUS SEGUERIANUS

    appears in a context that otherwise seems to demand an attribution to Neocles.One has to wonder whether AS succeeded in keeping his research notes in perfectorder. Confusion is perhaps more likely when he is collating observations from

    several sources into a list than when he is reproducing a single sources doctrine inextenso . Nevertheless, this example suggests that we need to be cautious whenharvesting rhetoricians fragments from AS.

    Where there is not a name, we may be tempted to put our trust in the authorshelpful habit of refreshing the introductory identification of a source by name withthe occasional parenthetic he says. To take a particularly striking example, hedoes this no less than thirteen times in the nine sections introduced underHarpocrations name at AS 243. By this criterion, the he at AS 142 should signalthat we are still working with the last named source, Alexander (AS 136). YetGraeven (xxxiv-xxv) and Patillon (xxviii-xxix, and 95 n.6) have convincingly

    argued that the source has switched unannounced to Harpocration at AS 138. Iffollows that fhs in 142 is the orphan of an introductory naming that has presumably been lost in the abridgement.

    That analogy suggests that in AS 89 and 94 fhs either looks all the way backto AS 62, where Alexander was named, or else is refreshing an introductorynaming lost in the abridgement. Dilts and Kennedy must take the later view, sincethey divide the intervening material between Neocles and Alexander. But that isnot plausible: AS 63-98 is a carefully structured exposition of the three virtues ofnarrative, and its tight integration suggests a single source. 19 Patillon attributes thewhole to AS himself. That requires emendation: fhs is changed to fas. Thechange is certainly a small one. But Patillons reference to the transmitted fas atAS 79 is not to the point: there it is not used to convey a generally accepted

    precept, but to re-emphasise the writers agreement ( e fasi ) with the generallyheld opinion that narrative should be clear before he goes on to give advice onconcision that may seem inconsistent with the advice on clarity. Since we knowthat source citations can be lost, this emendation seems to me unwise. 20 I shalltentatively suggest in 5.1 that there is evidence in the indirect tradition whichsupports an attribution to Alexander son of Numenius, and which is indicative ofAlexanders debt to Dionysius of Halicarnassus.

    19 I am thinking of the overall architecture, not necessarily of all the detail. Editorial insertions intoa context largely derived from one source are possible. I do not think we can assume, for example,that the appearance of the same definition of nrgeia at AS 96 and AS 111 proves that both

    passages are from the same source: it may have been inserted into one context from the other bythe compiler. When Neocles definition of emotion is given unattributed at AS 6 it is followed by alist of emotions different from that given at AS 223, where it is attributed by name; I infer that thecompiler has inserted Neocles definition into a non-Neocles context.20 At 115.1 Kennedys fas is better motivated: the subject established in 113 is theApollodoreans, refreshed in 114 with the plural rwtsi . The corruption to the singular occurred

    under the influence of the adjacent tj (which is the subject of paraleyei , not of the parentheticverb of saying).

    9

  • 8/10/2019 Anonymus Seguerianus - HEATH, M. (2005)

    10/30

    MALCOLM HEATH , N OTES ON THE A NONYMUS SEGUERIANUS

    I conclude this section with two prosopographical points. 21 Patillon (lviii)confidently identifies the Harpocration used by AS with the Suda s AeliusHarpocration ( A4013), who wrote (among other works) an Art of Rhetoric and On

    Ideas , and with the issue-theorist known from the scholia to Hermogenes On Issues ; he also, more tentatively, identifies him as the author of the second of thetreatises falsely attributed to Aristides (a suggestion already advanced in Patillonsedition of pseudo-Aristides). But we know of at least three different rhetoriciansnamed Harpocration; since the name was a common one, there may have beenmore than three. I have repeatedly had to confront the problem of how to share outthe testimonia among them, and have never found objective grounds on which to

    base a solution. 22 If Patillon has arguments to support his identification, I wouldlike to know what they are.

    Patillon suggests that the Athenodorus cited by Alexander in AC 84 may be

    the sophist Athenodorus of Aenos, attested by Philostratus ( VS 2.14). But thechronology is problematic. Alexander is generally placed around the middle of thesecond century (Patillon xliii); but Philostratus Athenodorus (who died young)taught in Athens at the same time as Pollux, dated to the reign of Commodus bythe Suda (P 1951). 23 An obvious alternative is the Athenodorus of Rhodes cited byQuintilian (2.17.15), about whom we know nothing more. A less obvious

    possibility might be the Stoic Athenodorus son of Sandon. 24 We do not have anyother evidence that he wrote on rhetoric; but since Alexanders citation refers tothe structure of the epicheireme (Athenodorus is reported as holding that it hasseven parts) it is perhaps not necessary to assume a strictly rhetorical context.Porphyry ( In Cat . 4.1, 86.20-24) cites Athenodorus Against Aristotles Categories

    alongside Cornutus Art of Rhetoric and his reply to Athenodorus ( prj'Aqhndwron ntigraf ). This hypothesis is consistent with the fact thatAlexander is the probable source of the quotation of Cornutus at AS 242: the briefcharacterisations of the style appropriate to different parts of a speech in AS 241-243 are a single unit, and the use of qrasutra at 241 is parallel to AS 136,explicitly attributed to Alexander.

    4. The original treatise (i): integrating ACLike Graeven, Patillon does not contaminate the edition of the abridged AS

    with material known only from the indirect tradition. Unlike Graeven, he does not

    propose a reconstruction even in his introduction. In view of the speculative andnot always convincing nature of Graevens reconstruction, this restraint may be

    21 It would perhaps be too egocentric to complain that Patillons discussion of Zeno does not referto a recent attempt to collect and analyse the (meagre) fragments: Heath 1994. I have revised theanalysis in Heath 2004, 24-32.22 See Heath 2003a, 147; Heath 2003b, 132f.; 2004, 76f.where I should have emphasised evenmore firmly than I did the purely exempli gratia nature of the speculation which I was (perhapsunwisely) airing. Graeven (lxix) is agnostic about the identification with Aelius Harpocration,though he advances a wholly uncompelling argument for identifying ASs Harpocration with theissue-theorist.23

    Avotins 1975, 321f. argues that Pollux held his chair in Athens from the early 180s.24 RE Suppl. 5 (1931), 47-55.

    10

  • 8/10/2019 Anonymus Seguerianus - HEATH, M. (2005)

    11/30

    MALCOLM HEATH , N OTES ON THE A NONYMUS SEGUERIANUS

    well-advised. But perhaps there is something more to be said. In this section Ishall concentrate on material from AC that can be assigned to AS with confidence,reserving for 5 more speculative forays into the remoter reaches of the indirect

    tradition.

    4.1 Material on the proem

    The task of integrating material from AC into the abridged structure of ASmay seem rather straightforward when we look at AC 1-2, from Alexander on the

    place of the proem. The obvious approach is to exploit an analogy with the sectionon narrative: AS 124-131, on the place of narrative, comes between AS 113-123,on whether narrative is always needed, and AS 132-135, on whether there can bemore than one narrative, which in turn is followed by AS 136-137, on style. All ofthis material comes from Alexander, except 125-128, which is attributed to

    Alexander and Neocles; but the inconsistency between 127 (which says thatnarrative is sometimes possible after the proofs) and 130 (in which Alexanderdenies this) shows that 126-128 must be from Neocles, not Alexander. Theanalogy suggests that AC 1-2, on the place of the proem, should come between AS21-35, on whether a proem is always needed, and AS 37-39, on whether there can

    be more than one. AS 36 (which also belongs to Alexander: cf. 196, 240) fits moreclosely with what precedes: whether one has proem is conditional on whether thenarrative needs to be prepared (AS 35), but if there is a proem one must be carefulnot to drift into narrative or argument (AS 36). AC 1-2 could then be insertedrelatively smoothly between AS 36 and AS 37.

    4.2 Material on argumentationThe material on argumentation presents a more obvious difficulty. The series

    of definitions of pardeigma in AS 154-156 looks like an indivisible unit. It is puzzling, therefore, that the definition in AS 156 is the same as that in AC 83a,where it is integrated into a completely different context, the exposition of

    Neocles classification of epicheiremes. It is, however, possible to see what hashappened. The redactor responsible for the abridgement has excised the sectionson epicheiremes and enthymemes almost in their entirety, retaining only thisdefinition of paradigm and three (out of five) definitions of enthymeme. Thedefinitions of enthymeme have been left to stand as a self-contained unit; thesalvaged definition of paradigm has been attached to another pair of definitions of

    paradigm. The adaptation to the new context is clear: in AC 83 Zenos definitionis introduced as if adding another item to the series ekn , parabol ...(pardeigma d stin, j Znwn fhs... ), but in AS 156 it is introduced as ifadding another item to the series Neocles, Alexander... ( j d Znwn... ); theillustration which accompanies it in AC 83 has also been lost in AS 156.

    This observation raises two questions. Why did the abridger decide toeliminate epicheireme and enthymeme? And why did definitions of paradigmoccur at two different points in the original text? I have no answer to the first ofthese questions: the deliberate excision of epicheireme from a treatise oninvention seems quite extraordinary. But an analysis (and hypothetical

    11

  • 8/10/2019 Anonymus Seguerianus - HEATH, M. (2005)

    12/30

    MALCOLM HEATH , N OTES ON THE A NONYMUS SEGUERIANUS

    reconstruction) of the section on argumentation may help to answer the second.The basic structure is as follows:

    1. Introduction and definition (AS 143-144)

    2. Artistic and inartistic proofs (AS 145-155)

    (a) Alexander distinguishes inartistic from artistic proofs (AS 145a); hisclassification of inartistic proofs (AS 145b); he divides artistic proofs into

    paradigm and enthymeme (AS 146).

    (b) Neocles divides artistic proofs into those based on pqoj and those based on prgmata (AS 147). Definition of proofs based on pqoj (AS148). Proofs based on based on prgmata are further subdivided (AS149a) into ekj (AS 149b-150), tekmrion (AS 150-3), 25 andpardeigma (AS 154). Alexanders definition of pardeigma (AS 155) is

    added as a supplement to that of Neocles.3. Epicheireme (AC 78-87)

    (a) Definitions of epicheireme (AC 78-80)

    (b) Division of epicheiremes into didaskalik and duswphtik , i.e.moiwmatik, known by some as paradeigmatik (AC 81)

    (c) Neocles subdivision of duswphtik into ekn , parabol andpardeigma : what are (presumably) Neocles accounts of ekn andparabol (AC 82) are followed by Zenos definition of pardeigma (AC83a = AS 156), and a comment on the difference between pardeigma

    and parabol (AC 83b).It is now possible to explain why definitions of paradigm occur in two places.

    In Neocles theory paradigm is both a subdivision of artistic proofs based onprgmata and a subdivision of duswphtik epicheiremes. The twoclassifications are not entirely uncoordinated: didaskalik epicheiremes effect

    proof from (i) common notions or (ii) underlying facts or (iii) the ekta ortekmria consequent on circumstance (AC 81), while duswphtik epicheiremesare paradeigmatik . Thus the three categories of artistic proofs based onprgmata are reproduced in the classification of epicheiremes. (A diagram maymake this clearer: see Figure 1.) I suspect that Neocles, having already defined

    pardeigma at the level of artistic proof (AS 154), did not do so again at the levelof epicheireme. Taking the definition of pardeigma as read, he progresseddirectly from introducing the new concept of parabol (AC 82) to explaininghow it differs from the already familiar pardeigma (AC 83b). The compiler ofthe original AS noted the absence of a definition, and inserted Zenos definition ofpardeigma (AC 83a = AS 156) to fill the gap. 26

    25 On shmeon see the discussion of AS 149 in 2 above.26

    AS regularly juxtaposes Neocles and Alexander (AS 116, 125, 146-7, AC 174-5, 177-8, AS 169-70, 222-3), so it would be natural for their two definitions to go together.

    12

  • 8/10/2019 Anonymus Seguerianus - HEATH, M. (2005)

    13/30

    MALCOLM HEATH , N OTES ON THE A NONYMUS SEGUERIANUS

    Figure 1: Neocles classification of proofs and epicheiremes

    prgmapqoj

    pokemenaprgmata

    parakolouqontat peristsei

    ikj tekmrion

    koinannoiai

    piceirmataduswphtik

    (paradeigmatik)

    piceirmatadidaskalik

    pardeigma parabol

    ekj pardeigma tekmrion

    ekn

    Inartistic Artistic

    Proofs

    We may complete the analysis with a much briefer outline, starting with theremainder of the discussion of epicheireme:

    (d) The parts of an epicheireme (AC 84)(e) Refuting epicheiremes (AC 85-7)

    4. Enthymeme (AS 157-9 = AC 170-2, AC 173-198, AC 225-7)

    (a) Definitions (AS 157-9 = AC 170-2, AC 173)(b) Difference between enthymeme and syllogism (AC 174-5)(c) Difference between enthymeme and epicheireme (AC 176-8)(d) Neocles on syllogism (AC 179-183)(e) Enthymeme and gnome (AC 184-7)

    (f) Alexander on kinds and use of enthymeme (AC 188-196)(g) Kinds of enthymeme (AC 197)(h) Style in enthymeme (AC 198)(i) Epenthymeme (AC 225-7)

    5. Proposition ( prqesij ) (AS 160-8)

    Neocles on the first principle of demonstration ( rc podexewj )

    6. Topics (AS 169-185)

    (a) Alexanders definition (AS 169)(b) Neocles on common and special topics (AS 170-1)

    13

  • 8/10/2019 Anonymus Seguerianus - HEATH, M. (2005)

    14/30

    MALCOLM HEATH , N OTES ON THE A NONYMUS SEGUERIANUS

    (c) Neocles classification of geniktatoi topics (AS 171-182)(d) Using the topics (AS 183-5)

    7. Refutation of proofs (AS 186-191)

    (a) Introduction (AS 186)(b) Refutation of paradigmatic proofs (AS 187)(c) Refutation of inartistic proofs (AS 188-91)

    8. Tactical disposition of arguments (AS 192-5)

    9. Style and delivery (AS 196-7)

    Note that the section on the refutation of proofs (7a-c) and the earlier sectionon the refutation of epicheiremes (4e in the analysis) are separate parts of theoriginal; the reorganisation attributable to the abridger is not as far-reaching (or,therefore, as inscrutable) as Patillon xxiii supposes. Note also that we seem at this

    point to be dealing with a classification of proofs much less elaborate than Neocles scheme. We have, in my view, turned from Neocles to Alexander, whoseclassification divided artistic proofs into paradigms and enthymemes (AS 146) though one might wonder why there is no mention here of the refutation ofenthymematic proofs.

    5. The original treatise (ii): the indirect traditionAC and AS derive independently from the original, unabridged version of the

    treatise. AC provides the indisputable minimum evidence that an unabridgedversion of AS once existed, and I argued in 4 that it is possible to see how thematerial from AC fits into the original structure of the treatise. There remains thequestion whether other traces of the unabridged AS are found elsewhere in thetradition. 27 Graeven argued that there are; Patillon argues that the parallelsadduced by Graeven can all be explained as deriving from AS in the form that wehave it. The problem is potentially intractable. On the one hand, agreement

    between the direct and indirect traditions does not disprove access to theunabridged version unless we have independent reason to think the two versionsdiverged at that pointwhich will rarely, if ever, be the case. On the other hand,disagreement between the direct and indirect traditions does not prove access tothe unabridged version unless other explanations, such as subsequent editorialintervention, can be excludedand that, too, will rarely, if ever, be possible to dowith certainty. However, the more (and the more complex) the editorialintervention we need to posit, the harder it will become to maintain that there wasno access to the unabridged version; conversely, persistent failure to finddisagreements between the two traditions that cannot plausibly be explained aseditorial will leave us with no reason to infer access to the unabridged version.

    Patillons reconstruction of the history of the text tentatively postulates athird- or fourth-century intermediary between the unabridged AS and the (fifth- or

    27 There is no problem in principle with using the indirect tradition as evidence for establishing the

    text of where the transmitted abridgement is corrupt. Here I am concerned with the possibility thatthe indirect tradition provides evidence for the text of the treatise before it was abridged.

    14

  • 8/10/2019 Anonymus Seguerianus - HEATH, M. (2005)

    15/30

    MALCOLM HEATH , N OTES ON THE A NONYMUS SEGUERIANUS

    sixth-century) commentary on [Hermogenes] On Invention that is the ancestor ofAC (xvii). I am unclear what the nature of the inferred mediating text, describedvaguely as commentaires utilisant lAn. I, would be (at that date, certainly not a

    commentary on [Hermogenes] On Invention );28

    and it would be circular to assumethat the commentator did not have direct access to the unabridged AS withoutindependent grounds for suspecting the relatively early loss of the unabridged AS.That would be a reasonable suspicion if fifth-century sources show no evidence ofaccess to the unabridged ASbut whether that is the case is precisely the point atissue here. If there is evidence of access to the unabridged AS in fifth-centurysources, there will be no problem in supposing that fifth- or sixth-centurycommentator acquired the AC material directly from the unabridged AS; and if weare willing to entertain that possibility, there is no reason to be sceptical in

    principle about the preservation of traces of the unabridged AS in other fifth-century sources.

    If the terminus post quem for the loss of the unabridged AS must be left openat this point, so must the terminus ante quem for the abridgement. In Patillonsreconstruction, parallels in Marcellinus commentary on Hermogenes On Issues show that the abridgement cannot be later than the beginning of the fifth century(xxiv, xcii). Only one these parallels genuinely belongs to Marcellinus, and I shallargue in 5.3 that there is no way of telling whether he drew on the abridged orthe unabridged AS. The other parallels which Patillon from RG 4 are not correctlyattributed to Marcellinus; I shall argue in 5.4 that they actually provide evidenceof access to the unabridged ASbut this fact is robbed of chronologicalsignificance by the impossibility of dating the obscurely headed section of scholia

    in which these parallels appear.

    5.1 The commentary on On Invention

    I begin with a passage in the anonymous commentary that Patillon does notinclude in his edition. There is a clear parallel to AS 240, on the style appropriateto a proem, in RG 5.382.8-10:

    AS 240 RG 5.382.8-12 (with apparatus)

    j m n on suntmwj epen, frsij stw tn mern to lgoutoiath. to m n prooimou tn d tn prooimwn rmhenean perergj teka kista shmeidhj, 29

    de t m n klog perergn tee nai ka kista shmeidh:gensqai d atn toiathn nkfgV tij toj kfanej trpoujka tj glwsshmatikj lxeij...

    ka spermatikj cous tina tnpragmtwn tj dihgsewj. 30

    28 On the history of commentaries on rhetorical technography see Heath 2004, 69-73.29

    Patillon retains the transmitted susshmeidhj ; but the absence of parallels makes me worry thatthis may be too difficult a reading.

    15

  • 8/10/2019 Anonymus Seguerianus - HEATH, M. (2005)

    16/30

    MALCOLM HEATH , N OTES ON THE A NONYMUS SEGUERIANUS

    The commentary adds the observation that a style that is suitably perergoj can be achieved by avoiding toj kfanej trpouj ka tj glwsshmatikjlxeij . Could that be, as Graeven thought (lxvi), a trace of the unabridged AS?

    There is supporting evidence in a set of anonymous prolegomena to On Invention ( RG 7.52-54). The text at RG 7.52.11-13 is shorter than RG 5.382.10-12, but thereis enough to show that it is (so to speak) an abbreviation of the same moreexpansive text:

    RG 7.52.11-13 RG 5.382.8-12 (with apparatus)

    rmhnea d to prooimou t m nklog perergoj stw:

    tn d tn prooimwn rmheneande t m n klog perergn tee nai ka kista shmeidh:

    gensqw d toiath, qenkfgV tij toj mfanej trpouj .

    gensqai d atn toiathn nkfgV tij toj kfanej trpouj

    ka tj glwsshmatikj lxeij...Of course, the fact that the prolegomena and the commentary run parallel in

    itself does nothing to show that either derives the additional text from theunabridged AS; the prolegomena could derive from the commentary, or both fromsome other common source. But as Graeven pointed out (xiii, noted by Patillonxix-xx), the immediately preceding part of the prolegomena contains materialfrom AS that is not in the commentary ( RG 7.52.3-8 = AS 36.1-5, and RG 7.52.8-10 is based on AS 24-5). Moreover, the immediately following part of the

    prolegomena ( RG 7.52.11-14, more fully at RG 7.715.7-10) comments on thedelivery of the proem:

    RG 7.52.11-14 RG 7.715.7-10to d prooimou pkrisij metraka tracoj ka toj pokeimnoijokea.

    pkrisin d prooimou t metranka tracon ka toj pokeimnoijrmzousan prgmasi katstasinparcesqai t tora.

    This can be compared with the advice in AS on the delivery of narrative andargument (AS 136-7, 196-7). It is therefore a tempting inference that the whole ofthis part of the prolegomena ( RG 7.52.3-14) derives from AS in its unabridgedform. That would in turn confirm the provenance of the comment on tropes andglossematic vocabulary in RG 5.382.10-12. 31 This conclusion gains added

    plausibility from the fact that glwsshmatikj also occurs in AS 85, again pairedwith tropes: xnoij ka tropikoj ka mfibloij ka glwsshmatikojnmas. The parallel is significant, because glwsshmatikj is not commonword in the rhetorical literature. However, there is one rhetorician who appears to

    be addicted to it: I have found 11 occurrences of glwss- or glwtthmatikj inthe critical works Dionysius of Halicarnassus, seven times in association with

    30 Patillon places a full stop after pragmtwn , and reads tj d dihgsewj in place of thetransmitted tj dihgsewj. d dighsij ; I do not think this is necessary31 But I suspect that the examples which follow ( RG 5.382.12-14) are supplied by thecommentator: the Demosthenes example is taken from On Invention itself (4.10), while

    stomfzein appears to have entered the tradition of rhetorical commentary under the influence ofHermogenes Id . 247.13: see Heath 1999, 64-6.

    16

  • 8/10/2019 Anonymus Seguerianus - HEATH, M. (2005)

    17/30

    MALCOLM HEATH , N OTES ON THE A NONYMUS SEGUERIANUS

    tropikj , seven times in association with xnoj .32 It is, perhaps, conceivable thatthe anonymous commentator has added material inspired by Dionysius to what heextracted from AS. But it is more likely that Alexander, ASs source in 240, was

    influenced by Dionysius. This is supported by other parallels: compare thecombination perergj te ka kista shmeidhj in AS 240 (above) withDionysius Letter to Pompeius 5.3, t m n shmeidej ka perergon (245.5f.); 33 and compare AS 19 tn te rmhnean sugkeimnhn k tropikj mllon kashmeidouj lxewj with Isocrates 2, shmeiwdn nomtwn... tropikn frsin(56.17-19).

    5.2 Demosthenes scholia

    Patillon does not list the passage from the prolegomena discussed in 5.1among the testimonia to AS 240. Another passage in the prolegomena ( RG

    7.54.18-21) is cited as a testimonium for AS 7. But a Demosthenes scholion (sch.Dem. A 3 54.1 (1) Dilts) which preserves a fuller text of the same version of AS 7(it shares a variation in the order of the points with the prolegomena) is notidentified as a testimonium, though one variant is mentioned in the apparatus. The

    procedure does not appear to be entirely systematic. Since the scholion preservesthe illustrative quotations in a less thoroughly mangled form than either RG 7.54.16-21 or AS 7, its evidence should have been presented more fully:

    AS 7 RG 7.54.16-21 sch. Dem. 54.1

    lambnetai d tproomia k tessrwn

    totwn:

    lloi d p tessrwnm n ka ato fasin:

    ll' o tn atnpnth, ll ka trwn:

    p tessrwn dlambnontai proomia,

    k to ato, k tontidkou, k tn

    32 The references are (with page and line in Usener-Radermacher): Lys. 3 (10.18f., with tropikj and xnoj) ; Dem . 4 (135.7, with tropikj ); Thuc . 24 (361.5, with tropikj and xnoj .), 35(383.9f., with xnoj ), 50 (409.19, with tropikj ), 52 (412.8, with xnoj ); Thuc. id . 2 (422.18f.,with tropikj ), 3 (425.9, with xnoj ); Comp . 25 (124.14, with tropikj and xnoj ), 26 (137.8,with tropikj and xnoj ); Imit . fr. 6.3.2, 209.2f.). None of these refers specifically to the proem. Inote one other occurrence of the pairing glwsshmatik ka tropik : Galen 18a.414.16-415.3,on Xenophon.33 This, however, is an extract from On Imitation , and the parallel passage in the epitome (= Imit .fr. 6.3.2, 209.2f.: see preceding note) reads t m n glwsshmatikn ka perergon instead of tm n shmeidej ka perergon . The relationship between the letter to Pompeius and the epitomehas been disputed. In Heath 1989 I argued that the letter was based on a draft of On Imitation , theepitome on the published text; Weaire 2002 has shown that I failed to establish that conclusion,and I am now inclined to accept Weaires conclusion that the differences between the letter and theepitome are best explained as the work of the epitomator. That is certainly possible for most of thedifferences in the passage in question ( Pomp . 5.3, 243.4-10 ~ Imit . fr. 6.4.2, 209.2-7). I am not surethat t brh ka t pqh (209.7) can be the epitomators summary of 243.10-244.1, as Weairesuggests (358); but I would be prepared to accept a lacuna at 243.10 ( ka kat toj schmatismoj ). I hesitate only over the replacement of shmeidej with a wordso distinctive to Dionysius as glwsshmatikn ; but it is possible that an epitomator steeped inDionysius vocabulary might do this. Discussion of the relationship between the two texts has

    concentrated up to now on material that is additional to or missing from one or other text; asystematic study of their respective critical vocabularies might be useful.

    17

  • 8/10/2019 Anonymus Seguerianus - HEATH, M. (2005)

    18/30

    MALCOLM HEATH , N OTES ON THE A NONYMUS SEGUERIANUS

    dikazntwn, k tnpragmtwn. k to ato, j

    Dhmosqnhj n t katKnwnoj: brewj.

    toi gr f' auto j

    n t katKnwnoj:

    p auto, j de

    Dhmosqnhj:brisqej, ndrej'Aqhnaoi,

    n d p r troulgVj, ka totopishmanesqai de,sper pepohke Lusajlgwn: pitdeij mostin Arcippoj otos, dikasta.p to ntidkou, jn t kat Meidou:tn m n slgeian, ndrej,

    p to ntidkou, jn t kat Meidou:

    k to ntidkou, jn t kat Meidou:tn m n slgeian.

    k tn kenJsunagoreuntwn, jDhmosqnhj: e m n tJplestoi sunepoien, boul. k d tn kroatn tn dikastn, j'Isokrthj:edtej mj,

    ndrej 'Aqhnaoi.

    p tn kountwn,jn t Platak'Isokrtouj:

    ptn dikastn, j'Isokrthj n tPlatak: edtej mj,

    ndrej dikasta, katoj dikoumnoijbohqen eqismnouj,

    k d tn pragmtwn,j Lukorgoj n tkat' Atolkou: pollnd ka meglwn gnwneselhluqtwn odpoteper meiznwn ketediksontej. 34

    p to prgmatoj, jLukorgoj n t katAtolkou.

    p to prgmatoj, jLukorgoj n t katAtolkou: polln kameglwn gnwneselhluqtwn odpoteper thlikotoudiksontej kete.

    Another Demosthenes scholion (sch. RT Dem. 20.5 (20)) provides an

    additional witness to part of the unabridged AS preserved by the anonymouscommentary at AC 197. Patillon does not mention it in the apparatus or among thetestimonia, but it provides a variant reading that merits attention:

    AC 197-8 ( RG 7.766.4-16, 5.410.5-7) sch. Dem. 20.5 (20)

    scmata d nqumhmtwn tssara:

    podeiktikn, legktikn,sullogistikn, mikton.

    tssara to nqummatoj tscmata:podeiktikn, legktikn,sullogistikn, mikton.

    podeiktikn m n on sti t x podeiktikn sti t x

    34 I have transposed this item to match the order in prolegomena and commentary.

    18

  • 8/10/2019 Anonymus Seguerianus - HEATH, M. (2005)

    19/30

    MALCOLM HEATH , N OTES ON THE A NONYMUS SEGUERIANUS

    koloqwn klwn tn potagncon, oon: sJ tonun eqhkretton doken, ponhrn e nai,

    ka xj.

    koloqwn klwn tn pagwgncon, j Dhmosqnhj de: sJ dkretton.

    legktikn d, t x nantwnklwn sunistmenon, oon:t laben on tdidmena mologn nnomon e nai,ka xj.

    legktikn d sti t x nantwnklwn sunistmenon, j n tparapresbeaj: 35 t laben on tdidmena nnomon e nai mologn.

    sullogistikn d t llo xllwn deiknmenon, oon: gr,oj n g lhfqehn, ka xj.

    sullogistikn d sti llo xllwn podeiknon, sper: groj n g lhfqehn.

    miktn d toi piceirhmatikn, met tn prqesin eqj ma tpiceirmati kfretai, oon ttoinde: sper n etij kenwnlw, s tde ok n grayaj, kaxj.

    miktn d:

    sper gr e tij kenwnlw, s tde ok n grayaj. kat xj.

    reta d nqummatojbracthj klwn ka eruqma kattn snqesin tn nomtwn.

    ret d nqummatj stibracthj klwn ka eruqma kattn snqesin tn nomtwn.

    Patillon adopts potagn (PaPc) in preference to pitagn (MS). The latter isevidently incorrect, but Patillons text needs a lot of help from explanatory

    paraphrase in the translation (Dmonstratif celui o un clon se subordonne lautre comme sa suite logique). The scholions pagwgn is easier to handle: a

    demonstrative enthymeme is one which draws its inference from consequentialkla (as in Demosthenes 20.6: in the degree to which it is better to be thoughtnave than unscrupulous, to the same degree it is more honourable to repeal thislaw than to enact it). This seems to me a preferable reading.

    Neither of these scholia can be identified as an integral part of any of the lateantique commentaries identified in my source-critical study of the scholia, 36 andthey most probable derive from AC material after its extraction from theunabridged AS. So they do not provide any chronological fix on the transmissionof the treatise.

    5.3 Marcellinus

    Excerpts from Marcellinus commentary on Hermogenes On Issues , generallydated to the first part of the fifth century, are preserved in the compositecommentary printed in volume 4 of Walzs Rhetores Graeci . This compilation issometimes known as the Dreimnnerkommentar , because it derives mainly fromMarcellinus, Syrianus and Sopater, though these are by no means its only sources.

    Patillon cites a number of passages in Marcellinus which are parallel to AS.This is the first:

    35

    In fact On the Crown 119.36 Heath 2004, 132-213

    19

  • 8/10/2019 Anonymus Seguerianus - HEATH, M. (2005)

    20/30

    MALCOLM HEATH , N OTES ON THE A NONYMUS SEGUERIANUS

    AS 203 RG 4.417.12-15

    diairetai d plogoj ej edhdo, ej te t praktikn ka tpaqhtikn: ka to m n praktikostin nakefalawsij, to dpaqhtiko t t pqhkataskeuzein ka wnnein tnlgon.

    diairontai d o plogoi dic: tm n gr atn sti praktikn, td paqhtikn: praktikn m n t ntaj nakefalaisesi,paqhtikn d t n tajleeinologaij:

    Both passages divide the epilogue into praktikn and paqhtikn . The shareddoctrine is commonplace, but the shared use of praktikn is not: mostrhetoricians used pragmatikn as the antithesis to paqhtikn .37 Paradoxically,Patillon eliminates the two occurrences of praktikn in AS 203, substitutingpragmatikn by conjecture. This emendation brings AS 203 into line with the

    terminology of AS 113 and 221, but it removes the only significant parallel withMarcellinus. Patillon may have overlooked this point, since his apparatus does notrecord Marcellinus support for the reading praktikn (and his list of testimoniahere gives an incorrect reference).

    In fact, if Patillons conjecture is right there is a closer parallel to AS 203 in RG 7.331.10-16, a passage which probably derives from an early fifth-centurycommentary by John of Caesarea: 38

    AS 203 (after Patillon) RG 7.331.10-16

    diairetai d plogoj ej edhdo, ej te t pragmatiknka t paqhtikn: ka to m npragmatiko stin nakefalawsij,to d paqhtiko t t pqhkataskeuzein ka wnnein tnlgon.

    diairontai d o plogoidic, ej te t pragmatiknka ej t paqhtikn, t m n onpragmatikn nakefalawsin ceika nmnhsin tn n toj gsinerhmnwn: t d paqhtikn louqran, lou kboln, t m n tofegontoj t d to kathgrou: ceid fqnou kboln.

    If that parallel were enough to establish a connection between John and AS, weshould have to conclude that ASs t t pqh kataskeuzein is an epitomatorsversion of the more elaborate wording preserved at RG 7.331.13-16. So if John

    37 [Apsines] 10.2; PS 212.11f. Rabe; sch. Dem. 21.5 (25a), cf. 21.21 (50b); Syrianus 2.89.17-20Rabe.38 On the sources of the RG 7 scholia see Heath 2003c, 29-32 (with 71-89), distinguishing Johnfrom what I describe as the patchwork commentary, assembled from snippets out of the RG 4scholia. In the present case, the section RG 7.329.1-336.27 comes from John, while the followingsection, RG 7.336.28-337.29, is from the patchwork: 7.336.28-337.2 ~ 4.415.18-21 (Sopater),7.337.2-13 ~ 4.417.14-26 (Marcellinus: 7.337.3, 7 secure the reading praktikn in Marcellinus),7.337.13-29 ~ 415.21-416.1 (Sopater). The Sopater of RG 4 used John, so RG 7 sometimes derivesthe same material from John and from the patchwork (Heath 2003c, 32). Moreover, the patchworkderives from and sometimes preserves a fuller text than the extant version of the

    Dreimnnerkommentar (Heath 2003c, 30 n.66). It is good that we have the resources to improve

    the text of RG 4; but to secure these resources it is necessary to hunt down scattered snippets in RG 7a less than enticing a prospect.

    20

  • 8/10/2019 Anonymus Seguerianus - HEATH, M. (2005)

    21/30

    MALCOLM HEATH , N OTES ON THE A NONYMUS SEGUERIANUS

    was using AS, he was using the unabridged version. However, this suggestion isnot one I would wish to pursue. First, one might well feel that the verbal parallelsare too slight, and the shared doctrine too routine, to establish any direct

    connection. Secondly, I am doubtful whether Patillons conjecture in AS 203 iscorrect: why should standard terminology have been corrupted into non-standardterminology? If it is assumed that the authors terminology must have beenconsistent (a risky assumption), why should AS 113 and 221 be exempt fromsuspicion?

    In sum: if Patillons conjecture is right, then it is possible (but not certain) thatJohn provides evidence for the unabridged AS, but there is no reason to think thatMarcellinus was drawing on AS at all. If, on the other hand, Patillons conjectureis wrong, then the point of contact between Marcellinus and AS is noteworthy, buttoo slight to give any indication of whether he knew AS in its original or its

    abridged form. So this parallel cannot help in providing a terminus ante quem forthe abridgement.

    5.4 ... and others

    Can we learn more from the other parallels in Marcellinus? There are no other parallels in Marcellinus. Walz printed the whole of RG 4.417.1-429.5 as a singleextract under Marcellinus name, with no indication of a change of source. But ithas long been known that this is incorrect: Rabe published a list of corrections toWalzs headings, from which we learn that a new section begins at RG 4.422.18under the heading Metrophanes, Athanasius, Porphyry, and Polemo. 39 The factthat this section incorporates material from both [Apsines] and AS simplycompounds the obscurity of the portmanteau heading.

    All the other parallels which Patillon attributes to Marcellinus occur in thissection. 40 But even if they are not parallels in Marcellinus, they are undoubtedly

    parallels. So what can we learn from them? Patillon says (xxiv) that these citationsof AS are rcrites, abrges, entre-coupes dillustrations, and adds: rien, mon sens, nindique que ce rheteur... ait eu sous les yeux une version de notretrait plus developpe que celle que nous connaissons. Certainly, if one explainsthe differences between the RG 4 parallel and the surviving version of AS in thisway, no indication will be left that the source was anything other than thesurviving version. But is that justified? There is no doubt that there has been much

    abbreviation and paraphrase, and it seems likely that some material has beeninserted from other sources. But it is not easy to identify an insertion withconfidence. For example, since Minucianus is not one of ASs regular sources one

    39Rabe 1909, 588.40 The parallels are as follows (the examples discussed below are marked with an asterisk): RG *4.422.27-423.3 ~ AS 188-200; 4.423.3-6 ~ AS 207-8; 4.424.24-27 ~ AS 202; 4.424.27-30 ~ AS235;4.426.17-19 ~ AS 204; 4.426.31-427.2 ~ AS 212; 4.427.8-10 ~ AS 205; *4.428.4-29 ~ AS 19-

    20; *4.428.29-429.2 ~ AS 237. All the texts are set out in parallel columns in Heath 2002b, 12-20;there is a brief commentary in Heath 2003a, 165.

    21

  • 8/10/2019 Anonymus Seguerianus - HEATH, M. (2005)

    22/30

    MALCOLM HEATH , N OTES ON THE A NONYMUS SEGUERIANUS

    must harbour suspicions about his introduction into the series of definitions ofepilogue at RG 4.422.30-423.3: 41

    AS 198-200 RG 4.422.27-423.3

    plogj stin, j m n Neoklj,lgoj p proeirhmnaij podexesinpilegmenoj, pragmtwn qroismnka qn ka paqn pericwn.j d tinej, mroj lgou statonpmenon podexesin.

    plogoj d stilgoj p proeirhmnaij podexesinpilegmenoj, qroismn pragmtwnka qn ka paqn pericwn,

    pnodoj erhmnwn, 42 j d 'Alxandroj,lgoj pirrwnnj t erhmna.

    lgoj pirrwnnwn t erhmna.kat d Minoukiann lgojdenwsin mewsin cwn tnpepragmnwn, lgoj gnsin cwn tnpepragmnwn, n ka kaloumnhdiatpwsij diaskeu pqojkinosa ka prj narg tnpepragmnwn xtasin tn dikastngousa.cousi d o plogoi kaparaklseij ka nakefalaiseijka pidihgseij.

    And yet there is no reason in principle why a third-century compiler should nothave supplemented his main sources with a reference to one of the leadingtheorists of the late second century, just as he made occasional reference to theslightly earlier Zeno. 43 Certainty is elusive.

    Moreover, the additional material in the scholia does not always fall intoisolated and easily excised blocks. At RG 4.428.4-29 we have a continuous textthat is parallel to AS 19-20 and 237, but substantially more expansive:

    AS 19-20 RG 4.428.4-29

    diafrei d to pilgou tproomion, ti n m n t prooimJ

    t scma ka tn rmhnean

    diforon d atn

    t scma ka rmhnea to lgou.

    mtrion e nai de katiqassn j n epoi tij,

    t m n gr to prooimou scmatamtria e nai de ka pia ka j ntij epoi tiqass.peid gr n rc nfousi mllono kroata, ka opw nakeknhtaiatn t pqoj, moiopaqen de

    41 Thus Graeven xix.42 In the note on pnodoj ( RG 4.422.29f.) in Heath 2002b, 14 I should of course have referred toPlato Phaedrus 267d3-6, and mentioned [Hermogenes] Meth . 427.22-428.6, Syrianus 2.89.17-25

    Rabe, as well as Aelius Aristides 4.21 and Longinus Rhetoric 48.86 Patillon.43 On Zeno and Minucianus see Heath 2004, 24-36.

    22

  • 8/10/2019 Anonymus Seguerianus - HEATH, M. (2005)

    23/30

    MALCOLM HEATH , N OTES ON THE A NONYMUS SEGUERIANUS

    toj koousi ka rmaprobibzein t te autn ka ttn kroatn pqoj. sti 44 d

    toto n toj te scmasi metroijka taj lxesi ka tajsunqsesin, ti d ka tajpokrsesin metraij crmeqa.

    n d' pilgoij t scmasugkekinhmnonka pollj m n mboseij con,polloj d scetliasmoj,

    d plogoj tonantonkekinsqai toj scmasin feleika pollj m n kboseij cein,polloj d scetliasmoj.

    tn te rmhnean sugkeimnhn ktropikj mllon ka shmeidoujlxewj, dunamnhj mntoi pesen ejpolitikoj lgouj.

    ka t m n proomia sustrofn ceitj lxewj, d plogojlelumnhn tn frsin.

    ti d ka totJdiafrei, ti poll tn n tojprooimoij okt'n pilgoij lekton.

    o mn ll ka lik tij stidiafor. poll gr tn n tprooimJ lecqntwn ok ngkhlgein n toj pilgoij,oon popteeta tij di periergan di polupragmosnhnesercmenoj toj gnaj: luqeshjtj poyaj n t prooimJ ok ti

    ngkh n toj pilgoij pertotou lgein.o mnon d tn prooimwn estinej lai,

    aj o crmeqa n toj pilgoij,

    ka llai tinj esi prooimiakalai, atinej tan diaperaiwqsinn toj prooimoij,perittn poiosin n toj pilgoijtn per autn lgon.

    ll ka tn pilgwn, n okstin n toj prooimoijcrea.

    es d ka n toj pilgoij laitin j aj n toj prooimoij ocrmeqa,

    poisomen t proomion, e tn

    kefalawn tn nagkawn ntotoij tj polyeij lhymeqa,

    oon per tn kefalawn ok ceikalj t tj polyeij n tojprooimoij lambnein:lkopoisomen gr t proomion.

    n d t pilgJ ngkh psa nti lambnein atn ej prrwsin parathsin.

    n d toj pilgoij ngkh labenti atn ej prrwsin.

    AS 237 RG 4.428.29-429.2

    diafrei d plogoj toprooimou ka kat tn lxin ka

    diafrei on,

    44

    stai RG 7.347.5. The patchwork commentary (see n.38) at RG 7.347.2-12 reproduces thesection from 428.6 ( peid gr n rc ) to 428.16 ( lelumnhn tn frsin ).

    23

  • 8/10/2019 Anonymus Seguerianus - HEATH, M. (2005)

    24/30

    MALCOLM HEATH , N OTES ON THE A NONYMUS SEGUERIANUS

    kat tn dinoian. kat m n tnlxin, ti n kenJ m n metrane nai de ka pan, n totJ d

    sugkekinhmnhn ka polljmboseij cousan kascetliasmoj. kat d tndinoian,ti ke m n pqoj mpoisai de,ntaqa d propn axsai kapirrsai.

    ti t m n pqoj paraskeuzei, d plogoj axei, d fegwn peirsetai n m n tojprooimoij meisai t pqoj katn diaboln (topon gr pnthnairen x rcj pr tnpodexewn), n d toj pilgoijmet tn pdeixin nairen kadiarrdhn kbllein piceirsei.

    We face a choice between two hypotheses: either additional material has beeninserted into the RG 4 parallels (as Patillon supposes), or these parallels and theabridged AS represent independent selections from a common source. It is verydifficult to identify things that could not possibly be due to editorial intervention;

    but there is nothing here, so far as I can see, that must be due to editorialintervention. It is worth noting that the imagery in RG 4.428.7 ( nfousi )corresponds with AS 137 ( nhflion ) and AS 239 ( kritj mequskmenoj tojpqesin ).45 I have argued in 2 that these scholia provides us with the solution toa textual problem in AS 20 ( RG 4.428.27f. ~ AS 20.6). If the transmitted version

    of AS had presented us with a text that was a combination of these two columns,is there anything that would have aroused our suspicions? 46

    There is, then, a respectable case for holding that the source for the parallelsin this section of RG 4 was the unabridged AS.

    5.5 Nicolaus

    Nicolaus Progymnasmata show a number of evidences of using AS. In mostcases, I agree with Patillon (xxiv) that there is no reason to suspect use of a fullertext than has survived in the direct tradition. 47 In one case, however, Graeven mayhave a point. He suggests (xx) that Nicolaus comments on the style of the

    prologue suggest a fuller text:AS 243 Nicolaus 46.12-14 Felten

    d tn pilgwn frsij paqhtiksti ka tolmhrotra toj nomasi

    dion d pilgwn t

    45 It does not seem to be a common image: but cf. Longinus F48.155 Patillon-Brisson: tojkoontaj, nfontoj to dikasto katarcj ka mgista, ka safj kosai qlontoj .Philostratus VS 573, in which Herodes describes Alexander Peloplaton as Scopelian sober, usesthe same language in a different way.46 To as an aid to readers who wish to try this experiment, I have given some (purely exempli

    gratia ) reconstructions in the Appendix.47 Felten xxxii adds another argument to Graevens case, but it is also unconvincing.

    24

  • 8/10/2019 Anonymus Seguerianus - HEATH, M. (2005)

    25/30

    MALCOLM HEATH , N OTES ON THE A NONYMUS SEGUERIANUS

    ka taj deinsesi ka tojscetliasmoj

    peripaqestra.

    deinseij paralambnein kascetliasmojka paqhtikn lwj rgzesqai tn

    frsin ka at d t pokrseikecrsqai peripaqestrv,

    Note that what is missing in AS is a comment on delivery. The delivery ofnarrative and argument is mentioned in AS (136-7, 196-7); we saw in 5.1 that atrace of the missing discussion of the delivery of the proem may be preserved inthe anonymous prolegomena ( RG 7.52.11-14, RG 7.715.7-10). Note, too, that RG 4.428.11f., part of the series of parallels discussed in 5.4, comments on thedelivery of the proem by contrast with the epilogue. So a cumulative case hasemerged: either comments on delivery have been inserted into material from ASindependently in three different sources, or the abridger had a tendency (notcompletely carried through) to eliminate or abbreviate references to delivery inthe original AS. The latter is the more economical hypothesis.

    5.6 John of Sardis

    There is another commentary on On Invention , of which the first part onlywas published by Walz ( RG 6.507-541, attributed to George the Divider). 48 The

    prolegomena were re-edited by Rabe ( PS 351-360), who identifies the author ofthe commentary as John of Sardis ( PS cii), but it has never been published infull. 49 Graeven reports (xvii) that, after the three definitions of epicheireme foundin AC 78-80 ( RG 7.752.5-9 + RG 5.395.13-15), discussed above, John adds pinhma kataskeuzei t prokemenon zthma . As Graeven points out, this

    fourth definition of epicheireme corresponds to the fourth definition ofenthymeme, just as the third definition of epicheireme does to the third definitionof enthymeme (see 3 above):

    picerhm sti... nqmhm sti...

    lgoj xwqen lambanmenoj prjpdeixin tn pokeimnwn.

    lgoj prj pdeixin lambanmenojtn pokeimnwn.

    pinhma kataskeuzei tprokemenon zthma.

    sugkataskeuzei t prokemenonkeflaion.

    This instance obviously does not raise a question of access to the original AS

    independent of the material preserved in AC; rather, if the definition was part ofthe original version of the treatise, its absence from AC is probably due anaccident of transmission. Admittedly, the additional definition of epicheireme isalso found in Rufus (27); so one might wonder whether it has been added in John

    48 George the Divider is, presumably, the fifth-century commentator on Hermogenes On Issues .The manuscript superscriptions of this commentary show considerable uncertainty aboutauthorship: see Rabes apparatus, PS 351.9-14.49 An edition of the commentary on book 2 is promised in the summary of Rosario Scaliasdoctoral thesis, Un commentario inedito al per ersewj pseudo-ermogeniano(http://www2.reggionet.it/filosofia/allegati/2000/Scalia.pdf). Scalia accepts the attribution to

    George, but does not appear to be aware of Rabes discussion (she reports Vat. gr. 901, alreadyused by Rabe, as a newly discovered witness to the text).

    25

  • 8/10/2019 Anonymus Seguerianus - HEATH, M. (2005)

    26/30

    MALCOLM HEATH , N OTES ON THE A NONYMUS SEGUERIANUS

    rather than lost in AC. But the parallel, if anything, supports its place in AS: Rufusalso includes definitions of narrative, paradigm and enthymeme (17, 31, 35)which are found in AS (48, 156, 158 = AC 171). 50 Patillon does not mention

    Graevens observation; but the definition certainly deserves at least to bementioned in an apparatus.

    Graeven also argued (xvi) that John ( PS 358.6-11 = RG 6.511.4-12) made useof an unabridged version of AS 1.

    AS 1 John of Sardis PS 358.6-14

    politikj toi dikanikj lgojej tssara mrh diairetai tprokemena:

    crzomen gr n atprooimwn m n prj tprosecestrouj poisai tojkroatj, dihgsewj d prj tdidxai t prgma, tn d pstewnprj t kataskeusai naskeusai t prokemenon: tojd pilgouj pgomen prj tpirrsai tn koonta ej tnp r mn yfon.

    ston d, j politikj lgojej tssara mrh diairetai,proomion, dighsin, psteij,plogon, per n nn `Ermognhj ntoj tris tmoij didskei:crzomen gr n t politik lgJprooimwn m n prj tprosecestrouj poisai tojkroatj, dihgsewj d prj tdidxai t prgma, tn d pstewnprj t naskeusai kakataskeusai t prokemenon, tnd pilgwn prj tpirrsai tn koonta ej tnp r mn yfon.

    We noted in 2 that Johns evidence may support the view that toi dikanikj isan interpolation. But that is disputed; and the interpolation might in any case

    postdate the abridgement. So this does not provide evidence that John usedanything other than the transmitted version. Nor does the fact that he lists the four

    parts of a speech: that could easily be editorial, as the reference to Hermogenesclearly must be. Such editorial activity is also clearly in evidence in the parallelwhich John provides ( PS 358.24-359.11) to AS 4: his text is much longer than theanonymous prolegomena or AS itself, but the main insertion is found also in RG 7.56.5-13, and has presumably been pasted in by John from another source. 51

    50 The first two are attributed to Zeno and the last to some. If Rufus uses Zenos definition of paradigm ( genomnou prgmatoj pomnhmneusij ej mowsin to nn zhtoumnou ), thenthe parallel definition of parabol (ntoj ka ginomnou prgmatoj pomnhmneusij prjmowsin to zhtoumnou Rufus 32) presumably belongs to Zeno also. It might then seemtempting to infer that Rufus has made extensive use of Zenos work on epicheiremes (attested inthe Suda ). But if Rufus definition of epicheireme ( pinhma kataskeuzei t prokemenonzthma ) is Zenos, then we should expect the corresponding definition of enthymeme in AC 173 ( sugkataskeuzei t prokemenon keflaion ) to belong to Zeno as well; but Rufus definitionof enthymeme (35) is that of AS 158 = AC 171. So we cannot assume that Rufus consistentlyfollows Zeno in his section on argument. 51

    Patillons apparatus to 4.4f. cites George from RG 6:511.27: consistency of practice would leadone to expect a reference to John of Sardis, PS 359.5.

    26

  • 8/10/2019 Anonymus Seguerianus - HEATH, M. (2005)

    27/30

    MALCOLM HEATH , N OTES ON THE A NONYMUS SEGUERIANUS

    So far, then, we have found no evidence in John of access to the unabridgedversion of AS independent of AC. But I am tantalised by one passage that has no

    parallel in AS. At PS 357.21-358.5 John addresses the definition of invention: 52

    eresij tonun stn pnoia nohmtwn proshkntwn ka rmodwnt pokeimnJ ka zhtoumnJ prgmati. eresj stin pnoiatn piqanj prosntwn toj prokeimnoij problmasin nomtwnte ka nohmtwn: nomata d esin nqumhmatik kaparadeigmatik. per d tj lxeij katagnetai eresij prj teren pithdeaj lxeij ka pepoihmnaj prj t pokemenonprswpon .

    This passage immediately precedes the parallel to AS 1. But it was not that juxtaposition which drew it to my attention, so much as a train of thought prompted one of Patillons conjectures:

    AS 154 pardeigma d, j Neoklj, mfer j ka moion ka ekjt zhtoumnJ prgmati.

    Patillon proposes reading prgm ti . My first thought was that the conjecture,though neat, produced an implausible word-order. My second thought, morefavourable to the conjecture, was a suspicion that t zhtomenon prgma wouldnot be a common turn of phrase, since rhetoricians usually find t zhtomenon sufficient. My impression was correct: t zhtomenon prgma is rare. But one ofthe very few examples that I found is in this passage from John of Sardis. Let ustherefore try a thought experiment: if the original version of the treatise containedmaterial before AS 1, what would it have looked like? The author speaks of own

    work as On Invention (AS 246),53

    so a definition of invention would not have been out of place; and it is characteristic of AS to give more than one definition most probably, those of his favourite sources, Neocles and Alexander. If the nameshad been lost in transmission it would be unreasonable to expect the definitions to

    be identified by unique features of language or doctrine; but if we were lucky, wewould find matches to what little we know of the two theorists. In this case, wehave found a parallel in the first definition of invention to Neocles use of tzhtomenon prgma at AS 154; and the second definitions division of thoughtinto nqumhmatik and paradeigmatik agrees with Alexanders division ofartistic proofs (AS 146). This division is, of course, by no means unique toAlexander. 54 But note, too, that the second definition, unlike the first, explicitlyextends invention to cover style ( nmata, lxij ) as well as thought ( nomata );this is consistent with Alexanders discussion of the styles appropriate to different

    parts of a speech. This passage, therefore, if we restore a naming formula (forexample, j m n Neoklj... j d 'Alxandroj... ), is exactly what we should

    52 Partial, and more or less distant, parallels: PS 236.1-8, PS 331.21f., RG 7.698.8-11.53 Cf. the cross-reference in AS 179 tn d mchn n toj per ersewj scolikoj, n ojper piceirhmtwn lgomen, didxamen . Like Patillon (107 n.9) I take the reference to be tothe discussion of mch at AC 190, 193-5though this is in the section on enthymemes, notepicheiremes: so the possibility that part of the original section on epicheiremes was omitted from

    AC must also be considered.54 See (e.g.) Aristotle Rhetoric 1.2, 1356a35-b17; 2.20, 1393a23f.

    27

  • 8/10/2019 Anonymus Seguerianus - HEATH, M. (2005)

    28/30

    MALCOLM HEATH , N OTES ON THE A NONYMUS SEGUERIANUS

    have expected in a hypothetical lost introductory paragraph to AS. Perhaps that iswhat it is.

    6. ConclusionI began describing Patillons edition as the default choice for serious study of

    AS. The doubts and disagreements I have expressed are not intended to cast doubtthat judgement. No edition of a technical compilation preserved in abridged formin a single manuscript can hope to answer all the questions conclusively. Therelevant criterion is the extent to which it enables discussion to move to a newlevel. Patillon amply satisfies this criterion. It is precisely because his work putsme in a stronger position to engage with this text and pose questions about it thanever before that I have been able to reach conclusions that sometimes go beyondand sometimes diverge from his. Not for the first time, I have come away from

    one of Patillons editions immensely stimulated and with a profound sense ofadmiration and gratitude.

    AppendixThese hypothetical reconstructions of sections of the original treatise, based

    on the abridged AS and the parallels in RG 4, are offered purely exempli gratia , toillustrate how easily the two sources can be read as independent selections from acommon source (see 5.4 and n.46 above).

    *AS 19-20 : diafrei d to pilgou t proomion kat t scma ka tnrmhnean. t m n gr to prooimou scmata mtria e nai de ka piaka j n tij epoi tiqass. peid gr n rc nfousi mllon okroata ka opw nakeknhtai atn t pqoj, moiopaqen de tojkoousi ka rma probibzein t te autn ka t tn kroatnpqoj. stai d toto n toj te scmasi metroij ka taj lxesi kataj sunqsesin, ti d ka taj pokrsesin metraij crmeqa. dplogoj tonanton sugkekinsqai toj scmasin felei ka pollj m nkboseij cein, polloj d scetliasmoj, tn te rmhnean sugkeimnhnk tropikj mllon ka shmeidouj lxewj, dunamnhj mntoi pesen ejpolitikoj lgouj. ka t m n proomia sustrofn cei tj lxewj, dplogoj lelumnhn tn frsin. o mn ll ka lik tij sti diafor.poll gr tn n t prooimJ lecqntwn ok ngkh lgein n tojpilgoij, oon popteeta tij di periergan di polupragmosnhnesercmenoj toj gnaj: luqeshj tj poyaj n t prooimJ ok tingkh n toj pilgoij per totou lgein. ka llai tinj esiprooimiaka lai, atinej tan diaperaiwqsin n toj prooimoij,perittn poiosin n toj pilgoij tn per autn lgon. es d ka ntoj pilgoij lai tin j aj n toj prooimoij o crmeqa, oon per tnkefalawn tn nagkawn ok cei kalj t tj polyeij n tojprooimoij lambnein: lkopoisomen gr t proomion. n d t pilgJngkh psa n ti lambnein atn ej prrwsin parathsin.

    *AS 198-200 : plogj stin, j m n Neoklj, lgoj p proeirhmnaijpodexesin pilegmenoj, pragmtwn qroismn ka qn ka paqn

    28

  • 8/10/2019 Anonymus Seguerianus - HEATH, M. (2005)

    29/30

    MALCOLM HEATH , N OTES ON THE A NONYMUS SEGUERIANUS

    pericwn. j d tinej, mroj lgou staton pmenon podexesin, pnodoj erhmnwn. j d 'Alxandroj, lgoj pirrwnnj t erhmna.kat d Minoukiann lgoj denwsin mewsin cwn tn pepragmnwn:

    lgoj gnsin cwn tn pepragmnwn, n ka kaloumnh diatpwsij diaskeu pqoj kinosa ka prj narg tn pepragmnwn xtasin tndikastn gousa. cousi d o plogoi ka paraklseij kanakefalaiseij ka pidihgseij.

    *AS 237-9 : diafrei d plogoj to prooimou ka kat tn lxin kakat tn dinoian. kat m n tn lxin, ti n kenJ m n metran e naide ka pan, n totJ d sugkekinhmnhn ka pollj mboseij cousanka scetliasmoj. kat d tn dinoian, ti t m n pqoj mpoisai de, d plogoj propn axsai ka pirrsai. d fegwn peirsetai nm n toj prooimoij meisai t pqoj ka tn diaboln (topon gr pnthnairen x rcj pr tn podexewn), n d toj pilgoij met tn

    pdeixin nairen ka diarrdhn kbllein piceirsei. d rmhneaka pnu pwj rmzei, e qrasea eh ka tetolmhmnh, ka te lxijpolln cousa tn tropikn paraskeun, te snqesij kekainwmnh kaparakinosa tn yucn. ka t scma d scetliasmoj ctw ka ecjka t toiata. gnwston d ti t pqh ka n toj kefalaoij msoijparalhfqsetai, mlista p tn sqenn poqsewn. sumperistlletaigr t sqen j to prgmatoj t tj paqhtikj daj perbol, ka kritj mequskmenoj toj pqesin okti t krib j dior to prgmatoj,per ka Dhmosqnhj pepohken n t parapresbeaj: ka grnekefalaioto n msoij ka tj axhtikj laj kstJ parspeiretn kefalawn.

    BibliographyAvotins, I. (1975) The holders of the chairs of rhetoric at Athens, HSCP 79, 313-24

    Boys-Stone, G. (2003) The Stoics two types of allegory, in G. Boys-Stone (ed.), Metaphor, Allegory, and the Classical Tradition (Oxford), 189-216

    Dilts, M.R. (1983-6) Scholia Demosthenica (Leipzig)

    (1992) Scholia in Aeschinem (Stuttgart)

    Dilts, M.R. and Kennedy, G.A. (1997) Two Greek Rhetorical Treatises from the Roman Empire: introduction, text, and translation of the Arts of Rhetoric,attributed to Anonymous Seguerianus and to Apsines of Gadara (Leiden)

    Felten, J. (1913) Nicolai progymnasmata (Leipzig)

    Graeven, J. (1891) Cornuti artis rhetoricae epitome (Berlin)

    Heath, M. (1989) Dionysius of Halicarnassus On Imitation Hermes 117, 370-3

    (1994) Zeno the rhetor and the thirteen staseis , Eranos 92, 17-22

    (1999) Longinus On Sublimity , PCPS 45, 43-74

    (2002a) Notes on pseudo-Apsines, Mnemosyne 55, 657-68

    29

  • 8/10/2019 Anonymus Seguerianus - HEATH, M. (2005)

    30/30

    MALCOLM HEATH , N OTES ON THE A NONYMUS SEGUERIANUS

    (2002b) Porphyrys rhetoric: testimonia and fragments, Leeds InternationalClassical Studies 1.5, 1-38

    (2003a) Porphyrys rhetoric, CQ 53, 141-166

    (2003b) Theon and the history of the progymnasmata, GRBS 43, 129-60

    (2003c) Metalepsis , paragraphe and the scholia to Hermogenes, Leeds International Classical Studies 2.2, 1-91

    (2004) Menander: a rhetor in context (Oxford)

    Kennedy, G. (2003) Some recent controversies in the study of later Greekrhetoric, AJP 124, 295-300

    Patillon, M. (2001) Apsins, Art rhtorique. Problmes faux-semblant (Paris)

    (2002) Pseudo-Aelius Aristide, Arts rhtoriques (Paris)

    (2005) Anonyme de Sguier. Art du discours politique (Paris)

    Patillon, M. and Bolognesi, G. (1997) Aelius Thon (Paris)

    Patillon, M. and Brisson, L. (2001) Longin. Fragments. Art Rhtorique. Rufus. Art Rhtorique (Paris)

    Rabe, H. (1908) Aus Rhetoren-Handschriften: 5. Des Diakonen und LogothetenJohannes Kommentar zu Hermogenes Per meqdou deinthtoj RM 63, 127-51

    (1909) Aus Rhetoren Handschriften: 11. Der Dreimnner Kommentar WIV, RM 64, 578-89

    (1928) Ioannis Sardiani Commentarium in Aphthonii Progymnasmata (Leipzig)

    (1931) Prolegomenon Sylloge (Leipzig)

    Vottero, D. (2004) Anonimo Segueriano: Arte del discorso politico (Alessandria)

    Walz, C. (1832-6) Rhetores Graeci (Stuttgart)

    Weaire, G. (2002) The relationship between Dionysius of Halicarnassus Deimitatione and Epistula ad Pompeium , CP 97, 351-359