University of Michigan 2009 ‘Theory into Practice: Aristotelian Principles inAristarchean Philology’ Schironi, Francesca ‘Theory into Practice: Aristotelian Principles in Aristarchean Philology’, CPh 104 (2009): 279-316.http://hdl.handle.net/2027.42/150117Deep Blue deepblue.lib.umich.edu279Classical Philology 104 (2009): 279–316[ç 2009 by The University of Chicago. All rights reserved] 0009-837X/09/10403-0002$10.00THEORY INTO PRACTICE:ARISTOTELIAN PRINCIPLES IN ARISTARCHEAN PHILOLOGYfrancesca schironihe peripatetic influence on many of the fields developed at Alex-andria is undeniable: chronology, ethnography, paradoxography, glos-sography, literary biography, and bibliography (pinakes), as well asmedicine and mechanics, 1 owe much to the Aristotelian school and itsapproach to Wissenschaft. In recent years, against Pfeiffer’s refusal to seeany link between Aristotle and the Alexandrian grammarians,2 scholars likeGallavotti, Nickau, Lührs, Porter, Montanari, Richardson, and Matthaios3have argued in favor of the influence of Aristotle on Alexandrian philologyand in particular on Aristarchus’ scholarship. Some parallels betweenAristotle and Aristarchus can be found in the distinction between Homerand the Cyclic poets; in the idea of the tevloÍ of a work (in particular, thefamous statement that the Odyssey had reached its tevloÍ at 23.296);4 inthe theory that the Iliad and the Odyssey are creations of one poet, Homer(schol. Il. 5.60a, 11.147a); and in the importance of the principle of consis-tency (Homer does not contradict himself ).In this paper, I would like to return to this issue, focusing in particular onthe intellectual relationship between Aristotle and Aristarchus. Passages fromthe Rhetoric and above all from the Poetics will be compared to the Aris-tarchean sources from the Homeric scholia. In order to proceed in my analysis,I will take into account only the scholia maiora to the Iliad and the Odysseyand, among them, only those by Aristonicus and, with more caution, byDidymus.5 These, I believe, are the only secure sources for Aristarchean1. A relationship between Aristotle and Alexandrian medicine, both sharing the principle of teleology,has been highlighted by von Staden (1997).2. Pfeiffer 1968, 67, 87–88, 95, 272. For a criticism of Pfeiffer’s view, see in particular Rossi 1976,110–14; and Montanari, in Montanari 1994, 2, 29–31.3. Gallavotti 1969; Nickau 1977, 132–83 (on Zenodotus and his analysis of the narrative contradictions);Lührs 1992, 13–17; Porter 1992, esp. 74–80; Montanari 1993, esp. 259–64; Richardson 1993, 35–36;1994; Matthaios 1999, passim; 2002, 174–77, 189–90; cf. also Podlecki 1969; and Montanari 2001.4. Cf. Gallavotti 1969 and Erbse 1972, 166–77.5. As a general principle, all the scholia quoted in the present study are by Aristonicus. I always alertthe reader in the few cases when I discuss a scholium by Didymus, a much more independent scholar, whohence is a much less reliable source for Aristarchean material.TThis is a revised version of a paper that I presented at the APA, in Montreal, January 2006. I would liketo thank Richard Janko, who read a first version of this paper, the two anonymous referees of ClassicalPhilology for their comments and suggestions, and Thomas Jenkins for proofreading the final version.Translations are mine unless otherwise noted.Francesca Schironi280material when the name of the grammarian is not expressly quoted. All theother scholia that do not explicitly mention Aristarchus are excluded in thepresent work, however “Aristarchean” they may sound.61. Some Preliminary Remarks: Aristotle’s Theory,Aristarchus’ Practice, and the Question of TerminologyThe comparison between Aristotle and in general the Peripatetic school, onthe one hand, and Aristarchus and the work done at the Museum, on the other,must be put in the right context, for it is clear that the ethos and essence ofthe work of Aristotle and that of Aristarchus were fundamentally different.Whereas Aristotle wrote theoretical treatises, Aristarchus did not produce asingle speculative work; all his theoretical background must be inferredfrom his own practice in dealing with the eßkdosiÍ of Homer, as witnessed bythe Homeric scholia. We can thus oppose Aristotle’s theory to Aristarchus’practice and see whether and, if so, how this Aristotelian conceptual frame-work fits Aristarchus’ methodology.As for terminology, many of the same terms are to be found both in Aris-totle and Aristarchean scholia. However, in approaching this topic, we mustemploy caution for various reasons. First, at least with Aristarchus, we aredealing not with his own work, transmitted by direct tradition, but with thescholia by Aristonicus. Though beyond a doubt derived from Aristarchus,these scholia are excerpts of his work, and thus may not necessarily preserveAristarchus’ ipsissima verba. This is especially true for grammatical termi-nology, since between the time of Aristarchus (third to second century b.c.e.)and that of Aristonicus (first century b.c.e.) grammatical and linguisticanalysis developed greatly. Thus Aristonicus had at his disposal a muchmore precise and extended vocabulary for grammatical categories, and thereis evidence that sometimes he rephrased Aristarchus’ original Wortlaut inorder to update it with the new terminology of the tevcnh grammatikhv.7However, for a different kind of vocabulary, that of literary criticism, thesituation is different. First, key words used by Aristotle and Aristarchus inthis field (for example muÅqoÍ, hßqh, prevpon, diavnoia, to name just a few) arepart of a common technical vocabulary, shared by all scholars discussingliterary texts (including Crates and the kritikoÇ too; see below) between thefourth and the first centuries b.c.e. Thus, in this field, unlike in the tevcnhgrammatikhv, the vocabulary available to Aristonicus was essentially that ofAristotle, and thus that of Aristarchus. When dealing with interpretation andexegesis in a broader sense, therefore, it is much more likely that Aristonicusdid not change the Wortlaut he found in Aristarchus’ hypomnemata. More-over, most of the terms in this field are not, strictly speaking, “technicalterminology,” since adjectives like eu˚telhvÍ, a˚pÇqanoÍ, and a˚duvnatoÍ and6. Interesting results in the analysis of exegetical scholia (which however are not derived from Aristarchus,and also probably represent a later stage in Homeric criticism) were developed by Schmidt (1976) and byRichardson (1980); see also Montanari 1995.7. See Matthaios 1999, 43–46, 520–22.One Line ShortAristotelian Principles in Aristarchean Philology 281adverbs like √dÇwÍ and kurÇwÍ are part of normal Greek vocabulary. Theywere also used by Aristotle as well as by other Greek authors in literary exe-gesis, but were not “invented” to express technical notions (as happened forthe parts of speech of the tevcnh grammatikhv). Even if the usage of thesemore common terms in Aristonicus’ scholia does not guarantee that theywere exactly the same terms used by Aristarchus, what really matters is notthe “form” but rather the “content” of these words, that is, the concept theyexpress. Since Aristonicus’ goal was to preserve Aristarchus’ opinions con-cerning a line, what matters most are the ideas Aristonicus conveys. Thus,in what follows, even if Greek nouns and adjectives are used to indicatecertain ideas common to Aristotle and Aristarchus (since these terms are tobe found in both Aristotle’s writings and in Aristonicus’ scholia), the focusis rather on the content they convey—Aristarchus’ ideas—than on Aris-tonicus’ Wortlaut.8 As will become clear, the affinity between Aristotle andAristarchus is evident in shared ideas and common approaches to literature,and this is the direction and the ultimate goal of the analysis that follows.2. A Preliminary Assumption:Tragedy and Epos Are Strictly ConnectedAt the beginning of the Poetics (1448b24–1449a6), Aristotle draws a famousdistinction between the two main “genres” of poetical works: serious andcomic. Among the former he counts tragedy, which, according to him, is de-rived from the serious epic represented by the Iliad and the Odyssey, just ascomedy is derived from the Homeric Margites (Poet. 1448b24–1149a2):diespavsqh de; kata; ta; o√ke∂a hßqh hJ poÇhsiÍ: o¥ me;n ga;r semnovteroi ta;Í kala;Í ejmimouÅntopravxeiÍ kaµ ta;Í tΩn toiouvtwn, o¥ de; eu˚televsteroi ta;Í tΩn fauvlwn, prΩton yovgouÍpoiouÅnteÍ, w§sper e§teroi u§mnouÍ kaµ ejgk∫mia. tΩn me;n ou®n pro; ÔOmhvrou ou˚deno;Í eßcomene√pe∂n toiouÅton poÇhma, e√ko;Í de; eπnai pollouvÍ, a˚po; de; ÔOmhvrou a˚rxamevnoiÍ eßstin, o∏onejkeÇnou oJ MargÇthÍ kaµ ta; toiauÅta. . . . kaµ ejgevnonto tΩn palaiΩn o¥ me;n hJrwikΩn o¥ de;√avmbwn poihtaÇ. w§sper de; kaµ ta; spouda∂a mavlista poihth;Í £OmhroÍ h®n (movnoÍ ga;r ou˚co§ti eu® a˚lla; kaµ mimhvseiÍ dramatika;Í ejpoÇhsen), ou§twÍ kaµ to; thÅÍ kwmådÇaÍ schÅma prΩtoÍuÒpevdeixen, ou˚ yovgon a˚lla; to; gelo∂on dramatopoihvsaÍ: oJ ga;r MargÇthÍ a˚navlogon eßcei,w§ster ∆Ilia;Í kaµ hJ ∆Oduvsseia pro;Í ta;Í tragådÇaÍ, ou§tw kaµ ou•toÍ pro;Í ta;Í kwmådÇaÍ.Poetry was split up according to their particular characters; the grander people repre-sented fine actions, i.e. those of fine persons, the more ordinary people representedthose of inferior ones, at first composing invectives, just as the others composed hymnsand praise-poems. We do not know of any composition of this sort by anyone beforeHomer, but there were probably many [who composed invectives]. Beginning with Homer[such compositions] do exist, e.g., his Margites etc. . . . Thus some of the ancients be-came composers of heroic poems, others of lampoons. Just as Homer was the greatestcomposer of serious poetry (not that he alone composed well, but because he alonecomposed dramatic representations), so too he was first to indicate the form of comedy,by dramatizing not an invective but the laughable. For his Margites stands in the samerelation to comedies as do the Iliad and the Odyssey to tragedies. (Trans. Janko 1987)8. The same point has been made by Lührs (1992, 16).Francesca Schironi282It is this “etiological” derivation that allows us to look at how Aristarchusanalyzes epic poetry in search of Aristotelian criteria, for ultimately thesetwo genres are not so far apart. Aristotle himself emphasizes this (Poet.1449b16–20):mevrh d’ ejstµ ta; me;n tau˚tav, ta; de; ≥dia thÅÍ tragådÇaÍ: diovper o§stiÍ perµ tragådÇaÍ oπdespoudaÇaÍ kaµ fauvlhÍ, oπde kaµ perµ ejpΩn: a¶ me;n ga;r ejpopoiÇa eßcei, uÒpavrcei t¬Å tragådÇç,a¶ de; au˚t¬Å, ou˚ pavnta ejn t¬Å ejpopoiÇç.As for their parts, some are the same, others are particular to tragedy. For this reason,whoever knows about good and inferior tragedies knows about epics too. Tragedy pos-sesses all [the parts] that epic has, but those that it possesses are not all in epic. (Trans.Janko 1987)It is this close relationship between tragedy and epic9 that allows for theidentification of Aristotelian criteria within the work of a Homeric scholarlike Aristarchus. If what I am going to argue is sound, Aristarchus knew whatthe philosopher had said about the affinity between these two genres andtherefore thought it legitimate to apply Aristotle’s criteria for a good tragedyto epic poetry. The Alexandrians knew some of the Aristotelian works, andwhether or not the Poetics was available to them, the dialogue On Poets, inwhich Aristotle discussed the same topics as in the Poetics, and the HomericProblems were both known.10For Aristotle tragedy is composed of six parts (Poet. 1450a9–10): plot(muÅqoÍ), characters (hßqh), diction (levxiÍ), thought (diavnoia), spectacle (oßyiÍ),and music (melopoiÇa). In the chapters on epic, he picks up this divisionagain, but rightly states that epic lacks the last two, music and spectacle(Poet. 1459b7–10).11 This is a very useful working distinction, which helpsAristotle to set out a systematic view of the main constituents of a tragedy(and of an epos too), as well as to refer to other works such as the Rhetoricfor elements that have already been treated elsewhere. This distinction ofepic into four elements seems, moreover, to operate also in Aristarchus’methodology, especially when he must decide about an athetesis, for a lineis generally judged with reference to its function for the plot, for the char-acters, for the thought it expresses, and in terms of style. Therefore we willfollow this division in our analysis and will see what Aristotle and Aristarchushave to say about the plot, the characters, the thought-element and the style.9. As proved also by the last chapters of the Poetics (chaps. 23–26), where Aristotle focuses on epicpoetry, drawing on the previous chapters where he analyzed tragedy. For an account of Aristotle’s viewson Homer, see Richardson 1992, and 1993, 31–35.10. Cf. Nickau 1977, 138–39, with n. 16; Lührs 1992, 14–15, Richardson 1994, 17–18, 27. On the debatedproblem about the destiny of the library and the books of Aristotle, see Moraux 1973, 3–31; Canfora 1988,34–37, 59–66; Richardson 1994, 8–12; Nagy 1998, 198–206; Barnes 1999; Canfora 2002. The Poetics inparticular does not seem to have enjoyed great popularity in antiquity: ancient soruces are silent, and theearliest quotation is in Porphyry (quoted by Simplicius [in Cat., p. 36.16–31 Kalbfleisch]): see Janko 1982,and 1991, 7 and n. 25.11. One problem is, of course, assessing whether this difference can be interpreted as a demonstrationthat tragedy is a more accomplished form of art than epic. However, there are also advantages in the lackof spectacle in the epos: see p. 286 below.One Line ShortAristotelian Principles in Aristarchean Philology 2833. Aristotelian Theory and Aristarchean Practice: MUQOSSince for Aristotle tragedy (and epic too) is an imitation of a complete andwhole action (mÇmhsiÍ teleÇaÍ kaµ o§lhÍ pravxewÍ; cf. Poet. 1450b24–25) andthe plot (muÅqoÍ) is defined by Aristotle as the mÇmhsiÍ thÅÍ pravxewÍ (Poet.1450a3–4), it follows that the plot is “the principle and as it were the soulof tragedy” (Poet. 1450a38–39: a˚rch; me;n ou® n kaµ o∏on yuch; oJ muÅqoÍ thÅÍtragådÇaÍ). One of the most important criteria for the plot is that it mustbe in accordance with probability and necessity (Poet. 1451a36–38 and1451b8–10):fanero;n de; ejk tΩn e√rhmevnwn kaµ o§ti ou˚ to; ta; genovmena levgein, touÅto poihtouÅ eßrgonejstÇn, a˚ll’ o∏a a˙n gevnoito kaµ ta; dunata; kata; to; e√ko;Í h˙ to; a˚nagka∂on.It is also obvious from what we have said that it is the function of a poet to relate notthings that have happened, but things that may happen, i.e. that are possible in accordancewith probability or necessity. (Trans. Janko 1987)eßstin de; kaqovlou mevn, tåÅ poÇå ta; po∂a aßtta sumbaÇnei levgein h˙ pravttein kata; to; e√ko;Íh˙ to; a˚nagka∂on, ou• stocavzetai hJ poÇhsiÍ ojnovmata ejpitiqemevnh.A universal is the sort of thing that a certain kind of person may well say or do in accor-dance with probability or necessity—this is what poetry aims at, although it assigns names[to the people]. (Trans. Janko 1987)For Aristotle, then, plots can contain what is “necessary” (a˚nagka∂on), butalso something which is kata; to; e√kovÍ, “according to probability.”12 In otherwords, a plot must consist of a necessary or probable sequence of events.This is due to the particular status of poetry, which distinguishes it fromhistory: poetry represents universals, not particulars, like history; hencepoetry is more philosophical (Poet. 1451b5–7):dio; kaµ filosof∫teron kaµ spoudaiovteron poÇhsiÍ ¥storÇaÍ ejstÇn: hJ me;n ga;r poÇhsiÍmaÅllon ta; kaqovlou, hJ d’ ¥storÇa ta; kaq’ e§kaston levgei.For this reason poetry is a more philosophical and more serious thing than history: poetrytends to speak of universals, history of particulars. (Trans. Janko 1987)Moreover, in poetry, elements that are impossible in reality are neverthelessadmitted because in this way the poet is able to astonish his audience andachieve amazement (to; qaumastovn: Poet. 1460a17), which is the tevloÍ ofpoetry (Poet.1460b23–26):a˚duvnata pepoÇhtai, hJmavrthtai: a˚ll’ ojrqΩÍ eßcei, e√ tugcavnei touÅ tevlouÍ touÅ auÒthÅÍ (to;ga;r tevloÍ e≥rhtai), e√ ou§twÍ ejkplhktik∫teron h˙ au˚to; h˙ aßllo poie∂ mevroÍ.12. The same kind of contrast is found again when Aristotle is dealing with episodic plots, where wehave one episode after the other without necessity or probability (Poet. 1451b33–35: tΩn de; aÒplΩn muvqwnkaµ pravxewn a¥ ejpeisodi∫deiÍ e√sµn ceÇristai: levgw d’ ejpeisodi∫dh muÅqon ejn å• ta; ejpeisovdia met’ aßllhlaoußt’ e√ko;Í oußt’ a˚navgkh eπnai [“Among simple plots and actions, episodic [tragedies] are the worst. By ‘epi-sodic’ I mean a plot in which there is neither probability nor necessity that the episodes follow one other,”trans. Janko 1987]).Francesca Schironi284One Line Long[If ] impossibilities have been produced, there is an error; but it is correct, if it attains theend of the art itself. The end has been stated [already, i.e.] if in this way it makes eitherthat part [of the poem], or another part, more astonishing. (Trans. Janko 1987)Since supernatural, impossible elements make the poetry more interest-ing,13 Aristotle judges the plot not according to whether it is possible or not,but whether it is piqanovÍ/a˚pÇqanoÍ, “believable or not” (Poet. 1460a26–27 and1461b9–12):proaire∂sqaÇ te de∂ a˚duvnata e√kovta maÅllon h˙ dunata; a˚pÇqana.Impossible [incidents] that are believable should be preferred to possible ones that areunbelievable. (Trans. Janko 1987)o§lwÍ de; to; a˚duvnaton me;n pro;Í th;n poÇhsin h˙ pro;Í to; bevltion h˙ pro;Í th;n dovxan de∂a˚navgein. provÍ te ga;r th;n poÇhsin a¥ret∫teron piqano;n a˚duvnaton h˙ a˚pÇqanon kaµ dunatovn: In general, the impossibility should be explained with reference either to the composi-tion, or to [making something] better [than it is], or to opinion. In relation to [the needsof ] the composition, a believable impossibility is preferable to an unbelievable possi-bility. (Trans. Janko 1987)What is “probable” (e√kovÍ) is thus also “believable” (piqanovn). Thus it isbetter for Aristotle to choose plots that are believable—though they maynot be possible in the real world (piqana; a˚duvnata)—than stories that couldhappen but are not believable (dunata; a˚pÇqana).14Since a poet, in order to achieve to; qaumastovn, has more freedom, thecriterion of “believability” becomes an internal one: something is believ-able if it follows from what has been stated before as a logical consequence.Within a work of poetry there are rules that are typical of poetry and, aslong as these rules are respected by the poet, the poetic work is good, nomatter how the muÅqoÍ in itself corresponds to truth in the real world. Thepremise behind these prescriptions is that poetry is a tevcnh that worksaccording to rules that are its own and different from those of other tevcnai(Poet. 1460b13–15):15pro;Í de; touvtoiÍ ou˚c hJ au˚th; ojrqovthÍ ejstµn thÅÍ politikhÅÍ kaµ thÅÍ poihtikhÅÍ ou˚de; aßllhÍtevcnhÍ kaµ poihtikhÅÍ.In addition, there is not the same [standard of ] correctness in the art of civic life as inthat of poetry, nor is there in any other art as in that of poetry. (Trans. Janko 1987)3.1 Aristarchus: The Criterion of BelievabilityAs for Aristotle, so also for Aristarchus, the main criterion for judging theplot is not the distinction between what is possible and what is not, but that13. In this light, Homer is a master of lies (Poet. 1460a18–19: dedÇdacen de; mavlista £OmhroÍ kaµ tou;ÍaßllouÍ yeudhÅ levgein wÒÍ de∂ [“Homer above all has taught the other [poets] to tell untruths in the rightway,” trans. Janko 1987]).14. This is because what is possible is believable, but not all that is believable is possible; hence believ-ability is a wider concept, as Aristotle explains at Poet. 1451b16–19: a≥tion d’ o§ti piqanovn ejsti to; dunatovn:ta; me;n ou®n mh; genovmena oußpw pisteuvomen eπnai dunatav, ta; de; genovmena fanero;n o§ti dunatav: ou˚ ga;r a˙n ejgev-neto, e√ h®n a˚duvnata. (“The reason is that what is possible is believable; we do not believe that what hasnever happened is possible, but things which have happened are obviously possible—they would have nothave happened if they were impossible,” trans. Janko 1987).15. Cf. Richardson 1992, 36.Aristotelian Principles in Aristarchean Philology 285between what is believable and what is not. Âduvnata are not necessarily ex-cluded if they help poetic goals. There are only two cases of lines athetizedby Aristarchus because they contain a˚duvnata, “impossibilities.” The first caseis in the Nekyia:16 the lines about Otus and Ephialtes’ project to put Olympuson Ossa and Pelium on top of them in order to reach the sky (Od. 11.315–16:¢Ossan ejp’ Ou˚luvmpå mevmasan qevmen, au˚ta;r ejp’ ¢Oss¬ / Phvlion e√nosÇful-lon, ªn’ ou˚rano;Í a˚mbato;Í e≥h) were rejected by Aristarchus wÒÍ a˚duvnatoi(schol. V Od. 11.315). The other case is at Odyssey 22.144–45, whereMelanthius, in order to help the suitors, takes out from the chamber twelveshields, the same number of spears and the same number of helmets, asEustathius testifies (Eust. Od. 1921.56):shmeÇwsai kaµ o§ ti to; perµ tΩn d∫deka sakevwn kaµ to; ejfexhÅÍ ÂrÇstarcoÍ a˚qethvsaÍkecÇaken, a˚duvnaton eπnai e√pøn tosauÅta bastavsai aßnqrwpon. 17Note also that Aristarchus athetized and marked with a chi the line about the twelveshields and the following one, saying that it was impossible that a person could carryall these [weapons].18Nevertheless, in the Iliad, for example, there are no cases of athetesis dueto a˚duvnata.19 Only at Iliad 2.667, when Homer tells the story of Tlepolemusarriving at Rhodes after killing his uncle Licymnius, Zenodotus’ readingaπya d’ o§ g’ ejÍ ÔRovdon (instead of au˚ta;r o§ g’ ejÍ ÔRovdon) was dismissedby Aristarchus because it would be impossible for a fugitive wandering inthe Aegean to arrive “at once” at Rhodes.20 Otherwise, elements that areimpossible from a rational point of view are allowed because of poeticlicense. For example, the fact that the Cyclops knows that ships exist orunderstands Greek is in itself absurd, but for Aristarchus it is to be kept, be-cause it is poetry (schol. HMQR Od. 3.71):oJ de; ÂrÇstarcoÍ o√keiovteron au˚tou;Í tetavcqai ejn tåÅ lovgå touÅ KuvklwpovÍ fhsin: . . .dotevon dev, fhsÇ, tåÅ poiht¬Å ta; toiauÅta. 21 kaµ ga;r nauÅn au˚to;n paravgei e√dovta, “a˚llav moie≥f’ o§ph eßsceÍ √øn eu˚ergeva nhÅa” (Od. 9.279), kaµ sunÇhsin ÔEllhnÇda fwnhvn.Aristarchus says that these lines are more properly included in the speech of the Cyclops[than here in Telemachus’ speech]. . . . But, he says, we must allow these [licenses] tothe poet. For Homer represents the Cyclops as aware [of the existence] of a ship, [as itis proved by] “but tell me where, arriving, you put your well-made ship” (Od. 9.279),and [the Cyclops] understands the Greek language.16. An episode that was particularly suspicious to Aristarchus; see p. 288 below.17. Eustathius is here probably quoting Aristonicus (see Carnuth 1869, ad loc.) from a collection of scholiaricher than the one that has reached us (and which does not present any scholium on Od. 22.144–45).18. The possibility that Melanthius might have carried all these weapons in more than one journey doesnot seem to have been taken into consideration by Aristarchus. 19. The only case of dunat- coming up in Aristarchean scholia is at Il. 21.475: ajpo; touv tou a˚qetouÅntaistÇcoi tre∂Í: ou˚ duv natai ga;r oJ a√douvmenoÍ “patrokasignhvtoio mighvmenai ejn palavm¬sin” (Il. 21.469) a˚eµprokale∂sqai to;n PoseidΩna ejn tåÅ ∆Oluvmpå pro;Í mavchn. This ou˚ duvnatai ga;r . . . , however, is due to in-ternal inconsistency or a˚prevpeia of a character; according to Aristarchus it is “impossible” that Apolloonce boasted that he was going to fight against his uncle Poseidon, if he is now afraid to face him. This isthus definitely not an athetesis due to something that is “impossible” from an objective point of view.20. Schol. Il. 2.667: au˚ta;r o§ g’ ejÍ ÔRovdon: o§ti ZhnovdotoÍ gravfei “aπya d’ o§ g’ ejÍ ÔRovdon.” ou˚ duvnatai de;tacevwÍ ejlhluqevnai ejpµ th;n ÔRovdon oJ provteron me;n nauÅÍ pephc∫Í, eπta a˚l∫menoÍ kaµ ou˚k †eu˚qu; ploikΩÍ†.21. This closely resembles the Aristotelian remark: dÇdomen ga;r tauÅta to∂Í poihta∂Í in Poet. 1460b13,apropos of metaphors and glossai.Francesca Schironi286A similar attitude is to be found in Aristotle when he comments on theepisode of the bath in the Odyssey: Eurycleia’s recognition of Odysseus isin itself aßlogon, but it is allowed because in this way the poet achieved to;qaumastovn (Poet. 1460a11–26). This is to be connected with the comparisonbetween tragedy and epic: according to Aristotle an advantage epic has overtragedy is that, lacking oßyiÍ, it can afford to be aßlogoÍ, because the audience,not seeing the plot performed, is likely to notice irrationalities and incon-gruities less (Poet. 1460a11–14):de∂ me;n ou®n ejn ta∂Í tragådÇaiÍ poie∂n to; qaumastovn, maÅllon d’ ejndevcetai ejn t¬Å ejpopoiÇçto; aßlogon, di’ o¶ sumbaÇnei mavlista to; qaumastovn, dia; to; mh; oJraÅn e√Í to;n pravttonta.[The poet] should put what is amazing into his tragedies; but what is improbable, fromwhich amazement arises most, is more admissible in epic because [the audience] doesnot see the person in action. (Trans. Janko 1987)The main criterion for Aristarchus, as for Aristotle, is thus that of prob-ability. Facts in the poems often receive comments along these lines:something takes place according to probability (e√kovtwÍ) and is thereforeacceptable. For example, in the Doloneia (Il. 10.447) Diomedes addressesDolon by name: “mh; dhv moi fuvxin ge Dovlwn ejmbavlleo qumåÅ” (“Do not, Dolon,have in your mind any thought of escape”), and some ancient scholars foundfault in the fact that Diomedes seems to know the name of Dolon, thoughthis is the first time that they meet. On this basis they read dolΩn, the par-ticiple of dolovw, “to deceive” (“Do not have in your mind any thought ofescape, trying to deceive me”). Aristarchus, however, defended the text: forhim, it was probable (e√kovÍ) that the Greeks knew the name of some of theirenemies after ten years of siege (schol. Il. 10.447a):Dovlwn: o§ti zhte∂tai, pΩÍ to; oßnoma eßgnw: diov tineÍ a˚nevgnwsan “dolΩn” wÒÍ noΩn. . . .e√ko;Í dev tinwn gin∫skesqai ojnovmata wÒÍ a˙n dekaetouÅÍ gegonovtoÍ crovnou, kaµ mavlistatouÅ DovlwnoÍ: h®n ga;r khvrukoÍ u¥o;Í “poluvcrusoÍ poluvcalkoÍ” (cf. Il. 10.314–15).Dolon: [the diple is] because there is a question about how [Diomedes] knew his name;hence some scholars read dolΩn, “deceiving,” perispomenon like noΩn. . . . But it is prob-able that they knew the names of some of them, since a decade had passed [with themthere], and in particular the name of Dolon. For he was son of a herald, “rich in gold,rich in bronze” (cf. Il. 10.314–15).This criterion of probability is mainly expressed in Aristonicus’ scholiaby the couple piqanovÍ / a˚pÇqanoÍ. It is one of the most common justifica-tions given for an athetesis or for rejecting Zenodotus’ readings which, accord-ing to Aristarchus, often lack believability. For example, in the assembly ofthe Achaean leaders at Iliad 2.50–86, Aristarchus did not find believableZenodotus’ reading according to which Agamemnon stands up to speak infront of only seven heroes (schol. Il. 2.55a: o§ti ZhnovdotoÍ gravfei “au˚ta;rejpeÇ rJ’ hßgerqen oJmhgereveÍ t’ ejgevnonto, / to∂si d’ a˚nistavmenoÍ metevfhkreÇwn Âgamevmnwn.” ajpÇqanon de; ejn eJpta; ojrqo;n dhmhgore∂n).22 Details that22. See Lührs 1992, 260–61 n. 365; cf. also schol. Il. 2.76a: a˚qetouÅntai stÇcoi ojkt∫, o§ti ou˚k ojrqouÅe√rhkovtoÍ touÅ ÂgamevmnonoÍ levgei hßtoi o§g’ w¶Í e√pøn kat’ aßr’ e§zeto, o§per a˚pÇqanon.One Line LongAristotelian Principles in Aristarchean Philology 287seem to go against human chronology are rejected on the same ground, as,for example, that Aethra, the maid of Helen, is to be identified with the motherof Theseus, who would have been extremely old by then (schol. Il. 3.144a):A≥qrh, PitqhÅoÍ qugavthr: e√ me;n th;n QhsevwÍ levgei mhtevra a˚qethtevon: a˚pÇqanon gavr ejs-tin ÔElevnhÍ a˚mfÇpolon th;n ou§twÍ uÒperarcaÇan, h¶n ou˚k ejkpoie∂ zhÅn dia; to; mhÅkoÍtouÅ crovnou. e√ de; oJmwnumÇa ejstÇ, kaqavper kaµ ejpµ pleiovnwn, duvnatai mevnein.Aethra, the daughter of Pittheus: if Homer means the mother of Theseus, [the line]must be athetized. For it is unbelievable that such a very old woman be the maid ofHelen. It is impossible that she has been living for such a long time. If instead it is acase of homonymy, as happens in many other cases, [the line] can remain.23Also the famous problem of the dual in Iliad 9 is solved by Aristarchuswith the criterion of credibility: there are only two people present, Odysseusand Ajax, because, if Phoenix were also present, it would not be believableto have Odysseus leading, since Phoenix was older (schol. Il. 9.192a):tø de; bavthn : o§ti ejpµ ∆OdussevwÍ kaµ A≥antoÍ to;du∑kovn: parovntoÍ ga;r touÅ FoÇnikoÍ a˚pÇqanon levgein “hJge∂to de; d∂oÍ ∆OdusseuvÍ.”The two of them came forward, and noble Odysseus led the way: [the diple is] becausethe dual is for Odysseus and Ajax. For if Phoenix had been present, it would have beenunbelievable to say “noble Odysseus led the way.”The criterion of believability plays a role even in the supernatural episodeswhere Aristarchus, like Aristotle, seems to put a limit to to; qaumastovn; thuswithin the microcosm of the epos, although the supernatural is allowed, be-lievability still applies. For example, in poetry it is fine for an animal tospeak, like Xanthus, the horse of Achilles. However, it is too much to havehim speak like a learned man or a seer foretelling his destiny to Achilles,as happens at Iliad 19.416–17, lines that Aristarchus rejected (schol. Il.19.416–17a):h§nper ejlafrotavthn : a˚qetouÅntai stÇcoi kaµ ou•toi o¥ duvo, o§ti . . .a˚pÇqanon ªppon levgein “fasÇn” w§sper aßndra poluÇstora.[we may be running together with the blows of Zephyrus] who they say is the lightestof all things; [yet still your destiny is] to be killed [in force by a god and a mortal]. . . .these two lines also are athetized because . . . it is unbelievable for a horse to say “theysay, etc.,” like a knowledgeable man.24Or, again, it is acceptable to have gods intervening in human affairsdisguised as human beings. However, at Iliad 21.290, when Poseidon andAthena, disguised as two men, go to Achilles, it is not believable for Poseidonto say “Athena and I will help you,” as Aristarchus remarked in schol. Il.21.290a: a˚qete∂tai, o§ti a˚pÇqanon e√Í a˚ndro;Í morfh;n wÒmoiwmevnon levgein“ejgø kaµ Palla;Í Âqhvnh”: tÇÍ gavr ejstin, ou˚ mh; nohvs¬ (“it is athetized be-cause it is not believable that disguised as a mortal he says ‘Athena and I.’For Achilles will not understand who he is”). How could Achilles know thatthese two men were divinities?23. On this athetesis, see Jenkins 1999.24. Cf. Lührs 1992, 46–48.Francesca Schironi288The most important case of a limitation of to; aßlogon25 and to; qaumastovnis the second part of the Nekyia (Od. 11.568 [565?]–627), where Odysseusclaims to have seen Minos, Orion, Tityus, Tantalus, Sisyphus, and Heracles.Notwithstanding the admissibility of a “marvelous” journey to the Under-world, and although the lines are not bad in style, this episode was rejectedas suspicious by Aristarchus. Odysseus never enters Hades but remains atthe gate of Erebus (cf. Od. 11.37, 150, 563), and therefore could never haveseen all these mythical examples of divine justice and wickedness beingpunished.26 The scholia by Aristonicus bear witness to a long series ofcomments by Aristarchus along these lines: how can Minos come to the sac-rificial blood? Does he go with all the people he is going to judge togetherwith his throne (schol. HQT Od. 11.570: ou˚k aßra uÒpexhÅlqen oJ MÇnwÍ ªnasunofq¬Å. aßlogon ga;r to; kaµ su;n dikazomevnoiÍ kaµ au˚tåÅ dÇfrå ejxelqe∂n)?How can Orion hunt in Hades? How could he come forward with all the beaststhat he has slain (schol. HT Od. 11.573: ou˚de; ejpµ touvtou tethvrhtai to; suvm-fwnon. aßlogon ga;r to; ejn £Aidou kunhgete∂n: pΩÍ te a§ma t¬Å tΩn qhrΩn a˚gevl¬prohÅlqe, kaµ dia; tÇ;)? How can Tityus come to the sacrifice, if he is lying onthe ground with his liver devoured by two vultures (schol. QT Od. 11.577:katagevlasta kaµ tauÅta, e√ katestrwmevnoÍ ejn tåÅ dapevdå prohÅlqen ejpµ to;sfavgion: au˚to;Í ga;r oJ ∆Odusseu;Í ou˚k hjduvnato diabhÅnai ejpµ to; eßreboÍ)? Orhow can Tantalus come to the sacrifice, together with the trees and the marshylake in which he lies? Or how did Odysseus see what was within from outside(schol. H Od. 11.588: ou˚de; ou•toÍ duvnatai su;n lÇmn¬ kaµ devndroiÍ ejxelhlu-qevnai ejpµ to; sfavgion, h˙ pΩÍ eßxwqen ta; eßsw ejqe∫rei;)? Or how can Sisyphus,who must push the huge stone up a hill, come to the sacrifice (schol. QT Od.11.593: pΩÍ duvnatai su;n tåÅ lÇqå kaµ t¬Å a˚krwreÇç ejf’ ¬• a˚nekuvlie to;n lÇqon,h§kein ejpµ ta; sfavgia;)? All this, according to Aristarchus, was too much andthe risk was that, instead of being astonishing, the episode became ridiculous(on which see pp. 298–99 below).3.2 Aristarchus: Internal ContradictionsThe conception of the work of poetry as a microcosm with its own rulesleads to the principle of noncontradiction. Avoiding inconsistencies withinthe plot is a cardinal principle for Aristotle, who argues repeatedly againstwhat is uÒpenantÇon (Poet. 1455a22–26):27de∂ de; tou;Í muvqouÍ sunistavnai kaµ t¬Å levxei sunapergavzesqai o§ti mavlista pro; ojmmavtwntiqevmenon: ou§tw ga;r a˙n ejnargevstata [oJ] oJrΩn w§ster par’ au˚to∂Í gignovmenoÍ to∂Í pratto-mevnoiÍ euÒrÇskoi to; prevpon kaµ h§kista a˙n lanqavnoi [to;] ta; uÒpenantÇa.In constructing his plots and using diction to bring them to completion, [the poet]should put [the events] before his eyes as much as he can. In this way, seeing them very25. On the criterion of to; aßlogon applied to athetesis of repeated lines in Aristonicus’ and exegeticalscholia, see Lührs 1992, 167–94.26. Schol. HT Od. 11.568: noqeuvetai mevcri touÅ “w¶Í e√pøn oJ me;n au®qiÍ eßdu dovmon ◊AidoÍ e≥sw” (627), kaÇtoiou˚k oßnteÍ a˚gene∂Í perµ th;n fravsin. uÒpe;r de; thÅÍ a˚qethvsewÍ au˚tΩn levgetai toiavde: (H) pΩÍ oπde touv touÍ h˙tou;Í loipou;Í eßsw tΩn £Aidou pulΩn oßntaÍ kaµ tΩn potamΩn…27. For Aristotle on uÒpenantÇon, cf. also Poet. 1461a31–1461b9, 1461b15–18.Aristotelian Principles in Aristarchean Philology 289vividly as if he were actually present at the actions [he represents], he can discoverwhat is suitable, and is least likely to miss contradictions. (Trans. Janko 1987)Internal contradictions and inconsistencies are constantly rejected by Aris-tarchus too, who seems to have developed Aristotle’s theory into a more com-plete system, where the philologist has to work on a text as a self-standingunity that must be purged of internal contradictions. Aristarchus takes ex-ception to contradictions in the text (e.g., schol. Il. 8.39–40: ejnantiouÅntai de;ejnqavde to∂Í uÒpokeimevnoiÍ; schol. BQ Od. 12.374; schol. HQ Od. 12.439) and asa rule, when this happens, he is in favor of athetesis.28 He also argues againstZenodotus for readings that are contrary to some data present in the poems.29The idea that something is consistent or inconsistent with the rest of the plotis expressed in Aristonicus’ scholia with comments like sumf∫nwÍ (“inharmony,” “in agreement with”) or a˚sumf∫nwÍ (“discordant,” “in disagree-ment with”). Âsumf∫nwÍ is used to argue against readings and interpre-tations by Zenodotus or other colleagues that Aristarchus does not sharebecause they contradict some other passage within the poem (schol. Il. 4.339b,8.19). On the contrary, a line is sumf∫nwÍ with the rest of the poem whenAristarchus wants to defend Homer against his detractors or against thediaskeuastaÇ (schol. Il. 3.230a), or defend his own readings (schol. Il. 7.330b,8.562). Alternatively, the same idea is expressed with mavcontai/-etai (“they/itcontradict[s]”), that is, that one or more lines are at odds (mavcetai/-ontai) withwhat has been said or known before, as in schol. H Od. 11.452: macovmenoito∂Í prokeimevnoiÍ. For example, at Iliad 20.269–72, ancient scholars werepuzzled that Aeneas’ spear reaches the golden plate of Achilles’ shield, havingpierced the two external plates, one of bronze and the other of tin, as if theshield had the golden layer underneath, hidden by those of bronze and tin.Aristarchus solved the problem by athetizing the lines, because according tohim these lines not only were odd in terms of content (why was the goldhidden by tin and bronze?), but, moreover, they were in clear contradictionwith what we know about the shield of Achilles as described at Iliad 18.478–607, where its surface is clearly made of gold (schol. Il. 20.269–72a: a˚qe-touÅntai stÇcoi tevssareÍ, o§ti dieskeuasmevnoi e√sµn uÒpov tinoÍ tΩn boulomevnwnprovblhma poie∂n. mavcetai de; safΩÍ to∂Í gnhsÇoiÍ [“four lines are athetizedbecause they were added by someone of those who want to create a question.For these lines are clearly in contradiction with those that are genuine”]).On the other hand, in Aristonicus’ scholia ou˚ mavcetai is used to solve aproblem by showing that in Homer there are no internal contradictions.30One famous question (zhvthma) was that of how many times Hector andAchilles ran around Troy, because at Iliad 22.208 we read: a˚ll’ o§te dh; to;tevtarton ejpµ krounou;Í a˚fÇkonto (“but when for the fourth time they cameto the springs”), whereas at Iliad 22.251 Hector says: “trµÍ perµ aßstu mevgaPriavmou dÇon” (“ ‘[Achilles, I do not flee you anymore, as when before] I ranthree times around the great city of Priam’ ”). Ancient scholars had taken28. See schol. Il. 5.838–39, 7.334–35, 11.767a, 19.407a, 21.570a, 22.199–200a, 24.45a.29. See schol. Il. 3.334–35a, 5.807.30. See schol. Il. 9.571a, 13.365a.Francesca Schironi290exception to this passage, as Aristarchus noticed in schol. Il. 22.208a2:shmeiouÅntai dev tineÍ dia; to; dokouÅn ejnantÇon eπnai “trµÍ perµ aßstu mevgaPriavmou dÇon” kaµ “a˚ll’ o§te dh; to; tevtarton” (“some mark this line becauseof the apparent contradiction between ‘I ran three times around the greatcity of Priam’ and ‘but when for the fourth time’ ”). Aristarchus, however,clarified and solved the problem by arguing that there were three full laps,but in the fourth they went as far as the fountain but did not go right aroundthe city. Therefore there was no contradiction between the two lines (schol.Il. 22.251a: ou˚ mavcetai de; tåÅ “a˚ll’ o§te dh; to; tevtarton” (Il. 22.208)· tre∂Í me;nga;r teleÇouÍ kuvklouÍ perievdramon, to; de; tevtarton e§wÍ tΩn krounΩn ejlqovnteÍou˚kevti perihÅlqon th;n povlin [“it does not contradict ‘but when for the fourthtime’ (Il. 22.208): for they ran in three full circles and in the fourth theyarrived at the fountains and did not go around the city”]).314. Aristotelian Theory and Aristarchean Practice: HQHAt Poetics 1448a1–18, Aristotle states that every mimetic art representshßqh, which can be either spouda∂oi, “serious,” or fauÅloi, “base.” Poetry too,then, has to choose which hßqh are to be the target of mimesis, and hence itis divided according to the characters that it is going to imitate. In this re-gard, epic and tragedy are identical, in that both of them represent “admirable,”“serious” people (Poet. 1449b9–10):hJ me;n ou®n ejpopoiÇa t¬Å tragådÇç mevcri me;n touÅ meta; mevtrou lovgå [em. Kassel: mevtroumeta; lovgou B, mevtrou megavlou A] mÇmhsiÍ eπnai spoudaÇwn hjkolouvqhsen.Epic poetry follows tragedy insofar as it is a representation of serious people whichuses speech in verse. (Trans. Janko 1987)Hence, what is valid for tragedy with regard to characters is to be con-sidered valid for epos too. Characters of tragedy and epic must be spouda∂oi,“serious,” and better than they are in reality. Hence they must not be cari-catures of real people, as happens in comedy, because everything that isridiculous is to be avoided in tragedy and epic. When Aristotle comes to adetailed account of tragic /epic characters (Poet., chap. 15), he first statesthat they must be crhstoÇ, “good,” that is, “of value,” aÒrmovttonteÍ, “appro-priate,” o§moioi, “(life)like,” and oJmaloÇ, “consistent” (Poet. 1454a16–28):e ¶n me;n kaµ prΩton, o§pwÍ crhsta; ¬®. e§xei de; h®qoÍ me;n eja;n w§sper ejlevcqh poi¬Å fanero;n oJlovgoÍ h˙ hJ praÅxiÍ proaÇresÇn tina [add. Vahlen] ¬®, crhsto;n de; eja;n crhsthvn.eßstin de; ejn eJkavstå gevnei· kaµ ga;r gunhv ejstin crhsth; kaµ douÅloÍ, kaÇtoi ge ≥swÍ touvtwnto; me;n ce∂ron, to; de; o§lwÍ fauÅlovn ejstin. deuvteron de; to; aJrmovttonta· eßstin ga;r a˚ndreÇanme;n to; h®qoÍ, a˚ll’ ou˚c aÒrmovtton gunaikµ ou§twÍ a˚ndreÇan h˙ deinh;n eπnai. trÇton de; to;o§moion. touÅto ga;r e§teron touÅ crhsto;n to; h®qoÍ kaµ aÒrmovtton poihÅsai wÒÍ proeÇrhtai.31. In analyzing plot, Aristotle deals also with the difference between simple and complex plots (Poet.1452a12–18), which are characterized by the presence of a˚nagnwrismovÍ and peripevteia. These conceptsare not present in Aristarchus, which is probably due to the fact that they are more part of a theoreticaldiscussion of a literary work than concepts that could be used in his philological activity. For the samereason, Aristarchus does not seem to have taken much from Aristotle’s sections on the different parts oftragedy (Poet. 1452b14–27), and on the different kinds of actions (pitiful, fearful, recognitions, Poet.1453b1–1454a15). This essential difference between the work of Aristotle (theory) and that of Aristarchus(practice) is fundamental.One Line LongAristotelian Principles in Aristarchean Philology 291tevtarton de; to; oJmalovn. ka˙n ga;r a˚n∫malovÍ tiÍ ¬® oJ th;n mÇmhsin parevcwn kaµ toiouÅtonh®qoÍ uÒpoteq¬Å, o§mwÍ oJmalΩÍ a˚n∫malon de∂ eπnai.First and foremost, the characters should be good. [The tragedy] will have character if,as we said, the speech or the action makes obvious a decision of whatever sort; it willhave a good character, if it makes obvious a good decision. [Good character] can existin every class [of person]; for a woman can be good, and a slave can, although the firstof these [classes] may be inferior and the second wholly worthless. Second, [theyshould be] appropriate. It is possible to be manly in character, but it is not appropriatefor a woman to be so manly or clever. Third, [the character should be life-]like. This isdifferent from making the character good and appropriate in the way already stated.Fourth, [the character should be] consistent. If the model for the representation issomebody inconsistent, and such a character is intended, even so it should be consis-tently inconsistent. (Trans. Janko 1987)If being crhstoÇ, “good,” is a necessary characteristic, due to the essenceof tragedy, which represents “serious” (spouda∂oi) actions and characters,the other three characteristics (to; aÒrmovtton, to; o§moion, and to; oJmalovn) aremore interesting to define. Characters must be aÒrmovttonteÍ, “appropriate,”in the sense that each character must fit the characteristics of the kind of indi-vidual it represents. This concept is later on coupled with that of prevpon (Poet.1454a28–31: eßstin de; paravdeigma . . . touÅ de; a˚prepouÅÍ kaµ mh; aÒrmovttontoÍo§ te qrhÅnoÍ ∆OdussevwÍ ejn t¬Å Skuvll¬. Cf. also Poet. 1458b14–15). Further-more, characters must be o§moioi, “(life)like,” “similar” to their real model(in “real life” or in the mythical tradition to which the poet refers).32 Finally,they must be oJmaloÇ, “consistent,” and not behaving in a contradictory way.The last three characteristics are thus sharply distinct. To; aÒrmovtton refersto the relationship between the “type” the poet has in mind and how thecharacter relates to it, so how “convincing” the representation of that par-ticular human being is (considered in terms of gender, age, social status,etc.). To; o§moion instead refers to the relationship between reality outside thework of poetry (i.e., the real life or the mythical tradition in the background)and the characters, in the sense that the audience has to recognize the characteras someone similar to and comparable to people from their own experience,whereas the idea of to; oJmalovn is an internal criterion to judge the develop-ment of the character within the poetical work. Aristotle then summarizesthese criteria with the principle of kata; to; a˚nagka∂on h˙ to; e√kovÍ: when thepoet portrays a character, as when putting the plot together, he has to aimat necessity and probability (Poet.1454a33–36):crh; de; kaµ ejn to∂Í hßqesin oJmoÇwÍ w§sper kaµ ejn t¬Å tΩn pragmavtwn sustavsei a˚eµ zhte∂n h˙to; a˚nagka∂on h˙ to; e√kovÍ, w§ste to;n toiouÅton ta; toiauÅta levgein h˙ pravttein h˙ a˚nagka∂onh˙ e√ko;Í kaµ touÅto meta; touÅto gÇnesqai h˙ a˚nagka∂on h˙ e√kovÍ.In the characters too, exactly as in the structure of the incidents, [the poet] ought alwaysto seek what is either necessary or probable, so that it is either necessary or probablethat a person of such-and-such a sort say or do things of the same sort, and it is eithernecessary or probable that this [incident] happen after that one. (Trans. Janko 1987)32. Cf. Janko 1987, 109: “the character should be ‘like,’ literally; but like what? This means either thatthe type should be recognisable to us as one we know, i.e. lifelike, or that it should be like us. As there isno explanation or example, Aristotle must have regarded the meaning as obvious. . . . Clearly a characterwho is not lifelike will also be unlike ourselves, so the first explanation brings the second with it.”Francesca Schironi292The analysis of characters plays a central role in the second book of theRhetoric (Rh. 2.12–17.1388b31–1391b6), because the knowledge of differentcharacters is essential for a good orator. Here Aristotle develops these ideasespecially in the direction of what is typical of different ages. In three beau-tiful chapters (Rh. 2.12–14), where he is explaining how a good orator shoulddepict various hßqh in order to be persuasive, Aristotle highlights the maincharacteristics of young people (rash, optimistic, generous), old people (proneto reflect, pessimistic, selfish) and mature people (a middle way between thetwo). If one wants to achieve a good mimesis, one cannot depict a characterwith the characteristic of another age, because this would go against thecriterion of to; aÒrmovtton. In this light the poet must be particularly carefulwhen putting words into the mouths of his characters, because their levxiÍmust be hjqikhv, consistent with the character that uses it, as Aristotle makesclear in the third book of the Rhetoric (1408a10–11):to; de; prevpon e§xei hJ levxiÍ, eja;n ¬® paqhtikhv te kaµ hjqikh; kaµ to∂Í uÒpokeimevnoiÍ pravgmasina˚navlogon.Your language will be appropriate if it expresses emotion and character, and it corre-sponds to its subject. (Trans. Roberts 1984)This is because each age and class has its own mode of expression,and a good representation of a character must consider these characteristics(Rh. 1408a26–32):. . . a˚kolouqe∂ hJ aÒrmovttousa [i.e., de∂xiÍ] eJkavstå gevnei kaµ e§xei. levgw de; gevnoÍ me;n kaq’hJlikÇan, o∏on pa∂Í h˙ a˚nh;r h˙ gevrwn, kaµ gunh; h˙ a˚nhvr, kaµ Lavkwn h˙ QettalovÍ, e§xeiÍ dev,kaq’ a¶Í poiovÍ tiÍ tåÅ bÇå: ou˚ ga;r kaq’ a§pasan e§xin o¥ bÇoi poioÇ tineÍ. eja;n ou®n kaµ ta;ojnovmata o√ke∂a levg¬ t¬Å e§xei, poihvsei to; h®qoÍ: ou˚ ga;r tau˚ta; ou˚d’ wÒsauvtwÍ a˚gro∂koÍ a˙nkaµ pepaideumevnoÍ e≥peien.Each class of men, each type of disposition, will have its own appropriate way of lettingthe truth appear. Under “class” I include differences of age, as boy, man, or old man; ofsex, as man or woman; of nationality, as Spartan or Thessalian. By “dispositions” I heremean those dispositions only which determine the character of a man’s life, for it is notevery disposition that does this. If, then, a speaker uses the very words which are inkeeping with a particular disposition, he will reproduce the corresponding character;for a rustic and an educated man will not say the same things nor speak in the sameway. (Trans. Roberts 1984)The interest in characterization in Homer is very well attested for Aris-tarchus too. In Iliad 1.117, when Agamemnon, angered by the response ofCalchas, agrees to give Chryseis back and says, “I prefer that the army besafe rather than destroyed” (bouvlom’ ejgø lao;n sΩn eßmmenai h˙ a˚polevsqai),Aristarchus rejected the athetesis of Zenodotus (according to whom the linewas simplistic) by noticing (schol. Il. 1.117a) that ejn hßqei ga;r levgetai, thatis, the line suits the character of Agamemnon (whom Aristarchus, probably,did not hold in great esteem).3333. Cf. also schol. Il. 15.505a, where Aristarchus notes that the line is according to Ajax’ character(hjqikΩÍ); see also Didymus in schol. Il. 16.50a (ÂrÇstarcoÍ “e≥ tina oπda” dia; touÅ ei, Aim b(BCE3E4)T ªn’¬® hjqik∫teron).Aristotelian Principles in Aristarchean Philology 293Aristarchus seems to have followed Aristotle in arguing for consistencyand credibility of characters. Characters, according to Aristarchus, shouldbehave according to appropriateness (to; aÒrmovtton), decorum (to; prevpon),and what is fitting (to; o√ke∂on). In this case, as for Aristotle, a˚prephvÍ becomesa synonym of ou˚c aÒrmovttwn, in the sense of “not convenient,” “unsuitable”to the human type at issue.34 This criterion actually embraces all the subtledistinctions of the Poetics as well as those of the Rhetoric: characters areaÒrmovttonteÍ if they behave as their social position, their status, their age,their present situation, or their “mythical model” require. Aristarchus thusdenies “unheroic” words to Homeric heroes. It was unacceptable to haveAgamemnon dwelling on the pleasure he was going to enjoy from Chryseisback in Argos (schol. Il. 1.29–31: a˚qetouÅntai, . . . a˚prepe;Í de; kaµ to; to;n Âga-mevmnona toiauÅta levgein) or showing himself too greedy (schol. Il. 1.133–34a˚qetouÅntai, o§ti . . . kaµ mh; aÒrmovzonteÍ Âgamevmnoni).35 The last words ofAchilles to Patroclus as he is about to go to battle, in which Achilles wishesthat every Trojan and every Greek may die so that only the two of themcould survive and sack Troy,36 were athetized by Aristarchus, because in hisview they were not in line with Achilles’ character (schol. Il. 16.97–100a):a˚qetouÅntai stÇcoi tevssareÍ, diovti kata; diaskeuh;n ejmfaÇnousi gegravfqai uÒpov tinoÍ tΩnnomizovntwn ejraÅn to;n Âcilleva touÅ Patrovklou: toiouÅtoi ga;r o¥ lovgoi “pavnteÍ a˚povlointoplh;n hJmΩn.” kaµ oJ Âcilleu;Í ou˚ toiouÅtoÍ, sumpaqh;Í dev.Four lines are athetized because in their construction they appear to have been writtenby one of those who believed that Achilles was in love with Patroclus. For such are thewords “[I wish] that all might die but the two of us” [i.e., these words support the ideaof an Achilles in love with Patroclus]. But Achilles is not like that, but is instead sym-pathetic [to the Greeks].Aristarchus also refuses to accept an Achilles who pettily insults Aeneas(schol. Il. 20.180–86a: a˚qetouÅntai . . . kaµ o¥ lovgoi ou˚ prevponteÍ tåÅ touÅÂcillevwÍ pros∫på). Base insults to Diomedes are denied to Hector too(schol. Il. 8.164–66a: a˚qetouÅntai stÇcoi tre∂Í . . . a˚navrmosta de; kaµ ta; le-govmena to∂Í pros∫poiÍ), a great hero who moreover, on another occasion,does not “hope,” but rather “boasts,” that is, “vaunts” that he will win (schol.Il. 14.366a: o§ti ZhnovdotoÍ gravfei “kaµ eßlpetai.” aÒrmovzei de; tåÅ pros∫på to;eußcetai, kaucaÅtai [“because Zenodotus writes kaµ eßlpetai, but kaµ eußcetai,‘he vaunts,’ is fitting to the character”]). Aristarchus also clearly deplores34. This does not mean that, with the word a˚prephvÍ, Aristarchus implied a moral judgment of poetry;however, I would not agree totally with Schenkeveld (1970, 167–68), who maintains that a˚prephvÍ inAristarchean scholia is used only to point out a contradiction in the text. When it refers to characters,a˚prephvÍ implies “contradiction” only in the sense of being contradictory to what a real individual in thesame condition would have done; hence a˚prephvÍ means “unfitting.” On a˚prephvÍ, see also n. 42 below.35. See also Didymus in schol. Il. 4.345–46a: a˚prepΩÍ kaµ para; ta; provswpa e√Í kreavdion ojneidÇzontoÍtouÅ ÂgamevmnonoÍ.36. Il. 16.97–100: ai˙ ga;r ZeuÅ te pavter kaµ ÂqhnaÇh kaµ ◊Apollon / mhvtev tiÍ ou®n Tr∫wn qavnaton fuvgoio§ssoi eßasi, / mhvtev tiÍ ÂrgeÇwn, nΩ∑n d’ ejkduÅmen oßleqron, / oßfr’ oπoi TroÇhÍ ¥era; krhvdemna luvwmen (“I wish, OZeus father and Athena and Apollo, that none of the Trojans, as many as they are, could escape death, norany of the Argives, but that only the two of us could avoid destruction, so that we alone could loose thesacred veils of Troy”).Francesca Schironi294Ajax’ characterization as a miles gloriosus who brags of his military su-periority (schol. Il. 7.195–99: stÇcoi pevnte a˚qetouÅntai, o§ti ou˚ kata; to;n A≥antao¥ lovgoi kaµ eJautåÅ a˚nqupofevrei geloÇwÍ [“five lines are athetized becausethis speech is not worthy of Ajax and it is ridiculous that he replies to him-self”]).37 By the same token there are words that may suitably be said to kingsand words that may suitably be said to subjects, like those that Odysseusspeaks to the soldiers to keep them from going back home (Il. 2.203–6: “notall of us Achaeans are to be kings; the rule of many is not a good thing; letthere be only one ruler, one king to whom the son of Cronos, crooked ofcounsel, has given the scepter and the laws in order that he can take counselfor them”).38 According to Aristarchus these lines should not be employedby Odysseus when Odysseus is addressing the rank and file (Il. 2.200–206),but should instead be transferred to a previous point, when he is addressingthe other leaders (Il. 2.190–97). His point of view is clear from Aristonicus’remark in schol. Il. 2.192a: e√sµ ga;r pro;Í basile∂Í aÒrmovzonteÍ, ou˚ pro;ÍdhmovtaÍ (“for these lines are suitable to kings, not to rank soldiers”).The same holds for women, who must behave properly, not give ordersto superior beings like gods, as in the case of Helen rebuking Aphrodite atIliad 3.406–7 (schol. Il. 3.395: kaµ blavsfhma [add. Friedländer] para;to; provswpovn ejsti ta; legovmena “h•so par’ au˚to;n √ouÅsa, qeΩn d’ a˚poveipekeleuvqouÍ, / mhd’ eßti so∂si povdessin” (Il. 3.406–7) [“and it is blasphemouseven for the character [i.e., Helen] to say: ‘Go and sit by his side, and giveup the way of the gods, and [do not go back to Olympus] with your feet’ ”];or, as in the case of Andromache, giving tactical advice to her husband,Hector, during their meeting (schol. Il. 6.433–39: a˚qetouÅntai stÇcoi eJpta; e§wÍtouÅ (439) . . . o§ti a˚noÇkeioi o¥ lovgoi t¬Å Ândromavc¬: a˚ntistrathge∂ ga;r tåÅ£Ektori [“seven lines are athetized down to line 439, because the argumentsare not fitting for Andromache, since she is giving strategic advice insteadof Hector”]. A young princess like Nausicaa was not allowed to dwell toomuch on the gossip people might have made upon seeing her enter the citywith Odysseus (schol. Od. HQ 6.275: a˚qetouÅntai stÇcoi idu . . . wÒÍ a˚noÇkeioitåÅ uÒpokeimevnå pros∫på [“fourteen lines are athetized . . . because they arenot fitting to the character in question”]. A mother like Thetis cannot giveexcessively explicit advice to her son (schol. 24.130–32a: a˚qetouÅntai stÇcoitre∂Í, o§ti a˚prepe;Í mhtevra u¥åÅ levgein “a˚gaqovn ejsti gunaikµ mÇsgesqai” (cf. lines130–31) [“three lines are athetized because it is inappropriate for a motherto say to her son: ‘it is good to have intercourse with a woman’ ”]).As between men and women, there is a clear distinction between what isallowed to gods and to humans (schol. Il. 1.204b: touÅto de; t¬Å ÂqhnçÅ aÒrmovzeimaÅllon diabebaiouÅn [“but this sense of certainty about the future is moresuitable to Athena [than to Achilles]”]; schol. Il. 2.791: a˚qetouÅntai stÇcoipevnte: . . . PolÇt¬ a˚noÇkeion. maÅllon de; ¢Iridi aÒrmovzei ejpitavssein [“Five37. Cf. also schol. Il. 9.612b: o§ti ZhnovdotoÍ gravfei “ojdurovmenoÍ, kinurÇzwn,” o∏on qrhnΩn. eßsti de; ou˚cÔOmhriko;n kaµ para; to; provswpon [i.e., Phoenix].38. Il. 2.203–6: ou˚ mevn pwÍ pavnteÍ basileuvsomen ejnqavd’ ÂcaioÇ: / ou˚k a˚gaqo;n polukoiranÇh: e∏Í koÇranoÍeßstw, / e∏Í basileuvÍ, å• dΩke Krovnou pav∑Í a˚gkulomhvtew / skhÅptrovn t’ hjde; qevmistaÍ, ªnav sfisi bouleuv¬si.Aristotelian Principles in Aristarchean Philology 295lines are athetized. . . . [it is] not appropriate for Polites. To give commandsis more fitting to Iris”]). Also, the verbs suitable to the immortal gods arein the present, not the past tenses (schol. Il. 2.448c: ZhnovdotoÍ de; gravfeiparatatikΩÍ “hjerevqonto,” o§per ou˚c aJrmovzei ejpµ a˚qanavtwn). Even amonggods there is a hierarchy: what is permitted to Zeus is not allowed to Iris(schol. Il. 8.406–8: o§ti tåÅ touÅ Dio;Í pros∫på aÒrmovzousin o¥ lovgoi, tåÅ de; thÅÍ¢IridoÍ [sc. 8.420–22] ou˚kevti).39The distinction between Greeks and barbarians and what was fitting fora Greek hero to say played an interesting role in the athetesis of Iliad16.237. Here Achilles prays to Zeus and says: “you did me honor, andgreatly oppressed the army of the Achaeans” (timhvsaÍ me;n ejmev, mevga d’≥yao lao;n ÂcaiΩn). According to Aristarchus, the line was wrongly re-peated from Iliad 1.454, where it was at the right place, when Chryses ispraying to Apollo. For it was not likely for Achilles to rejoice at the defeatof the Greeks, whereas this was fitting for a barbarian enemy of the Greeks,like Chryses (schol. Il. 1.454: ou˚ ga;r e√kovtwÍ Âcilleu;Í ejpicaÇrei t¬Å h§tt¬tΩn ÔEllhvnwn. oJ de; CruvshÍ bavrbaroÍ kaµ misevllhn [“for it is not plausiblefor Achilles to rejoice at the defeat of the Achaeans, but Chryses is a bar-barian and enemy of the Greeks”]).40 The words of Aeneas to Achillessuggesting that they should not insult each other like women in the streetswere athetized by Aristarchus for the same reasons, because they were moresuitable to barbarians than to civilized Greeks (schol. Il. 20.251–55a1: a˚qe-touÅntai stÇcoi pevnte. . . . kaµ ta; legovmena a˚navxia tΩn pros∫pwn: kaµ para;barbavroiÍ dev, eßsti to; ta;Í guna∂kaÍ proercomevnaÍ loidore∂sqai wÒÍ par’A√guptÇoiÍ [“five lines are athetized. . . . and what is said is unworthy of thecharacters. Among barbarians it is possible to have women come out andhurl abuse, as among the Egyptians”]).41An important point is that words and actions must be suitable to theage and the social level of the characters. A typical case is Iliad 3.156–60,when the old Trojans see Helen arriving on the tower and remark that it isshameful to fight for a woman, no matter how beautiful she is. The scene isintroduced as follows (lines 154–55): oi ¶ d’ wÒÍ ou®n e≥donq’ ÔElevnhn ejpµ puvrgon√ouÅsan, / h®ka pro;Í a˚llhvlouÍ eßpea pterovent’ a˚govreuon (“and when they sawHelen arriving upon the tower, they softly spoke winged words to oneanother”). Aristarchus argued against a variant suggested by Zenodotus,≈ka, “swiftly,” instead of h®ka, “softly,” “in a low tone,” because that adverbwas a˚prepevÍ if it referred to Helen (a noble woman cannot come “quickly,”“run”: ejpµ puvrgon √ouÅsan ≈ka) and a˚navrmoston if it referred to the old Trojans(old men are bradulovgoi, they cannot speak “swiftly”: ≈ka pro;Í a˚llhvlouÍeßpea pterovent’ a˚govreuon).42 Similar to this case is that of Odyssey 15.45,39. Cf. schol. Il. 8.420–24a.40. Cf. also schol. Il. 16.237a.41. Cf. Lührs 1992, 117–20, esp. 119.42. Schol. Il. 3.155a: h®ka {pro;Í a˚llhvlouÍ}: o§ti ZhnovdotoÍ gravfei “w® ka.” e≥te de; ejpµ thÅÍ ÔElevnhÍ ejstÇn, o§tiw®ka ejporeuveto, a˚prepe;Í eßstai: e≥te ejpµ tΩn dhmogerovntwn, o§ti w® ka dielevgonto, a˚navrmoston: bradulovgoigavr e√sin o¥ gevronteÍ. This scholium raises, moreover, the question of a possible distinction between the mean-ing of a˚prepevÍ and a˚navrmoston/ou˚c a§rmotton: whereas the first seems here to point to a specific appropriate-ness, related to the character Helen (running is not fitting to Helen), the latter seems to hint at a genericFrancesca Schironi296where Telemachus wakes up Pisistratus by touching him with his foot (la;xpodµ kinhvsaÍ). According to Aristarchus, this way of waking someone up ismore suitable to Nestor, who is old and does it at Iliad 10.158 to wake upDiomedes.43 These comments seem to have behind them the same ideas aboutold age as we read in the Rhetoric, Book 2.13, though, admittedly, Aristotledoes not talk about the “slowness” and lack of strength of old people, butfocuses on a psychological description. The same idea of old age as measuredand never inclined to excess suggests that measured and dignified languageis fitting to old kings like Priam44 and Alcinous.45 The reverse is true foryoung people, who should respect the old and restrain themselves from speak-ing too openly.46 Also unfitting are references to marriage and family forwarriors who are too young for it (schol. Il. 15.439a: πsa fÇloisi tokeuÅsin:o§ti ZhnovdotoÍ gravfei “πsa fÇloisi tevkessin.” ou˚c aÒrmovzei de; tou;Í perµ to;nA≥anta nevouÍ oßntaÍ levgein “tevkessin”: prµn ga;r paidopoihÅsai ejstrateuv-santo [“because Zenodotus writes πsa fÇloisi tevkessin [we honour him] ‘likeour sons’ [instead of πsa fÇloisi tokeuÅsin, ‘like our parents’]. But it doesnot fit those around Ajax, who are young, to say ‘like our sons.’ They servedin the army before begetting children”]).A case where the contacts between Aristotelian theory and Aristarchus’practice are particularly strong and interesting is Odyssey 4.156–60. Here,Pisistratus speaks to Menelaus and confirms that Telemachus is Odys-seus’ son:“Âtrei?dh Menevlae diotrefevÍ, oßrcame laΩn,keÇnou mevn toi o§d’ u¥o;Í ejthvtumon, wÒÍ a˚goreuveiÍ:a˚lla; saovfrwn ejstÇ, nemessaÅtai d’ ejnµ qumåÅw•d’ ejlqøn to; prΩton ejpesbolÇaÍ a˚nafaÇneinaßnta sevqen, touÅ nΩ∑ qeouÅ w¶Í terpovmeq’ au˚d¬Å.”Divine Menelaus, son of Atreus, leader of people, he is indeed that man’s son, as you say.But he is prudent and feels shame in his heart that on his first coming he might showhimself hasty in speaking in the presence of you, in whose voice we both take delightas in a god’s.43. Schol. HVind.133 Od. 15.45: noqeuvetai wÒÍ diapeplasmevnoÍ ejx hJmisticÇou thÅÍ k ∆IliavdoÍ (line 158):ejke∂ ga;r proshkovntwÍ Nevstwr koim∫menon Diomhvdhn a˚nÇsthsi, kuvyai katoknhvsaÍ dia; to; ghÅraÍ.44. Schol. Il. 24.556–57: a˚qetouÅntai, o§ti a˚navrmostoi tåÅ pros∫på a¥ eu˚caµ kaµ ejpautovfwroÍ hJ uÒpovkrisiÍ.Cf. also Didymus at Il. 24.636a: ou§twÍ ÂrÇstarcoÍ a˚ntµ touÅ a˚napaus∫meqa: a˚prepe;Í ga;r to;levgein to;n PrÇamon “tarp∫meqa.”45. Schol. P Od. 7.311: tou;Í e¶x ÂrÇstarcoÍ distavzei ÔOmhvrou eπnai. e√ de; kaµ ÔOmhrikoÇ, e√kovtwÍ au˚tou;ÍperiaireqhÅnaÇ fhsi. pΩÍ ga;r a˚gnoΩn to;n aßndra mnhsteuvetai au˚tåÅ th;n qugatevra kaµ ou˚ protrepovmenoÍ,a˚lla; liparΩn… Cf. also schol. Il. 9.56–57, about Nestor saying to Diomedes: “You might even be my son”:o§ti ou˚c aÒrmovzei tåÅ tre∂Í genea;Í bebiwkovti levgein “ejmo;Í a˙n pa∂Í e≥hÍ” (cf. line 57) kata; th;n hJlikÇan. pro;ÍejpitÇmhsin touÅ polevmou.46. In this sense, one could quote the case of Od. 2.316–17, lines probably athetized by Aristarchus onthe grounds that Telemachus here is threatening the suitors too strongly (schol. EM Od. 2.325: h® mavla] bebaiw-tika; tauÅta ta; eßph touÅ mh; e√rhÅsqai uÒpo; Thlemavcou tou;Í prohqethmevnouÍ stÇcouÍ (lines 316–17) “peirhvsww§Í k’ ußmmi kaka;Í ejpµ khÅraÍ √hvlw hje; Puvlond’ ejlqøn h˙ au˚touÅ tåÅd’ ejnµ dhvmå.” a˚porouÅnteÍ ga;r levgousin “h®mavla ThlevmacoÍ,” ou˚k a˙n a˚porhvsanteÍ o¥ proakhkoovteÍ).appropriateness, related to the characteristics of the human type depicted (speaking quickly is not fitting toold people in general). However, the evidence from the other scholia, which use the two terms inconsis-tently, prevents the acceptance of such conclusions.Aristotelian Principles in Aristarchean Philology 297Lines 158–60 were athetized by Aristarchus because they were consideredcontrary to the traditional usage (ta; pavtria) and not appropriate for the char-acter of Pisistratus; moreover, they were superfluous and utterly unsuitableto be spoken by a young man (para; ta; pavtria kaµ ou˚c aJrmovttonta tåÅ Pei-sistravtou pros∫på . . . perittoµ kaµ uÒpo; nevou pantavpasi levgesqai a˚prepe∂Í).This comment by Aristarchus finds a striking overlap with what Aristotlehad stated in the Rhetoric (1395a2–6 and 1404b15–16):aÒrmovttei de; gnwmologe∂n hJlikÇç me;n presbutevrwn, perµ de; touvtwn w•n eßmpeirovÍ tÇÍ ejstin,w§ste to; me;n mh; thlikouÅton oßnta gnwmologe∂n a˚prepe;Í w§ ster kaµ to; muqologe∂n, perµde; w•n aßpeiroÍ, hjlÇqion kaµ a˚paÇdeuton.The use of maxims is appropriate only to elderly men, and in handling subjects inwhich the speaker is experienced. For a young man to use them is—like telling stories—unbecoming; to use them in handling things in which one has no experience is silly andill-bred. (Trans. Roberts 1984)ejpeµ kaµ ejntauÅqa [i.e., in poetry], e√ douÅloÍ kalliepo∂to h˙ lÇan nevoÍ, a˚prepevsteron, h˙perµ lÇan mikrΩn.For even in poetry, it is not quite appropriate that fine language should be used by aslave or a very young man, or about very trivial subjects. (Trans. Roberts 1984)A young man cannot speak in gnΩmai either for Aristotle or for Aris-tarchus.475. Aristotelian Theory and Aristarchean Practice: DIANOIAAristotle defines diavnoia as the thought-element, everything transmitted andexpressed with words (Poet. 1450a6–7 and 1456a36–1456b2):diavnoian dev [levgw], ejn o§soiÍ levgonteÍ a˚podeiknuvasÇn ti h˙ kaµ a˚pofaÇnontai gn∫mhn.By “reasoning,” I mean the way in which they use speech to demonstrate something orindeed to make some general statement. (Trans. Janko 1987)eßsti de; kata; th;n diavnoian tauÅta, o§sa uÒpo; touÅ lovgou de∂ paraskeuasqhÅnai. mevrh de;touvtwn tov te a˚podeiknuvnai kaµ to; luvein kaµ to; pavqh paraskeuavzein (o∏on eßleon h˙ fovbonh˙ ojrgh;n kaµ o§sa toiauÅta) kaµ eßti mevgeqoÍ kaµ mikrovthtaÍ.All [the effects] that have to be produced by speech fall under reasoning. The types ofthese are demonstration and refutation, the production of emotions (e.g., pity, terror,anger, etc.), and again [arguments about things’] importance or unimportance. (Trans.Janko 1987)Thus, with diavnoia we are dealing with the content of speeches, eitheruttered by characters or by the poet himself when he intervenes in the nar-rative. In this sense it is opposed to levxiÍ, which is the form in which thethought takes shape. In particular, Aristotle states that the two main parts ofthe diavnoia are tov te a˚podeiknuvnai kaµ to; luvein (“demonstration and refuta-tion”) and kaµ to; pavqh paraskeuavzein kaµ eßti mevgeqoÍ kaµ mikrovthtaÍ (“the47. The attention toward characterization and what is fitting for each character is also used in the so-calledluvseiÍ ejk touÅ pros∫pou, which are typical of Aristarchus’ exegesis; on this principle, see Dachs 1913.Francesca Schironi298production of emotions, and again importance or unimportance”). Aristotlein the Poetics does not treat diavnoia extensively, but refers his readers to theRhetoric, where the opposition between the thought-element versus the formalelement in speeches is well stated (Rh. 1404a18–19):o¥ ga;r grafovmenoi lovgoi me∂zon √scuvousi dia; th;n levxin h˙ dia; th;n diavnoian.Speeches of the written kind owe more of their effect to their language than to theirthought. (Trans. Roberts 1984)48The thought expressed by the characters or by the poet must thus reflectthe general “ethos” of a poetic work. In particular, the diavnoia of tragedyand epic must be in keeping with the “solemnity” (semnovthÍ) and seriousnessthat characterize both genres. Therefore, anything that is gelo∂on, “ridiculous,”must be avoided, because this is typical of comedy and low genres and ex-traneous to tragedy and epic (Poet. 1449a32–37):hJ de; kwmådÇa ejstµn w§sper e≥pomen mÇmhsiÍ faulotevrwn mevn, ou˚ mevntoi kata; paÅsankakÇan, a˚lla; touÅ a√scrouÅ ejsti to; gelo∂on movrion. to; ga;r gelo∂ovn ejstin aJmavrthmav tikaµ aπscoÍ a˚n∫dunon kaµ ou˚ fqartikovn, o∏on eu˚qu;Í to; gelo∂on provswpon a√scrovn ti kaµdiestrammevnon aßneu ojduvhnÍ.Comedy is, as we have said, a representation of people who are rather inferior—not,however, with respect to every [kind of ] vice, but the laughable is [only] a part of whatis ugly. For the laughable is a sort of error or ugliness that is not painful and destruc-tive, just as, evidently, a laughable mask is something ugly and distorted without pain.(Trans. Janko 1987)49Aristarchus certainly shared the idea that epic is mainly concerned with“serious” content, for he often rejects lines expressing a ridiculous thought.In particular, there are some interesting cases where Aristarchus arguesagainst Homer himself. At Iliad 12.176, describing the battle at the Achaeanwall, the poet says: “it would be too much toil for me, as if I were a god,to tell all this” (a˚rgalevon dev me tauÅta qeo;n w¶Í pavnt’ a˚goreuÅsai), a statement48. A correspondence with this dichotomy between form (levxiÍ) and thought (diavnoia) can be found inthe three books of the Rhetoric, where the first two deal with the thought, whereas Book 3 deals with thestyle. In particular, Book 1, where Aristotle analyzes the different kind of speeches and arguments, seemsto correspond to the first part of diavnoia (“proof and refutation”), while Book 2, dedicated to the analysisof the different emotions and characters that the good orator must imitate in order to persuade, reflects thesecond and third parts (“the arousing of feelings and then again exaggeration and depreciation”). Book 3,instead, is about the pezh; levxiÍ, the prose style, extraneous to poetry. This is probably why in the PoeticsAristotle refers his readers to the Rhetoric as far as diavnoia is concerned (the thought-element obeys thesame rules both for prose and for poetry); whereas he needs to have a proper section on style in the Poetics(poihtikh; levxiÍ is not the same as pezh; levxiÍ). That this dichotomy of form vs. content is operating in Aris-totle’s Rhetoric is shown by the end of Book 2 (1403a34–1403b3): ejpeµ de; trÇa eßstin a¶ de∂ pragmateuqhÅnaiperµ to;n lovgon, uÒpe;r me;n paradeigmavtwn kaµ gnwmΩn kaµ ejnqumhmavtwn kaµ o§lwÍ tΩn perµ th;n diavnoian, o§qente eu˚porhvsomen kaµ wÒÍ au˚ta; luvsomen, e√rhvsqw hJm∂n tosauÅta, loipo;n de; dielqe∂n perµ levxewÍ kaµ tavxewÍ(“Three points must be studied in making a speech and now we have completed the account of examples,maxims, enthymemes, and in general the thought-element—the way to invent and refute arguments. We havenext to discuss language and arrangement,” trans. Roberts 1984).49. On the opposition between gelo∂on (typical of comedy) and semnovn (typical of tragedy), cf. Rh. 1406b6–8: e√sµn ga;r kaµ metaforaµ a˚prepe∂Í, a¥ me;n dia; to; gelo∂on (crΩntai ga;r kaµ o¥ kwmådopoioµ metafora∂Í),a¥ de; dia; to; semno;n aßgan kaµ tragikovn (“Metaphors like other things may be inappropriate. Some are sobecause they are ridiculous (they are indeed used by comic poets too). Others are too grand and tragic,”trans. Roberts 1984, modified).One Line LongAristotelian Principles in Aristarchean Philology 299that Aristarchus rejected as gelo∂on.50 For the same reason he found fault inthe concluding line of the long catalogue of the Nereids mourning Patroclus(Il. 18.38–49). Here, after giving a list of thirty-three Nereids, the poet con-cludes: “and all the other Nereids that were in the depth of the sea” (aßllaiq’ ai ¶ kata; bevnqoÍ aJlo;Í Nhrhi?deÍ h®san)—as if, Aristarchus commented,Homer first intended to list them all by name, but then got tired (schol.Il. 18.39–49: gelo∂ovn te ejx ojnovmatoÍ proqevmenon e√pe∂n pavsaÍ, w§spera˚pokamovnta e√pe∂n “aßllai d’ai ¶ ktl.”). Also, the consoling example ofNiobe that Achilles tells to Priam is considered ridiculous, because a straight-forward paraphrase of the exemplum would give: “eat, because Niobe tooate and then was turned into stone” (schol. Il. 24.614–17a: kaµ hJ paramuqÇageloÇa: favge, ejpeµ kaµ hJ Niovbh eßfage kaµ a˚peliq∫qh)—indeed a rather poorencouragement.51Aristarchus found fault also with the words uttered by the characters. Forexample, at Iliad 8.189 Hector talking to his horses and inviting them todrink was ridiculous: geloiovtatoÍ ejpµ ªppwn oJ stÇcoÍ, o§ti oπnon ªppoi ou˚pÇnousi. kaµ o§ti “qumo;Í a˚n∫goi e√Í mevqhn” gelo∂on.52 Other atheteses byAristarchus due to a gelo∂on thought in the words of characters are attestedat Iliad 10.409–11 (Odysseus to Dolon);53 Iliad 14.376–77 (Poseidon tothe Greeks); Odyssey 4.158–60 (Pisistratus to Menelaus); Odyssey 4.553(Menelaus to Proteus); Odyssey 11.157–59 (Anticlea to Odysseus); Odyssey14.495 (Odysseus to the Greeks at Troy). Alternatively, Aristarchus notessomething gelo∂on in the readings of predecessors (Zenodotus in particular),which are therefore to be rejected, at Iliad 3.74 (Paris to Hector) and Iliad23.94 (Achilles to Patroclus’ ghost).54The simile comparing Achilles and Hector running around Troy to a dreamin which the pursuer cannot reach the one who runs away at Iliad 22.199–201 (wÒÍ d’ ejn ojneÇrå ou˚ duvnatai feuvgonta di∫kein: / oußt’ aßr’ o ¶ to;n duvnataiuÒpofeuvgein oußq’ o ¶ di∫kein: / w¶Í o¶ to;n ou˚ duvnato mavryai posÇn, ou˚d’ o¶Í a˚luvxai)was athetized by Aristarchus, as he found the lines “cheap” both in styleand in thought (schol. Il. 22.199–201a): kaµ t¬Å kataskeu¬Å kaµ tåÅ nohvmatieu˚tele∂Í.55 Eu˚telhvÍ, “cheap,” “of no value,” is thus used as a synonym ofgelo∂oÍ in the scholia of Aristonicus to convey the idea that the contentor the style of a passage are not consonant with the serious content of theepos. Aristotle likewise employs the adjective eu˚telhvÍ when comparing two50. Gelo∂on is surely part of the Aristarchean vocabulary since it appears in a direct quotation of Aristarchusby Didymus in schol. Il. 2.420a1.51. Cf. also the first part of schol. Il. 24.614–17a: a˚qetouÅntai stÇcoi tevssareÍ, o§ti ou˚k a˚kovlouqoi tåÅ “hJ d’aßra sÇtou mnhvsat’, ” (Il. 24.613)· e√ ga;r a˚peliq∫qh, pΩÍ sitÇa prohnevgkato…52. Cf. Lührs 1992, 46 n. 92.53. On this athetesis, see Lührs 1992, 226–28.54. A rather close synonym of gelo∂on is eu˚hvqhÍ, “foolish.” Aristarchus athetized lines because of their“foolishness” in the thought expressed by the character: at Il. 1.139 (Agamemnon to Achilles); Il. 2.80–81(in schol. Il. 2.76a, Nestor commenting on Agamemnon’s dream); Il. 8.185 (hJ prosf∫nhsiÍ, i.e., the addressof Hector to his own horses, eu˚hvqhÍ). This criterion was also at the basis of an athetesis by Zenodotus inschol. Il. 1.117a (a scholium already mentioned): bouvlom’ ejgø lao;n sovon: o§ti ZhnovdotoÍ au˚to;n hjqevthkenwÒÍ thÅÍ dianoÇaÍ eu˚hvqouÍ oußshÍ. ou˚ de∂ de; au˚to;n √dÇç profevresqai, a˚lla; sunavptein to∂Í aßnw: ejn hßqei ga;rlevgetai.55. Though for different reasons, the dÇwxiÍ of Hector by Achilles in Il. 22 was problematic for Aristotletoo (cf. Poet. 1460a11–18).Francesca Schironi300contrasting types of authors: the semnoÇ who will write tragedies and theeu˚tele∂Í who will prefer comedies (Poet. 1448b25–27):o¥ me;n ga;r semnovteroi ta;Í kala;Í ejmimouÅnto pravxeiÍ kaµ ta;Í tΩn toiouvtwn, o¥ de; eu˚televsteroita;Í tΩn fauvlwn, prΩton yovgouÍ poiouÅnteÍ w§sper e§teroi u§mnouÍ kaµ ejgk∫mia.The grander people represented fine actions, i.e., those of fine persons, the more ordinarypeople represented those of inferior ones, at first composing invectives, just as the otherscomposed hymns and praise-poems. (Trans. Janko 1987)Therefore, in Homer we must avoid everything that is eu˚televÍ. And indeed,we find many Aristarchean atheteses on the basis that some lines are eu˚tele∂Ít¬Å sunqevsei kaµ t¬Å dianoÇç (schol. Il. 1.133–34, Agamemnon to Achilles;schol. Il. 15.212a, Poseidon to Iris; schol. Il. 16.93a, Achilles to Patroclus;schol. HP Od. 5.94–95, probably referring to lines 97–98, Hermes toCalypso), some eu˚tele∂Í kata; th;n diavnoian (schol. Il. 3.395 on Il. 3.414,Aphrodite to Helen), some eu˚tele∂Í t¬Å kataskeu¬Å kaµ to∂Í nohvmasi (schol.Il. 20.180–86a, Achilles to Aeneas).56On the other hand, the Aristotelian principle that epic diavnoia is concernedwith or must express importance (mevgeqoÍ; cf. Poet. 1456a36–1456b2 quotedabove) seems to be at the basis of Aristarchean practice too. Arguing againstan athetesis of Zenodotus in Iliad 17.260, Aristarchus maintained that thoselines containing a comment by the poet himself (tΩn d’ aßllwn tÇÍ ken ¬•sifresµn ou˚novmat’ e≥poi, / o§ssoi dh; metovpisqe mavchn hßgeiran ÂcaiΩn; [“Butof the rest, what man of his own wit could name the names of all that cameafter these and aroused the battle of the Achaeans?”] were actually increas-ing the importance (to; mevgeqoÍ) of the battle for Patroclus’ corpse (schol. Il.17.260a: to; mevgeqoÍ thÅÍ uÒpe;r Patrovklou mavchÍ).6. Aristotelian Theory and Aristarchean Practice: LEXISThe analysis of poihtikh; levxiÍ is carried out by Aristotle (Poet. 1457b1–3)with reference to the oßnoma: each word can be normal (kuvrion), a gloss(glΩtta), metaphorical (metaforav), ornamental (kovsmoÍ), invented (pepoih-mevnon), lengthened (ejpektetamevnon), truncated (uÒf¬rhmevnon), or altered(ejxhllagmevnon). Moreover, whereas the kuvrion usage is typical of the prosestyle, the other schemata are characteristic of poetic style. In particular,at Poetics 1459a9–10, glossai are said to be the most apt to epic; this, asalready mentioned, is one of the points of contact between Aristotle andthe work done at Alexandria, where collections of glossai were widely pro-duced. However, poetic style cannot be reduced only to figures of speech;to be a good poet one has also to be clear (Poet. 1458a18–34):levxewÍ de; a˚reth; safhÅ kaµ mh; tapeinh;n eπnai. safestavth me;n ou®n ejstin hJ ejk tΩn kurÇwnojnomavtwn, a˚lla; tapeinhv: . . . semnh; de; kaµ ejxallavttousa to; √diwtiko;n hJ to∂Í xeniko∂Íkecrhmevnh: xeniko;n de; levgw glΩttan kaµ metafora;n kaµ ejpevktasin kaµ paÅn to; para; to;kuvrion. a˚ll’ aßn tiÍ a§panta toiauÅta poihvs¬, h˙ a≥nigma eßstai h˙ barbarismovÍ: a˙n me;n ou®nejk metaforΩn, a≥nigma, eja;n de; ejk glwttΩn, barbarismovÍ. . . . de∂ aßra kekraÅsqaÇ pwÍtouvtoiÍ. to; me;n ga;r to; mh; √diwtiko;n poihvsei mhde; tapeinovn, o∏on hJ glΩtta kaµ hJmetafora; kaµ oJ kovsmoÍ kaµ ta®lla ta; e√rhmevna e≥dh, to; de; kuvrion th;n safhvneian.56. On this athetesis, see Lührs 1992, 191 n. 153.One Line LongAristotelian Principles in Aristarchean Philology 301The virtue of diction is to be clear and not commonplace. Diction made up of standardnames is the clearest, but it is commonplace. . . . Diction that uses unfamiliar names isgrand and altered from the everyday. By “unfamiliar,” I mean the exotic [name], metaphor,lengthening and everything that is contrary to what is standard. But if someone makesall [the names] of this sort, [his poem] will be either a riddle or gibberish. If [it is com-posed] of metaphors, it will be a riddle; if of exotic [names], gibberish. . . . [The poet],then, should mix these [two kinds] in some way. The first (i.e. the exotic name, metaphor,ornament and the other kinds we mentioned) will produce that which is not everydayand commonplace, and the standard name will produce clarity. (Trans. Janko 1987)To achieve clarity one must use “normal” words (kuvria); however, this canlead to tapeinovthÍ, “meanness of style,” which must be avoided in poeticstyle. Therefore poets also use words that get away from common language(ejxallavttousa to; √diwtiko;n hJ to∂Í xeniko∂Í kecrhmevnh), for example, glossaiand metaphors. However, they must not be excessive, otherwise the resultwill be an a≥nigma (if there are too many metaphors) or a barbarismovÍ (ifthere are too many glossai).57 The aim is thus to achieve an equilibrium be-tween norm and novelty, where the true poet is able to mix the two and thusbe at the same time both clear and striking.Attention to both form and style is evident in Aristarchus; in particular,the definition of what is Homeric style and what is not seems to be thecoherent development of the Aristotelian distinction between poetic andprosaic levxiÍ.58 As for Aristotle, for Aristarchus too language must first besafhvÍ, “clear,” and Homer is a champion of safhvneia: he always builds hissentences in the clearest way, e§neka safhneÇaÍ (cf. schol. Il. 15.8a). For thisvery reason, Aristarchus argued against Zenodotus at Iliad 14.169: eßnq’ h§ g’e√selqouÅsa quvraÍ ejpevqhke faeinavÍ (“And she [Hera] entered there [thechamber], closed the bright doors”), where Zenodotus changed the finite verbejpevqhke into a participle ejpiqe∂sa (so that the main verb was to be found inline 171, kavqhren). Homer—Aristarchus argued—prefers to have anothermain clause with a finite verb at line 169 and thus to be clear and avoid apostponed principal clause.59Aristotle in particular maintained that safhvneia was due to the usage ofkuvria ojnovmata, “standard names.” It is interesting how many times Aristarchusnotes that a term in Homer is used kurÇwÍ, that is, according to the normalusage, hence “clear.”60 On the other hand, Aristarchus seems to be well awareof the characteristic of poetic language, which is allowed to depart fromkuvrion usage. Schol. Il. 5.266b describes as √dÇwÍ, that is, proper to Homer,the usage of the word poinhv in this passage (in the sense of “price paid,”57. For a similar analysis of prose style, see Rh. 1404b1–1408b20.58. In this case, an important change in terminology happened, since for Aristarchus levxiÍ meant “word”and not “diction,” “utterance,” “style,” as in Aristotle; cf. Matthaios 1999, 198–200; 1996, 68–69. To mean“diction,” “composition of words” in Aristonicus’ scholia we find suvnqesiÍ and kataskeuhv.59. Schol. Il. 14.169a: quvraÍ ejpevqhke faeinavÍ: o§ti ZhnovdotoÍ gravfei “quvraÍ ejpiqe∂sa faeinavÍ,” ªna sunafh;ÍoJ lovgoÍ gevnhtai. oJ de; £OmhroÍ aßllaÍ a˚rca;Í lambavnei, ªna mh; a˚safh;Í hJ perÇodoÍ gevnhtai hßtoi uÒsteroperÇo-doÍ. On the concept of safhvneia in Homer, cf. also Didymus in schol. Il. 6.76b1 and schol. Il. 8.349a1.60. Cf. schol. Il. 4.141a, 7.146b, 7.255a, 10.75b, 11.523, 22.31, 22.319a, 22.489b. Aristarchus couldnot adopt the very word of Aristotle, kuvrion oßnoma, which, by that time, meant something different, denotingthe grammatical category of “proper names” (or just “names”), as opposed to that of epithets; see Matthaios1996, esp. 69–70; 1999, 218–25.Francesca Schironi302“recompense”) and opposes it to the standard usage (kurÇwÍ) of the word, asat Iliad 9.636 (in the sense of a “fine paid by the slayer to the kinsmen ofthe slain”).61 Sometimes when a word is used ou˚ kurÇwÍ the purpose is to givea deeper meaning, as at Iliad 2.670 where the image of Zeus pouring goldis said to be kurÇwÍ in Pindar (Ol. 7.34) and metaphoric in Homer to suggestthe aboundance of gold (pro;Í eßmfasin touÅ plouvtou). Or the epithets a˚rgu-rovhlon “with silver nails,” at Iliad 2.45, and crusovhlon, “with goldennails,” at Iliad 11.29–30—both referring to the sword of Agamemnon—arenot to be taken in the standard way (kurÇwÍ). Otherwise, Homer would con-tradict himself and depict Agamemnon’s sword once as silver nailed, onceas golden nailed. Rather, they are used ornamentally as a poetical device.62Aristarchus is also keen to note metaphorical usages in Homer, as in schol.Il. 1.37e (kata; metafora;n ejk tΩn tetrapovdwn); schol. Il. 11.632b (kata; meta-fora;n a˚po; tΩn ejmyuvcwn); schol. Il. 22.356a (a˚po; tΩn oßsswn hJ metaforav).63Just as Aristotle allows poetic diction to depart from common languagein order to avoid tapeinovthÍ, “meanness,” so Aristarchus athetizes linesor rejects readings because the words (levxeiÍ), the style (kataskeuhv), orthe composition (suvnqesiÍ)64 are eu˚tele∂Í, as happens in schol. Il. 2.314b(eut˚elh;Í de; hJ levxiÍ); schol. Il. 8.164–66a (eu˚tele∂Í e√si t¬Å kataskeu¬Å);schol. Il. 10.497a (t¬Å sunqevsei eu˚telhvÍ); schol. Il. 11.130a (eu˚tele∂Í t¬Åkataskeu¬Å); schol. Il. 11.413a (eu˚telh;Í gÇnetai hJ suvnqesiÍ); schol. Il. 11.767a1(eu˚telh;Í de; hJ suvnqesiÍ); schol. Il. 15.56a (kata; th;n suvnqesÇn e√sin eu˚tele∂Í).65However, notwithstanding this shift from common language, Homer neverbarbarÇzei. Hence, anything against grammatical correctness must be rejected,as in schol. Il. 12.34 (o§ti ZhnovdotoÍ gravfei “w¶Í hßmellon oßpisqe.” eßsti de;bavrbaron) and in schol. Il. 15.716 (w§ste barbarÇzein to;n £Omhron), bothagainst Zenodotus’ readings.Finally, some interesting parallels with the Rhetoric. For prose, one of themost serious vices is to be yucrovn, “frigid,” and this can arise from fourcauses: the usage of glossai, of excessive epithets, of strange compounds,and of improper metaphors, all devices that pertain to poetry rather thanprose (Rh. 1405b35–1406a13):ta; de; yucra; ejn tevttarsi gÇgnetai kata; th;n levxin, eßn te to∂Í diplo∂Í ojnovmasin, . . .pavnta tauÅta ga;r poihtika; dia; th;n dÇplwsin faÇnetai. mÇa me;n ou®n au§th a√tÇa, mÇa de; to;crhÅsqai gl∫ttaiÍ, . . . trÇton d’ ejn to∂Í ejpiqevtoiÍ to; h˙ makro∂Í h˙ a˚kaÇroiÍ h˙ pukno∂ÍcrhÅsqai: ejn me;n ga;r poihvsei prevpei “gavla leuko;n” e√pe∂n, ejn de; lovgå ta; me;n a˚prepevstera.Frigidities in language may take any of four forms: the misuse of compound words, . . .The way all these words are compounded makes them, we feel, fit for verse only. This,then, is one form in which bad taste is shown. Another is the employment of strange61. Schol. Il. 5.266b: poinh;n <–ou§nek’>: o§ti √dÇwÍ tåÅ poinhv kevcrhtai: kurÇwÍ ga;r ejpµ fovnou, “poinh;ndexamevnå” (Il. 9.636). . . . Similarly, the distinction between a usage kurÇwÍ and a usage that is not standardis noted by Aristarchus in schol. Il. 8.439a and schol. Il. 10.528b.62. Cf. schol. Il. 2.45a.63. Cf. also schol. Il. 1.51c, 2.49b, 4.521a, 5.21b, 5.299a, 11.390a, 13.147a, 13.317, 13.420, 13.745–46a,23.226c, 23.273a. On the expressions kurÇwÍ, a˚kuvrwÍ, and metaforikΩÍ, cf. Matthaios 1996, 66, with n. 44.64. In particular, on suvnqesiÍ as a technical term, see Schironi, in Bottai and Schironi 1997, 1058–62.65. On the athetesis of Il. 15.56, see Lührs 1992, 129–32. The word eu˚telhvÍ is used by Aristotle(Poet. 1458b19–22) as a negative term to compare two different results in composing iambi: a good one byAeschylus and a bad one by Euripides.One Line LongAristotelian Principles in Aristarchean Philology 303words. . . . A third form is the use of long, unseasonable, or frequent epithets. It is appro-priate enough for a poet to talk of “white milk,” but in prose such epithets are sometimeslacking in appropriateness. (Trans. Roberts 1984)In Aristarchus, we find the same principle applied to poetry, which mustavoid too many prose elements: some atheteses are due to the style, whichis pezovÍ, “prosaic” (schol. Il. 2.252a: a˚qetouÅntai stÇcoi pevnte, o§ti pezovteroit¬Å sunqevsei).66On the other hand, Aristarchus, like Aristotle, takes exception to epithets“out of place” (ejpÇqeta aßkaira).67 In this case, he seems to have appliedAristotle’s principles to poetic style, because sometimes he rejects a linebecause the epithet there is aßkairoÍ. This happens at Iliad 21.331, whenHera, addressing her son Hephaestus, calls him kullopodÇwn, “club-footed”(schol. Il. 21.331a: aßkairon to; ejpÇqeton: hJ ga;r filanqrwpeuomevnh kaµlevgousa “ejmo;n tevkoÍ” ou˚k wßfeilen a˚po; touÅ ejlatt∫matoÍ prosfwne∂n [“theepithet is out of place. For since she regularly shows kindness and says ‘mychild,’ she ought not to have addressed him by mentioning his defect”]), orat Iliad 21.218, when the Scamander, talking to Achilles, calls his streamsejrateinav, “pleasant” (schol. Il. 21.218a: o§ti aßkairon to; ejpÇqeton: pefoÇniktaiga;r uÒpo; touÅ aªmatoÍ [“because the epithet is out of place: for [the river] is redwith blood”]). Iliad 23.581 was athetized because Menelaus, angry with Anti-lochus, address him as diotrefevÍ (schol. Il. 23.581a: a˚qete∂tai, o§ti a˚kaÇrwÍlevgei diotrefevÍ, ojrgizovmenoÍ au˚tåÅ [“it is athetized because he says ‘fosteredof Zeus’ inappositely, since he is angry with him”]). Often Aristarchus findsthat the epithet d∂oÍ, “divine,” is used a˚kaÇrwÍ: for example, when Menelausaddresses his enemy Paris (schol. Il. 3.352a: a˚qete∂tai, o§ti . . . kaµ “d∂on”a˚kaÇrwÍ oJ MenevlaoÍ to;n ejcqro;n levgei [“The line is athetized because . . .and Menelaus calls his enemy ‘divine’ inappositely”]), or for the voluptuousAnteia (schol. Il. 6.160a), or for Hector when he is talking to himself (schol.Il. 7.75a1). Aristarchus’ principle is that epithets must be aÒrmovttonta, “fitting,”to the names and contexts to which they refer, just as Aristotle explains (Rh.1405a10–13):de∂ de; kaµ ta; ejpÇqeta kaµ ta;Í metafora;Í aJrmottouvsaÍ levgein. touÅto d’ eßstai ejk touÅ a˚nav-logon: e√ de; mhv, a˚prepe;Í fane∂tai dia; to; par’ aßllhla ta; ejnantÇa mavlista faÇnesqai.Metaphors, like epithets, must be fitting, which means that they must fairly correspondto the thing signified: failing this, their inappropriateness will be conspicuous: the wantof harmony between two things is emphasized by their being placed side by side. (Trans.Roberts 1984)If instead ejpÇqeta are not aÒrmovttonta the result is a˚prepevÍ; this leadsAristarchus to an athetesis.6866. Cf. also schol. Il. 3.432 (a˚qetouÅntai stÇcoi pevnte, o§ti pezovteroÇ e√si kaµ to∂Í nohvmasi yucroµkaµ a˚katavllhloi); schol. Il. 9.688–92a (a˚qetouÅntai stÇcoi pevnte . . . , o§ti . . . to∂Í nohvmasi kaµ t¬Å sunqevseipezovteroi); schol. Il. 11.767a1 (a˚qetouÅntai de; a˚po; touvtou stÇcoi ejnneakaÇdeka, . . . o§ti hJ suvnqesiÍ au˚tΩnpezhv).67. Cf. also Rh. 1408b1–2: to; d’ eu˚kaÇrwÍ h˙ mh; eu˚kaÇrwÍ crhÅsqai koino;n aÒpavntwn tΩn e√dΩn ejstin (“allthe variations of oratorial style are capable of being used in season or out of season,” trans. Roberts 1984).68. Cf. Matthaios 1999, 237–38.Francesca Schironi3047. Aristotle, Aristarchus, and theHellenistic Literary CriticismThe parallels between Aristotle’s theoretical statements in the Poetics andin the Rhetoric and Aristarchus’ practice in his eßkdosiÍ of Homer must notbe taken, however, as evidence of a “direct” dependence of Aristarchus onAristotle. Aristarchus is not a Peripatetic, or, better, is only a Peripatetic tothe extent that the other scholars and scientists working at the Museum orin Alexandria were Peripatetic. Aristotle seems to have shaped their rigor-ous and systematic approach to knowledge, also providing them with a setof methodological tools to proceed in their studies, such as the principle ofcause and effect, analogical reasoning, the combination of teleological andmechanistic views of natural phenomena, and the analysis of concrete datain order to offer a systematic view of the kosmos.69 This is what, mutatismutandis, we find in Aristarchus, who, among the “philologists” of his owntime, such as Crates of Mallos and the so-called kritikoÇ, is certainly themost concrete, systematic, and anti-speculative. This is an important point,because all of these similarities between Aristotle, Aristarchus and, in gen-eral, the Alexandrian philological school70 in the field of literary criticismare not particularly meaningful if it cannot be demonstrated at the same timethat this is a unique case and that the majority of the other contemporarygrammarians, philologists, and critics were adopting other views. It is thusnecessary to focus on Aristarchus’ colleagues.One of the most important sources for Hellenistic poetic theories is OnPoems by Philodemus of Gadara. In Books 1 and 2, Philodemus, using Crates’work surveying the literary views of the “critics” and of the “philosophers,”attacks the views of the so-called kritikoÇ.71 Then in Book 5 (cols. 29–39Mangoni), based on the previous work of his teacher Zeno of Sidon, he sur-veys thirteen different views of poetry.72 Unfortunately, these two accountsare not completely consistent with each other and, moreover, it is not alwayseasy to judge who held particular views. However, some general points anda certain degree of detail for some of these critics and literary theorists canbe reached. A brief review of them will show how Aristarchus’ approach topoetry is different from (and more Aristotelian than) all these other theories.7.1 The kritikoÇ (Heracleodorus, Pausimachus, and the Others)and CratesThe term kritikovÍ, as Janko has pointed out, is in itself very ambiguous be-cause by the time of Philodemus it meant generically “literary critic.”73 How-69. The same relationship with Aristotle can be envisaged for Erasistratus; cf. von Staden 1997.70. I am focusing on Aristarchus because among the Alexandrians he is the one about whom we knowmost and also he is considered the acme of Alexandrian philological activity. However, Zenodotus, Eratos-thenes, and Aristophanes of Byzantium share, at least in part, the same Aristotelian approach with Aristarchus.On Zenodotus, see Nickau 1977, 132–83; on Eratosthenes, see p. 309 below. The contrast thus is betweenthe Alexandrian school (with Aristarchus as the main and final representative of it) and the other schoolsflourishing in other parts of the Hellenistic world, like Pergamum and Athens.71. The best survey of these views is that by Janko (2000, 120–89).72. For an analysis of these columns, see Asmis 1992b.73. Cf. Janko 2000, 126.One Line LongAristotelian Principles in Aristarchean Philology 305ever, we tend to use this name to label a group of scholars that consideredsounds and composition the primary (if not sole) criterion for judging poetry,74as Philodemus explains in PHerc. 1676, col. 6, 2–9 (from On Poems, Book 2):kaµ to; “th;n me;n | [ejpifai]nomevnhn [e]u˚fwnÇ|an ≥dion [eπ]nai, ta; de; nohv{i}|mata kaµ [t]a;ÍlevxeiÍ ejkto;Í | eπnai kaµ koina; sunavges|qai de∂[n,” pa]ra; paÅsi me;n wÒÍ | ejn [sthv]lhi mev[n]eito∂Í kri|tiko∂[Í].that “the supervenient euphony is particular [to poetry], but the contents and the words areexternal [to it] and must be considered common [to all]” is fixed as if in stone amongall the critics. (Trans. Janko 2000, 124–25)75According to these critics, then, the content, the thought-element, and thewords, were something outside poetry, and the poets must only take careof the form and verbal composition. The extraneousness of thoughts anddiction from the tevcnh poihtikhv was thus the common ground and the sharedbasic principle of all the kritikoÇ. There were, however, some differences inthe details. Heracleodorus (late third century b.c.e.)76 maintained that onlyeuphony that supervenes upon word-order (suvnqesiÍ) mattered, whereasgenre, style, and meter had nothing to do with poetry.77 In fact, accordingto him, even content was superflous: if the line sounded good, it did notmatter if the verse was unintelligible.78 Similar but more extreme views wereheld by Pausimachus of Miletus (c. 200 b.c.e.),79 who considered sound theonly source of poetic pleasure (that is, the sounds of vowels in particular),and denied a role to content, genre, and even to the choice (ejkloghv) of words.The first three views quoted in Zeno’s list of Book 5 are along these lines:the first group calls for a “composition that delights the hearing or movesalong beautifully and expresses the thought powerfully”; the second for a“verbal composition that signifies the underlying thought vividly and sug-gestively,” and the third for a “composition that makes clear the underlyingthought clearly and concisely along with preserving a poetic style.”80 Allthese views focus on the verbal composition (suvnqesiÍ), which is presentedas the main task of a poet. The thought (diavnoia) is indeed there but its roleis secondary; it is necessary only in the sense that poetry, qua poetry, mustexpress something.A particular case is represented by Crates, quoted by Philodemus amongthe kritikoÇ and considered the rival of Aristarchus in Homeric criticism.The assessment of the real nature of this opposition is still under debate. If,on the one hand, it is undeniable that Crates was in many respects pursuinga different kind of scholarship, more interested in the philosophical aspectsof the text and in its allegorical meaning (especially in the light of Stoic cos-mology), on the other hand the methodology he uses (etymology, attentionto Homeric style, grammatical analysis, and even analogy) is largely shared74. On the kritikoÇ, see Porter 1995.75. Cf. also Asmis 1992a, 142.76. On Heracleodorus, see Janko 2000, 155–65.77. Janko 2000, 162, frag. 29.78. Janko 2000, 161, frag. 22.79. On Pausimachus, see Janko 2000, 165–89.80. Asmis (1992b, 397–99) thinks that the first opinion is Crates’ (on Crates, see below).Francesca Schironi306with the Alexandrians. Yet the similarity of methods used by these two schoolshighlights better their fundamentally different approaches to the topic, forthere are frequent cases where Crates and Aristarchus use the same pro-cedure (etymological analysis or even analogy) to reach opposite results.81This fundamental difference between Aristarchus (a grammatikovÍ, dealingmostly if not only with the text) and Crates (a kritikovÍ, with a pronounced“philosophical” approach to Homer) was already recognized by Crates him-self, who proudly maintained that he was able to “judge” a poem, unlike thegrammarians concerned only with the limited analysis of words, syntax, andquestions of authenticity (Sext. Emp. Math. 1.79):kaµ ga;r ejke∂noÍ [i.e., KravthÍ] eßlege diafevrein to;n kritiko;n touÅ grammatikouÅ, kaµ to;nme;n kritiko;n pavshÍ, fhsÇ, de∂ logikhÅÍ ejpisthvmhÍ eßmpeiron eπnai, to;n de; grammatiko;naÒplΩÍ glwssΩn ejxhghtiko;n kaµ prosådÇaÍ a˚podotiko;n kaµ tΩn touvtoiÍ paraplhsÇwne√dhvmona: paro; kaµ ejoikevnai ejke∂non me;n a˚rcitevktoni, to;n de; grammatiko;n uÒphrevt¬.[Crates] said that the “critic” was better than the grammarian and that while the criticwas experienced in all of logical science, the grammarian was simply an interpeter ofrare words (glossai), establisher of accents, and knower of things like these; hence thecritic was like an architect and the grammarian like his servant. (Trans. Blank 1998)82Further evidence of non-Aristotelian elements in Crates comes fromPhilodemus, On Poems, Book 5, cols. 24.25–29.18 Mangoni = frag. 101Broggiato, where Philodemus dwells on Crates’ view, and in particular fromthe following excerpts:to vd’ e≥per ejt’ eπpe, to ; m[h;] | piqano;n eπnai th ;n di|avnoian ejpaine∂n, a˚tev|cnou ge dh; touÅtoiouvtou | oßntoÍ. (Col. 27.3–7 Mangoni)For he also said that it is not plausible to praise the thought, since this sort of thing isnon-technical. (Trans. Asmis 1992a, 151)dia; to; | favskein di[a]gin∫skes|qai th;n uÒpavrcousan ejn | to∂Í poihv[m]as[i]m fusikh;n | dia-fora;n thÅªiº [a˚k]ohÅi. (Col. 27.17–21 Mangoni)Because he claims that the natural difference that exists in poems is discerned by thehearing. (Trans. Asmis 1992a, 153)kaµ tovde | m[hvte ta; a√]sqhvsei ejpi|t[erphÅ] mhvte th;n diav|n[oian d]e∂n krÇnein | tΩ[n] poih-mavtwn, a˚lla; | ta; logika; qewrhvmata | ta; fuvsei uÒp[avrco]nta di’ a√s|qhvsewÍ kr[Çnein],kaµ ou˚|k aßneu tΩn [noo]umevnwn, | ou˚ mevntoi ta; noouvme|na . . . (Col. 28.19–29 Mangoni)As to the claim that one must judge neither what is delightful to perception nor the thoughtof poems, but [one must] judge through perception the rational theorems that exist bynature, not without thoughts—not, however, the thoughts, . . . (Trans. Asmis 1992a, 155)From these fragments, a major point in Crates’ literary views emergesclearly.83 As for the kritikoÇ, for Crates too hearing alone can judge whethera poem is good or not. However, for Crates the judgment of euphony is not81. On Crates and Stoicism, cf. Broggiato 2001, xxxi–xxxiii, lii, lvii–lviii, lx–lxi, lxiv–lxv; and Asmis1992a, 139–40, 156–57, 161.82. Cf. Blank 1998, 140–41; see also Broggiato 2001, 249–50 (on frag. 94).83. See also Janko 2000, 122–23, esp. V 7, V 8, V 11.One Line LongAristotelian Principles in Aristarchean Philology 307subjective, but happens on the basis of objective criteria that are recognizedby the hearing. Also, whereas Heracleodorus and Pausimachus did not takethe content into consideration at all, Crates allows for it, in the sense thatcontent is embedded in poetry. However content (diavnoia, ta; noouvmena) isnot the object of judgment.84 For diavnoia is a nontechnical (aßtecnon) partof poetry.85In sum, the constant and sole focus on suvnqesiÍ and eu˚fwnÇa86 as the onlycriteria to judge poetry, on the one hand, and the firm point that content andargument are outside the art, on the other, mark all these views as not Aris-totelian87 and also not Aristarchean. As we have seen, Aristarchus very oftenopposes (or couples) form (suvnqesiÍ or kataskeuhv) and thought (diavnoia ornovhma), as if they were the constitutive and polar parts of poetry, especiallywhen giving reasons for an athetesis. If one decides on a line on the basisof the form or the thought-element (or both), obviously he considers thesetwo elements as fundamental (and also technical) parts of the poetry, as didAristotle. This is exactly what Crates and the kritikoÇ were arguing against.It is moreover interesting to note that the vocabulary adopted by the kritikoÇand Crates is identical to that we find in Aristonicus’ scholia, with terms likediavnoia, nohvmata, and suvnqesiÍ.88 This means that we are dealing with tech-nical terminology that indeed started first with the Peripatos and was fullydeveloped during Hellenistic times and still used in the first century b.c.e.However, within this common terminology, the principles are radically dif-ferent: for the critics and Crates, diavnoia is not part of poetry, and sound alone(either as verbal composition, suvnqesiÍ, or as pure euphony) is the sole cri-terion by which to judge poetry. For Aristotle and Aristarchus the thought(diavnoia)—but we could speak more generally of content, including thus alsothe muÅqoÍ and the hßqh—is at the core of poetry.7.2 Zeno’s List: Theories on Diction (suvnqesiÍ)There is, however, more. In the list of Zeno, a second group of critics main-tains that (Book 5, col. 30.6–10 Mangoni) the main point of a poem is asuvnqesiÍ levxewÍ ejnargΩÍ kaµ ejmfatikΩÍ th;n uÒpotetagmevnhn diavnoian sh-maÇnousa (“a composition expressing the underlying thought vividly [ejnargΩÍ]and suggestively [ejmfatikΩÍ]”).89 In rhetorical handbooks, ejnavrgeia and84. Cf. Asmis 1992a, 160: “Crates does not take the extremely radical position that what makes a poemgood is simply the sound, considered apart from any meaning. . . . His point, which is radical enough, isthat what makes a poem good is the sound in relation to the meaning, regardless of what the meaning is.”85. Cf. Porter 1995, 93–99.86. The theory of euphony stemmed from Pythagoras and the atomists and then was developed by theStoics. Aristotle and the Alexandrians, however, rejected it; see Janko 2000, 173–82, 189.87. Similar views are argued by Arrighetti (2001, 138–46).88. For a full list of this technical terminology, overlapping with the Aristotelian one, see Mangoni1993, 79–103.89. On this opinion, see Asmis 1992b, 401–3. ∆EmfatikΩÍ is translated here with “suggestively,” becauseeßmfasiÍ in ancient literary criticisms does not correspond to the modern “emphasis,” but is a technical termthat indicates the ability that a poet has to express something that is not clearly stated, as Asmis (1992b, 402)explains: “There is no good English equivalent; ‘suggestive’ comes close in meaning. Although the termcan mean ‘expressive’ and this sense can slide off into ‘forceful,’ it is misleading to translate ejmfatikovÍas ‘emphatic’ or ‘forceful,’ as it is often done.” I have thus preferred to leave the Greek term eßmfasiÍ in-stead of translating it into “suggestion,” which may be equally confusing.Francesca Schironi308eßmfasiÍ are considered among the best qualities of style;90 this is in contrastwith Aristotle, who maintains that the major virtue of style (both in proseand in poetry) is safhvneia.91 We have seen that Aristarchus most highly praisessafhvneia while also esteeming suntomÇa. 92 True, Aristarchus pays someattention to Homer’s ability to achieve außxhsiÍ and eßmfasiÍ,93 but this issomething that Aristotle as well admires in Homer (Rh. 1413b32–1414a7):oJ ga;r suvndesmoÍ e ¶n poie∂ ta; pollav, w§ste eja;n ej xaireq¬Å, dhÅlon o§ti tou˚nantÇon eßstai to;e ¶n pollav. eßcei ou®n außxhsin: . . . touÅto de; bouvletai poie∂n kaµ £OmhroÍ ejn tåÅ “Nireu;Í au®Suvmhqen,” “Nireu;Í Âglai?hÍ,” “Nireu;Í o¶Í kavllistoÍ.” perµ ou• ga;r polla; levgetai, a˚navgkhkaµ pollavkiÍ e√rhÅsqai: e√ ou®n [kaµ] pollavkiÍ, kaµ polla; doke∂, w§ste hußxhken, a§paxmnhsqeÇÍ, dia; to;n paralogismovn, kaµ mnhvmhn pepoÇhken, ou˚damouÅ u§steron au˚touÅ lovgonpoihsavmenoÍ.Just as the use of conjunctions makes many statements into a single one, so the omissionof conjunctions acts in the reverse way and makes a single one into many. It thus makeseverything more important. . . . This is the effect which Homer seeks when he writes“Nireus likewise from Syme, Nireus the son of Aglaia, Nireus, the comeliest man.” Ifmany things are said about a man, his name must be mentioned many times; and there-fore people think that, if his name is mentioned many times, many things have been saidabout him. So that Homer, by means of this illusion, has made a great deal of Nireus,though he has mentioned him only in this one passage, and he has preserved his memory,though he nowhere says a word about him afterwards. (Trans. Roberts 1984)It is interesting to note that the criterion of the ejnargevÍ seems to be absentfrom the scholia by Aristonicus.94 Moreover, the fact that those criticspraised only the “excessive” qualities leading to amplification and eßmfasiÍand did not pay any attention to clarity separates them from Aristotle andAristarchus.95 The third opinion, instead, demands only clarity and concise-ness (Book 5, col. 31.7–8b Mangoni: suvnqesiÍ safΩÍ kaµ suntovmwÍ ); this too is not in line with Aristarchus,who recognized that Homer had a particular poetic manner of expression,for example with glossai, metaphors, and other figures of speech.90. Cf. Dion. Hal. Lys. 7 (ejnavrgeia); Quint. Inst. 6.2.32 and 8.3.61 (ejnavrgeia), 8.3.83 and 9.2.3(eßmfasiÍ); Rhet. Her. 4.67 (significatio = eßmfasiÍ), 4.68 (demonstratio = ejnavrgeia).91. On Aristotle’s virtues of diction, see Ax (1993, esp. 27–31), who outlines the virtues of dictionamong Peripatetics and Stoics as follows: for Aristotle they are safevÍ, eJllhnÇzein, prevpon; for Theophrastus:ÔEllhnismovÍ, safhvneia, prevpon, kataskeuhv; and for the Stoics: ÔEllhnismovÍ, safhvneia, suntomÇa, prevpon,kataskeuhv (see Diog. Laert. 7.59).92. Cf. schol. Il. 1.110a, 3.352a, 8.108a, 8.528.93. Cf. schol. Il. 2.299b, 2.670 (quoted above, for the distinction between kurÇwÍ usage and the meta-phorical one), 2.809, 3.80b, 6.169a, 8.108a, 9.44a, 15.470a, 15.622, 16.161a, 17.172, 23.16a, 24.6–9a1,24.205b1; cf. also Didymus in schol. Il. 19.386a. In another instance (schol. Il. 9.14b) Aristarchus was ar-guing that lines 15–16 of Book 9 of the Iliad, where Agamemnon is weeping “like a fountain of dark waterthat pours down murky water from a steep cliff” and that Zenodotus rejected, were instead a˚nagka∂a dev . . .e√Í außxhsin, i.e., for the amplification of the tragic character of the moment.94. Out of forty-nine occurrences in the Iliad and Odyssey scholia of the words ejnavrgeia, ejnargevÍ, andejnargΩÍ, none goes back to Aristarchus. See also Zanker (1981, 307–8), who maintains that ejnavrgeia isnever used by Aristotle, but that he uses ejnarghvÍ in Poet. 1455a24 and so does Theophrastus. However,this is an Epicurean concept (ibid., 309–10).95. According to Asmis (1992b, 405–6), the recognition that eßmfasiÍ is the most important constituentof poetry leads to the adoption by Hellenistic scholars of allegorical reading, a way of reading poetry thatwas totally rejected by Aristarchus.Aristotelian Principles in Aristarchean Philology 3097.3 Zeno’s List: Theories on Thought (diavnoia), Imitation (mÇmhsiÍ),and MoreOf course, other Hellenistic theorists gave great importance to the thoughtof poetry. For example, the diavnoia is central to the fourth, fifth, and sixththeories listed by Philodemus (himself a great supporter of the importanceof the thought versus sound) in Book 5 of On Poems.96 Some (col. 31.33–34 Mangoni) demanded wise thought (sofh; diavnoia) and some (col. 32.3Mangoni) useful thought (w˚fevlimoÍ diavnoia). All these definitions are equallyextraneous to Aristarchus, who never defines Homeric poetry as “useful” orremarks on the sofÇa of the poet. This can be seen in conjunction with theidea that Homer does not aim at didaskalÇa but at yucagwgÇa, a statementascribed to Eratosthenes, who maintained that in Homer one should not lookfor any technical knowledge, geography in primis.97 This idea was shared byAristarchus as well, and many are the scholia where, commenting on Homericgeography, Aristarchus claims that we do not have to seek for accuracy, ora true depiction of the cosmos.98Then (col. 33.1–3 Mangoni) Philodemus mentions the opinion of thosewho call for a composition able to teach “something more” (suvnqesin levx-ewn prosdidavskousavn ti perittovteron dia; poihvmatoÍ). This statement isvery interesting when compared with Aristarchean evidence, and not onlybecause again it calls for a didactic aim in poetry. More importantly, perissovÍand its derivatives are some of the most typical expressions to be found in thescholia of Aristonicus, but they are always used in a negative sense, meaning“superfluous.” Everything that is perissovn is rejected in Homer. Often thiscriterion is the deciding ground for an athetesis.99 This is a very interestingpoint because the negative sense for perissovÍ (in itself a vox media, sig-nifying “extraordinary” but also “superfluous,” “useless”) is shared with96. See Asmis 1992b, 406–8.97. Strabo 1.1.10.31–35: ou˚de; ga;r a˚lhqevÍ ejstin, o§ fhsin ∆EratosqevnhÍ, o§ti poihth;Í paÅÍ stocavzetaiyucagwgÇaÍ, ou˚ didaskalÇaÍ: ta˚nantÇa ga;r o¥ fronim∫tatoi tΩn perµ poihtikhÅÍ ti fqegxamevnwn pr∫thntina; levgousi filosofÇan th;n poihtikhvn (“for Eratosthenes is wrong in his contention that the aim of everypoet is to entertain, not to instruct; indeed the wisest of the writers on poetry say, on the contrary, thatpoetry is a kind of elementary philosophy,” trans. Jones 1969). And also Strabo 1.2.3, in particular 1.2.3.37–42: oJ d’ ejphvnegken o§ti yucagwgÇaÍ movnon, didaskalÇaÍ d’ ouß. kaµ prosexergavzetaÇ ge, punqanovmenoÍ tÇsumbavlletai pro;Í a˚reth;n poihtouÅ pollΩn uÒpavrxai tovpwn eßmpeiron h˙ strathgÇaÍ h˙ gewrgÇaÍ h˙ rJhtorikhÅÍ h˙o∏a dh; peripoie∂n au˚tåÅ tineÍ ejboulhvqhsan… (“but his [of Eratosthenes] words were ‘mere entertainment andnot instruction.’ And Eratosthenes gives himself quite unnecessary pains when he asks how it contributesto the excellence of the poet for him to be an expert in geography, or in generalship, or in agriculture, or inrhetoric, or in any kind of special knowledge with which some people have wished to ‘invest’ him,” trans.Jones 1969).98. Cf., for example, Aristarchus’ criticism of those (like Crates; cf. Strabo 3.4.4.1–15 (= frag. 75Broggiato) who interpreted the voyages of Odysseus as a true geographical decription of the o√koumevnh(schol. PQ Od. 5.55: pro;Í ta; perµ thÅÍ plavnhÍ, o§ti povrjrJw pou ejn ejktetopismevnoiÍ tovpoiÍ a˚orÇstoiÍ . . . ,schol. BQT Od. 10.189: ejk touvtwn de; ejktetopismevnh faÇnetai hJ plavnh touÅ ∆OdussevwÍ). Cf. Lehrs 1882, 241–46; Buonajuto 1996. The link between yucagwgÇa and poetry is present in Aristotle too (Poet. 1450a33–35): pro;Í de; touvtoiÍ ta; mevgista o∏Í yucagwge∂ hJ tragådÇa touÅ muvqou mevrh ejstÇn, aª te peripevteiai kaµa˚nagnwrÇseiÍ (“In addition, the most important things with which a tragedy enthralls [us] are parts of plot—reversals and recognitions,” trans. Janko 1987). Here however the point is not so much about the goals ofpoetry but rather about the means by which the yucagwgÇa is achieved. Hence, the parallel between Aristotleand the two Alexandrians is not so close.99. On this topic, see Lührs 1992.Francesca Schironi310Aristotle. The philosopher, especially in his works on biology, sees natureas purposive (i.e., with a tevloÍ) and not doing anything in vain, perittovn.100The same is, according to Aristarchus, the fuvsiÍ of Homer. Instead, thesetheorists quoted by Philodemus demand a poet who teaches us somethingperittovteron, where the adjective is obviously used in a positive meaning,incompatible with the meaning that this key concept has for Aristarchusthroughout our sources.The seventh view in Zeno’s list calls for the mimesis of other poets.101This is one of the bases of Roman literature, and we have no earlier evi-dence for it apart from this hint in Philodemus. To present Homer as a modelthat must be imitated can indeed be seen as a development of Aristotle’sview of Homer as a master of the craft and of Aristarchus’ distinctionbetween Homer and the later poets (o¥ ne∫teroi). However, the exception-ality of Homer compared to all other poets is so deep according toAristarchus that a mimesis is virtually impossible: the ne∫teroi can only tryto imitate “the poet” but their results are so openly inferior that Aristarchuscannot but notice their bad outcome. In the Aristarchean scholia thene∫teroi’s attempts to imitate Homer are often commented on with remarkslike ejplanhvqh /ejplanhvqhsan, “he was/they were misled,” that is, he/theymissed the point, he/they got it wrong.102The other opinions (8–13) are extremely generic, and therefore theycannot be referred back to any particular school.103 The opinions, whichjudge poetry according to whether it has a levxiÍ prevpousa to∂Í ejxagomevnoiÍpros∫poiÍ (eighth opinion),104 or on the basis of its effects on the audience(ninth and tenth opinions), of its beauty (eleventh opinion), of its goodness(twelfth opinion), or of prevpon (thirteenth opinion) are not comprehensivetheories of the tevcnh poihtikhv, as are those of Aristotle and of Aristarchus.For those about whom we know more, however, the kritikoÇ and Crates, thepoints of discrepancy are so many that the affinity between Aristarchus andAristotle becomes even more evident.7.4 The Peripatetics: Neoptolemus of Parium, Heraclides of Pontus, Andromenides, and Megaclides of AthensAmong Hellenistic scholars dealing with poetry, Neoptolemus of Pariumshows an interesting overlap with the views we have found in Aristotleand Aristarchus.105 Neoptolemus was a Peripatetic living in the third cen-tury b.c.e., whose work is entirely lost and can be recovered mainly (again)through Philodemus On Poems 5 (cols. 13.32–16.28 Mangoni).106 As iswell known, Neoptolemus divided the art of poetry into three parts: the poet100. See von Staden 1997.101. See Asmis 1992b, 408–10.102. Cf. schol. Il. 2.659, 4.439–40, 14.500, 15.119. On the neoteroi, cf. Severyns 1928. For an example,see Aristarchus’ criticism of Antimachus in Schironi 1999.103. See Asmis 1992b, 410–14.104. This is from Andromenides (see p. 312 below); cf. Janko 2000, 147, F 8.105. Cf. Asmis 1992c.106. Cf. Brink (1963, 145–49), who calls Neoptolemus’ approach a “revised Aristotelianism”; andJanko 2000, 152. The fragments of Neoptolemus are collected by Mette (1980).Aristotelian Principles in Aristarchean Philology 311(poihthvÍ), the poem (poÇhma), and poesy (poÇhsiÍ). The second and third partscorrespond to form (suvnqesiÍ or poÇhma) and content (uÒpovqesiÍ or poÇhsiÍ).This opposition suvnqesiÍ/uÒpovqesiÍ as outlined by Neoptolemus (he himselfseems to have used these terms together with poÇhma and poÇhsiÍ) is statedin Book 5 (col. 14.26–28 Mangoni):q[a]uma[sto;]n d’au˚|touÅ kaµ [to;] thÅ[Í] pohvsew[Í] | eπnai t[h;]n uÒpovqesin [m]o v|non [. . .](col. 15.1–3 Mangoni) kaµ to; pohv[matoÍ mov]|non th;n [suvnqesin thÅÍ] | levxewÍ m[etevcein . . .It is astonishing of him to claim that only theme belongs to poiesis. . . . Also, [it is astonish-ing] that only verbal composition [participates] in a “poem.” . . . (Trans. Asmis 1992c, 210)The technical terminology is the usual one; however, a link betweenNeoptolemus and the Alexandrians can be seen in the opposition betweenform and content as expressed with suvnqesiÍ /uÒpovqesiÍ. We have alreadymentioned the vast usage of the term suvnqesiÍ to indicate “composition,”“style,” among the Alexandrians. This terminology is actually missing inAristotle who, as was observed, used instead levxiÍ to indicate “form” and“style.” It seems as if the usage of suvnqesiÍ for “form” (in opposition, evenfrom a morphological point of view, to uÒpovqesiÍ for the “content”) is a laterdevelopment, which might have been due to Neoptolemus himself.107 ThisPeripatetic scholar, like Aristotle and Demetrius of Phalerum before him,was known at Alexandria108 and exercised some sort of influence amongthe grammatikoÇ there. While in the scholia derived from Aristarchus thereseems to be no trace of the opposition suvnqesiÍ / uÒpovqesiÍ as formulated byNeoptolemus, the use of the word uÒpovqesiÍ to indicate the “content” of a lit-erary work comes from Aristophanes of Byzantium, who used the term torefer to the summaries of the “content” of the dramas. However, apart fromthis idea of poetry as based on form and content, the tripartite divisionof poetry as devised by Neoptolemus is absent in Aristarchus as well as inAristotle.109 Also different is the idea that the poets, according to Neoptole-mus, should both delight and benefit, a criterion that seems closer to Platothan to Aristotle.110 The importance of didaskalÇa as the goal of poetry hadalready been advocated by Heraclides of Pontus (fourth century b.c.e.), apupil of Plato and Aristotle,111 who, besides writing “grammatical” treatiseson Homer and other poets in the best Aristotelian tradition, maintained,more in line with the euphonistic school, that ejmmevleia, “musicality” andligurovthÍ, “sonority,” were primary virtues for poets.Andromenides (third century b.c.e.?) is perhaps the closest to Aristotleand Aristarchus. His case, like that of Heraclides, is interesting evidence ofhow the Hellenistic critics were influenced by many different doctrines.107. See Schironi, in Bottai and Schironi 1997, 1058–62.108. He was quoted by Aristophanes of Byzantium in his glossographical works. This, however, doesnot mean that Neoptolemus was indeed working at Alexandria, a hypothesis supported by Mette (1935,esp. 2467). For a more cautious view on the relationship between Neoptolemus and the Alexandrian schol-ars, see Brink 1963, 135–50.109. See Asmis 1992c, where she tries to find other traces of this theory in rhetorical treatises.110. See Asmis 1992c, 218.111. On Heraclides, see Janko 2000, 134–38.Francesca Schironi312Philodemus in Book 1 quotes him among the kritikoÇ;112 however, he hasmany views in common with Neoptolemus of Parium. Like Neoptolemusand the Peripatetic tradition, he collected glosses. Moreover, he seems tofollow the same division of poetics into poet (poihthvÍ), poem (poÇhma), andpoesy (poÇhsiÍ). He is more Aristotelian than Neoptolemus, however, whenhe maintains that poetry aims to please (and not to impart truth like prose)and when he holds the view that each genre has its own subject matter andappropriate style (to; prevpon) and that each character has a distinctive dic-tion. On the other hand, Andromenides, like Crates, maintained that the earwas the ultimate judge of poetry (and this is why he is considered a kritikovÍby Philodemus), and also that diction and word choice (ejkloghv) were par-amount. This emphasis on phonetic beauty, not so prominent in Aristotle,is at the core of Theophrastus’ theory of style113 and can be seen as a laterdevelopment of the Peripatetic school.The case of Megaclides (early third century b.c.e.) is similar; he, whilequoted by Philodemus in Book 1 among the kritikoÇ, is defined as a Peri-patetic in other sources.114 His work on Homer (he wrote a perµ ÔOmhvrou)and on his language (he believed that Homeric dialect was Attic, like Aris-tarchus) is purely in line with Alexandrian interests. Moreover, his distinction,especially in terms of mythological tradition, between Homer and Hesiod,on the one hand, and post-Homeric and post-Hesiodic poets, on the other,makes him an Aristotelian with ideas very close to those of Aristarchus.However, he was listed by Crates among the advocates of the ear (and notthe intellect) as the best judge for poetry. And even if his precise poeticaltheories are still under debate, due to the fragmentary state of the evidencefrom On Poems 1, he seems to be close to Heraclides of Pontus and An-dromenides, as Janko has concluded.115 The euphonistic approach to litera-ture shared by all these Peripatetic scholars makes them pupils more ofTheophrastus, with his theory of style and of word choice (ejkloghv), than ofAristotle, with his content-based approach to texts.ConclusionAs I hope to have shown, Aristarchus seems to have been aware of Aristo-telian reflections on poetry. In his work on Homer, he uses Aristotelian cate-gories and critical concepts. This is particularly evident when Aristarchushas to deal with atheteses or argues against Zenodotus’ readings. Interestingsimilarities between Aristotle and Aristarchus are to be found in the handlingof the plot (it can contain “impossible elements,” but they must be “accordingto necessity or probability”), of the characters (they are of necessity “serious”and their behavior must be according to what is considered “proper”), of thethought-element (epos is a serious genre, hence all the comic elements must112. On Andromenides, see Janko 2000, 143–54.113. Cf. Dion. Hal. Isoc. 3.1; and Ardizzoni 1953, 70–72.114. Frag. 2 Janko. On Megaclides, see Janko 2000, 138–43.115. Cf. Janko 2000, 143.One Line ShortAristotelian Principles in Aristarchean Philology 313be avoided) and of style (which must be clear, but also poetic, i.e. rich inglosses and metaphors).The affinity between Aristotle and Aristarchus is further proved by a com-parison with other Hellenistic views on poetry. Here, while the technicalvocabulary is almost identical, the views held by Crates or the other Helle-nistic kritikoÇ are opposite or, at best, far in spirit from what was expressedby Aristotle. In particular, their continuous focus on the sound-element andeuphony against content are in striking opposition to Aristotle’s theory andAristarchus’ practice.116There is, however, a fundamental distinction between Aristotle and Aris-tarchus: whereas the philosopher theorizes these principles, the philologistapplies them. In this, Aristarchus is different from colleagues like Cratesand the kritikoÇ as well. The latter had an “active” philosophical background;they claimed to be critics because they aimed to give prescriptive views onhow one should write a poem. Aristarchus does not claim anything like that,but tries to make a better text of Homer by editing and commenting on it.In this sense, Aristarchus (and his Alexandrian predecessors) are differentfrom all the other scholars who were taken into account, who like Crates (and,in this view, Aristotle too) wanted to develop theories out of the study ofHomer. For Aristarchus, Aristotelian philosophy, like grammatical categories,is, instead, just a tool to use in his job: working on texts, preparing editionsand writing commentaries. The contrast could not be greater: on the onehand, literary critics like Crates and the other kritikoÇ, having a particularagenda, or, as in the case of Crates, influenced by Hellenistic philosophy,versus grammarians like Aristarchus, on the other. The latter had a “scientificapproach” to the text, looking at the bare data on the basis of a very clearaccount: that of Aristotle, the founder of scientific inquiry.Harvard University116. I am not arguing that Aristotelian concepts are present only in Aristarchean scholia. As Richardson(1980) has demonstrated, they are abundantly present in the exegetical scholia. For example, in schol.QV Od. 23.310–43 (ou˚ kalΩÍ hjqevthsen ÂrÇstarcoÍ tou;Í tre∂Í kaµ triavkonta: rJhtorikh;n ga;r pepoÇhkena˚nakefalaÇwsin kaµ ejpitomh;n thÅÍ ∆OdusseÇaÍ), the scholiast argues against Aristarchus for his athetesis ofOd. 23.310–43 with arguments that recall Arist., Rh. 1417a12–15 (eßti pepragmevna de∂ levgein o§sa mh; prat-tovmena h˙ oπkton h˙ deÇnwsin fevrei: paravdeigma oJ ÂlkÇnou a˚povlogoÍ, o¶Í pro;Í th;n Phnelovphn ejn eJxhvkontaeßpesin pepoÇhtai), where the philosopher praises the story Odysseus tells Penelope as a good way of nar-rating a story cutting off all the most terrifying details. However, my claim here is that, even if we findmore Aristotelian concepts in the exegetical scholia (which are a product of later scholarship), we do findAristotelian concepts even in Aristarchus, and in opposition to the other Hellenistic theorists.Francesca Schironi314AppendixAristotle and Aristarchus on the Four Parts of Epic PoetryAristotle AristarchusMuÅqoÍ plot must be kata; to; e√ko;Í h˙ to; a˚nagka∂ona˚duvnata are allowed in poetry for the sake of to; qaumastovna˚duvnata e√kovta preferable to dunata; a˚pÇqanakata; to; e√kovÍ = piqanovnagainst internal inconsistencies (uÒpenantÇa)poetic licence for to; qaumastovn § but piqanovthÍ is necessarykata; to; e√kovÍ = piqanovnlines piqanoÇ § to be keptlines a˚pÇqanoi § athetesiswhat is a˚sumf∫nwÍ or mavcetai with the rest is rejected¢Hqh characters must be kata; to; e√ko;Í h˙ to; a˚nagka∂on § they must behave and speak according to the criterion of appropriateness (to; prevpon)character differences include age, sex, nationality, etc.lines a˚prepe∂Í, mh; aÒrmovzonteÍ, ou˚ prevponteÍ, a˚navrmostoi, para; to; provswpon, a˚noÇkeioi with reference to the characters § athetesisbehavior codes for heroes, women, old and young people, Greeks, and barbarians, et al.Diavvnoia thought-element § in tragedy and epic it is serious and not ridiculous (fauÅloÍ, gelo∂oÍ, eu˚telhvÍ)lines gelo∂oi, eu˚tele∂Í t¬Å dianoÇç § athetesisLevxiÍ poetic diction must be clear and not commonplacesafhvneia is achieved through the use of kuvria ojnovmatato; para; to; kuvrion to avoid commonplaceRhetoric: language must not be “poetical” in proseRhetoric: against ejpÇqeta aßkairasafhvneia and kurÇwÍ usages in Homerou˚ kurÇwÍ usages in Homerlanguage must not be prosaic in poetryejpÇqeta aßkaira § athetesisAristotelian Principles in Aristarchean Philology 315LITERATURE CITEDArdizzoni, A. 1953. PoÇhma: Ricerche sulla teoria del linguaggio poetico nell’antichità. Bari.Arrighetti, G. 2001. Filodemo, Lucrezio e le poetiche dell’Ellenismo. WürzJbb, n.s., 25: 133–55.Asmis, E. 1992a. Crates on Poetic Criticism. Phoenix 46: 138–69.. 1992b. An Epicurean Survey of Poetic Theories. CQ 42: 395–415.. 1992c. Neoptolemus and the Classification of Poetry. CP 87: 206–31.Ax, W. 1993. Der Einfluß des Peripatos auf die Sprachtheorie der Stoa. In Dialektiker undStoiker, Zur Logik der Stoa und ihrer Vorläufer, ed. K. Döring and T. Ebert, 11–32. Stuttgart.Barnes, J. 1999. Roman Aristotle. In Philosophia Togata: Plato and Aristotle at Rome, ed.J. Barnes and M. Griffin, vol. 2, 1–69. Oxford.Blank, D. 1998. Sextus Empiricus, “Against the Grammarians (Adversus mathematicos I).”Translated with introduction and commentary. Oxford. Bottai, F., and F. Schironi. 1997. Sull’uso di suvnqesiÍ nella critica letteraria antica. SCO 46:1049–77.Brink, O. 1963. Horace on Poetry: Prolegomena to the Literary Epistles. Cambridge. Broggiato, M. 2001. Cratete di Mallo. I frammenti. Edizione, introduzione e note. La Spezia.Buonajuto, A. 1996. L’EXWKEANISMOS dei viaggi di Odisseo in Cratete e negli Alessandrini.AeR 41: 1–8.Canfora, L. 1988. La biblioteca scomparsa. Palermo.. 2002. Aristotele “fondatore” della Biblioteca di Alessandria. In Scritti in onore diItalo Gallo, ed. L. Torraca. Pubblicazioni dell'Università degli Studi di Salerno. Sezione atticonvegni miscellanee 59, 167–75. Naples.Carnuth, O. 1869. Aristonici Perµ shmeÇwn ∆OdusseÇaÍ reliquiae emendatiores. Leipzig.Dachs, H. 1913. Die luvsiÍ ejk touÅ pros∫pou: Ein exegetischer und kritischer Grundsatz Aris-tarchs und seine Neuanwendung auf “Ilias” und “Odyssee.” Inaugural Dissertation. Erlangen.Erbse, H. 1972. Beiträge zum Verständnis der “Odyssee.” Berlin.Gallavotti, C. 1969. Tracce della Poetica di Aristotele negli scoli omerici. Maia 21: 203–14.Janko, R. 1982. A Fragment of Aristotle’s Poetics from Porphyry, concerning Synonymy. CQ,n.s., 32: 323–26.. 1987. Aristotle “Poetics” I, with the “Tractatus Coislinianus,” Recontruction of“Poetics” II, and the Fragments of the “On Poets.” Indianapolis.. 1991. Philodemus’ On Poems and Aristotle’s On Poets. CronErcol 21: 5–64.. 2000. Philodemus, “On Poems,” Book One. Edited with introduction, translation,and commentary, Oxford.Jenkins, T. 1999. Homêros ekainopoiêse: Theseus, Aithra, and Variation in Homeric Myth-Making. In Nine Essays on Homer, ed. M. Carlisle and O. Levaniouk, 207–26. Lanham, Md.Jones, H. 1969. The Geography of Strabo. Vol. 1. London.Lehrs, K. 1882. De Aristarchi Studiis Homericis 3. Leipzig.Lührs, D. 1992. Untersuchungen zu den Athetesen Aristarchs in der “Ilias” und zu ihrer Be-handlung im Corpus der exegetischen Scholien. Hildesheim.Mangoni, C. 1993. Il quinto libro della “Poetica” (P. Herc. 1425 e 1538). Naples.Matthaios, S. 1996. Kuvrion oßnoma. Zur Geschichte eines grammatischen Terminus. In AncientGrammar: Content and Context, ed. P. Swiggers and A. Wouters, 55–77. Leuven.. 1999. Untersuchungen zur Grammatik Aristarchs: Texte und Interpretation zurWortartenlehre. Hypomnemata 126. Göttingen.. 2002. Neue Perspektiven für die Historiographie der antiken Grammatik: DasWortartensystem der Alexandriner. In Grammatical Theory and Philosophy of Language inAntiquity, ed. P. Swigger and A. Wouters, 161–220, Leuven.Mette, J. 1935. Neoptolemus (11). RE 32: 2465–70.. 1980. Neoptolemos von Parion. RhM 123: 1–24.Francesca Schironi316Montanari, F. 1993. L'erudizione, la filologia e la grammatica. In Lo spazio letterario dellaGrecia antica, ed. G. Cambiano, L. Canfora, and D. Lanza, 1.2: 235–81. Rome., ed. 1994. La philologie grecque à l’époque hellénistique et romaine. Entretiens surl’Antiquité classique 40. Geneva.. 1995. Termini e concetti della Poetica di Aristotele in uno scolio a Odissea IV 69.In Studi di filologia omerica antica, vol. 2, 21–25. Pisa.. 2001. Gli studi omerici di Demetrio Falereo. SemRom 4: 143–57.Moraux, P. 1973. Der Aristotelismus bei den Griechen: Von Andronikos bis Alexander vonAphrodisia. Vol. 1. Berlin.Nagy, G. 1998. The Library of Pergamon as a Classical Model. In Pergamon: Citadel of theGods, ed. H. Koester. Harvard Theological Studies 46, 185–232. Harrisburg, Pa.Nickau, K. 1977. Untersuchungen zur textkritischen Methode des Zenodotos von Ephesos.Berlin.Pfeiffer, R. 1968. History of Classical Scholarship. Vol. 1, From the Beginnings to the End ofthe Hellenistic Age. Oxford.Podlecki, A. 1969. The Peripatetics as Literary Critics. Phoenix 23: 114–37.Porter, J. 1992. Hermeneutic Lines and Circles: Aristarchus and Crates on the Exegesis ofHomer. In Homer's Ancient Readers: The Hermeneutics of Greek Epic’s Earliest Exegetes,ed. R. Lamberton and J. J. Keaney, 67–114. Princeton, N.J.. 1995. O¥ kritikoÇ: A Reassessment. In Greek Literary Theory after Aristotle, ed.J. G. J. Abbenes, S. R. Slings, and I. Sluiter, 83–109. Amsterdam.Richardson, N. 1980. Literary Criticism in the Exegetical Scholia to the Iliad: A Sketch. CQ,n.s., 30: 265–87.. 1992. Aristotle's Reading of Homer and Its Background. In Homer’s Ancient Readers:The Hermeneutics of Greek Epic’s Earliest Exegetes, ed. R. Lamberton and J. J. Keaney,30–40. Princeton, N.J.. 1993. The “Iliad”: A Commentary. Vol. VI, Books 21–24. Cambridge.. 1994. Aristotle and Hellenistic Scholarship. In Montanari 1994, 7–38.Roberts, W. Rhys. 1984. Rhetoric. In The Complete Works of Aristotle. The Revised OxfordTranslation, ed. J. Barnes, vol. 2, 2152–2269. Princeton, N.J.Rossi, L. E. 1976. Umanesimo e filologia (a proposito della Storia della filologia classica diRudolf Pfeiffer). RivFil 104: 98–117.Schenkeveld, D. 1970. Aristarchus and £OmhroÍ filovtecnoÍ: Some Fundamental Ideas of Aris-tarchus on Homer as a Poet. Mnemosyne 23: 162–78.Schironi, F. 1999. Aristarco studioso di Antimaco. RivFil 127: 282–90.Schmidt, M. 1976. Die Erklärungen zum Weltbild Homers und zur Kultur der Heroenzeit inden bT-Scholien zur “Ilias.” Munich.Severyns, A. 1928. Le cycle épique dans l’école d’Aristarque. Liège.Staden, H. von. 1997. Teleology and Mechanism: Aristotelian Biology and Early HellenisticMedicine. In Aristotelische Biologie: Intentionen, Methoden, Ergebnisse. Akten des Sym-posions über Aristoteles’ Biologie vom 24.–28. Juli 1995 in der Werner-Reimers-Stiftung inBad Homburg, ed. W. Kullmann and S. Föllinger, 183–208. Stuttgart.Zanker, G. 1981. Enargeia in the Ancient Criticism of Poetry. RhM 124: 297–311.

‘Aristarchus, Greek Dialects and Homer’

Abstract

Extracted data

‘Aristarchus, Greek Dialects and Homer’

Abstract

Extracted data

Related items

Related items