Harmonia and ethos in ancient greek music

MATE = UNIVERSETY OF SSUES © BGPP AR PAPEL ES AE dh Harmonia and Ethos in Ancient Greek Music Author(s): Thomas J. Mathiesen Source: The Journal of Musicology, Vol. 3, No. 3 (Summer, 1984), pp. 264-279 Published by: University of California Press Stable URL: http://www. jstor.org/stable/7638 16 Accessed: 27/01/2010 15:08 Your use of the JSTOR archive indicates your acceptance of JSTOR's Terms and Conditions of Use, available at http://wwwjstor.org/page/info/about/policies/terms.jsp. JSTOR's Terms and Conditions of Use provides, in part, that unless you have obtained prior permission, you may not download an entire issue of a journal or multiple copies of articles, and you may use content in the JSTOR archive only for your personal, non-commercial use. Please contact the publisher regarding any further use of this work. Publisher contact information may be obtained at http://www jstor.org/action/showPublisher?publisherCode=ucal. Each copy of any part of a JSTOR transmission must contain the same copyright notice that appears on the screen or printed page of such transmission. JSTOR is a not-for-profit service that helps scholars, researchers, and students discover, use, and build upon a wide range of content in a trusted digital archive. We use information technology and tools to increase productivity and facilitate new forms of scholarship. For more information about JSTOR, please contact support@jstor.org. Universityof California Pressis collaborating with JSTOR to digitize, preserve and extend access to The http://www jstor.org

Author(s): Thomas J. Mathiesen Source: The Journal of Musicology, Vol. 3, No. 3 (Summer, 1984), pp. 264-279 Published by: University of California Press Stable URL: http://www.jstor.org/stable/763816 Accessed: 27/01/2010 15:08 Your use of the JSTOR archive indicates your acceptance of JSTOR's Terms and Conditions of Use, available at http://www.jstor.org/page/info/about/policies/terms.jsp. JSTOR's Terms and Conditions of Use provides, in part, that unless you have obtained prior permission, you may not download an entire issue of a journal or multiple copies of articles, and you may use content in the JSTOR archive only for your personal, non-commercial use. Please contact the publisher regarding any further use of this work. Publisher contact information may be obtained at http://www.jstor.org/action/showPublisher?publisherCode=ucal. Each copy of any part of a JSTOR transmission must contain the same copyright notice that appears on the screen or printed page of such transmission. JSTOR is a not-for-profit service that helps scholars, researchers, and students discover, use, and build upon a wide range of content in a trusted digital archive. We use information technology and tools to increase productivity and facilitate new forms of scholarship. For more information about JSTOR, please contact support@jstor.org. University of California Press is collaborating with JSTOR to digitize, preserve and extend access to The http://www.jstor.org

THOMAS J. MATHIESEN study of ancientGreekmusic has held the interestof modem scholarship since the Renaissance, when the widely attestedpowers of ancient music began to attractthe attentionof a numberof the humanists. From the very first, two subjects within the field of ancient Greek music were seen as especially troublesome and were given particularscrutiny. These are the subjects of harmoniaand ethos. The difficulty with these subjects lies in part in the fact that ancient authoritiesdisagree among themselves, and it has thereforeseemed that the subjectsmust remain forever in the realm of philosophical dispute, eluding all attempts at systematization. This is not to say that scholars ignored the subjects. In fact, there have been a number of notable works dealing with these matters, including Girolamo Mei's unpublished De modis musicis antiquorum (autographin Vaticanuslatinus 5323)-dating from the sixteenthcentury, the beginning of modem scholarshipon ancient Greek music-and, more recently, studies by HermannWiegandt, A. F. Walter, HermannAbert, E. M. von Hornbostel,L. P. Wilkinson, JamesRiley, EdwardLippman,Warren Anderson, J. Garcia L6pez, and Lukas Richter.1All these works are incisive and combine technical remarksabout modes, scales, pitches, and so on, with general philosophical treatmentsof music and character,but none of these-nor any of the other studies that might be cited-attempts to systematizethe ethical characteristicsof ancientGreek music or to show a close relationshipbetween harmoniaand ethos. Nonetheless, it is possible to develop and demonstratea reasonablyprecise system for the analysis of the ethos in pieces of ancient Greek music. This system leans heavily on the De musica of AristidesQuintilianus,which unquestionablyprovidesthe most detailed and comprehensiveancient treatmentof the subjects of har'HermannWiegandt, "De ethico antiquorumrhythmorumcharactereauctoreAristideQuintiliano" (Ph.D. diss., Halle, 1881); Anton FriedrichWalter, "Die ethisch-padagogischeWiirdigungder Musik durch Plato und Aristoteles," Vierteljahrsschrift fiir MusikwissenschaftVI (1890), 388-415; Hermann Abert, Die Lehre vom Ethos in der griechischen Musik (Leipzig, 1899: reprinted., Tutzing, 1968); Erich M. von Hornbostel, "Tonart und Ethos," Festschriftfir Johannes Wolf zu seinem sechzigsten Geburtstag, ed. W. Lott, H. Osthoff, and W. Wolffheim (Berlin, 1929), pp. 73-78; L. P. Wilkinson, "Philodemus on Ethos in Music," Classical QuarterlyXXXII (1938), 174-81; James Riley, "Ethos in Greek Music" (M.M. thesis, College-Conservatoryof Music of Cincinnati, 1948); EdwardLippman, Musical Thoughtin AncientGreece (New York, 1964; reprinted., New York, 1975);WarrenAnderson, Ethos and Educationin GreekMusic (Cambridge,Mass., 1966);J. GarciaL6pez, "Sobre el vocabulario etico-musicaldel Griego," Emerita, boletin de linguisticay filologia cldsica XXXVII (1969), 335-52; and Lukas Richter, "Antike asthetischeTheorienzur gesellschaftlichenFunktionder griechischenTragodie," Die griechische Tragodie in ihrer gesellschaftlichenFunktion, ed. HeinrichKuch, Ver6ffentlichungen des Zentralinstitutsfur Alte Geschichte und Archaologieder Akademie der Wissenschaften der DDR, XI (Berlin, 1983), 173-92.

monia and ethos, and which pulls together all the various disciplinesphilosophy, harmonics,metrics, psychology, and cosmology-that bear on these subjects. In order to explain the system, it will be necessary first to treat the concept of harmonia.The notion of mimesis, especially as it is formulated by Aristides Quintilianusand the Platonists, must then be reviewed. With this as background,the analyticalsystem may be extractedfrom the treatise of Aristides Quintilianusand applied to three early fragmentsof ancient Greek music. Some preliminaryconclusions will then emerge. Harmonia A full and systematic discussion of the term harmonia-even limited to its use in ancient Greek music theory-would extend far beyond the limits of this paper. Because my discussions have already appearedelsestatements where,2it will be most expeditiousto make some straightforward about harmoniathat will be necessary for our understandingof the larger subjects in view. Harmoniain ancient Greek music theory must be understoodon two levels, depending on the approach(or, perhaps, on the sophistication)of the theorist. Theorists primarilyconcernedwith music as an analogue for higher philosophical truths conceive the harmoniaas manifesting certain basic proportions,orders, and charactersmimeticof higheruniversals,even when applying the term to certain scales or genera, which are seen as analogues. This more comprehensiveusage appearsin the treatisesof Aristoxenus (e.g., Harmonica 1.2, 4, 16, and 26), [Ps.-] Plutarch(e.g., De musica 34 [1143E]), Ptolemy to some degree (e.g., Harmonica 3.3-7, 1011), and especially AristidesQuintilianus(De musica 1.1, 3, 8, 11-12, 19; 2.4, 6, 7, 12, 14, 16-17; and 3.6, 9, 12, 22, and 27). It must be stressed that these usages occur in proximityto usages of terms like enharmonios, tonos, eidos, tropos, and so on; thus, harmoniamay not be considered a simple synonym for these other technical terms, though it does subsume these technical concepts. On the other hand, theoristsrepresentingthe second-centuryrevival of ancient theory (such as Cleonides and Gaudentius) who seem to have viewed music essentially as a system ratherthan as a "consort of philosophy"-as Aristides Quintilianusrefers to it-tend to concentrateon the limited technical sense of harmonia, while remaining aware (to a greateror lesser degree) of the largerperspective.3 2ThomasJ. Mathiesen, "Problemsof Terminologyin Ancient GreekTheory: 'APMONiA," Festival Essays for Pauline Alderman,ed. BurtonKarson(Provo, 1976), 3-17; and idem, Aristides Quintilianus on Music in Three Books, Music Theory TranslationSeries (New Haven, 1983), pp. 42-57. See also useful articles by KathleenSchlesinger, "The Harmonia,"Music Review V (1944), 7-39 and Classical QuarterlyXXXVI (1942), 119-41; Isobel Henderson,"The Growthof the Greek&pJov(oaci." 94-103; and Olof Gignon, "Zum antikenBegriff der Harmonie," Studiumgenerale XIX (1966), 53947; and the still indispensablemonographof BonaventuraMeyer, APMONIA.Bedeutungsgeschichte des Wortesvon Homer bis Aristoteles (Ziirich, 1932). 3Mathiesen,"Problems," pp. 3-4, 17.

It is thereforenecessaryto keep in mind when studyingancientGreek music theory that a harmoniais not a harmonia.The same may be said, by the way, for some of the other technical terms that appearthroughoutthe Greek theoreticaltradition. Greek musical theorists are not separatefrom their philosophicalor technical context, and their treatisesmust be studied in a larger intellectual context if one is to understandthem. Historiansof music theory have realized this for many years in connection with terminology in medieval and Renaissancemusic theory, but the lesson has been all too rarely applied to ancient Greek music theory. Harmoniain its fullest sense, then, is a unificationof things thatappear on a lower level to be dissimilaror unrelatedor lacking in order.This sense of the term appearsclearly in the Phaedo and Timaeusof Plato, in Politica 8.5 (1340b18-19) and De anima 1.4 of Aristotle, in Plutarch'sDe animae procreatione in Timaeo, and in the musical treatises as well, for instance in Aristides QuintilianusDe musica 3.6. Thus, to choose three examples. the octave is a harmonia(cf. Aristides QuintilianusDe musica 1.8; 2.12). a proportionlike 2:4:8 (which may also produceoctaves) is a harmonia(cf. Aristides QuintilianusDe musica 3.6), and a tonos is a harmonia(cf. Aristides QuintilianusDe musica 1.10; 2.14). Each of these provides a unification and an order for lower level entities such as dissimilar numbers, groups of notes, or rhythmic patterns. From this definition, it should be easy to see two things: (1) how less sophisticatedtheorists(and some modem commentators)began to think of tonos and octave species as synonymous with harmonia,and (2) while an octave, a proportion,or a tonos are harmoniai,the converse is not true:harmoniais not an octave, a proportion, or a tonos. Given this general definition for harmonia,it will not be surprisingto discover that the term is also used in connection with cosmology, ethics, metaphysics, and other branchesof Greekphilosophy. Moreover, the term often appears in passages where musical parallels are explicit, as for instance in the Phaedo, noted above. This is importantbecause it might otherwisebe assumed that the term is being used metaphoricallyor at least in a sense unrelatedto music. But if harmoniaas a concept might relate divergentbranchesin Greekphilosophy, it may be possible to discoverhow that much-vauntedpower of ancient Greek music-the ability to convey and affect ethos-actually worked. In fact, this relationshipof divergent philosophical branchesthroughharmoniais one of the central features of the treatiseof AristidesQuintilianus,4and the treatiseprovidesa means for distilling a system to analyze the ethical quality of a composition. Aristides Quintilianus and Platonic notions of mimesis The treatiseof AristidesQuintilianusis quiteunlikeotherancientGreek musical treatises in that it is not a technical work but ratheran elaborate 4This matteris discussed at length in my Introductionto Aristides Quintilianus,pp. 14-57.

work of philosophy that attempts to show how music through harmonia providesa comprehensiblemodel for understandingphysical and metaphysical realities and thus for attaininghigher knowledge throughphilosophy. The treatise begins with a study of the technical details of harmonics, rhythmics, and metrics;proceeds to a considerationof the effect of music on character;and concludes with an exegesis of number,the soul, and the order of the universe. Since all of these sections are related cumulatively in the most intricateway, the concludingsections naturallyprovidereasons for the powers and phenomenaobserved in the earlier sections.5 It has long been observed in general terms that the Platonic notion of mimesis is in some way relatedto music's power to affect and effect ethos. Indeed, one might even say that the notion of mimesis is fundamentalto the notion of musical ethos, at least in the ancient Greek sense. It is therefore necessary at this point to review mimesis as the process throughwhich harmoniaimbues music and throughwhich ethos is developed in response to music. The concept of mimesis is exceedingly difficult to condense, despite the importantwork by S. H. Butcher, HermannKoller, Paul Moraux, and Gerald F. Else.6 Nevertheless, Aristides Quintilianusonce again provides a great deal of help in his ratherspecific remarksabout mimesis in music. De musica 2.4 points out thatsince the soul is naturallyattractedto whatever is like one or anotherof its parts, mimetic arts may affect the soul. The notion of the soul's attractionis based-perhaps-on Plato's Timaeusand Sophista, but it is also consideredin Aristotle'sMetaphysica1.6. The parts of the soul, of course, are the rationaland irrationalparts-the latterfurther divided into the thymic (or, aggressive) and epithymetic (or, indulgent) parts (these are elegantly characterizedin Plato's Phaedrus 245-248 and Timaeus 41E-42D). Musical mimesis is especially powerful, according to Aristides Quintilianus,because it is not a simple imitationof things but is ratheran imitation of life itself, capable of raising the soul once again to the harmoniaof the universe: The reasons for music's actuality are apparent. For inasmuch as our first learning comes through similarities, which we conjecture by attending to our sensory perceptions, painting and sculpture teach only through vision and the likeness both excites and amazes the soul; how then could music fail to captivate, since it makes its mimesis not through one sensory perception but through many? Even poesy uses hearing alone through pure diction, but it does not always move passion without melody or adapt it to the underlying matters without rhythms. For instance, if ever it is necessary to move passion in accord with the interpretation, such a thing does not result without slightly altering the voice in the 5Ibid., pp. 14-15. 6S. H. Butcher, " 'Imitation' as an Aesthetic Term," Aristotle's Theoryof Poetry and Fine Art with a Critical Text and Translationof the Poetics, 4th ed. (London, 1911; reprinted., New York, 1951); HermannKoller, Die Miinesis in der Antike (Bern, 1954); Paul Moraux, "La 'mimesis' dans les th6ories anciennes de la danse, de la musique, et de la po6sie," Etudes classiques XXIII (1955), 3-13; and Gerald F. Else, " 'Imitation' in the Fifth Century," Classical Philology LIII (1958), 7390. Note also the useful article by W. Tatarkiewicz, "Mimesis," Dictionary of the History of Ideas, 5 vols., ed. Philip P. Wiener (New York, 1973), III, 225-30.

THEJOURNAL OF MUSICOLOGY direction of melody. Only music teaches both by word and by the counterparts of actions, not through motionless bodies or those fixed in a single form, but throughanimatebodies, of which it altersboth the figure and the motion to the kindredform in accord with each of the actions recited. These things are evident both from the dance of the ancient choruses, the trainerof which is the science of rhythmic, and from the things writtendown by many authorsaboutdelivery. Those arts having materialsthat are special may not lead quickly to a notion of the action, for to some the colors, to others the mass, and to still others the phrase have suggested things alien to the truth. But music persuadesmost palpably, for it makes its mimesis with such means as happen also to carry out truthfullythe actions themselves. Since in the usual orderof things, will leads, word follows, and thereafteraction is accomplished, music imitatesthe ethoses and passions of the soul in the notions, the words in the harmoniaiand the molding of the voice, and the action in the rhythmsand motion of the body.7 In De musica 3, Aristides Quintilianusshows how all the parts of music-pitch, scale, tonos, rhythmicpattern,and so on-are like the order of the universe, and thereforethroughmimesis, music may make the order of the soul like the orderof the universe. Thus, the harmoniaof music may create a like harmoniain the soul, and this in turncreatesa particularethos. In De musica 2.14, Aristides Quintilianusstates: By using the harmoniaiin the aforesaidways-either presentinga harmoniato each soul by similarityor contrariety-you will disclose the inferiorethos lying hidden, heal it, and instill a better one. You will persuade, if ignobility and stiffness should lurk, by leadingthrougha middlestateto the oppositecondition; and if refinementand usefulness shouldbe present,by augmentingthem through similarity in symmetricalproportion.... As I said, the harmoniaiare like dominatingintervalsor surroundingnotes, and these are like the movements and affections of the soul. That notes, even of continuous melody, mold throughsimilaritya nonexistentethos in children and in those already advancedin age and bring out a latent ethos, the disciples of Damon showed. In the harmoniaitransmittedby him, it is possible to discover that sometimes the feminine, sometimes the masculine of the movable notes either dominate or have been employed to a lesser degree or not at all, since it is evident that a harmoniais utilized in accord with the ethos of each soul.8 With this sense of the mimetic power of music, it is now possible to see the sort of analytical system that might be distilled from Aristides Quintilianus' treatise. The descriptions of ethos, harmonia, the technical parametersof music, and the musical psychology are so consistentthroughout the treatise that they may be arrangedand classified as shown in Table 1 below. The table illustratestwo of the indispensablecomponentsof any piece of Greek music: rhythmand melos.9 These componentsare furthersubdi7AristidesQuintilianusDe musica 2.4 (trans. from Aristides Quintilianus,pp. 118-19). 8AristidesQuintilianusDe musica 2.14 (trans. from Aristides Quintilianus,pp. 144-45). 9Melos is the combinationof words, melody, gesture, and rhythm(see Aristides QuintilianusDe musica 1.12 and Mathiesen, "Problems," pp. 5-6).

vided into masculine, feminine, and medial characteristics.By applying these characteristicsto the musical fragments, a dominant ethos may be determinedin the fragment, and it may then be considered whether this ethos would match the characterand function that might be assumed from a study of the text of the fragment. Application to the fragments Any numberof fragmentsmight be chosen for analysis, but it would surely be best to concentratefirst of all on the two earliest surviving fragments, since they are most likely to exhibit the classical conceptions of harmoniaand ethos described by Aristides Quintilianus.These two fragments are P. Wien G2315 and P. Leiden Inv. 510, both dating from the third century B.C. and both ostensibly preservingmusic from Euripides' Orestes and Iphigenia in Aulis. 10 Figure 1 exhibits a transcriptionof P. Wien G2315, 1 which preserves Orestes338-344; a table of the notes found in the fragmentandtheirrelation to the harmoniaigiven in Aristides Quintilianus'De musica 1.9; and a full analysisof the fragment,following AristidesQuintilianus'system, as shown in Table 1. In the analysis, the melos has been shown on two levels: that of melos vowels and that of the notes themselves and the tonoi, scales, and octave species they imply. The rhythmic area reveals fifty-five masculine elements, sixty-nine feminine elements, and twenty-ninemedial elements (note the summaryto Figure 1). This would seem to provide generally a medial ethos, with perhapsa slight feminine emphasis. In rhythmicterms, the composition is ambiguousin ethos. A differentpictureemerges from the melos. Ignoringfor the moment the notes themselves, one finds fifty masculine and medial masculine elements (i.e., thirty-fourmasculine melos vowels and sixteen medial masculine melos vowels), with only twenty-one feminine and medial feminine elements. This would seem to convey a decidedly masculineethos. Only a small portion of the notationremains, but as a randomsample, it may be statistically valid. Here again, a review of Table 1 will show that the stationary notes are medial in character,while the movable notes are either l?The matterof the propertranscriptionof these fragmentsis, of course, significant to any study of ethos, but it must be assumed for the present that the published transcriptionsof the fragmentsare at least reasonably accurate. The Orestes fragment appears in Egert Pohlmann, Denkmaler altgriechischer Musik, ErlangerBeitrage zur Sprach- und Kunstwissenschaft,XXXI (Niirnberg, 1970), pp. 78-82; the Iphigenia in Aulis fragmentappearsin Thomas J. Mathiesen, "New Fragmentsof Ancient Greek Music," Acta musicologica LIII (1981), 14-32. 11n the transcription,I have added the rhythmicvalues (Pohlmanntranscribesthe piece with note heads alone) indicatedby the rhythmicnotationand the rhythmicvalues of the text and have changed the low instrumentalnote (- ) from the F# of Pohlmann'stranscriptionto an E, which is the proper readingfor this Greek notationalsymbol.

TABLE 1 Characteristicsof Ethos in Ancient Greek Music Based on the System of Aristides Quintilianus RHYTHM (masculine) Feminine Masculine Long vowels (r, o) x), X) Rough mutes (, Double consonants (4, 4, 4) and spirant(r) Feet dominatedby long syllables Short vowels (e, o) Smooth mutes (x, r,,r) Liquids (X, p, v, p) Feet dominatedby short syllables Medial Dichrona(a, L,v) Medial mutes (3, /, 6) Composite feet MELOS (Feminine) Feminine Medial Masculine Vowels a (more masculine), Land Vowels E and aq Vowels o and w v (more feminine) Lichanoid notes (lichanos hypa- Parhypatoidnotes (parhypate ton, lichanos meson, paranete hypaton, parhypatemeson, Stationarynotes (proslambanometrite synemmenon,tritedienos [more feminine];hypate hysynemmenon, paranete diepaton and hypate meson [more zeugmenon, paranete hyperzeugmenon, trite hyperbolaion) masculine]; mese followed by bolaion) Scales bounded by masculine Scales bounded by feminine diezeugmenon [more feminine]; notes mese followed by synemmenon notes [more masculine]; and parGapped scales with predomi- Gapped scales with predomiamese, nete synemmenon, nete nantly feminine notes nantly masculine notes Dorian mode diezeugmenon, and nete hyperLydian mode bolaion [more masculine]) Instrumentspitchedfrom F to Bb Instrumentspitchedfrom bbto bb' Scales boundedby medial notes Gapped scales with predominantly medial notes Phrygianmode Instrumentspitched from Bb to ab Figure 1. Analysis of P. Wien G2315 (EuripidesOrestes 338-344) According to the System of Aristides Quintilianus Alypius symbols instruvocal mental Dynamic names g' paranetediezeugmenon trite diezeugmenon f' e' paramese d' mese chromaticlichanos meson b bb parhypatemeson a hypate meson lichanos hypaton g hypate hypaton e Z E zZ I 7T P C D Aristides Quintilianus Dorian Phrygian a' 0 g' f' A e'* E e' Z d' I a# II a* P a C g c Lr A E zZ I n

PC Text uactox\oppoiJa(u p'jLTposT Rhythm D F F D FM D F M D M Melos m MM f MM m F M1 M m M F M Fm Notes actS, M --I I = V, 'l o6pfTsoLPos a ' havo La*XE tIL. Text ayIa o0 Rhythm IF D D M MM F FD M F M Melos M mm m FF M Fm M M M m m m f m F TT P c Notes Text p6VL,OS V; Rhyth F D F Melos M f M Notes I z va 6E AOL 0S W"!S M F M DD F M F M F M M m n F M M1. M M F m f m c P T CpL c- 0oac tLVcQtas6Ciijuv "L BpOTOi,, Text t6S axaro) Rhythlm D I D M FM D M M M Melos f m m M Mn fm m MM m F 1 rF Mm Notes 1TT P rT I J,e Text xdtK(XuaoEv Rhythm D M Melos m Notes ,MM D M fF f F M 7 ) -13 6ELVSV

7 Text J6Vuv') Rhythm F M MM M M Melos Notes M!n - r mf m M 1 3 sjs )OVTOu P C PpZ -: Text Acpo oACpoL oAS L av tV Rliythm M M FM DM D FF Melos m M M F f M f FF f m f MM m Fm ? M Notes u'KUiarv Key D = medial dichrona M = masculine F = feminine m = medial masculine f = medial feminine Summary Masculine 5 rough mutes 19 double consonants and spirants 31 masculine syllables - 34 masculine melos vowels 12 masculine notes Feminine 69 16 smooth mutes 34 liquids 19 feminine syllables 12 feminine melos vowels 8 feminine notes Medial 29 8 medial mutes 21 medial syllables 16 medial masculine melos vowels 9 medial feminine melos vowels

14 medial masculine notes 3 medial feminine notes Tonos: Lydian Octave species: Phrygian Scale boundedby masculine notes Rhythmicpattern:dochmiacs (mixed) masculineor feminine. The medial notes may have a masculineor feminine inflection.12The notationin this fragmentshows twenty-six masculine and medial masculineelements, and only eleven feminine and medial feminine elements. It immediatelystrikes one that the notationalproportioncomplements almost perfectly the ratio of melos vowel elements. The proportion 26:11 is almost identical to the proportion50:21. In terms of the melos, therefore,the ethos of this part of the chorus is decidedly masculine. How does the rhythmicpatternrelate? The rhythm(or meter, if one perfers) of the fragment is dochmiac. Aristides Quintilianus'De musica considers this to be a "mixed rhythm" and states (2.15) that these types of rhythmscreatea disorderedaffect. AristidesQuintilianusdoes not assign specific masculine, feminine, and medial characteristicsto rhythmicpatterns (except in the general terms shown in Table 1); he ascribes instead certainaffects to them. Thus, rhythmicelements convey characters,while patternsconvey affects. So, althoughthe rhythmicelements provide a medial ethos, the rhythmicpatternconveys an affect of disorder. Finally, there is the matter of tonos. The tonos of the fragment is generally considered to be Lydian. The octave species, which has not received much attentionin publishedtreatmentsof this fragment,is probably Phrygian (though Dorian and Mixolydian are also possible, if the instrumental notes are included in the pattern), and-as an interestingcoincidence-the notes of the fragment also match the Phrygian and Dorian harmoniaigiven by Aristides Quintilianus'De musica 1.9, where he states that these are "exceedingly ancient." If the tonos dominates the ethos, there is a feminine ethos here. But if it is the harmoniaand if the harmonia in this fragment is Aristides Quintilianus'Dorian or Phrygian (the latter matchingthe octave species), there is either a masculineor medial ethos. Some conclusions may now be drawn from the data. The rhythmic elements are medial in ethos, the melos elements are decidedly masculine, the rhythmicpatterncreates a disorderedaffect, and the harmoniais masculine or medial. Thus, the fragmentwould seem to convey a masculine ethos with some mediationand an agitatedaffect. How does this charactermatchthe characterandfunctionof this chorus in the tragedy?The fragmentcomes from a point near the beginningof the tragedy.Electrahas dominatedthe first scene, along with a chorusof Argive 12Thecharactersassigned in Table 1 come from the various charactersassigned to vowel sounds in Aristides QuintilianusDe musica 2.11-13 and the use of some of these vowels in a solmization system to sing each note of the PerfectImmutableSystem, as describedin De musica 2.14. See Aristides Quintilianus,pp. 33-34 and 139-46.

maidens. Orestes is also present in a frenzied state. Electra tells Orestes that Menelaus has arrived and is coming to see him. After she exits, the chorus begins, a chorus of anguish and weeping, full of disorderedimages of struggle and question that fit the agitated dochmiacs with their many successive shortsyllables. But the chorusalso serves to introduceMenelaus. The masculine ethos would therefore seem appropriateto the appearance of Menelaus, and the disorderedaffect foreshadowsMenelaus' own agitation expressed in his opening speech. It would be prematureon the basis of one example to draw a firm conclusion aboutthe relationshipof musical and dramaturgicalethos in this chorus, but one might at least say that the ethoses seem reasonably cohesive. Another fragment may provide additional evidence. Figure2 exhibits a transcriptionof P. Leiden Inv. 510, which preserves Iphigenia in Aulis 783-796; a table of the notes found in the fragment;and a full analysis of the fragment,once again following AristidesQuintilianus' system as shown in Table 1. The analysis follows the same patternused for the Orestes fragment. Figure 2. Analysis of P. Leiden Inv. 510 (EuripidesIphigenia Aulidensis 783-796, According to the System of Aristides Quintilianus Alypius symbols Dynamic names chromaticparanetediezeugmenon nete synemmenon diatonic paranetesynemmenon chromaticparanetesynemmenon trite synemmenon paramese mese diatonic lichanos meson parhypatemeson diatonic lichanos hypaton chromaticlichanos hypaton parhypatehypaton hypate hypaton bb' jpnrt eVOt Rhytlhm MF Melos F F F Notes M F M M PITi I g#' g' t U f#' e' d' b bb, a g# A or / I I O I 0 A 5A }yperaeolian Text Hyperphrygian Hyperaeolian bb' b ab5' I g' f' U bb, f Il C T . epotaL T?VWVV TLeXVOLS OwLts C&bL or MM F F MF MM M F F F M f F F M m F m M F e^eOOL M M F D F F F M M M F f

CT,T .T T aoraoVoL Iop&. La;toriZ ivJuCW.tL Text OOLVaA roXVxpuoLL Au6ai Kt Rhythm MI) M F fI D M D M M D M D F M M M D D D rl M D i; r; Me los Mln M M f f M f 1' M f M m M M F M f m m f M f F M Notes FF F M in ml m m X\ . . .,1 aiXoxo puvy&ZV Hyperphrygian Text rXXnra S' TCS apot p Rhythm D F F M M D D D FM Melos m F F m F m f m m F Notes M m m m m m m Trd6e is _ _I Text tQa& TaLS Rhythm DD M Melos mm M Notes mm m rr . bpa 6abpu6Ev FM MD M DFM M m f m m fMF euil,Aoxdjpouv xOpias F D F m Mm M IHyperaeolian F r I CTTTT 6dX)opJas aXtoXTLet; 6L&a TCfv (, DD F F F D D F M D M DD F M M M F D F M F F m fm M M F m m M M f F fm F m f M M f M F M M M m Mm 1-1In aT'pisL D M F m f Key = medical dichrona = masculine = feminine = medial masculine = medial feminine Summary Masculine -7 rough mutes 20 double consonants and spirant 43 masculine syllables 33 masculine melos vowels 10 masculine notes U'KVOU 6oAtLXauIXCVyOSYo),

Feminine r 28 smooth mutes 95 41 liquids L 26 feminine syllables 24 feminine melos vowels 5 feminine notes Medial 36 9 medial mutes 27 medial syllables 22 medial masculine melos vowels 17 medial feminine melos vowels 19 medial masculine notes Tonos: Hyperaolianand Hyperphyrygian Octave species: Mixolydian Scale boundedby medial notes Rhythmicpattern:choriambic(composite) In the area of rhythm, there are seventy masculine elements, ninetyfive feminine elements, and thirty-six medial elements (note the summary to Figure 2). Here again, this would seem to imply a medial ethos, with a slight feminine emphasis. In rhythmicterms, the ethos of the fragmentis ambiguous, but somewhat feminine. The melos, as was the case with the Orestes fragment,shows a masculine ethos in the fifty-five masculineand medial masculineelements (i.e., thirty-threemasculine melos vowels and twenty-twomedial masculinemelos vowels) and forty-one feminine elements (i.e., twenty-fourfeminine melos vowels and seventeen medial feminine melos vowels). This is certainly less markedlymasculinethanthe Orestesfragment.In this fragment, too, only a small portionof the notationsurvives, but it may claim statistical validity. There are twenty-nine masculine and medial masculine notes, as againstonly five feminine notes. There are no medial feminine notes at all. If one considers the melos vowels and melos notes together, eighty-four masculine and medial masculine elements and forty-six feminine elements appear.So far the ethos of this fragmentis less clear thanthatof the Orestes fragment. The rhythmicpatternof the fragmentis choriambic. Aristides Quintilianus considers this patternto be composite by conjunction, which-as can be seen from Table 1-creates a medial ethos. The predominanceof long syllables, however, conveys a more masculinequality(noteDe musica 1.23, 26; 2.15), and this matches, as will be shown, the solemn nature conveyed by the prayerin this part of the chorus. The tonos of the fragment is Hyperaeolianand Hyperphrygian;the octave species appears to be Mixolydian, which according to Aristides Quintilianus'De musica 1.9 is threnodicor feminine. The conclusions that can be drawn about this fragmentare less consistent than those for the Orestes fragment. The rhythmic elements are medial in ethos with a feminine emphasis, the melos elements are more masculine by far (though the relationshipof the melos vowels and melos

notes is not as precise as it was in the Orestes fragment), the rhythmic patterncreates a medial quality with some masculine emphasis, and the harmoniaand tonos are clearly feminine. Thus, the fragmentwould seem to be medial because of the dominantfeminine harmoniaand the masculine melos combined with the medial rhythmicpattern. A considerationof the functionof this chorus in the tragedymay help in understandingthe ambiguityof the ethos. The chorus' function is problematic in this work because the chorus appearsin one place in the manuscripttraditionandanotherplacein the papyrus,unlessthe papyrusrepresents some sort of anthology with the solo and choralpartsnot necessarilyin the properorder. Adding to the problemis the fact that it is not certainwhether Euripidesfinished this tragedyor whethersome of it is spurious.Iphigenia in Aulis was producedposthumouslyand scholars have questionedthe authenticityof some parts of the work.13 In the manuscripttraditionfor Iphigenia in Aulis, the chorus precedes the first appearanceof Achilles; in the papyrus, the chorus follows the departureof Iphigeniato be sacrificed.If the manuscripttraditionis correct, one might expect by analogy with the Orestes fragment that the chorus would have a masculine ethos to introduceAchilles while at the same time conveying a disorderedaffect in keeping with the knowledge of Agamemnon's treachery.On the other hand, if the papyrusrepresentsan alternate version of the tragedy(or Euripides'intendedorder), one might expect the chorus to convey a calmer, neutralethos in keeping with the grandeurof Iphigenia's decision to accept her sacrifice willingly. Moreover, the text is one of supplication, and the dominanceof the long syllables, as has been noted, conveys a quality of solemnity suited to supplication. Iphigenia's heroic quality at this point in the tragedycould thus be emphasizedin the masculineethosof the melos, the lamentationappropriate to the text foretelling the fate of Phrygia in the harmonia, and the dispassionateacceptance of Iphigenia's fate in the rhythmicpattern.If this analysis is correct, it may serve to suggest that this chorus does indeed belong at the end of the tragedy, following line 1509, and to support the appropriatenessof the medial ethos conveyed by all the elements of the setting. It may be objected at this point that the evidence for the ethoses of these two choruses is conflicting and the system thereforedoes not work. But it must be rememberedthat Euripideswas noted for his unusual and complex approachto tragedyand for his avoidanceof stereotypedcharacters and moods. It should thereforebe surprisingif a Euripideanchorus conveyed one and only one ethos in all its parts. Moreover, Aristides Quintilianus (De musica 2.16) makes it clear that complementaryethoses were the norm ratherthan a uniformethos in every part: 13SeeG. M. A. Grube, The Drama of Euripides (London, 1961), pp. 421, 438; Denise JourdanHemmerdinger,"Un nouveau papyrusmusical d'Euripide(presentationprovisoire)," Comptesrendus de l'Academie des Inscriptions et Belles-Lettres(1973), 292-302; Giovanni Comotti, "Words, Verse and Music in Euripides' Iphigenia in Aulis," Museumphilologum Londiniense2 (1977), 69-84; and Mathiesen, "New Fragments," pp. 24-25.

In the perfect actuality of music, a suitable notion is determined, and seemly diction, a scale akin, harmonia of notes, qualities of rhythm, and use of instrument are made homologous. We must apply the perfect types of musical creation when the extreme is in no way harmful. But sometimes we must mix and beware of otherness, lest somewhere, because of the extreme, we unknowingly lead the underlying ethos in the opposite direction. . . . We must make a mixture not through bare opposites (for this is unsuitable and repellent) but rather through harmoniously arranging the means with the extremes: for example, with a masculine scale, not a feminine but rather a medial rhythm; or with a more feminine rhythm, a medial scale and instrument, not only an opposite one; and in like manner for the rest, producing out of masculinity and the middle state or femininity and the middle state a mixture of one with the rest or two or even more with more.14 It should also be noted that the two fragmentshappen to come from tragedies that are generally regardedas unusualin their structureand particularly in their use of chorus. This may, incidentally, also explain why these particularfragmentsexist. The peculiarityof theircharactermay have necessitated their notation and thus insured their survival. In Iphigenia in Aulis, the chorus is unusuallypassive and plays a largely neutralrole. The ethos discovered in the fragmentseems to matchthat neutralityquite well. In Orestes, the chorus fills its more usual role in Euripides'work of being directly involved and sympatheticto the main characters.Thus, the more definite ethos in the Orestes fragmentis to be expected. The conflicting evidence of the two Euripideanchorusesmay also lead one to wonder if all the fragmentswould demonstratesuch complexity or ambiguity. A summaryof one additionalcomposition may serve to dispel this question. The first segment of the Delphic hymn by Limenios (Delphi Inv. Nr. 489, 1461, 1591, 209, 212, 226, 225, 224, 215, 214)15 exhibits an altogether feminine ethos. In the rhythmic area, it presents fifty-nine masculine elements, 118 feminine elements, and only eleven medial elements.16 In the area of melos, it has only seven masculinenotes as against thirty-onefeminine notes and thirty-sixmedial notes. It also has fifty-two feminine and medial feminine vowels and fifty-threemasculineand medial masculine vowels. It is in a Lydian tonos with a Phrygianoctave species, and the rhythm is dominated by short syllables. Except in the area of melos vowels and octave species, which are medial in ethos, the ethos is feminine, which surely fits a text invoking the Muses. 4Aristides QuintilianusDe musica 2.16 (trans. from Aristides Quintilianus, pp. 150-51). This excerpt reflects the theoryof contraries,which operatein mimesis (cf. AristotleMetaph. 10.8 [1058a828]); on this excerpt, cf. PlutarchDe musica 33. 15Thehymn is published in P6hlmann.pp. 68-76. 16I.e.,there are thirteen rough mutes, thirteen double consonants and spirants, and thirty-three masculine syllables (totaling fifty-nine masculine elements); and. on the other hand, there are thirtyone smooth mutes, fifty-two liquids, and thirty-fivefeminine syllables (totaling 118 feminineelements).

Conclusion It may be argued that an analysis of ethos based on so many minute details has no chance of illuminatinga truly functionaleffect in pieces of music because audiences could not possibly grasp or be affected by such minutiae. Yet this argumentmay easily be defeated. It is known from early treatises such as Dionysius of Halicarnassus'De compositioneverborum, Ps.-Longinus' De sublimitate, and various works of Hephaestion-not to mention the technical observationsof Plato and Aristotle-that the Greeks had an extraordinarilysubtle sense of their language. This subtletyhas also been admirablydemonstratedby modern scholars, especially by W. B. Stanford and W. Sidney Allen.17 It is also known from the writings of Plato, Aristotle, Philodemus, Sextus Empiricus,and othersthat the Greeks did perceive the ethos conveyed by theatricalworks and music and that they were concernedwith this subject.18 Finally, it might be noted that the affectiveor ethicalqualityof such musicalfeaturesas keys, modes, rhythms, andeven pitch vibrationswas still observedand seriouslytreatedby modern philosopherslike Leibniz-who proposes that the soul, in listening to music, unconsciously counts the beats of the tones, compares their mathematical ratios, and appreciates the mathematical relationships as "un arithmetiqueocculte"'9--and by systematic theorists at least as late as Daniel Gottlob Turk and JohannPhilipp Kirnberger.20It is not, of course, the individualminuteelements of any composition, ancientor moder, that convey ethos, but rathertheir collective effect.21 Harmoniaand ethos in ancient Greek music are inextricably linked and overarchthe whole body of mousike. By studying the fragmentswith the aid of AristidesQuintilianus'system, it may be possible to come closer to an understandingof the special power and beauty of both the musiceven in its fragmentarystate-and the theory. Brigham Young University 17W. B. Stanford, The Sound of Greek: Studies in the Greek Theoryand Practice of Euphony, Sather Classical Lectures, XXXVIII (Berkeley, 1967); and W. Sidney Allen, Accent and Rhythm: Prosodic Features of Latin and Greek: A Study in Theoryand Reconstruction,CambridgeStudies in Linguistics, XII (Cambridge, 1973). 'lNote especially Aristotle Politica 8.5 (1340al-b19); Philodemusand Sextus Empiricustake a skeptical view of ethos, but their denial nonetheless affirms the prominentplace of musical ethos in Greek thought. 9'"Remarks on an Extractfrom Bayle's Dictionary," PhilosophischenSchriften, ed. C. J. Gerhardt, 7 vols. (Berlin, 1880), IV, 550. 20DanielGottlob Turk, Klavierschule 6.3 (School of Clavier Playing, trans., introduction,and notes by RaymondH. Haggh [Lincoln, 1982], pp. 337-53); JohannPhilipp Kirberger, Die Kunstdes reinen Satzes in der Musik 2.2, 4 (The Art of Strict Musical Composition,trans. by David Beach and JurgenThym, introductionand notes by David Beach, Music TheoryTranslationSeries, 4 [New Haven, 1982], pp. 314-46, 375-417). 21NotePlutarchDe musica 33-36.