TSITSIBAKOU-VASALOS, EVANTHIA, The Meter of the Lille Stesichorus , Greek, Roman and Byzantine Studies, 28:4 (1987:Winter) p.401 The Meter of the Lille Stesichorus Evanthia Tsitsibakou- Vasa/os S TESICHORUS' POSITION in the literary tradition-until recently a matter of guesswork based on indirect references from later periods-has become appreciably clearer thanks to the papyri. The discovery of the Lille papyrus attests to Stesichorus' importance in the history of metrics in particular, for the use of an early form of the dactylo-trochaic verse offers some insight into the puzzling issues of ancient lyric rhythm and perhaps its corollary, ancient Greek music. 1 Attempts to analyze the meter of P.Lille 76 have nevertheless ex­ posed once more the limitations of our knowledge, due mainly to the fragmentary nature of the direct evidence. Metrical sequences at­ tested also in other Stesichorean fragments have been characterized in different ways, and differing inferences have been drawn regarding both the technique of the poet and his significance in the development of metrics and music. As this controversy is not restricted to termi­ nology but, more significantly, affects our understanding of the very nature of the Stesichorean verse,2 my purpose here is first to describe and characterize the P.Lille meter and second to review and evaluate the three major analyses it has received. Emphasis will be placed on the structure presented in the editio princeps, both because its under­ lying principles remain controversial and because its implications for the questions of rhythm and music are far-reaching. 3 1 The following will be cited in abbreviated form: G. P. ANCHER, "Etude metrique des P.L. 76 a,b, et c," Cahier de recherches de I'Institut de Papyrologie et d'Egypto­ logie de Lille 4 (1976) 311-23 and 350f (the editio princeps); A. DAIN, Traite de metrique grecque (Paris 1965); A. M. DALE, "Greek Metric 1936-1957,'" Lustrum 2 (1957), and The Lyric Metres o/Greek Drama 2 (Cambridge 1968); M. HASLAM, "The Versification of the New Stesichorus (P.Lille 76abc),'" GRBS 19 (1978) 29-57, and "Stesichorean Metre," QUCC 17 (1974) 7-57; J. IRlGOIN, Recherches sur les metres de la lyrique chorale grecque. La structure du vers (Paris 1953); W. J. W. KOSTER, Traite de metrique grecque suM d'un precis de metrique latine (Leiden 1966); B. M. PA. LUMBo-STRACCA, "Osservazioni metriche al nuovo Stesicoro (Pap. Lille 76 a,b,c),'" Boll.d.c/ass. 25 (1977) 31-43. 2 See especially Haslam, GRBS 54. 3 This paper is an abbreviated version of my dissertation, Stesichorus and His Poetry (Chicago 1985) 150-227. 401 TSITSIBAKOU-VASALOS, EVANTHIA, The Meter of the Lille Stesichorus , Greek, Roman and Byzantine Studies, 28:4 (1987:Winter) p.401 402 THE METER OF THE LILLE STESICHORUS The metrical scheme of the Lille poem, in strict relation to the colometry presented on the papyrus, is the following:4 . C" -Str.lant.: 1 _vv_vv_: vv_~_vv_vl Epode: 2 _vv_vv_: v:::.v_vl 3 _vv_vv_1 V-::vv_vv_vl 4 _vv_vv_1 __ 5 -- . v -I ~_vv_vv_ :VV"_V_V 6_vv_ vv_1 7 c:..v_ V-::v_vl I_vv_vv_1 2 i:v~vv_~v __ 1 3 _v_~_v(~)1 --4 _vv_vv_: :::v _~_vv_vl 5 c::.v_vl ---. 6 _~_vv_:~_vv_vv_vl ·vv 7 ___ v __ 1 The structure consists of units of dactylo-Iength and epitrites, the association of which has been made possible with the help of certain metrical rules that apply with distinct regularity. The most important are the~strictobservanceofword-end at the end of the colon _vv_vv_ (D in Maas' notation), with a few but justifiable exceptions, as we shall see: the chaining of the anceps element (x) with the metrical unit that follows it; and the indifference of the medial anceps, which can be bisyllabic (biceps), long or short. These characteristics (especially the employment of an indifferent anceps '::v which provides a salient difference from later dactylo-epitrites) suggest that we are dealing with an early attempt to compose in the dactylo-epitrite meter. These innovations had been observed already in the Geryoneis and lliou Persis, among other poems of Stesichorus, and thus provided an essential criterion for attributing the Lille poem to him. The first important metrical analysis was that presented by Ancher in the editio princeps, based on the principles laid down by lrigoin. For a better understanding of Ancher's structures, it will therefore be helpful to summarize and evaluate the main features of lrigoin's theory. lrigoin observed in Pindar and Bacchylides certain metrical phe- 4 This is Haslam's tabulation (GRBS 33). ~ indicates short and long syllables each at~ested; ~, presumed anceps element, short in all attested instances; I n, word-end in all (most) attested instances; -- zeugma (no word-end). TSITSIBAKOU-VASALOS, EVANTHIA, The Meter of the Lille Stesichorus , Greek, Roman and Byzantine Studies, 28:4 (1987:Winter) p.401 EV ANTHIA TSITSIBAKOU-V ASALOS 403 nomena whose recurrence, statistically confirmed, he raised to the status of a virtual law, although he refrained from using the term.s Essential for understanding these postulates is his preliminary defi­ nition of two terms: (a) synapheia, which designates the chaining together, or liaison, of rhythmical elements grouped in cola, and (b) generalized caesura, which signifies the caesura produced at the same point in the verse each time this verse is repeated in a poem with responsion.6 Within the lyric verse, the poets avoid coincidence of word-end and end of a rhythmical element. This entails the absence of diaeresis, which is almost always replaced by liaison, or zeugma, be­ tween two consecutive rhythmical elements within the verse. The contacts, i.e., the points where two cola meet, are concealed by being bridged together in synapheia.7 This procedure is frequently empha­ sized through the use of a "coupe" (caesura, in this context), shifted one syllable before or after the point of contact, or "suture," of two rhythmical elements within the verse. In the lyric verses there is, consequently, neither diaeresis nor a generalized caesura, but only this flexible and mobile non-generalized caesura. This tendency re­ veals the poet's desire to assure the unity of the verse by avoiding the dislocation of its elements.8 Regarding the dactylo-trochaic (i.e., dactylo-epitrite) verses, lrigoin repeats and emphasizes the same conclusions: the diaeresis appears only rarely and exceptionally between the rhythmical elements that constitute the dactylo-trochaic verses. The appearance of the caesura moving one syllable "en avant ou en arriere" the point of "suture" of two consecutive rhythmical elements is irregular and rarely genera­ lized. The mere fact that it is flexible and movable proves that it is only a secondary feature. 9 The contention that there is neither diaeresis nor word-end between two consecutive metrical elements but, on the contrary, a metrical en­ jambment (synapheia), leads to consequences that may be seen in the analysis of the same sequence of long and short syllables in two or more different ways. Verses of identical appearance, such as _v __ _ vv_vv ___ v_, are analyzed as either _v_::=::"vv_vv::=::"_v_ (i.e., trochaic dipody, hemiepes feminine, catalectic trochaic dipody under 5Irigoin 16; cf Dale, Lustrum 25. 6 lrigoin 12 with n.2, where he quotes the definitions of L. Havet and L. Duvau, Cours eiementaire de metrique grecque et latine 4 (Paris 1896) §5. 7 Koster 18; see also Dain 34. 8 lrigoin 16. 9 Irigoin 40.4: Les consequences de la synaphie. TSITSIBAKOU-VASALOS, EVANTHIA, The Meter of the Lille Stesichorus , Greek, Roman and Byzantine Studies, 28:4 (1987:Winter) p.401 404 THE METER OF THE LILLE STESICHORUS the form of cretic) or _v'::::::"_vv_vv'::::::"_v_ (i.e., acephalous iambic dipody, prosodiac, iambic dipody). One might, for that matter, ana­ lyze it further as catalectic trochaic dipody, prosodiac, iambic dipody. The practical application of lrigoin's theories may be observed in his analysis of the verse __ v ___ vv_vv ___ v_, which, although it could be viewed as iambic dipody, prosodiac, iambic dipody, is divided into iambic dipody, enoplios, cretic (representing a catalectic trochaic dipody). The verse __ vv_vv ___ v __ , found in Bacchyli- des and in the Lille poem (str.lant.5: 201 [with initial x], 215, 222), is regarded by lrigoin as an enoplios and trochaic dipody, not a prosodi­ ac and catalectic iambic tripody (rufulianum ~_v_~::9.IO The decisive factor for the choice of the one or the other analysis is to be found in lrigoin's theory of avoidance of word-end at the end of a metrical unit and in the concept of synapheia and non-generalized, movable cae­ sura. This aspect of lrigoin's theory has produced some controversy. The main point of disagreement has been the doctrine of synapheia, II which, according to critics, is undermined by two shortcomings. First, the anceps is denied the status of a link-unit; as a result, dactylo­ epitrites are analyzed into trochees, enoplians, prosodiacs, and so on. Second, there is a lack of conclusive evidence for lrigoin's thesis that the alternative metrical analysis of identical verses depends on the pattern of word-division, quite simply because he has not proved sufficiently that word-division is avoided between metrical elements. The assumption cannot be proved from dactylo-epitrites because it is seldom certain where element-end occurs. Thus lrigoin's argument tends, as Dale and Parker conclude, to be circular, for it would have been necessary for lrigoin to identify the metrical units before he could pass judgment on the pattern of word-division; but the identifi­ cation of metrical units in the dactylo-epitrites is close to impossible because the anceps functions as a link-element that cannot be de­ tached from the one unit and attached to the other. It belongs to both simultaneously, and this makes it difficult to mark the end and the beginning of two consecutive metrical units; as a consequence, it defies every effort to distinguish the pattern of word-di vision between the units. Both Dale and Parker deny that word-division can be regulated by a law, and believe that the avoidance of diaeresis be­ tween the longs in ... v_, _v ... , for example, is due either to "the 10 lrigoin 22f. II Dale, Lustrum 25f; L. P. E. Parker, "Some Recent Researches on the Versifi­ cation of Pindar and Bacchylides," BICS 5 (1958) 13-24, and "Greek Metric 1957- 1979," Lustrum 15 (1970) 91f. TSITSIBAKOU-VASALOS, EVANTHIA, The Meter of the Lille Stesichorus , Greek, Roman and Byzantine Studies, 28:4 (1987:Winter) p.401 EY ANTHIA TSITSIBAKOU-Y ASALOS 405 necessity of fitting in words containing consecutive long syllables in this, the only available place in certain types of metre" (Dale, Lus­ trum 26) or to "the poet's taste and his aesthetic sense."12 The most impressive aspect of Irigoin's theory, however, concerns reversal of rhythms. Although it is passed over completely by Parker and only fleetingly mentioned by Dale, this theory is of great impor­ tance for us, for it provides the basis for Ancher's notion of Stesicho­ rus as a melodic innovator. lrigoin illustrates his theory of the rising and falling rhythmic movement and rhythmic reversal (41-44) with two verses: __ vv_vv ___ v_ and _v ___ vv_vv_~_v_. The first type can be analyzed as prosodiac and iambic dipody ( __ vv_vv-=::' -v-),in which case it is of ascending rhythm. But if this very sequence is analyzed as enoplios and catalectic trochaic dipody ( __ vv_vv~ '_v_), the rhythm reverses: while in the beginning it is ascending (enoplios), it is then transformed, thanks to the enoplios, to descending rhythm; the other type one might expect-namely, the prosodiac and rufulianum (of ascending rhythm)-is not found (lri­ goin 41). This is of importance for our analysis of the Lille poem because this specific sequence is found there (str./ant.5, ep.2) and is characterized by Ancher as enoplios and trochaic dipody.13 A second type of verse (_v ___ vv_vv ___ v_) that lrigoin (20f) finds common in Pin dar can be divided as: (a) trochaic dipody, hemiepes feminine, trochaic dipody catalectic under the form of cretic, in which case the rhythm is dactylo-trochaic with catalexis, and therefore descending; (b) cretic (an acephalous iambic dipody), pro so­ diac, iambic dipody, in which case the rhythm is anapaesto-iambic with acephaly, therefore ascending; (c) more rarely cretic, enoplios, cretic, in which case there is a reversal of rhythm: from ascending with acephaly it becomes descending with catalexis. Although the sequence of Zonga and brevia is identical for all three types, the produced verses are distinct from the viewpoint of rhythm, lrigoin maintains, and their raison d'etre is to be sought in the poet­ metrician's desire to create a rhythmical variety, compensating thus for the monotony of the dactylo-trochaic series. The reversal of rhythm is accomplished with the aid of the cretic, enoplios, rufuli­ anum, and hemiepes. The mechanism of the rhythmical reversal is ap­ parently activated by the presence of a central common value (lrigoin 42-44) in the first two metrical units (i.e., the cretic and enoplios), 12 Parker (supra n.ll), BICS 24, Lustrum 92. 13 Italian analyses (see 421f infra and n.57) present these as prosodiac and reizianum (~_v_~), which, externally at least, seems to be the equivalent of rufulianum. TSITSIBAKOU-VASALOS, EVANTHIA, The Meter of the Lille Stesichorus , Greek, Roman and Byzantine Studies, 28:4 (1987:Winter) p.401 406 THE METER OF THE LILLE STESICHORUS while it is outside the meter in the last two (i.e., the rufulianum and hemiepes). An example of the first type of reversal with central common value is to be found in the verse_vv_vv ___ v ___ vv_v v - from Pindar Pyth. 4.18 (ep.l), avla T' clVT' ipfTP.iiw alcppov~ Tf vwp.acroLCnv (LfAAO'JTOaa~: hemiepes feminine (descending rhythm), cre­ tic (element producing the reversal), prosodiac (ascending rhythm). It is the short v of the cretic that makes the reversal of the rhythm possible because it is common to both halves of the verse, which are symmetrical in relation to this central value. From the beginning of the verse up to the central value (the short v of the cretic) included, the rhythm is descending. From this central value to the end of the verse, the rhythm is ascending, and we find the same metrical sequence (i.e., the same alternation of short and long syllables, but reversed, according to lrigoin 42). The common central value, consequently, functions as a pivotal point around which the reversal of rhythm takes place: ascending ...---... _vv_vv ___ v ___ vv_vv_ ~ descending A similar phenomenon is noticed by lrigoin (43) in the enoplios: descending ~ ~_vv_vv_~ ~ ascending The central value belongs to both halves and reverses the rhythm. The pivot in this mechanism of ascent and descent is the central common value, because it enables the cretic to hold the place of a trochee and iambus, and the enoplios that of an anapaestic dipody and dactylic dipody. A comparable procedure is claimed to be at work in the case of rufulianum and hemiepes, with one difference; the central value is not common, not included in both halves, but is isolated in the middle, excluded from them. Its isolation allows the ambivalence of the surrounding metrical elements to emerge. The reversal of rhythm occurs, therefore, at either side of this central value: descending ~_Ivl_~ --- ascending and ascending ~ _vvl_lvv_ descending This allows the rufulianum to hold the place of an iambus and a trochee, and the hemiepes that of a dactyl and an anapaest. The TSITSIBAKOU-VASALOS, EVANTHIA, The Meter of the Lille Stesichorus , Greek, Roman and Byzantine Studies, 28:4 (1987:Winter) p.401 EVANTHIA TSITSIBAKOU-VASALOS 407 ambivalent rhythmical value of these four elements, obtained through the artifice of the central value (either "commune" or "hors metre," according to lrigoin 44), assures rhythmical reversal in all the verses where these four elements are inserted, whether placed at the begin­ ning or at the end of the verse. To the extent that they apply to the Lille poem, it must be noted that certain ofIrigoin's theories are obscure. It is not immediately clear, for example, why the prosodiac is an ascending rhythm, when he maintains (42) that the type (b) of the verse _v ___ vv_vv ___ v_ (cretic, prosodiac, iambic dipody) is anapaesto-iambic with acephaly, of ascending rhythm. There are several possibilities here. First, lrigoin views the initial cretic as an acephalous iambic dipody (P:::~· ]_v_), in which case the initial rhythm is ascending and, for unspecified reasons, extends to the following prosodiac; or he regards it as one of the four elements that produce a rhythmic reversal, in which case the ascending rhythm starts from its pivotal point, the medial short, and affects the rhythm of the prosodiac, which becomes ascending. Alternatively, Irigoin takes the prosodiac itself to be of ascending rhythm that continues the ascent already begun with the initial cretic. The latter seems to be true, since he considers the prosodiac "souvent, equivalent a une tripodie anapestique," and thus, independently of its environment, an ascending element equivalent to vv_vv_vv_. lrigoin also regards the hemiepes, - v v - v v -, under certain circum­ stances, not as an element in rhythmical reversal, but as an acephalous prosodiac, that is, as an acephalous anapaestic tripody of hidden rising movement. 14 The origin and rhythm of the prosodiac-a prominent metrical sequence in Italian analyses of the P.Lille poem-remain uncertain. Genetically the prosodiac is consistently associated with the dactylic metron. 15 Its rhythm is seldom discussed explicitly with the exception of Dain (who, like lrigoin, takes it to be ascending) and of Dale (who defends its descending rhythm).16 Koster compares the prosodiac the first syllable of which is long, with a catalectic anapaestic dimeter, or paroemiac, the first foot of which is a spondee,-I-vv-vv-v:--vv_ 14Irigoin 18 n.3; cf Dale's criticism (Lustrum 25f): "when there are further hints of 'acephalism' which disguises rising as falling movement, so that _vv_vv_ can be 'anapestique, avec acephalie' (what have anapaests to do with dactylo-epitrites?), realities seem to have evaporated in tenuous subtleties." 15 D. S. Raven, Greek Metre (London 1962) Index A, 91; Dain 37 §41 and 39 §43; cf 113f §§175f regarding Irigoin's thesis; Koster 57f; cf B. Snell, Griechische Metrik (Goettingen 1962) 31 n.l, on its origin in an "Urvers." 16 Dale, LM 17 and ch.IO passim, esp. I 58f, and Lustrum 25; cf M. L. West, Greek Metre (Oxford 1982) 199. TSITSIBAKOU-VASALOS, EVANTHIA, The Meter of the Lille Stesichorus , Greek, Roman and Byzantine Studies, 28:4 (1987:Winter) p.401 408 THE METER OF THE LILLE STESICHORUS I vv __ ; but he does not specify if this resemblance in meter entails a resemblance in rhythm as well. It is not immediately clear if what is characterized as "dactylic" in origin and, implicitly, as descending in rhythm, becomes automatically ascending if characterized as anapaes­ tic. This phenomenon appears in the case of vv_vv_, which can be analyzed as acephalous hemiepes,17 although it is identical with an anapaestic metron (or dipody). It is undeniable that the fluctuation in terminology may affect the characterization of a rhythm; the form Y of a hypothesized rhythm can be analyzed as Z of an antithetical rhythm, depending on the presence or absence of the initial or final anceps. What appears to be an anapaestic meter becomes, with acephaly, dactylic, and vice versa; a dactylic meter with procephaly or anacrusis18 appears as anapaestic, while their rhythms change respec­ tively. The analysis depends, in each case, on what each analyst considers to be the basic unit. 19 In general, rising meters are con­ sidered those that begin with x- or vv _.20 The anapaestic and ascending prosodiac, consequently, plays an important role in lrigoin's theory of rhythmic reversal. In the scheme- v v _ v v - - _v ___ vv_vv_ (hemiepes feminine, cretic, prosodiac) the initial descending rhythm is said to reverse into ascending in the middle of the verse, at the single short of the cretic. In the middle of this verse we notice the strange scheme ... ___ v ___ ••• , in which the central single short seems stifled among the surrounding tonga. One may wonder, without help from music or phonetics for its eluci­ dation, how this isolated single short, of almost fleeting, elusive pres­ ence, can counteract the impressive heaviness of its immediate en­ vironment-how it can overturn the imposing, weighty, and 'drag­ ging' tonga, the succession of which in certain instances has justified the description 'dying fall'. It is not clear, to speak metaphorically, how this weak, single short can shake off the load of the preceding, and 17 So Dain 37. 18 For anacrusis see Koster 21; Dale, Lustrum 23. Cj, however, the plea of Haslam, QUCC 15f n.16, "against the current term 'rising' or 'lyric dactyls', 'steigende Dak­ tylen', as applied to such structures [i.e., the dactylo-anapaestic runs of the Geryoneis and the Suotherae]. The name is question-begging, insofar as it carries the impli­ cation that the basic unit of composition is _vv (or a compound thereof), so that such lines as ~_vv_vv __ (Ger. str.l) are conceived of as being docked of an element at the beginning-the ancient concept of acephalea-or as having in some wayan introductory biceps that does not really count-the modem concept of metrical anacrusis. n 191t is incomprehensible, e.g., why Dale characterizes the enoplios (xDx) as ascending, and the prosodiac (x D) as descending. She admits, nonetheless, the arbitrary character of the terms. 20 West (supra n.16) 199; Dale, LM 23 n.l. TSITSIBAKOU-VASALOS, EVANTHIA, The Meter of the Lille Stesichorus , Greek, Roman and Byzantine Studies, 28:4 (1987:Winter) p.401 EVANTHIA TSITSIBAKOU-VASALOS 409 to a greater extent the succeeding, longa to produce the claimed reversal of rhythm. Finally, one may ask why the internal rhythmical reversal that Iri­ goin discerns within the four elements referred to above, along with the internal rhythmical tension created by the successive ascent and descent (which is a domestic affair, so to speak), should be transferred outside its colon-limits and influence the rhythm of the adjacent cola and, eventually, the rhythm of the entire verse. 21 Acceptance of this theory produces in practice a complicated situation. If we deal with a sequence like that of P.Lille str.lant.4fDxDxe-, or e1se-uu-uu-~­ u u _ u u - ~ - u - - (hemiepes masculine, enoplios, and trochaic dip­ ody, in the terminology of lrigoin and Ancher), we should expect-if we decide to apply their theories to the letter and take into consideration both the rhythmic reversal of the four elements and that produced in the entire verse-to find a successive breaking and reversal of rhythm, since both hemiepes masculine and enoplios dis­ playa dual action: they reverse the rhythm of the verse while at the same time incorporating in themselves their own internal rhythmic reversal. The verse DxDxe- may then look like this: ascending descending _uul_l.:::;----u _uu~ _u __ -------- ~ '----"" descending ascending descending ~/ or:/~ ~ hem. rnase. enoplios troeh. dip. It is a confusing picture rendered more so by inconsistent use of the notations for ascent and descent. 22 To conclude, lrigoin's theory of reversal of rhythms rests on a tenuous presupposition that the alleged rule of absence of diaeresis (existence of synapheia, avoidance of a generalized caesura and use of a movable non-generalized caesura instead) is operative and valid. The effort to accommodate this theory and prove its applicability re­ sults in a fluidity in the metrical structures as well as in the proposed rhythmical tendencies. The cretic _u_, for example, is analyzed either as iambic dipody with acephaly, of ascending rhythm by necessity, or 21 For the terminology see Irigoin 12; for the use of 'period' and 'verse' interchange­ ably, see Haslam, QUCC 10; cf GRBS 34: "periods (or verses, in the continental European terminology)"; Parker, BICS (supra n.11) 13: "Verse, the 'minor period' in lyric (as French overs', German 'Vers')." 22 In the form c~ the symbol ___ marks the descent (trochee) and--the ascent (iambus), while in ~ __ v~~~ --marks ascent (anapaestic dipody) and---descent (dactylic dipody). Similarly in Slvl~. TSITSIBAKOU-VASALOS, EVANTHIA, The Meter of the Lille Stesichorus , Greek, Roman and Byzantine Studies, 28:4 (1987:Winter) p.401 410 THE METER OF THE LILLE STESICHORUS as trochaic dipody with catalexis, of descending rhythm in this case. In one instance, type (c) of _v ___ vv_vv ___ v_, it is said to be both in the same verse: acephalous iambic in the beginning (ascending) and catalectic trochaic at the end (descending). The rufulianum xex or ~_v_~ is subjected to the same arbitrary treatment: depending on how the preceding colon is analyzed, it becomes either trochaic dipody, with its initial x adjunct to the previous metrical colon, or iambic tripody catalectic. A characteristic example appears in the enoplios and trochaic dipody (xDxe-), ~_vv_vv_~_v __ (a verse also found in the P.Lille poem), which displays the alleged reversal of rhythm: ascending in the beginning becomes descending after the middle long of the enoplios up to the end. If however we analyze this verse as prosodiac and rufulianum, as the Italian scholars do, the rhythm must be ascending, according to Irigoin, because the meter is anapaesto-iambic. Clearly this theory has no permanent and stable basis, for its structures and rhythms are dependent on too many variables, above all on the metrician's divisions into cola. But these very cola are in dactylo-epitrites indistinguishable because of the nature and function of the link element. There is a pervasive subjec­ tivity and arbitrariness in the colon-division and the criteria on which it is based, as well as in the interrelated and interdependent reversal of rhythms. Moreover, two unknowns undermine the theory: first, our uncertainties regarding the pronunciation of ancient Greek, which render us unable to verify whether or not a given metrical pattern corresponded to an analogous sound-pattern that would have carried through the feeling of ascent and descent; and second, our ignorance of ancient Greek music. We are therefore in no position to confirm the claim that reversal of rhythm within a verse is important, because the possibility (on the performance level) of interpreting differently the same metrical sequences introduces an important element ofvariety.23 Keeping in mind these difficulties in Irigoin's theories, we may tum to Ancher's analysis (312) of the P.Lille meter, which is based on the papyrus colometry and time values: Str./ant.: 1. hemiepes feminine+hemiepes feminine (3+ 3) 2. hemiepes feminine+trochaic dipody (3+2) 3. hemiepes feminine+hemiepes feminine (3+3) 4. hemiepes masculine 5. enoplios+trochaic dipody (4+2) 6. hemiepes masculine 7. rufulianum+trochaic dipody (3+2) 23 See Dain 113. 6 5 6 3 6 3 5 34 TSITSIBAKOU-VASALOS, EVANTHIA, The Meter of the Lille Stesichorus , Greek, Roman and Byzantine Studies, 28:4 (1987:Winter) p.401 EVANTHIA TSITSIBAKOU-VASALOS 1. hemiepes masculine 2. enoplios+trochaic dipody (4+2) 3. trochaic dipody+trochaic dipody (2+2) 4. hemiepes feminine+hemiepes feminine (3+ 3) 5. rufulianum 6. hemiepes feminine+hemiepes feminine (3+3) 7. spondaic measure+trochaic dipody (2+ 2) 3 6 4 6 3 6 4 32 411 Regarding the status of str.lantAf, 6f, ep.l f and 4f, Ancher considers various alternatives-for instance, whether or not these lines can be conjoined. If we conjoin them, he maintains (313), we may avoid the "'trou' rythmique" that is created when a catalectic element (the D is taken as the catalectic form of the hemiepes feminine) is followed within a period by an acephalous element, namely, the enoplios or rufulianum. Consideration of the above metrical sequences as com­ plete forms eventually makes this alternative unnecessary (316). Ancher also explores the implications of the contact of two anceps syllables in epAf, namely, the final anceps of the second hemiepes feminine (4) and the initial of the rufulianum (5).24 Taking the seman­ tic continuity into consideration, he tentatively conjoins epA+5 into a single verse with one anceps, the final of epA, although he recognizes the value of anceps iuxta anceps as indicative of period end.25 The non­ generalized caesura allows the final anceps of ep. 4, if it is long or single short, to move to the head of ep.5; but if it is double short, the anceps splits: its first short goes with the preceding, its second short with the succeeding line (in my scheme, epA:vlv:ep.5). EpAf is analyzed as two hemiepes feminine and a trochaic dipody (314), while ep.5 retains the appearance of the rufulianum.26 This supposition introduces enjambment into the poem and creates the need for supplements capable of accommodating the postulated dis­ location or split of anceps. Ancher (298) in fact supplements 207 (epA) with lLfdpa[L l1AAO Kat l1AAo, thus creating a bisyllabic anceps l1AAOI8£OL, in contrast to the restorations by Parsons and Haslam l1AAovI8£OL, in which the last syllable of l1AAoV becomes long by position. The same 24 What the French metricians designate "rufulianum," i.e., catalectic iambic trip­ ody, is tenned "reizianum" by the Italians, "pentad: iambic trimeter segmented at caesura point'" by Haslam, GRBS 56, "penthemimer" by West (supra n.16) 198. 25 B. Snell and H. Maehler, Bacchylidis carmina (Leipzig 1970) xxx: "est autem finis periodi certus ... cum duo ancipitia interposita se excipiunt.'" 26 Cf the criticism of R. Pretagostini, "Sticometria del Pap. Lille 76 a, b, c (11 nuovo Stesicoro)," QUCC 26 (1977) 56, and Palumbo-Stracca 41. TSITSIBAKOU-VASALOS, EVANTHIA, The Meter of the Lille Stesichorus , Greek, Roman and Byzantine Studies, 28:4 (1987:Winter) p.401 412 THE METER OF THE LILLE STESICHORUS procedure is observed in 228 (ep.4); the supplements crawcrfL,