Bryn Mawr Classical Review 2004.09.09
Bruno Gentili, Liana Lomiento, Metrica e ritmica: storia delle forme poetiche nella grecia antica. Milano: Mondadori università, 2003. Pp. xvi, 332. ISBN 88-882-420-8. €21.80.
Reviewed by Joel B. Lidov, Queens College and the Graduate Center, City University of New York (Lidov@qc.edu)
Word count: 6375 words
(I apologize for the lateness of this review. References by author's name and date or short title are spelled out at the end.)
Were I to begin by explaining or illustrating the actual method of metrical analysis put forward in this book, readers trained in what passes for the common teaching of metrical analysis in the English-speaking world, indeed in most of the world of Classical scholarship, would quickly remove it from their shopping carts and move on to the next review. But the book's assembly of information, its explanations, and the questions it addresses make it valuable for anyone seeking to engage with the ways of understanding what is, after all, the difference between poetry and prose. It will also shed some light on the rarely acknowledged problems of the common teaching, and in particular on the largely unnoticed deviations from that by M. L. West. Moreover, Italian scholarship has been marked by an enthusiasm for the study of meter as an essential feature of the experience of reading Greek poetry -- a recent Italian introductory text (one in fact more representative of the common teaching than this one) is invitingly titled "Gli strumenti del poeta" (Martinelli) -- and Bruno Gentili has been at the center of this activity, through his editorship of QUCC, his articles, and his coedited conference publications (such as Musica and Mousike). Notwithstanding the double authorship by Gentili and Lomiento (= GL) "Metrica e ritmica" is a summa of the work Gentili has been pursuing for decades (references to his publications occupy 2 of the 14 pages of the bibliography). Although the book does not live up to its claim to present its subject as "un valido sussido per la globale comprensione storico-antropologica della civilità greca" ("Premessa" p. ix), it is a comprehensive overview of poetic form and practice in which everything is subordinated to a single system. In this review I will try to portray GL's account of that system and place it in the context of modern studies.
The first part of the book, "Una revisione storica" (91 pages), contains five general chapters; I will discuss them more fully below. Part II (150 pages) is comprised of eleven chapters devoted to the various meters. Part III (35 pages) covers the recited iambic, trochaic, and dactylic forms in three chapters; I will not discuss these here. GL have not written an introductory work; they assume a reader who is familiar with reading a metrical analysis and aware of general concepts. They also do not present a continuous exposition of the arguments for Gentili's beliefs. For a discursive overview one is better off with his essay "Metro e ritmo nella dottrina dei antichi e nella prassi della 'performance'" in Musica, substantial portions of which are cut and pasted into this volume (paragraphs also show up in Gentili's Poetry and Public). Many of the individual points were elaborated in his earlier articles. Problems of theory and method are taken up piecemeal and partly interwoven, so I will begin by assembling a summary of the theory.
Gentili holds that Damon ("Socrates' teacher," except on p. 197 where he is only his contemporary) was the first to recognize the character of lyric verse when he identified fundamental rhythms based on thesis-arsis (strong and weak tempi) pairs -- these obviously are the feet -- with the two parts normally having either a 1:1 (the genos ison) or a 2:1 (diplasion) ratio (the 3:2 of cretics enters the system after Damon). GL explain that these ratios derive from the "conventional" 2:1 ratio of the long and short syllables whose alternation is the basis of meter. ("Alternanza" here is used to mean that syllables have either of two lengths as alternatives, not to describe a feature of the pattern. In adhering to ancient Greek sources GL have not assimilated W. S. Allen's distinction of syllable weight and vowel length, which he takes from ancient Indic sources; "conventional" covers some of the same ground, not, however, without prejudice.) Particular patterning of these conventional lengths creates a metrical rhythm. The actual syllable lengths produce a verbal rhythm that is distinctive in individual spoken lines. In sung lyric, there is also a musical rhythm that is the product of the contrast between thesis and arsis; this distinction belongs to the notes accompanying the verse. When the metrical patterning is homogeneous, it is easy to see the rhythmical feet that correspond to the music, but often the sequence of longs and shorts is ambiguous. Gentili maintains that this description of verse rhythms represents the uniform system of theory and analysis that the ancients built up from Damon's teaching, that this system is preserved in ancient writings on rhythmics and metrics (including metrical scholia), and that we can learn from those writings a correct understanding of both the meter and the rhythm of the transmitted texts. So "storia" in the subtitle has a double sense: the book claims to be a reconstruction of the practices of Greek poetic composition based on the ancient descriptions of it.
Two essential features of this presentation are GL's separation of the analyses of musical and metrical rhythm and their claim that the ancient system was uniform. Both rest on a radical reinterpretation of the evidence for the common view that there were two doctrines in antiquity: the theory of metrical prototypes, which reduced all verses to sequences of the eight or ten different metra (the distinct metrical rhythms that realize the musical rhythm in Gentili's view), and derivation theory, which described all verse as segments, variously manipulated, of iambic trimeter and dactylic hexameter (I suspect that most of us use the one to describe the first line of the elegiac couplet, the other to describe the second). GL argue that there was no separate derivation theory; that the manipulations attributed to it -- especially transposition, addition and subtraction -- are those of epiplokê, the process referred to (in the A scholia to Hephaestion) in the context of prototype theory to show the family relations of metra of the same length (e.g, choriambic and iambic differ by transposition of the first two tempi). What modern scholarship calls derivation theory was no more than the practice of using epiplokê to explain the origins -- rather than the rhythms -- of the metra within the equal (dactylic) and double (iambic) genos. But epiplokê itself is also an essential part of rhythmic analysis. Gentili holds that the colon, perceived as a dimeter, and the stichos, which is a trimeter or tetrameter, were the fundamental structures of composition, and it is epiplokê that explains the relationships among the different metrical rhythms (represented by the prototypes) that can be combined in one colon or stichos in the realization of a single musical rhythm. In particular, the joining of two metra related by transposition (metathesis, now usually called anaclasis) explains the make-up of a large group of closely related dodecaseme cola (cola of twelve morae or chronoi) in the genos diplasion (where each metron contains two feet of three chronoi -- this is Gentili's reinterpretation of the Vierheber and choriambic dimeter of the German historicist metricians).
In Part I, the first chapter, "Metrica, ritmica, e musica," is largely devoted to discussions of the some of the main points and implications of this theory. It includes an argument that will be crucial to the working method, that the Alexandrian editors had an awareness of the original musical structures, and therefore of the correct analysis of the colometry. GL conclude the chapter by explaining that even though the move away from the subordination of music to words after the latter part of the fifth century led to the separation of rhythmical theory as a subject for Aristoxenus and his successors, from the start the music was always the vehicle for expressing the rhythm of lyric.
In the second and fourth chapters, there is less of theoretical interest, but each shows GL's concern for ancient doctrines. The second, which covers prosody, shows a nice balance between the rules offered in introductory texts and the catalogue of details in West's Greek Metre, but GL adhere to ancient terminology and organization and wind up making the presentation of the factual data discovered by modern scholarship confusing. This is not a problem in the fourth chapter, "Strutture della versificazione," where GL simply review the terminology (mainly from Heph., Peri poiêmatôn) that the ancients devised for all the various larger units or systems of recited or sung verse, from kata stichon to the parts of the comic parabasis. No attempt is made to connect this to their analyses elsewhere in the book or to the role of these categories in modern scholarship. The completeness, examples, and economy of the presentation makes it useful for reference.
Chapter Three -- perhaps the most valuable -- is a glossary of terms (plus a brief explanation of the sigla the manuscripts use for poetry). GL explain the ancient vocabulary of the study of meter and rhythm. They are especially full here, since they are careful to indicate all the various meanings of significant terms in the ancient texts (listing the six meanings of "synthetos" for example, or explaining the different uses of periodo first in the metrical, and then in the rhythmical scholarship) and the various alternative terminologies (names of cola are not included in the glossary but are presented with equal thoroughness in the specific chapters). GL are often helpful about how an ancient term was taken over in nineteenth-century scholarship, and what its fate was there (e.g., logaedico), but they do not systematically explain the terminology of modern theories (which they generally regard as mistaken or unnecessary when they differ from the ancient), and some modern terms are completely missing (e.g., any term corresponding to "overlap" or "dovetailing," although the question of the coincidence of colon and word end is discussed in the book in other contexts). The explanation, of course, assumes that GL's understanding of the continuity and common concerns of the ancient tradition is correct, but major areas of controversy are usually signalled, at least in the notes, by a statement of the error which is to be corrected or of the opinion disagreed with. GL are not reliable as a bibliographic guide to contrary opinions, but even so, by regularly alerting the readers to the existence of disagreement, they go beyond most general books on metrics as a guide to the field.
In addition, many glossary entries are the take-off point for brief essays that expand the presentation of a principle of theory or analysis, especially when GL differ significantly from other modern scholars, so that entry takes on subjects more general than the lemma would suggest. Under anceps GL rehearse their objections to Maas's system for describing the so-called dactylo-epitrite (it is also discussed in the appropriate chapter later on -- nothing else in fact seems to irritate Gentili so much1). Under procephalo, GL make clearer than they did in Ch. 1 the importance of recognizing that changes of metrical rhythm are a typical feature of Greek verse, decrying the persistence of the "obstinate conviction" that the cola of a passage should be homogeneous. These entries take a somewhat adversarial tone, and it is one of the problems, throughout the book, that the choice of adversary can sometimes leave the impression that there only two points of view to choose between. At their worst, GL's criticisms devolve into circular reasoning, as when, under "metron", they criticize Snell's distinction between verses constructed kata metron and kata kôlon on the grounds that the colon, in ancient doctrine, is in fact a dimeter. The argument assumes what Snell was denying, that anything which could possibly be analyzed in terms of metra must have metra as its structural principle and that the colon is a dimeter. But the fullness and thoroughness of Gentili's construction (or, in his terms, reconstruction) of metrical theory can be grasped by putting together groups of related entries: "alogos," "adiaphoros," "anceps," "brevis-in-longo"; "ictus," "arsis," piede; "metron," colon, "periodo," iato, cesura, catalessi. (The Italian style of rendering phonetically most but not all Latin transliterations of Greek terms can lead to some confusion here and in the index; look for hypercatalectic as ipercataletto and for sinalefe but also "synthetos." )
The fifth chapter, "Mode dell'esecuzione," looks at the ways lyric was delivered by song, instrument, and dance, and returns to the issue of the relation of text and song. GL recognize the priority of words to music in the early tradition, but here they mostly concern themselves with the developments which began with the dithyramb and spread, through the New Music of the late fifth century, into the Hellenistic practice of adjusting the words to the musical accompaniment. They then claim that some use of these practices can be retrojected to the general practice of the earlier periods fifth century. It is all too apparent that the argument is constructed a priori (I will return to such difficulties below). The chapter concludes with a survey of dances and instruments. Although one must still turn to more specialist literature (including the chapter on instruments in West's Ancient Greek Music) for more critical discussions, this survey is a very convenient guide and a welcome addition to a book on meter.
We might pause here, at the end of Part I, to deal with a red herring. GL insist that we must treat the texts as evidence for, not records of, the cultural context for which and in which they were presented. It is not just a question of how they were delivered; rather we must understand that they belong to "performance." Performance does not really mean the same thing as "performance." It is a technical term for something peculiar to a culture based on orality, a distinctive praxis, which not only encompasses the delivery (resa) of the song but defines how we should conceptualize that. For GL, however, the emphasis on oral culture involves no particular claim for the role of the audience, as it does, for example, in Gentili's rejection of generic meanings in the odes of Pindar in favor of particularized historical references,2 and it implies no particular mode of composition or context of reception, as it can in Homeric studies. In the end, it merely glamorizes the observation that the text was composed for delivery -- performance -- with musical, and sometimes choreographic, accompaniment. Innumerable writers on the subject have made this point without reference to theories of oral culture, and here it serves only to deflect any questioning of the particular relation that Gentili posits between the text and the music.
The descriptions of the individual meters in Part II show us the analytic method in practice. I should now explain how GL analyze verse, but, again, to forestall any reader's immediate rejection of their project, I will emphasize first the importance of what they seek to accomplish. Rhythmic analysis earned a bad name from the nineteenth-century assumption that all verses should be interpreted by varying the syllable lengths to produce equal musical bars with an initial beat, but simpler forms of it persisted into the twentieth century. Their adherents sought uniform rising or falling movements within cola and even stanzas, and, by positing only prolongation and rests, uniform musical measures in either single or double time.3 A different concept of rhythm supported theories in which the cola were measured off in such a way that counts of theses or of metra would produce symmetrical patterns within a stanza; such methods are, whether acknowledged or not, forms of counting beats.4 At a minimum, many metricians held that the different meters were rhythmically distinguished because they were sung or spoken with different ratios for the lengths of long and short syllables; this was Maas's belief, even though he thought the actual differences were unrecoverable and restricted discussion to the observable metrical sequences.5 George Thomson, on the other hand, abandoned any form of analysis by metrical unit in favor of isolating a small group of cola, used in sequence and often overlapping, as self-evident rhythmical phrases.
Modern discussions tend to follow Snell in treating the metrical units -- metra and cola -- as self-sufficient entities. But anyone who says that there exist both dactyls (l s s [I use l and s for long and short]) and choriambs (l s s l), and that the double-short sequence in a glyconic (x x l s s l s l) belongs to a choriamb and not a dactyl (the position taken by Koster, Dale, Raven, and Korzeniewski, following Wilamowitz, but carefully avoided by Snell) implies a theory that makes the distinction meaningful. Dale (1968) adapted Snell's methods but she nonetheless relied on general rhythmic typology to define the character of the cola and made clear her belief that quantitative variations could establish the actual rhythm of ambiguous cola (1968:8). Her system makes use of a pragmatic laxity: sometimes rhythm (iambo-trochaic) obviates colometry; many cola, in transitions, are rhythmically ambivalent. In any case, as she noted (1969), the system works only for dramatic choral lyric. T. Georgiades offered a defense of Snell's view in its pure form by arguing that differing patterns of always identical longs and shorts compiled into cola of differing lengths by themselves constitute differing rhythms; this was meant and taken as an attack on the notion of a musical rhythm.6 But the absence of an explicit account of rhythm has made modern treatments of colometry look like nothing more than a problem, or game, of notation, dependent on implicit criteria of taste or unacknowledged principles.7 This subordination, or avoidance, of rhythm has been confronted in recent books. M. L. West, radically deviating from the common teaching's reliance on Maas and Snell, reinterpreted Georgiades's definition of the colon as a unit of variable length to make it a vehicle for reasserting the arsis-thesis rhythm of the foot as a beat; he has produced a system very close to that in GL.8 Thomas Cole's Epiplokê, one of the two fullest modern attacks on the problem of explaining colometry in terms of rhythm, adapted Thomson's method by abandoning the ancient tradition of preexisting cola and metra in favor of a small number of continuously repeating rhythmic patterns whose segments, often interwoven, appear as cola in actual verses. Taking a totally different approach, Sicking (following Dale 1969) restricted metrical analysis to the recurrence of just two units, one for the single- and one for double-short rhythm (in isolation, equivalent to the cretic and choriamb), and described all verse as types of repetition of these.
GL also bring the problem of rhythm back to center stage, but with the terms of traditional metrical analysis. Because they insist that the metrical pattern allows us to perceive the metrical and musical rhythm separately, their method avoids the pitfalls that have undone earlier modern theories: it is predictable and consistent; it does not force rhythms into uniform rising or falling patterns; it does not require a metrical ictus or expect temporally equally measures; and it does not impose artificial symmetries of construction on stanzas. Furthermore, the changeful patterns over the course of a dramatic stanza offer an opportunity to appreciate the relation of meter and content and to free metrical esthetics from a dry formalism. And there is no ambivalence: instead of the "gliding transitions" of Snell and Dale, GL's method yields a specific rhythm for every colon; instead of the Snell's theme-and-variation approximations of aeolic cola in choral lyric, it gives every verse a specific metrical-rhythmic shape; and, of course, it has no truck with a "link anceps" isolated from any rhythmic role.
So what, finally, is the method? In a footnote, Dale (1968, p. 69) dismissed Hephaestion's "habit of starting at the beginning of a colon and numbering off in fours." The frequency of eight-syllable (dimeter) lengths has always made some room for analysis by fours, of course, but GL resume the habit as a rigorous method. Each quadrisyllabic unit (with allowances for dactyls and anapaests) is one of the prototypes; each metrical prototype can be understood in terms of the rhythmic feet within it. This method yields the antispast (nominally s l l s) as the first four syllables of the glyconic, and restores the choriamb to its eponymous origin as a dipody (l s + s l) with rhythmic reversal. Ionics a maiore appear as a metron of archaic verse. The system can work because of the variability -- the equivalences -- built into the prototypes. Most prototypes are realized in more than one form, by incorporating truly free positions, or rhythmically irregular ones (alogoi -- a long where a short is expected -- and occasional ataktoi; I will use x for free positions and a or b where an alogos or brevis-in-longo are possible.) The ionic a maiore is x l s s; the antispast is actually x x l s, not s l l s, and consequently l s l s is a perfectly good antispast (the standard objection that antispasts and ion. mai. do not occur serially before the Hellenistic period does not persuade GL, but they do defend their revival of these two metra under "metron" in the glossary and acknowledge some ambiguity with choriambic). Prototypes closely related by epiplokê in the same rhythmic genos can substitute for them in responsion or combine with them in dimeters and trimeters. In the genos ison a dactyl can stand for an anapaest; in the genos diplasion metathesis can allow an iambic s l s l to respond to a choriamb. In all such cases the musical rhythm preserved the necessary uniformity. Furthermore, every (quadrisyllabic) metron has catalectic (three syllable), brachycatalectic (two-syllable), and hypercatalectic (five syllable) forms. (The hypercatalectic monometer is a penthemimer; acephaly and procephaly occur only when the context strictly demands them, since the count always starts from the first syllable). In consequence, any sequence whatsoever can be described in quadrisyllabic units. So the glyconic is an antispastic dimeter acatalectic (x x l s | s l s b, with an iambic foot by metathesis in the last two positions). A choriambic trimeter catalectic (more strictly an epichoriambic form of it) that combines a choriambic metron with both its generic equivalents, a trochaic metron and an iambic catalectic metron -- that is: l s l a | l s s l | s l b -- yields the sapphic hendecasyllable. The alcaic hendecasyllable belongs to the same generic category (cola made of metra composed of two arses and two theses related by epiploke, with a change of rhythmic direction) but its structure is quite different, for it is an ionic a maiore trimeter catalectic, and begins with an iamb: a l s l | x l s s | l s l. These examples illustrate three different metrical analyses of the double short in aeolic verse, and show that what we are used to calling the anceps in the Lesbian hendecasyllables is an alogos when it belongs to a trochee but a free position when it belongs to an ionic.
In Part II each of the first seven chapters (Six through Twelve) is devoted to the cómmata (= lengths shorter than a full dimeter), cola and stichoi formed by one of the metra based on feet of the equal (anapaest, dactyl) or double genos (iambic, trochaic, choriambic, antispastic [= metri gliconici] and ionics, which are apparently included as dipodies of six chronoi). Each chapter lists all of the lengths that occur (monomoters, penthemimers, catalectic dimeters, etc.) and their frequent combinations and associations. Traditional nomenclature is explained. For each type, GL provide examples in chronological order, often including Latin and Italian adaptations. Then Chapters Thirteen and Fourteen on strutture polischematiche and metri κατ' ἐνόπλιον deal with cola marked by much freer substitutions of different metra, so that each gathers in one rhythmic category cola otherwise very diverse. Chapters on cretic-paeonic (hemiolic rhythm except when prolonged to six-time) and dochmiacs conclude the survey.
You can dispute the reasonableness of GL's mode of analysis (the argument would be too long to fit in this parenthesis), but you cannot say it does not work. And reasonableness is not the issue here. The point that cannot be overstressed is that for GL the method is based on authority (as interpreted by Gentili). Some time ago I proposed that it was best to understand all archaic lyric verse in terms of an alternation much like the arsis-thesis contrast essential to their method (Lidov 1988). Gentili would not have regarded this as a step in the right direction of restoring such a rhythmic basis to our understanding of verse. Quite the contrary, as a proposal based on direct observation of the data, he would have seen it as a totally wrong-headed turn away from the ancient sources on meter and music, just as he dismisses Sicking (p. 6 n. 21), while ignoring his reconceptualization of the problem of rhythm and meter. The importance of establishing and then following authority explains why GL insist so strongly on a few points that might otherwise seem quite arguable.
Damon's doctrines are referred to several times; he is the bedrock for the authoritative teaching that GL claim to represent. From the descriptions here a reader would conclude that we actually possess at least a fragment. But in each case the footnotes lead back to Plato, Rep. 400b-c, where Socrates mentions Damon's theories while refusing to elaborate on them. Gentili interprets this passage with the aid of Arist. Nub. 649ff., where Socrates appears to refer to the same teaching. (Gentili's treatment is fuller in Musica than here. The possibility the passage in Aristophanes may be humorous and the one in Plato ironic -- where else are the two Socrateses in such agreement? -- does not arise.) He manages to attribute to Damon not only the distinction between dactylic and iambo-trochaic types that most readers find here, but also the method of characterizing cola as combinations of rhythmically variable metrical prototypes whose effective rhythm is that of the music in performance.
The term monochronos ("single-element") in the Aristoxenian fragment P.Oxy. 2687 (which incorporates P.Oxy. 9; included also in Pearson) unusually signifies not the length of a short syllable but a single syllable lengthened to the extent necessary for the rhythmic context. GL make it a key term, because it validates as part of theory what is recognized by the triseme (etc.) symbols in later musical texts. Since GL take the theory as a guide to the whole poetic tradition, the presence of the term here (along with the accompanying examples) confirms for them that the separation of scansion (metrical rhythm) and musical rhythm and the consequent practice of adjustment between these that they claim were fundamental to Greek practice.
Because of the ambiguities created by the variability of some metra, as well as by the free substitutions in some types of cola, GL frequently return to the theme that the only certain guide we have to the correct analysis of specific lines is the colometry in the ancient sources. Many passages in the individual chapters are selected to demonstrate the value of the colometry transmitted by the manuscripts. But, just as significant are free responsions, the passages where the text of the manuscripts shows responding passages that do not respond. Gentili is strongly set against the practice of emending in these cases. The equation of (to use standard terms) glyconic and choriambic dimeter that we find once or twice in Sappho (and throughout Corinna) is but the most obvious of a whole range of equivalences among prototypes. In Py. 1 str. 6 (p. 206, to take an example almost at random) there are nine instances of ... l l s s l s s l -- i.e., x D in Maas's terms -- but line 52 has ... l l s s l s l l. Nothing here signals an error to Gentili; for the substitution of a trochaic for a choriambic metron following an ionic mai. within a colon does not present a problem. (A fortiori, a mismatch within a single strophic pair in drama can almost never be a metrical problem.) The all but absolute reliance on the authority of the manuscripts and the theory of rhythmically equivalent metra are two sides of the same coin: the theory justifies the texts and the texts show how the theory works in practice. Large numbers of examples are drawn from the critical literature to show that texts deemed metrically misaligned are perfectly acceptable.
This preference for authority over any form of rationalization sets GL apart from other texts that look back to ancient doctrines. Superficially, this book most resembles Koster's Traité (1966), which also begins with an overview of theoretical possibilities (with the added benefit of a short history of modern theories) and definitions of ancient terms. But although Koster's descriptive system stays close to the categories and often the concepts of ancient metrics and uses its terminology by preference, he is imposing his own judgment of what produces the best classification of the data. West (1992), with a reliance on ancient rhythmicians similar to GL's, accepts the antispastic glyconic and the choriamb as a dipody, and his extensive use of anaclasis (using a notation which has found much favor with recent editors, who seem unaware of its theoretical basis) reflects the theory of epiplokê as Gentili describes it, but for his own reasons West chooses not to accept the choriambo-ionic quadrisyllabic measurement of the dactylo-epitrite that follows from it.9
In the end, of course, even GL have to fudge. In the chapter on trochaics, defending a dactylic first foot in the metron (pp. 121-2), GL allow that short syllables could be shortened to adjust to the rhythm of the foot (allowing two to equal one elsewhere), a position unanticipated by the earlier theoretical definitions and reminiscent of nineteenth-century metrics. And GL must also admit that the ancients lacked both a sign and a term for this necessary phenomenon (the shortening also appears to fly in face of the fragments of Aristoxenus's "On the primary chronos" [Pearson 32-35]). In the chapter on meters kat' enoplion the usefully polymorphic reizianum, as defined by Wilamowitz, suddenly makes an appearance and plays an important part, but GL connect it to no ancient source. Perhaps these specifics, like all the details in the book, could be argued this way or that, but they reflect the underlying problem of circular or tendentious argumentation that makes the whole exposition unconvincing. The authors assume that if they can demonstrate that all the ancient material can be understood in a way that conforms to their reconstruction of ancient doctrine, then that must be the correct interpretation of both. Too often, the possibility of usually easier, alternative readings presents itself. The treatment of derivation theory10 and of Damon are only the most obvious. It seems to me that P.Oxy 2687 is evidence for a change of practice, not continuity, since it appears to instruct contemporary musicians in the methods by which traditional cola could be made to conform to contemporary tastes for more uniform rhythm.11 One does not need a continuity in theory or written instructions to explain the colometry of the Alexandrians; they could also have worked from observation and their own hypotheses.
Once we recognize that the theory is constructed from, not found in, ancient writings, it cannot be accepted as the grounds for a priori arguments about the nature of meter and rhythm. One can reject musically-based rhythm and still allow irregularly occurring and unsystematic prolongation. In the study of equivalences among cola, the persistence of syllable-counting cannot be simply wished away.12 I have already mentioned Cole's and Sicking's different ways of incorporating a rhythmic basis. Devine and Stephens (1993, 1994) show that the assumption that meter is a stylization of speech rhythm produces results for speech that are consistent with findings about human speech drawn from comparative linguistics. Their work opens up the prospect of treating meter in terms of the underlying alternation in speech rather than in terms of music, an approach already taken in Nagy's analysis of lyric cola (1996: 101-2 and earlier).
By the time I finished this book, I found myself wondering whether it is really a book about rhythm and meter or even the rediscovery of performance at all. The devotion to justifying the ways of Hellenistic scholarship to modern encompasses all areas of belief. Corinna was an archaic poet; Midas of Acragas broke the mouthpiece of his aulos as he began to compete (otherwise Py 12 makes no sense); the magadis was a type of lyre (pace Barker's contribution to Musica); the cretic was introduced into Greek poetry by Thaletas of Crete; Damon was Socrates' teacher. Readers of Gentili's Pindar commentaries will find all this quite familiar. For Gentili, metrics is just one piece of the argument for a scholarly method based on complete subservience to ancient authorities read with minimal regard for their own contexts and history.
This devotion, however, even if it does not yield directly a more convincing understanding of the meters, does produce a helpful understanding of the terms with which we engage with them. Just as it is often helpful to read interpretive scholia not for the final answer to a problem, but for the material of the subsequent discussions of it, so the ancient methods have left a long spoor in the modern scholarship. GL's thoroughness, and their interest in explicating the ancient terms as they were used, provide one of the best starting points I have seen for entering into the disputes over metrical analysis. The explanations here seem to me fuller and more usable than those in Schroeder, no more distorted than his by the authors' penchant for declaring what is right or wrong, and more attentive to variations among ancient sources. Similarly, their emphasis on ancient colometry allows us to discover the questions that are so often left unstated when modern scholarship supplies answers. It becomes especially clear that we would be better served if modern editions found a way to note the relation of their own colometry to what has been transmitted (as noted by Anne Mahoney in BMCR 2003.06.49) and if, at the very least, editors gave some indication of the ideas, not just the sigla, informing their practices. In his justifications of free responsions Gentili often applies a deep knowledge of the language and the literature, and an outlook that values diversity in expression. Individually disputable, they add up to an argument not for his metrical method, but for restraint in applying any theory or method that hopes to solve problems of text or meter by the application of a predetermined, rigorous rule.
The book is a sturdy paperback, handsomely formatted to present metrical analysis. There are occasional problems of editing13 and some typographic lapses in the schemes that can be puzzled out. It concludes with a generous bibliography of items cited, and three indexes: metrical terms, names (including all modern scholars mentioned in the footnotes), and passages cited (over 50 columns in small type).
Dale, A. M. 1968. The Lyric Metres of Greek Drama. 2d ed. Cambridge: Cambridge Univ. Press.
Dale, A. M. 1969. "The Metrical Units of Greek Lyric Verse, I, II, III." In Collected Papers, 41-97. Cambridge: Cambridge Univ. Press.
Danielewicz, J. 1966. The Metres of Greek Lyric Poetry: Problems of Notation and Interpretation. Pomoerium Supplementa, vol. 1. Bochum: Druck und Verlag von Pomoerium.
Devine, A. M., and L. D. Stephens. 1993. "Evidence from experimental psychology for the rhythm and metre of Greek verse." TAPA 123:379-403.
Devine, A. M., and L. D. Stephens. 1994. The Prosody of Greek Speech. New York: Oxford University Press.
Gasparov, M. L. 1996. A History of European Versification. Ed. G. S. Smith and L. Holford-Strevens. Trans. G. S. Smith and M. Tarlinskaja. Oxford: Clarendon Press.
Gentili, B. 1988. Poetry and its Public in Ancient Greece: From Homer to the Fifth Century. Translation of Poesia e pubblico nella Grecia antica. Trans with introd. by A. Thomas Cole. Baltimore; London: The John Hopkins University Press.
Gentili, B., and R. Pretagostini, eds. 1988. La musica in Grecia. Storia e società. Rome and Bari: Editori Laterza.
Gentili, B., and F. Perusino, eds. 1995. Mousike: Metrica, ritmica e musica greca in memoria di Giovanni Comotti. Studi di metrica classica, no. 11. Pisa: Instituti Editoriali e Poligrafici Internazionali.
Gentili, B. 1999. Polemichetta metrica: "Anceps-biceps" nella Tebaide di Stesicoro? QUCC n.s. 61 (90):89-91.
Georgiades, T. 1973. Greek Music, Verse, and Dance. Translation of: Der griechische Rhythmus. Trans. E. Benedict and M. L. Martinez. Da Capo Press music reprint series. New York: Da Capo Press.
Irigoin, J. 1995. "La composition métrique de la IXe Pythique de Pindare." In Gentili, Mousike.
Korzeniewski, D. 1968. Griechische Metrik. Die Altertumswissenschaft. Darmstadt: Wiss. Buchgesellschaft.
Kitto, H. D. F. 1942. "Rhythm, metre, and black magic." Classical Review 56:99-108.
Koster, W. J. W. 1966. Traité de métrique grecque suivi d'un précis de métrique latine. 4th edition revised. Leiden.
Leonhardt, J. 1989. "Die beiden metrischen Systeme des Altertums." Hermes 117:43-62.
Lidov, J. B. 1989. "Alternating rhythm in archaic Greek poetry." TAPA 119:63-85.
Maas, P. 1962. Greek Metre. Trans. H. Lloyd-Jones. Oxford: Oxford Univ. Press.
Martinelli, M. C. 2001. Gli strumenti del poeta. Ristampa reveduta e corretta. Bologna: Capelli Editore. Originally published in 1995.
Nagy, G. 1996. "Metrical convergences and divergences in early Greek poetry and song." In Struttura e storia dell'esametro greco, vol. 2, ed. M. Fantuzzi and R. Pretagostini. Studi di metrica classica, no. 10, 63-110. Roma: Gruppo editoriale internazionale.
Parker, L. 1997. The Songs of Aristophanes. Oxford: Oxford Univ. Press.
Pearson, L. 1990. Aristoxenus: Elementa Rhythmica: The Fragment of Book II and the Additional Evidence for Aristoxenean Rhythmic Theory. Texts edited with introduction, translation and commentary. Oxford: Clarendon Press.
Pöhlmann, E. 1995. "Metrica e ritmica nella poesia e nella musica greca antica." In Mousike.
Pohlsander, H. A. 1964. Metrical studies in the Lyrics of Sophocles. Leiden: Brill.
Sicking, C. M. J. 1993. Griechische Verslehre. Handbuch der Altertumswissenschaft. Munich: C. H. Beck.
Thomson, G. D. 1961. Greek Lyric Metre. 2d ed. Cambridge: Heffer.
West, M. L. 1992. Ancient Greek Music. Oxford: Oxford Univ. Press.
West, M. L. 1982. Greek Metre. Oxford: Oxford Univ. Press.
1. The disagreement with Maas's system arises from the essential differences of his theory; for a fuller exposition see his "Polemichetta metrica".
2. See T. Cole's explanation in his introduction to Poetry and Its Public.
3. For example, in a notorious article by Kitto and in Pearson's Aristoxenus, esp. xlviii-lv.
4. See Irigoin's contribution to Mousike, the schemes in Pohlsander, and some analyses in Korzeniewski.
5. One can see this by putting together Section 1, p. 1; Section 4, p.4; Section 28, p. 23; Section 31; p. 24; ﾧ80, p. 56.
6. See Pöhlmann's polemic in Mousike. For Georgiades I am using the English version, but the German of 1949 appears fuller.
7. In this spirit, Danielewicz is an excellent short guide to metrics, despite his unacknowledged standards of judgment.
8. The theory first appears in his Greek Metre, despite the general prevalence there of Snell's style of analysis; among the reviewers, only Sicking (Mnemosyne 39:423-32) seems to have noticed West's unexpected revival of older methods. It is fully developed in Chapter 5 of Ancient Greek Music; see the criticism by O. Steinmeyer in his Arion review (3rd ser. 4: 223-236, esp. 229-30).
9. Pp. 151-2; he prefers a 5/4 phrase (= Maas's D), and without explanation defends only his decision not to prolong the final long to make it 6/4 (= 3/2), rejecting also the extra lengthening, which "people generally assume," of the long before the caesura of the elegiac pentameter. Note also on p. 161 his unusual theory of "impacted bars" (perhaps influenced by Cole or Thomson), adopted to preserve his rhythmic system against the actual practice of the poets.
10. They acknowledge M. Palumbo-Stracca as the source of their argument; Leonhardt 1989 deals with it, and gives a fuller picture of the doctrines and their roles.
11. The primary example, iii.15-18 = PMG 926e, divided syntactically, presents two pairs of an ithyphallic and a lecythion. But since all the examples appear to be Dionysiac or more broadly dithyrambic, I would be wary of using them for general rules; invocations of Dionysus seem to be one of the biggest sources of problems in metrical analysis (cf. the fifth stasimon of S. Ant.).
12. Dale's accepted it to explain irregularities (1968: 64-65, 78, 89); Parker's rejection (1997:117) states a principle. Gasparov's (1996) assumption of it as continuing feature illustrates how the history and practice of Greek meter complicate the problem more than is acknowledged.
13. The important discussion under asinarteto depends on the reader recognizing that "tomé" refers to section, not caesura, but the explanatory cross-reference ("vedi cesura") leads only to a discussion of caesura vs. diaeresis.