I thought it would be helpful to readers of the BMCR if I responded to Mr. Blankenborg’s review of my book, The Dance of the Muses: Choral Theory And Ancient Greek Poetics. The thing was a life’s work (so far), and naturally there is a good deal of indefensible defensiveness in such cases; but all the same, I believe I can see my way clear to making some points that could be useful to my reviewer and to this readership.
The fundamental criticism is that the book exhibits a circular reasoning. This was in fact the original criticism of W. Sidney Allen’s ‘discovery’ of a stress pattern in ancient Greek, leveled by M. L. West. Both Mr. West and Mr. Blankenborg seem not to have thought through the nature of descriptive accounts of phenomena, whether in linguistics, where the results have of course been admirable, or in description generally. One intuits a pattern in the phenomena; one then looks for it and for evidentiary consequences of it. The more and more disparate phenomena that seem to answer to a proposed rule, the more persuasive the rule will be, to the users, academic or otherwise, who validate it. The process and the argumentation are therefore necessarily circular. (The circularity can be seen in the very concept, ‘descriptive rule’, if one unpacks it. Obviously that does not mean that there are not descriptive rules, or that our grammars are disqualified. The circle is indeed a divine figure.)
What distinguishes genuine descriptive accounts from normative accounts, masked or otherwise, is demonstration, not by logical deduction—where a charge of ‘circular reasoning’ would certainly carry its necessary weight—but by example. Individual and specific example is the only recourse for a descriptive account, and there is no substitute for judgment in this kind of analysis. But analysis it is. To read Mr. Blankenborg’s review, one would think that there were no such demonstrations in the course of my argument. The new theory of the Greek accent, after its historical and synchronic exposition, is ultimately demonstrated by samples of epic and lyric poetry, analyzed by quantitative patterns overlaid by positions of stress, which disclose, for the first time in modern history, that they are musical. Ancient Greek, alone of all known languages, living or dead, is supposed to have displayed no relation between its prosody and its poetical performance texts. This has not made modern classical scholars shy, however, when they interpret performance texts from Homer to Euripides. Unfortunately there are no set criteria for what constitutes a ‘musical’ pattern, but the charts I drew up in graduate school for this book seem to show something obvious, and more importantly, something compelling. I am no Latinist, by the way, but it turns out that there is a new theory of the Latin accent as well (see my pp. 75-9), which is also based on the analogy of the Vedic svarita and so buttresses convincingly the only sort of demonstration possible in all such descriptive claims: a breadth of cover for the account shown by individual and specific example. My account vindicates those ancient grammarians who described the Latin accent in tonal terms. I hope that readers will look to these demonstrations, unmentioned by the reviewer, if they wish to look to the heart of the matter. Naturally I would welcome discussions of the technical aspects: this review says virtually nothing about the linguistic substance of my argument.
Mr. Blankenborg lists a number of ‘assumptions’ that he claims to be mine, before asserting that he subscribes to none of them. He does not argue for any counter-assumptions, presumably because one cannot in principle do so. But all the same, to review an argument critically is not simply to say, ‘I disagree’. I thought it would be instructive if I worked through his examples, even given what I have said above about the process of description. One wonders, is there anything but a political difference between hypotheses and ‘assumptions’? Descriptive accounts use only the former, and genuine ones advertise themselves accordingly. Let me stress at the outset: Mr. Blankenborg is completely consistent in a) paraphrasing me inaccurately and misleadingly, b) never giving citations so that the reader can check for herself, and c) never once showing how or where I use any of these alleged assumptions as conclusions.
1) ‘Rhythm is bodily movement.’ This is Mr. Blankenborg’s simplification, not mine, of a direct and absolute claim of Plato’s, cf. Philebus 17c-d (my translation p. 29). Hence it constitutes the closest thing possible to an informant’s data available to us in a descriptive project on ancient Greek usage. Data from authoritative native witnesses are not ‘assumptions’.
2) ‘the caesura is an audible pause’. This is nowhere claimed by me, let alone assumed by me. Stephen Daitz doubts this. I think the solution is different for bardic performers of catalogue poetry, and for rhapsodic performers of Homeric poetry.
3) ‘a modern dactylic dance is to put on a par [sic] with the Homeric hexameter’. This is not an assumption, but as the reviewer himself indicates, a comparison, which in my view illuminates the things compared. In the absence of any historical or contemporary parallel for the unusual structure of the epic hexameter, let alone the Homeric hexameter, a dactylic dance from Greece, indeed the national dance of modern Greece, known to be at least two thousand years old, which exhibits the trochaic caesura and the bucolic diaeresis as integral physical moments, ought to get the attention of a descriptivist. Having so done, perhaps it will also get the attention of BMCR readers.
4) ‘ictus is an audible beat’. Ictus means‘beat’. This is not an assumption. I cannot even formulate a counter-assumption. The burden is on those who would find some other sense for so blunt a word as ‘ictus’, than ‘beat’. Inaudible beats belong to Zen metaphysics (and perhaps Dutch Classics). The only reason that we in the recent world have not looked for ‘the beat’ in ancient Greek poetry, as we would in every other conceivable form of music and musical performance, was because we could not see the connection between Greek prosody and Greek metre. Problem solved, IMHO. Have a look.
5) ‘pitch equals stress in the foot’s thesis (as in Latin)’. I”m not sure what this means, but as I understand it, it is quite obviously false for Greek. I present a theory of how and when for Greek change of pitch results in stress. Obviously I do not apologize for any complexity in my description, or any sophistication in Greek poetic texts. The reviewer’s claim, attributed to me, is as false about Latin as it is about Greek. Music depends on counterpoint. Rhythmic counterpoint—syncopation, more familiarly, stressing the arsis—depends on an established pattern, which is established by (usually) stressing the thesis. But the harmonia of the iambic trimeter, for example, depends on a regular mid-line syncopation, cadencing on the arsis of the third foot.
6) ‘Aristotle’s remarks concerning the hexameter deal with the rhythmical pause and the division of the period in hemistichs’. Mr. Blankenborg in his review erroneously refers us to the Rhetoric. The text is Metaphysics 1093a29-b1. Read it. It says what it says. There is extensive argument in the modern era that attempts to make it say something different. But my argument is descriptive not normative, and Aristotle’s statement is not subtle. Neither Aristotle nor my reading of his passage says anything about ‘rhythmical pause’ or ‘hemistichs’. There is some ambiguity, as I mentioned above, about a dynamic pause at the caesura, but this passage from Aristotle describing a division in the hexameter does not point to the caesura as the point of division. Earlier in the review, Mr. Blankenborg repeats a persistent misinterpretation that I took pains to correct in my book (see p. 97): he says, apparently intending to paraphrase me, ‘[i]n dance, the first hemistich starts from the left foot, the second from the right.’ This is exactly backwards, about a key part of my reconstruction. Aristotle actually says that the epos ‘is stepped on the right with nine syllables, and on the left with eight’. He divides the hexameter between the first three feet and the last three. Since there is almost never a word division at this location, it is impossible that there was a pause here in mid-word. Neither does Aristotle describe these divisions as ‘half-lines’. In point of fact, such a description would be highly misleading, as the second segment does not parallel the first: it contains the bucolic diaeresis.
7) ‘all early Greek poetry took the shape of a catalogue’. This is not true and was never claimed to be so by me. I point to the fact that the earliest recorded hexameters record epigrams, not catalogues. What might be worth thinking about is the likely oral nature of catalogues and their transmission—paralleled not by oral improvisatory storytelling, but by oral genealogies—in contradistinction to the choral nature of Homeric and other hexameter poetry, recorded in texts.
8) ‘every syllable in (lyric) poetry represents one step of the dancer’. This is openly presented by me as a speculative assumption; it is perverse to suggest that it is an unquestioned one. Where does Mr. Blankenborg suspect that I use it as a conclusion? When I apply it, I explicitly refer to it as an hypothesis (see p. 259). Obviously I feel that there is a descriptive and an interpretive payoff—the only possible kind of demonstration in this context—including significant implications for the editing of lyric texts (p. 233). But there can be no substitute for judgment. I conclude: ‘[t]o assume one syllable one step for lyric is the only way to recover the dance from the words, because it turns the sequence of syllabic quantities into a choreography. That it is the only way, regardless of whether it is the right way, is small comfort to a truth-seeker. But perhaps the economy of expression that this assumption implies, where the lyric texts preserve not just words but choreia, was all one with the imagination of the culture and the poets who produced them.’ (pp. 235-6)
9) ‘as a result of tradition (still visible in the syrtos), the dactyl is the fundamental element of all lyric rhythm’. This is a highly inaccurate report of my reasoning. The suggestions I do make about the generative significance of the dactyl are purely descriptive and synchronic, based on the extant forms of Greek lyric metre. They have absolutely no basis in tradition or the survival of dactylic dancing in modern Greece. For a discussion of the seminal relationship between dactyl and cretic in the generation of all the major lyric forms, see my pp. 236-245.
10) ‘dactylic rhythm is strictly isochronal’. Again, the ‘strictly’ is all Blankenborg and no part David. It is disconcerting, to say the least, when, repeatedly, words are put in one’s mouth and claimed to be one’s own ‘assumptions’. Of course there are ancient descriptions of ‘irrational’ dactyls in lyric. But my authorities for the isochrony of the (presumably) epic dactyl are Plato (Republic 400b) and Antoine Meillet (see p. 158).
Mr. Blankenborg then appears to sophisticate his review by distinguishing between the above ‘assumptions’ and a second set also found to be unacceptable, which, he claims, ‘are not presented as conclusions elsewhere in the study’. (Oddly, the assumptions that are supposed to be presented elsewhere as conclusions, were said to have been done so ‘simultaneously’.) The distinction is entirely rhetorical and without merit or basis in fact. The reviewer is on the trail of circles, and does not appear to understand how assumptions are supposed to sound, or how descriptive accounts are supposed to persuade. I shall discuss these because they are in fact fascinating talking points:
1) ‘D[avid] states that the epic bodily movement resembles the movement of the planets’. The visible outer planets (Mars, Jupiter, Saturn) exhibit a constant circular movement in the ecliptic, but with periods of retrogression tied to their oppositions to the sun (as seen from the earth). The regular retrogression is also the most distinctive feature of the syrtos (along with its dactylic rhythm). Perhaps this is why Plato himself refers to the orbit of a planet as a choreia (Timaeus 40c). At any rate, Mr. Blankenborg’s objection is to Plato’s usage, not mine. I myself speak of an ‘epic movement’, not an epic bodily movement; I demonstrate the influence of the shape of this movement on all levels of Homeric composition, including the narratological.
2) ‘the focus on rhapsodic presentation (rather than dance) of Homer can be seen in the staff of Athena/Mentes and Chryses’. Again, this cannot be construed to be an ‘assumption’, by anyone in command of the word. It is a passing suggestion, clearly parenthetical to my argument—credited to someone else!—that the reader may take or leave. See p. 143.
There then are listed some statements which, confusingly, are described as conclusions—again, completely without warrant from my actual presentation, where they are nothing of the sort. Perhaps the reader also smells a rat. A critic of an argument does well by his readers in pointing to assumptions that appear to become conclusions; it lends him credibility to point also to assumptions that do not become conclusions; but what on earth is a conclusion without assumptions? Game’s up, I think. The reviewer has circled himself. These also are worth taking up:
1) ‘Without counter argument, Aristotle and Dionysius of Halicarnassus are said to err on matters of prosody’. I do not recall ever making this claim, as a ‘conclusion’ or otherwise, about Aristotle; naturally his whole corpus (including the Problems) is a primary source. As for Dionysius, I show that this late author was not describing accents, but interpreting accent marks. The descriptive acuity that resulted in this diacritical system, where with great economy only the pitch rise in the voice was marked, appears to have led Dionysius to conclude that rising pitch was the only accentual feature in Greek. The more ancient of the Greeks, however, clearly refer to two features—not just ‘sharp’ ( oxus), but also ‘heavy’ ( barus). I identify these two features in the new theory. It is fair to say that Dionysius’ misinterpretation of the accent marks which he received in texts has thwarted the study of Greek prosody for two millennia.
2) ‘Demodocus (in the Odyssey) is repeatedly seen as a performer who adapts his words to the dance performed to his music’. This is relatively accurate, except that it is misleadingly put. The dance is not performed to the music; rather, the (word-) music is performed to the dance. As I often say, the dance comes first. There is also no question of a kind of living adaptation of song to a changing dance. The epic dance was a circle dance with a fixed and constantly repeating pattern—an hexameter. Many of the curiosities of epic phonology and diction, I claim (supported, obviously, by Pierre Chantraine’s evidence), are a result of adapting Greek to a fixed, endlessly repeating, isochronous and isometric—and hence non-linguistic—rhythm. It is evident from Homer’s text that the Phaeacian boys give Demodocus a ‘back beat’.
Earlier versions of the manuscript submitted for review were hailed by some as revolutionary and likely to inspire research, but by others as heretical. I do hope that Mr. Blankenborg’s attempt at trivialization does not dissuade people from looking for the book in their libraries. I would also be happy to send readers audio recordings of passages, and also a video recording of an Homeric catalogue being danced to an hexametric syrtos. (Some remuneration for postage and media would be welcome.) I can also e-mail audio samples. Such demonstrations are the most pleasing, and for some perhaps the most convincing; but unfortunately they do not fit in a book. Mr. Blankenborg’s review sets unenviable standards for poor editing, carelessness and inaccuracy in reporting another’s words. I urge the reader not to yield to this clumsy and demeaning hatchet job: please read the book for yourself.