Approaching the Problem

Interpreting the Sounds

It was in 1996 that I put on my website my first essay on the importance of reading Homer in an authentic mode of delivery. I felt it was important preserve the grace and sheer loveliness of the Classical Greek language, and I was concerned with the traditional academic style of reading Greek verse using Stress in place of the diacritic pitches and indicated vowel lengths. I have been thinking about problems relating to the sound and interpretation of the Homeric poems for many years, and I have now prepared this paper for a University setting where a group of people with a wide array of talents can work on the development of a serious way of doing a Homeric Performance Project . This will be done in a scholarly and interpretative workshop aiming to produce a CD for answering to some degree the question of what a Homeric bard might have sounded like. Many talents are needed for this, since it involves specialized sound recording and editing staff to produce a CD with possible internet MP3 selections, as the results of this group effort. Let me outline the nature of this Project from several points of view.


A :      First an outline-score must be written, in the style of a cinema score with separation of parts and instructions for the acoustic and background rendering of specific sections. Type of voice, quality of intonation, rate and rhythms must be laid out clearly so a singer-chanter can prepare for performance much as a stage actor works up a dramatic part. Certain sections will follow spoken parts of the Iliad, which account for almost half of the text of the Iliad, in an artistic mode of delivery suitable to both speaker and situation. This is especially important since much of the Iliad is a leading thread into the Greek Drama.

B :      Second, a score in standard music notation is to be prepared with tones indicated for acute pitch location and the perispomenic micro-melismata, as well as tempo and dynamics. Along with this score the text of the Greek is to be written as in standard song score, words divided in accord with the notes, and the text printed below with both Greek and Roman letters. This is critical since it may be necessary to use voice-readers who do not know Greek but are suitable for parts in terms of their voice quality and musical training. If there are performers who have at least a basic knowledge of Greek, that would be best; but since one purpose of this project is producing listenable recordings, the choice of a talented singer-reader comes first, and selection must be done with juried try-outs.

C :       The immediate product of this workshop cooperation is to be the recorded sound, publicly available as an introduction and a propaideutic to the art of the Homeric epic. There is much interest in Homer in this age, both in the college classroom and also by private persons learning enough Greek to read the Iliad and Odyssey. Recorded results from the Project will go far to amplify current interest in Greek studies and must be pre-planned for quality and distribution as a serious end in sight.

D :      However the production of a set of recordings in just the first step. The recordings should serve as a learning tool for those studentrs of Greek who want to do interpretative readings on their own. Reading a few pages of the Iliad brilliantly a recording not enough; we want to set forth good examples for continuing in the same vein, so the whole Epic can become a living performable text. For this a commentative printed Manual on the means and tools involved in the Project can be furnished along with the audio CD or available as published on the internet. The interpretative infrastructure of the Project can become available as a sort of open source plan for future work in a similar direction.

E :      For such a project, a University site with a Classics Department is natural, one which has talented personnel interested in the Homeric performance as a project, with time and equipment for a continuing workshop, and access to a source of funding to cover working costs of the project. When the project is completed, it should be made available on a CD embracing sound-recordings as well as papers on methods for the interpretations of Greek accentuated texts. This should be available to the scholarly public on a CD purchased at an appropriate price through the University Press or Publications.

A list of the Problems

First we should consider some of the general considerations which such a Project would entail. The kernel of this project can be found in the essay I wrote for the internet in 1996 on the importance of reading Homer in an linguistically and acoustically proper mode of delivery. I felt it was important preserve the natural sound of the Classical Greek language, and I was concerned with the traditional academic style of reading Greek verse using Stress in place of the diacritic pitches and indicated vowel lengths. There were many encouraging responses to that essay, often suggesting that I record a selection the way I was prescribing, and put it on the web as a .wav file. The better part of a decade has passed and I have demurred, but now in 2006 I again take that project in hand in response to continued interest. However there are many problems.

By now the world has become much interested in the sound of Greek verse and there are a dozen recordings of Homer done in a variety of styles and modes. But evoking the sound of what the Homeric Bard might have been singing is a complex business. I am encouraged by the appearance of the recordings done to date, but none seems to satisfy the variant criteria of a bardic mode of performance. This paper defines some of the problems which beset current work on the sound of an authentic Homeric verse, as follows:

I:       Reading ancient documents, we use the sounds of our native speech even when we try to adopt a modified or Purist pronunciation of Greek. Speech sounds which are learned before the age of four are part of our automatic response mechanisms; we know when we try to learn a modern foreign language how hard it is to adopt another set of phonic elements. We do it by continual practice in careful imitation of the sounds and the lip and facial movements of our language teachers. We do not learn new sounds by following euphonic rules or by describing lines on a scope, even by directions for positioning of parts of the mouth. When we try to modify our speech sounds to suit an ancient language where there is no living audio tradition, we find ourselves with no guidance at all. So it is natural to use the basic features of our normal speech-patterning, and we read our Greek with an British accent, or a French accent, or a modified mongrel accent with elements borrowed from here and there. If this sounds phonetically phony when reading Plato or Homer, that is understandable, because it is a phony quasi-reconstruction of a tradition which has left us with no auditory tracks.

II:       There is no excuse for reading Homeric verse with the acoustics of our daily speech. The spoken-sound of a daily usage is completely unsuitable to artistic Greek poetry, whether lyric or epic. When Homer states "The anger, O Goddess SING, of Peleus-son Achilles", there is no question about a special voice quality which is quite different from daily communicative prose. The Greek verb "aeido" demands a singing and musical sound. However, a singing-voice is a cultural variable, it varies in discrepant cultures and in different parts of the world. I feel that for our purposes "chant" would be a better translation for this verb, certainly far more suitable than either operatic or country-music voicing, while we look for further examples of ethnic folk-chant to serve as our singing guides.

III:      In any age there arises for each set of speakers a roughly defined "National Voice" along with its dialect variations. This voice tone can have hollow-sounding or highly nasalized vowels, and the sounds can be fast in pace or very slow, clear and distinct or heavily slurred, and there are sounds in many languages which foreign speakers cannot even hear. At the other end of the acoustic spectrum there is a set of auditory responses, something like a "National Ear" which knows what to register and what to reject as incomprehensible in regular communication. But all this changes when a singing voice is used. Years ago when I was working with a language orientation program at Yale with foreign students coming to this country, we discovered in an evening's entertainment that the wide range of our students' "accents" almost completely disappeared when they were singing. Applying this to our teaching we experimented successfully with the sung voice as a language teaching device. I maintain that "chanting" will obviate some of the phonetic problems of Homeric verse as pronounced by international readers.

IV:       In ancient Greece there was a wide variety of dialects with different sounds and allophones. The Homeric poems were recited everywhere in the Post- Archaic period across political and dialect lines. Singing or better chanting the verses in the diverse Archaic period would have avoided phonetic localisms, and I stress this point as an indication of the universality of the chant-sung Epic in ancient times. If music is the universal art, then chanted or sung speech will share a part of this universality by making the speech-sounds accessible to variously oriented ears.

V:       Greek verse is academically overloaded with a panoply of terms for Feet, Pauses, Substitutions and statistical irregularities. Our traditional metricization of Greek verse comes from the work of Aristophanes of Byzantium and the Alexandrian academicians, and shows the formalized approach of the ancient academy in dealing with the epic language as a written text, rather than as a live Homeric performance. Trying to recreate a lively Homeric style from the ancient grammatical writers in the Teubner "Scriptores Metrici Graeci" is as fruitless as trying to write a new Mozart concerto from a college textbooks on standard l8th century practice. Performance must be based on skill and experience along with a certain amount of acoustic intuition.

VI:       The parallels of the Serbian guslars and Parry's fascinating ethno-poetic work are so well known in academic circles that we hardly need to register their significance. On the other hand there is a two thousand year gap in such a Serbian-Hellenic tradition, where few would believe that there is a line of historical continuity. Too little of the Serbian recordings are still available for us to hear, but the little I heard years ago from the original Parry discs did give me a sense of what a similar ancient epic poetry could be like. Yet the voice quality and style of the guslars seemed to me distant from the elegantly manicured technique of the long Homeric tradition. Less well known that the recordings of Milman Parry is the research and transcriptions of several Serbian scholars in the l9th century whose work is not yet available to the West.

VII:      We have had for years a narrow tendency to think of Greece as the foundation of the West, ignoring the Near Eastern and Egyptian cultural threads which have been subsumed into the fabric of Hellenic culture. Remember that Troy was a city in Turkey ! Should we be thinking more carefully of Homeric poetry as part of a Near Eastern tradition, using musical and voice styles which we find still alive there? The only really old traditions of musical sounds which we can trust are that of the Jewish services which may date from the early Hellenistic period, and are possibly earlier and better documented than the Arabic musical tradition. Should a reconstructed sound of Homer intentionally sound somewhat "Near Eastern" even if perhaps foreign to modern Western ears? Surely an authentic Homer would not be sung in the Western harmonic style of 16th to 19th century opera.

VIII:       Homer was read aloud as Performance at one time, so there must have been sufficient voice projection to carry over open listening areas before the ancient sound-designed theaters of Vitruvius were built. It is said that a chorus of girls' voice was used in annual recitations, probably for the sound carrying quality of their voices which are in the best human hearing range. Perhaps a tenor to alto voice would be good for chanting Homer, or in some parts a lower male voice might seem appropriate; but at any rate a male voice in monotone American style would be inadvisable. We need to think about the right voice fopr chanting Homer, but performance styles have varied considerably with the time and culture. The fast running pace of the Serbian guslars and the rapid reading of parts in Shakespeare's Globe Theater were dependent on the listeners' acoustic habits, and might not suit a society which expected a more esthetic and leisurely rate for reading poetry. Then there is the matter of an individual performer's skill and intent in his reading. James Joyce did uninspired readings of Finnegans Wake while Sioban McKenna read the same passages with high performance art. What is authentic and what is artistically right?

IX:       Several of the Homeric recordings which one hears are literal readings rather than performances, and proceed at an even pace with little adaptation of the voice to changes in the speakers' parts or the storyline. This comes from the history of the Iliad being read for millennia as a book, a text written on a scrolling papyrus or leather codex or now printed on paper in an OCT edition. If we think of the Greek text both as written poetry, and also as an exact score to be interpreted textually and emotionally as the reading progresses, we can look to a much more exciting experience than a correct classroom reading. But many of us are trained in the scholarly traditions of the West, more at home with an Apparatus Criticus than the idea of footnotes for a live Performance. A full interpretative re-creation of the Iliad will be initially very difficult to consider without considerable thought and preparation. In a sensitive reading, the pace and tone of the voice must change to suit the evil Agamemnon, his nasty brother, the lordly Hekebolos or the sniveling priest Chryses. Do we want to read Helen's lines with the classroom voice of a middle aged gentleman Associate Professor? The final question is: What do we want our reading to do? And how far do we want to go for the sound, without losing sight of the actual words of the ancient texts? This is difficult since the academic world is so often resistant to re-evaluation and change.

XIII:      Chanting is indicated in the first line of the Iliad with the verb "aeide". As I tried to deal with some of the above problems in making my early Iliad recordings, I began to see how well I could actually chant lines of the Iliad in a lifelike performance mode. That was harder than I thought when I first started, I found it takes time and effort to get a recording made and edited. This was for me a first step, but after all it is just ten years from the publication of my first Homeric paper to the discussion of this audio interpretative project , and this is not an unreasonable delay in the long run. It took Homer fifteen hundred years to get printed, so my request for patience in generating an acoustically authentic performable Homer in the next few years is quite in order.

But it is clear that realizing this Homeric Project will have to involve multiple readers, a lot of decision making about scholarly and esthetic matters, and must come out of a coordinated workshop experience with people who know their various businesses. I have hopes that the idea of a group doing the Project can interest others to progress faster and more effectively than I have been able to go on my own. If my encouragement and precautions can be of use to others in a full Homeric Performance Project project, that will fulfill my purpose in writing this paper.

Interpreting the Sounds

Durations and Lengths

Ancient Greek has Durations or lengths of syllables, in three classes. There are short vowels like e-psilon and o-micron facing their long partners eta and o-mega, while alpha, iota and u-psilon can be either long or short. Now the reading of Greek requires careful attention to the length of the vowels in a syllable, as well as a sort of compensatory lengthening of a short vowel before two consonants. The rules covering all this are complex and cannot be laid out here. But you must understand that if a long vowel or syllable is analogous to our musical quarter-note, then a circumflexed vowel will be somewhere between a dotted quarter and a half note. Greek is a metrically organized language and the Durations must be observed carefully in reading poetry or everything goes awry.

Combining the sound systems described in this section on Durations with those in the next paragraph on the diacritic Accents, is not something you can attempt when reading a Greek line of verse, without a great deal of practice and a sure musical ear. Many professional Classics teachers will fail in this important combinatory process, and few beginners can get beyond the first ten lines of the Iliad before desperation sets in.

Then there is the problem of Stress. Somewhere in the historical development of Greek, there occurred a process of substituting Stress, meaning loudness or amplitude, for the pitches of discussed above, and this became the standard for pronunciation of Greek as taught in the West since the Renaissance. The carefully indicated musical Pitches which are a special adornment of the sound of ancient Greek, were converted into thuds , and this is the way we have all learned our classical Greek. The result is that we place Stresses where we see "accents", which automatically increases loudness , even on short vowels where lengthening is impossible! When we read prose in our Herodotus or Plato, we find ourselves ignoring the Durations completely while adding Stresses to places where the pitch-marking Diacritics or accents are written. An "accented" short vowel will be pronounced loud adn long !

We have learned to accept this as the normal lumpy sound of prose. But when we turn to Verse we have to reverse the whole process and re-establish Lengths in order to read metrical line of poetry. No language can have entirely different ways of reading the languages of Verse and Prose. The way we handle the situation is a sure way to confuse the student who is facing Verse after learning his Greek out of a Prose textbook. Having first obliterated the delicately cadenced and configured phraseology of beautifully wrought prose like Plato's by stressing the Accents, he now has to stress the long vowels as loud beats in a rocking-horse manner for his reading of Greek verse.

For some so when we start reading verse we find we have been on an entirely wrong track. Now we have to peer at each line, scanning for the "longs" which the grammar book requires for poetry, usually ending up reading with a crude rocking-horse effect. It is easy to trip over our previous prose introduction, and make a short vowel Loud and Long (because it has an acute accent), and we end up with a Homeric dactylic line which has disintegrated into worthless metrical garbage. How much better to have started with Homeric verse in our first Greek lessons, and then use the lengths of vowels and syllable as properly understood and taught, when reading prose. Bad traditions here as elsewhere, are always the hardest to break.

Then how do we start reading dactylic verse at all? A first stage is to ignore the "accents" completely, and read for long and short vowels in the two metrical patterns (dactyl and spondee) which the Homeric line employs. At this point I suggest looking at a topractical approach to the Dactylic Hexameter, with advice for determining a number of the patterns on a logical basis. It is never worthwhile to crib in long and short marks above a dactylic line, the traditional schoolroom procedure of yore. Whether hearing some who reads Homer well, or figuring out lines for a correct-sounding rhythm on your own, you develop an ear for the verse configuration and the dactylic sound, at which point, with some practice, you should be able to read lines off as quickly and directly as you read a line of Shakespearean pentameter. With practice and experience, you must be able to read the Iliad line by line, with the metrics and durations residing largely in the background. In all reading in every language, Meaning is in the foreground, with grammatical forms, syntax and the metrics available on need from the background layer. But we must become more familiar with correct usage of the pitches and durations before we can safely consign their operation to the unseen but available linguistic background level.

The Pitch Diacritics

Greek text as we have it now, shows a number of other diacritic which once had a meaning but are largely pointless to us today. The aspiration indicated by the "rough breathing" is clearly a legitimate breath-sound, but the "smooth breathing" is apparently a caution to barbarian Romans like Catullus' (H)arrius, who had a bad Cockney habit of aspirating initial vowels. This caution has no meaning to us now, any more than the rough breathing on rho which was probably a throat -r- like French -er-, which is still a problem for Americans with our lateral -r- and equally Britons even more with their vanishing grade of the r-sound. Unneeded diacritics are harmless if we ignore them, but useless decorations to a diacritic crowded text . I mention these only as preface to the next paragraph where serious confusion reigns.

The so-called "accents" of classical Greek are not accents or indicators of stress at all. They refer to a system of tones in which the acute accent (Gr. oxu 'high') calls for a rising musical Pitch near to a musical fifth, while the circumflex demands the same uprising fifth which then slides down at similar rate to a consistent base level. The grave (Gr. baru 'heavy, low') accent, which by its shape point "down", indicates the lowering of a previously raised pitch to "base level", and some student papyri actually place a grave over all syllables that are not otherwise marked with a diacritic! Graves just mean there is no rising of pitch, a warning signal for non-Greek learners who tended to raise the voice from base-level incorrectly. There is another use of the grave which means that there was here a rising sound, but it was lowered by euphonics conventions of the line or sentence structure. I only need to note here that Grave means a null pitch-wise, hence must be ignored when reading aloud.

We must also think carefully about the pitches and what they actually mean. Ancient Greek scholarly citations only give bare outlines, but this one clearly marks the Grave diacritic as "illegitimate and stupid", referring apparently to the use of this mark on all syllables not noted by Acute or Circumflex. Such use was for elementary texts used by children or foreigners who didn't know the Greek intonations, and ancient school papyri grave-ize all otherwise unmarked syllables. We must also disregard Graves as replacements for devalued Acutes, yet many will still read them as stressed syllables!

The perispomene or circumflex name comes from a verb "spao" meaning "twist, wrench, sprain (Med.)" which I think refers more to the sound than the diacritic's shape which, whether Porson or Teubner font, is modern. The Acute can be a tone raised from the Grave base by a fifth or in inversion a fourth interval; but the perispomene is much harder to define. The usual statement that it is an "up-over-down" on three metrical "morae" or eighth-notes, does not tell us much about the sound. It must very similar to what we call in Western music a melisma, but that has in modern usage a different musical connotation.

There may be one linguistic clue to the sound of the circumflexed vowel. I believe insight can be drawn from the onomatopoetic word "bA^" or circumflexed beta-eta, which is well known as the sound of a sheep bleating. This is one of the few onomatopoetic cases in which we know what an ancient sound was by how it was represented in letters. The sound of a 'bleat" is very much a rising, holding and falling sound, and the fact that this rare word "ba:^" is capped with the circumflex accent, does point to the diacritic defining itself as what we might call "the ovine melisma". Compare further the word for sheep "mE^lon" in Attic, which does not convert to alpha in Doric dialect as expected, although an identical word meaning "apple" does convert. (This m/b alternation is well attested, e.g. Skt. mrtas: Gr. brotos.) To summarize: The sheep-word was apparently spoken with a certain bleaty sound , as remarked by the presence of a circumflex diacritic in all dialects where in fact, then as now, sheep bleat that way.

XII:      I do not follow the above phonetic argument regarding the Perispomene to the absurdum, but am sure that Homer knew what I mean with his phrase "mE^kadas aigas" use of goats, with the same telling diacritic. Using this indirectly, I have modified my pronunciation of a natural long vowel under a circumflex to give a hint of a bleaty tone of lengthened triplet duration, and once I have removed the humor of the situation, I find this intonation quite nice and natural. As I said above in learning a foreign language, this gives a phonetic target to aim at; how we handle it in this case is arbitrary, but far better than directions like "up-over-down", which is better suited to the world of gymnastics.

Consider here the spelling of the rho letter, which is internally doubled and always furnished with an aspirating rough-breathing. Since such use of onomatopoetic words is rare and unusual, let me cite a different but parallel example from the Roman satirist Lucilius who speaks of the letter -r- as better pronounced by a dog than by a man. This means that the Roman -r- was throat-sounded, to some degree like the French -r- , which the British with their tongue flapped or disappearing --r- always find difficult to reproduce. For ancient Greek, without further information, I would have to use the -r- of my own language for any place which showed cautionary spelling of this rho sound. Unfortunately there are very few cases in which we can make a deduction as in this example, perhaps linguists should be on the watch for more!

Pronunciation of the Characters

Students of classical Greek know that Greek writing was usually read aloud, and this was especially important in the case of the poets where sounds of the words were a critical part of the poetic art. So the question quickly arises: What were the actual sounds and how are we to articulate them when we read Pindar or Homer, either reading silently with the sounds in our imagination, or reading aloud with real cadences of the articulated sounds?

The problem is that we have very little chance of reconstructing the sounds of any language which has been out of use for more than a couple of centuries. English, has gone through multiple sound-changes as the language came into use throughout the world, and has continued to alter itself in the course of each generation. We can hardly know how the familiar lines of Shakespeare's plays sounded in the Elizabethan Globe theater. The vowels, when charted on an sound spectrograph machine, are seen as the most "musical" of the phonemes, but they are unstable and liable to change and shift pitch in relation surrounding sounds. There are a few traces of the Greek vowels in borrowings into other languages, but these are often unclear; yet they are all we have to work with.

For example. when the Romans tried to define the Greek bilabial u-psilon as spoken with puckered lips, they had to create the Latin -y- from Greek capital upsilon (not sounded like our -y-) for transliterating Greek words. In a similar situation Americans might devise a new character with the "umlaut diacritic" or - ü- for representing a similar almost bilabial French -u- which is not native to English. This is just what the Romans did, and it is our main evidence for their evidence as to pronunciation of the Greek vowel. Then there is a Roman spelling "thensaurus" for Greek the:sauros, which was never pronounced with an -n- at all, but represented an open long vowel eta, since in Latin popular usage the syllable -ens- had become -es- with long open -e-. Cross-language detective work can help with a few of the Greek vowels, but the arguments are complicated and hard to prove conclusively.

Among the consonants we are fairly sure that the pi had no aspiration, because the phi was clearly defined as aspirated, and never pronounced as the -f- which we use reading our Greek! The zeta may be more like -dz- or even -sd- in Aeolic, witness Latin rosa indirectly representing a zeta. Considering that it was Kadmos who borrowed the Greek alphabet from the Semitic Tyrian coast, we might find further traces in Semitic phonology, but again as we go wider our arguments become less convincing.

In short, we can make a few corrections about aspiration for a throat-guttural chi and aspirated phi in our reading of Greek, as a first step toward authenticity, but we can't go much further without producing a new Greek phonology, which would be incomprehensible beside the "corrupted" sounds we have been traditionally using all along. And remember that we are neurally and experientially linked to the sounds of our native languages. Most of us will speak French with a distinct American accent even after years of study, and we will read our Greek with the same American flat tones, while an Italian will read Greek with clear Italian vocalism, with double consonants always separated and pronounced as distinct sounds. So when one asks about an authentic Greek "pronunciation" of the vowels and consonants, the best answer would be that there is a double problem. There is the loss of phonic details in a two thousand year span; but this is less critical than the use of a native pronunciation of the modern person who is articulating the ancient text.

Homeric Singing-voice as Chant

There is good evidence for the Chant in recital of Homeric verse, from the first line of the Iliad, which specifically states : " Sing, O goddess, the wrath ......". When Virgil in literary fashion says "Arma virumque cano..." he copies the Homeric word 'aeide', but no Roman was ever serious about chanting the Aeneid in a musically bardic style. There is a wide gap between urban Augustan Rome and the 9th century bardic precursors of the Homeric poems.

When we speak of "singing" we will think first of the modern conservatory trained voice, but this is just one part of the world-wide uses of the human voice in a range of styles. The gap between singing and chanting is hard to define, but when one hears a genuine chant from an Islamic call to prayer or the traditional Hebrew singing of an Orthodox Synagogue, one hears something quite different from the voice of a Schubert song or a Country Western concert. The chanting techniques the Serbian guslars' historical storytelling, which operate within a five hundred year historical window, are too well known from the work of Professors Parry, Lord, Nagy and others to need comment here. Many years ago as a student at Harvard I had a chance to spend some time with the aluminum field recordings which Parry made in the l930's, and the chant-sound of the guslars is still clear in my memory. In the mixed South Slavic, Albanian and Turkish Islamic society of the Balkans after l450, the Chant had a voice of its own. I believe that the sound of Homer's "pre-guslars" must have had a mixed sound stemming from the culturally fused world of Hellenic, Semitic and Hattic-Trojan worlds which were the backdrop against which the Iliad was set. Chanting with improvisation can be found worldwide in such diverse cultures as the Basque bertsos, Cuban décima. The Dozens, (ritual rhyming insults among African American ghetto youths), Norse and Germanic flyting, Provençal and Catalan Jocs Florals, Arabic naqa'id, Argentinian payadores, the partimen and tenso of troubadours and Lebanese zajal.

Without question we should "chant" the Iliad, and at long last cease reading it as a level sounding un-metrical kind of prose. If chanted sounds sound somewhat "Near Eastern" to our ears, that should be no surprise, since the war against Troy was in fact a Near Eastern historical event. There are so many features which Greece derived from the older societies of the East, that we can no longer claim that Greece was the sole origin of what we like to call our "Western world". Threads from Hittite (the name Alexander) or phrases like King of Men (Akkad. lugal lugalu), and even Tyrian writing with a phonetic alphabet (the name of the inventor Cadmus is typical tri-consonantal Semitic word: k-d-m- meaning "East") turn up everywhere, and more are bound to appear as our ignorance becomes less complete.

Chanting however must involve bringing together a great deal of other things we have been discussing, into a coherent and workable form, so that we can use diacritic pitches along with the long-short Durations consolidated for an integrated acoustic reading. But this is not easy. Once we determine how to interpret the pitch diacritics, which are visible in printed texts for real-time reading, we have to determine on the fly the lengths of the syllable clusters. Experience with the complicated length-rules, coupled with a musical sense of how the dactylic line should sound, will be needed to make this level of our acoustic reading functional. The question remains how a progressing student of the Greek language can do all this at the same time and come out with a sensible chanted product. One might initially suspect that the advice "For professionals only, please do not try to do this at home..." might be applicable here.

Among the various professional scholars who have published recorded versions of Greek poetry read aloud, there is a wide spectrum of disappointing results. A review of the dozen well known readings is beyond the scope for this study, and it would be unappreciative to document the sincere efforts of those who have tried to make the reading of Greek better understood. One of the earliest of the "authentic readers" does the singing part of his recordings well, but the articulation of the words, while integrating all of the other factors, is so unclear that it is impossible to follow the text even with book in hand. Another who does a Pindar reading with clear word articulation, follows the old system of stressing acute accents, even when they are on a short vowel ! One reading from Germany is heavy and unmusical if technically accurate, but the chant is totally absent and the voice follows regular German speech patterns. Since we all have our own particular problems, some technical with the pitch diacritics, some with durations and the length of syllables, some of a musical nature, it might be felt (after reading this long and complex paper) that it would be best to toss authenticity to the winds and continue as we have been doing if only for our peace of mind. But continuing in the dark when there is a possibility of shedding light on a world of brilliant artistic writing, would be unfortunate indeed.

Scoring on a Standard Musical Staff

In order to control accurately the varied parameters of vowel Length and also diacritic musical Pitch, along with an awareness of special relationships among Feet and line Breaks, while using a carefully monitored Chanting voice which sings the Phonemes of the text clearly enough to make the meaning come through - - - - these things can only be done with use of a prepared, multi-level standard musical Score. A Bard with a life of practice and preparation can do all this things together, even while improvising with the text itself. But for us, it has to be done a different way.

The traditional staff score for Western music has a long history, signs of which are evident in the intricacies and peculiarities of the system which we use in writing and reading music. There have been dozens of scoring schemes devised to replace the traditional system, many of them quite sensible and clear, but the use of our traditional score in all levels of musical education is so widespread that change is most unlikely. However this means that we have access to an almost universal music-writing system as familiar if inconvenient in some ways , stretching from church hymnals and elementary music lessons to high-end professional composition. This system has everything which we need for purposes which I will here outline for the reading of Greek metrical-musical texts.

We can write out on a single treble staff the three simultaneous levels of sounds by note placement on the staff as tone-pitches, indicating sound lengths by the conventional note shapes, with a syllable-separated text written out below the staff. This familiar layout will answer to representing everything which we have been discussing above. A person with musical instrumental experience can score lines of Homer in this manner, and anyone with an absolute minimum of experience in reading church or popular sheet music, should be able to read this score with a little practice.

Rather than try to acoustically imitate the various recordings of Greek texts as interpreted by classical professionals, we must be able to construct our own interpretation of the functioning of a line of Homeric verse, and then practice our own personal performance from the score, in terms which suit our sense of acoustic taste. Coordinating the word and sounds with sufficient articulation of the words to make the meaning clear for an audience, is always a difficult procedure. But having pitches, durations and wording viewable together on a staff of score, we can concentrate on the various levels simultaneously until we reach a level of performance which satisfies us. The recorded sessions of the Iliad as discussed above on a CD will provide a sense of how a modern bardic Homeric performance should sound, but it is also important for a student to learn to work out from Score an independent performance on his own.

The whole purpose of reading Homer's dactylic lines as chanted text is for personal enlightenment and enjoyment. Probably none of us will want to do a publicly read performance of Homer's Iliad. On the other hand we all would like to have the acoustics of authentic reading in our mind when reading the Iliad for our personal pleasure. If we write out a scored performance of a dozen lines of the Iliad, and practice that until we feel it become familiar and acoustically alive, we have in hand the basics of what we need for reading Greek poetry intelligently. The next fifty lines should come more easily, and when we have mastered a hundred more we are well on our way to understanding the acoustic orchestration of Homer's language. What works for Homer will also work for Pindar and Sophocles, and this opens the door for perceiving the artistic detailing which every ancient educated reader expected when reading the flowing and musical sentences of his Plato aloud.

Preparing a Chant for Iliad I

There are many interesting passages in the Iliad which would be ideal for a chanting reading; but since most students will be reading Iliad I as their introduction to Homer, it seems appropriate to make a recording for the entire book so it can be used by students as they work their way into Homer's world. Learning vocabulary and grammar along with the recorded chanted sound will put an authentic sound into their memory, and they should be able to read on in Homer with these sounds firmly established in their memory. This Project is an educational as well as esthetic program, and it is the only way to get a new generation of students reading Homer in a correct and sensitive manner.

A few words of comment are necessary in preparing to score the first lines of the Iliad as an example of what is done with this approach. First, a suitable range must be selected for the comfortable center of the voice range of the reader. A typical low male voice on treble staff can be written in a D - A range for some passages, while A - E will center for a higher voice. There is no need to furnish conventional information for a Key since only two tones, the Base (grave) and Raised (acute) tones are required and we don't have to deal with key-modulations of the diatonic scale beyopnd raising of the range by a tone or two. Some scholars have objected to the fifth range for the diacritic-marked acute, feeling that this is too strained a jump for a normal speaking voice. But remember that we are now Chanting (Gr. aeidein) and using a mixed musical -speaking range for our voice. For singing-chanting the Fifth is a natural jump and one which seems built into the acoustic tuning of the human ear. Listen to any popular Country or Rock music and you will find an approximate Fifth continually used, along with its inversion as the Fourth. There are, however, many possible uses of stretched or quarter tone intervals, so we are not bound to chant in Perfect Fifths, although we write them thus in the traditional musical staff we are using. Flexibility is always the key word in matters of the arts.

Tempo will of course be adjustable to the situation. There are lines of grave importance which want slow movement, others will reach for a tight or even frantic pace at times. These are things which can be worked out later in reading performance from score. Marking approximate tempi might be helpful here, but I would depart from standard Italian terminology since we are in an area so far removed from the Post-Renaissance musical tradition. The Greek measure of the dactylic line is in a sense four-square, since the Dactyl and Spondee both will have in combinations four basic "units" of length, so we can register in standard measures with the conventional 2/4 staff indication which Bach so often used to house his complex rhythms.

But there is no reason to count out the measures with the repugnant "One-and-two-and..." counting which teachers have forced on thousands of young instrumental students for decades. Since our Chant readings will be solo and not need synchronization with another instrument, the tempo can vary with meaning and serve as an artistic device to modulate lines with emotional feeling. Some interpreters have thought that the lyre was used with Epic readings, surely a mistake drawn from use of the lyre with Aeolic Lyric poetry, although the singer Demodokos in the Odyssey does carry an unused lyre. In the early days of the 9th century the Epic was described as read annually by a chorus of clear girls' voices, probably to match the best in an open field celebration, the reception range of the human ear. Some of us may want to use a detuned guitar stroke as a modern pseudo-lyre for a secondary voice line beside the chanted text, in which case the chanter's guitar can follow the tempo changes of the words naturally and very easily.

The same dynamic changes of volume which we assume for a Schubert Lied will be used here along with time changes and pauses in the recital where felt suitable. But one things must be noted as mandatory, the separation of verse lines by a pause. This can be slight as some places or occasionally over-ridden as an artistic device, as when in Iliad I Apollo "shoots" (balle...) as a startling first word on an overlapped line. This artistically used 'over-the-line' writing device (which Vergil imitated e.g. impulit, consciously), makes it clear that individual lines as normally chanted must have paused before going on. We will want to follow this separation of lines for chanting in our writing out of our integrated score.

A treble staff should be laid out with pitches in 2/4 time and the text in transliteration for a vocalist who does not read Greek, as well as the Greek text underneath. This should enable a vocalist to read words with timed tones simultaneously and at the same time get the pitches as musical intonations. But the Greek text will be there to be read by the student or scholar who can watch the scored details of the other acoustic parameters, this is of course the btter way and essential.

If all of these things can be done carefully and with scholarly forethought, and recorded in a modern audio studio for production of a well mastered CD recording, this should answer the question which is always coming up abou the sound of Homeric verse. Homer may have sounded somewhat different in the atmosphere of the Archaic world, one can never estimate what the sounds of lost centuries were like. But we can use something from the voice timbre of modern Near Eastern singing, which proceeds from an ancient traditional base. It does not have to sound like the singing in a Mosque or a Synagogue or an Eastern Church, but it should not sound like a Professor reading a paper at an annual TAPA meeting. Doing these things right takes time. The combined efforts of a closely associated group of people working with attentive classical, musical and technical backgrounds, will be needed before there can be a serious answer to the problem of that eternally teasing question:

How did a Homeric poem actually sound ?

William Harris
Professor Emeritus Classics
Middlebury College