Sing it like you would speak it?
  • A common criticism of the rhetorical performance of chant is that directors who instruct their choirs to "sing it like you would speak it" really mean to "sing it like I would speak it," as if the interpretation were up to the whim of the director. This disregards the phonological truth that Latin syllables are of variable length, and have what Columba Kelly refers to as syllabic value.

    I use the words tempo, templo, and templum to illustrate this concept. Read more about The long and short of it over at
  • 16 Comments sorted by
  • Thanks for this topic, Incantu. It's very timely for me to read this thread, because our schola will be singing for Missa Cantata on Sept. 29, the feast of St. Michael (EF Mass), and I do have a question.

    Propers of Introit, Gradual and Communio, they all start with the same word 'Benedicite," which has the stress on the third syllable, 'di.' In Introit, The first two syllables, 'Be-ne,' have two punctums. In Gradual, one punctum for 'Be' and five note neum for the second syllable, 'ne.' In Communio, three note neum for "Be" and two note neum (pes) for the next syllable 'ne'. So they seemed to vary the length of the first two syllables before they reach to the stressed third syllable 'di.' So how would you interpret the length of the first two syllables of the INTROIT that are on the two punctum of the two same note (sol, sol). Thanks. Mia
    (maybe Graduale Novum might have more info on this than Graduale Triplex?)
  • Good question, mia.

    For one, if you read my article in the context of the rest of the blog, you will see it does not concern modern square-note notation or the Solesmes method of interpretation, but performing from the earliest manuscripts.

    Also, I need to go back to my article and make it more clear that what I'm talking about applies only to syllables set to a single note. A single syllable on a single note is normally the length of that syllable, but there are some cases where two syllables need to be fit into the length of one because of adaptation. I don't mean to suggest that a 40-note melisma will be sung in the time it takes to speak the syllable!

    When it comes to multiple notes on one syllable, the rhythm is determined by the neume, which can be long, short, or a combination of the two. For example, three notes might be long-long-long or short-short-long. In either case, it's the final note of that group (where the syllable is pronounced) that has the length of the syllable, or full syllabic value. Short notes are half as long as the syllable to which they are attached.
  • Thanks, Incantu, but my final question was about the two punctums of the Introit which is syllabic, not melismatic.

    " how would you interpret the length of the first two syllables of the INTROIT that are on the two punctum of the two same note (sol, sol)"

    (In the Triplex, both Laon and the Gall have the same signs for those two punctums, Laon has two dots and Gall has two short lines. So are those two punctums the same length?)

    (I quoted those melismatics of Gradual and Communio to get to the question on Introit of the first two puctums. I know it's hard to describe here without showing the actual music.)
  • Your article, Incantu, makes sense. The key appears to be "clear enunciation" which takes a bit longer with some vowels and consonants--and combinations--than with others.

    But that's still "as though you were speaking it" (obviously, not quite as quickly), right?
  • I'm not necessarily advocating "sing as you speak," myself (it's perhaps a little too simplistic). I mention that maxim as something cited by critics of the rhetorical approach in general. However, the issue of tempo that you bring up is related to rhetoric as well. Many chant groups sing so slowly that the ornamental notes no longer sound like ornaments. Yes, in a large church you would need to speak more slowly to be understood, but some chant performances unnecessarily prolong the notes beyond that which is necessary for the words to be pronounced, and the music suffers.

    In our Sunday Masses, our priest wonderfully chants "Let us pray with confidence to the Father in the words our Savior gave us" in a perfectly natural speech rhythm, followed by our cantor singing OUR FA-THER in absolutely rigid quarter notes. The effect is quite jarring. What is often natural for the musically untrained is a stumbling block for the professional. Actually, a close examination of the manuscripts reveals evidence of many performance practices that today would be considered vocal faults or deficiencies. But that' an article for another day...
  • Mia, to answer your question, individual notes are the length of the syllables to which they belong. "Ben" is going to be longer than "e" because of the n. That doesn't mean that you prolong the first note. It just means that you don't shorten it unnecessarily to fit the "beat." "E" doesn't need to be sung faster. Just don't stretch it out to the same length as "Ben."

    In your example, however, these notes come at the beginning of a phrase -- even the beginning of a chant. Because of that, they will probably move a little faster anyway. If you get to the point where you're trying to figure out the duration of every single note (this note is .7 seconds, that one 1.2 seconds) then that is the opposite of singing rhetorically. While I might look at something in a very technical way to illustrate a concept about performance practice, there is a necessary difference between analysis and performance.
  • Yes, we do compromise the analysis when we perform, but the analysis certainly gives a good foundation of the interpretation so we don't sing chant as I would speak as you mentioned :-)
    Awhile ago there was a beautiful youtube clip that showed Solesmes monks (was it Dom Saulier?) explaining on singing chant. It was in French with some English subtitle, does anyone have the link? (Couldn't find it through the 'search.')
  • It sounds like you're talking about minor (but potentially important for singing) differences in sound of the syllables, as naturally pronounced by the choir. Or do you actually work with vowel length, like pronouncing regina with the line over the final a?

    But what I'm wondering is.... Well, one of the markers of classical Latin pronunciation, and one of the important poetic factors, was the difference between vowel lengths. Medieval Latin poetry generally was interested more in alliteration, rhythm, and rhyme, and the belief is that vowel length pretty much dropped out of living pronunciation. (Very scholarly poets with good education would take vowel lengths into consideration anyway, but it was a pretty nerdy thing to do during most of the Middle Ages, until the Renaissance guys tried seriously to revive classical pronunciation and poetic forms.)

    But most liturgical texts being sung are from the Bible, and thus came either from Jerome's Vulgate or the Vetus Latina, which isn't exactly the classical period but did still care about such things. So one would expect (come to think of it) that vowel lengths might come into play.
    So one wonders... do chants reflect either the medieval or the classical vowel length pronunciation in how they are set up? Is there some way to tell composer usage and intent?

    Because, come to think of it, you could just as easily sing "Salwe reggeen-awww" as the other, so the classical vowel length thing does work there, whether or not you're pronouncing everything else in one of the medieval ways. * So I'm sure some Victorian guy did a big study of this somewhere; it was the kind of thing they liked to do. Where should I look it up?

    * Not a valid phonological representation of Classical Latin. Void where prohibited by choir director.
  • "Ben" is going to be longer than "e" because of the n.

    We have a snag. Benedicite would be sung Be ne di ci te--at least, if I'm to believe the Jesuits who taught me Latin and the several well-respected choral conductors under whom I've had the privilege to learn and sing. I'll grant you that the Jebbies are not an unimpeachable source, but the choral guys?.....They were.

    (Similarly, the 's' is always attached to the FOLLOWING syllable, e.g., e sto (nobis)...)

    Both of these are for clarity in sung enunciation; keeps all the singers unified.
  • Dad29, the approach you were taught works well for singing metered music and is widely accepted as a part of good choral technique. However, this is a relatively recent innovation and does not reflect the practice in place when the chants were first written down.

    You do not even need to examine the manuscripts to find evidence of this. Look at the offertory Ave Maria (or better, in the Graduale Novum). In the phrase "in mulieribus" the sol of "in" is repeated, which is a vestige of the practice of voicing the final n along with the following m on the first of the two notes (in other words, inm-ul-i-er-i-bus, not i- nmu-li-e-ri-bus as most of us were taught).

    The same thing happens on the l of mulieribus, where an extra note in unison with the preceding pitch has been add to the pes on the second syllable, making it look like a torculus. In fact, it is a torculus with a weak initial note (torculus initio debilis). Although the classical Solesmes method would place the ictus on the first note, the vowel actually comes on the second note. In many similar places, the Vatican edition has simply dropped the first note completely, as is the case with "benedicta" (aha! the syllable in question).

    At the beginning of that bar, "benedicta" could have been notated with a pes connecting the la of "ben" to the do of "e" (as you can see on page 18 of the Laon manuscript). But since in that place St. Gall has two notes in unison, the Vatican edition follows the latter and omits the weak note.

    Spend some time working with amateur singers, or listen to vocally untrained priests, and you'll find that this scooping up from the previous note when voiced consonants are involved is a natural tendency. Although it has been polished out of our conservatory trained singers and banished from our current choral techniques, I am quite certain this was part of the vocalism heard by the scribes who first recorded the chant on paper.

    While I'm not necessarily advocating a return to the former technique, understanding of this concept is crucial if one wants to avoid misinterpreting the signs used to notate the music.
  • Maureen, I'm not sure I completely understand your question. But I will say that the difference in vowel length between the modern Spanish "casa" and Italian "casa" is obliterated as soon as those words are sung for more than a beat. The duration of those words in a syllabic chant (hypothetically speaking, of course) would have more to do with the acoustic than the vowels lengths of the language in question.

    In the chant melodies what we call the accented syllable in Latin is as a rule placed on a note that is longer and higher than the the other syllables in a word or phrase. But there are exceptions to this general rule as well, such as at the beginning (where the melody generally ascends) and the end (where is generally descends) of a chant or phrase, in intonations (which speed up), and at cadences (which slow down).
  • and banished from our current choral techniques

    ....not entirely. Salamunovich likes to use the liquescents--in fact, he darn near exults in them, e.g., in his "Christmas" album w/the LAMastersingers singing the Introit of Christmas/Midnight Mass.

    But yes, for most purposes, the technique is 'banished'; I think it has to do with rhythmic precision, as you state; and for practical purposes with amateur chanters in typical parish settings, I'm a little leery of allowing 'nature' to take its course.

    But that's the diktator persona, eh?
  • I think the chants' rythmns are not always generated by mere pronunciation of conversational speech patterns but by a more dramatic declamation or proclamation. Like trying to create a score that captures an auctioneer's cadence or the drama in a carnival barker. "Herrrrrrrrrrrrrre's Johnny."

    It is difficult to perceive a systematic approach to what seems to be a suigeneris view. But both speech rythmn methods and the Solesmes method honor a rythymn that unifies the text,the word, the neumes, and the phrase.
    Ignoring speech rythmns completely could really destroy chant.
  • I think the simple answer (rightly or wrongly - although I believe the prevailing winds at Solesmes are similar...) is that it depends on the specific type of chant.

    For example, your average office antiphon should usually give more deference to spoken rhythms, while the balance of text-to-music in a melismatic gradual or offertory verse is obviously heavier to the music side. Introits and Communions fall somewhere in between - and some are more melsmatic and some are more syllabic.

    I think this consideration makes the biggest difference than a one-size-fits-all method. For example, the more melismatic a chant, the more I find myself relying on the "old Solesmes" markings, the more syllabic a chant, the more I find myself disregarding them.

    Of course, now I'm starting to get into semiology, which could change everything...
  • I have to admit that the title of this post deliberately used a buzz phrase. I didn't think it would get as many clicks if I had called it "Syllabic value: phonological support for the graphic variations in the uncinus of Laon 239."