Saturday, September 21, 2013

How to Read Chant in Dominican Notation

As this is the beginning of the school year and choirs, especially those in Dominican Houses of Study are reforming.  I think this a good time to republish this post on the interpretation of Dominican chant notation. In response to a question I have also added an explanation of the use of the flat.  I also urge those who have strong historical interests in the history of Dominican chant to read the comments to this post by Mr. Richard Llewellyn explaining medieval practice in executing the "plica"--as far a I know these embellishments where not used by Dominicans in the 20th century, but choirs might consider using them.


Those who know Gregorian music, especially with the so-called Solesmes marks, are often surprised when they first encounter printed versions (or on-line manuscript) versions of Dominican chant. The first thing they notice is the complete absence of the Solesmes marks: epistema, dot, ictus, etc. The next thing they notice is that there is no quilisma. The so-called "expressive neumes" are totally absent. Then there is that funny Dominican quarter bar: it is not on the top line of the staff, but moves up and down to the height of the note it follows. And the double bar appears where in the Roman books one expects an asterisk. Finally they notice that the liquescents often appear on diphthongs and liquid consonants where they are lacking in Roman music.

Modern Dominican books reproduce the shape of the neumes in the medieval manuscripts of our music, which were based on the Humbert of Roman Codex exemplar of 1254, which still exists as MS 1 in the Domincan Archives in Rome. In order to sing our music properly, cantors need to know the system of interpretation explained in the treatises of Jerome of Moravia (d. after 1271. probably Scottish), a Dominican musicologist, who codified traditional practice. An image of his treatise De Musica, copied at the beginning of every medieval Dominican Antiphonal, is at the top of this post. It is from the fourteenth-century Poissy Antiphonal.

His system for rhythm is presented in an accessible, albeit Latin, form in the Dominican Processionarium of 1949. The Dominican system of adapting psalm tones to the psalms found in that book was suppressed in 1965 with the publication of the post-Vatican II Regulae Cantus in favor of the Benedictine method, although the Dominican melodies of the tones were left unchanged. But for all other chants, the thirteenth-century system continues in use today, in continuity with the living tradition of chant extending in an unbroken line back to the earliest days of the Order. The most important section of the Regulae Cantus for interpreting the chant is Section III, "De pausis et earum signis et proportione." I have adapted that section for use by choirs in section two below, but first singers need to know certain things.

1. Dominican and Gregorian neumes mean the same thing as to relative pitch on the staff as indicated by the clef. Those who do not know how to read neumes and what their names are, should consult the many books and on-line aids that explain this before reading any further.

2. In our chant, the two liquescents (the Ephiphonus and the Cephalicus) normally indicate a diphthong or liquid consonant. Since, in normal speech, the two sounds of the diphthong slur into a single vowel and the liquid is a semi-vowel merging into the following or preceding vowel, the two notes in these neumes do not get equal length because of a palatalization (touching tongue to roof of mouth). It is often said that the small note is "shorter" than the bigger one, but strictly speaking they are really about the same--it just seems shorter because of the palatalization. The same is kind of neume (but without the little note) is used for liquescents even when they do not indicate a diphthong or two distinct notes. A careful student will notice that we regularly use liquescents where Benedictine and Roman music has the two sounds of a diphthong sung on two distinct puncta, e.g., the separation of the "e" and "i" of the "ei" in Eleison.

3. The double bar in our system has two uses. They are very different and have some variations. Here they are:

3A. Most commonly, the double bar indicates the end of a major phrase, at which one side of the choir stops singing, and the other side takes over, following the traditional system (not only Dominican) of "antiphonal singing." This meaning is especially common in Kyries where the last Kyrie is sung by one choir, then the other, then both together. This shift is indicated in Roman books by asterisk and double asterisk. We use double bars. This use of the double bar also indicates a lengthening of the last syllable(s) of the phrase according to a system I will explain later.

3B. The double bar also appears at the beginning of a chant, after the first word or couple of words: there it indicates where the choir comes in to join the cantor who has intoned the piece. It functions like the asterisk in Roman chant. In this case, it does NOT indicate any lengthening of the previous notes.

3C. An interesting variant of this is found in Dominican books for the Gloria. There is one double bar to indicate the end of the priest's intonation phrase: "Gloria in excelsis Deo." But then there is another double bar at the end of "et in terra pax," an odd place to put it. This phrase belongs to the cantor. The choir comes in after that second double bar on the word "hominibus." Why? Many priests cannot sing for beans. They will foul up the intonation, but in the old rite they had to do it. So, the choir ignored what the priest sang and waited to hear the correct pitch and melody from the first cantor. Many priests simply recto-toned their intonation. The Gloria is the only place where this system is used.

4. Expressive neumes and quilisma do not exist in Dominican music. They did not exist as far as I know in any thirteenth-century chant manuscript. The expressive neumes are the product of Solesmes theorists. The quilisma is founded on the appearance, at certain places in the pre-staff notation, of a kind of fuzzy squiggle. There have been long debates over what it meant. My personal opinion is that it originally indicated the half-step and had no phrasing significance. Prof. William Mahrt of the CMAA, with whom I have discussed this, says that is very possible I am right, although he thinks it indicated a light vibrato. Good news! None of this matters for those singing Dominican chant because we don't have any of these flourishes, at least in "normal" execution of the chant.

5. Those who know the history of Dominican chant know that generally Dominican versions of chants avoid as much as possible the use of the flat. We also use it differently than the Solesmes books. When the flat appears in Dominican chant it remains in force until the end of the printed line of chant unless canceled by the natural sign. It is not voided by a bar line of any type. If a note is flatted in the next printed line of chant, the flat must be repeated before that note. Those who have second hand Dominican books will find that owners have often penciled in extra flats in a line of chant after a bar because they do not understand our use of the flat.


The most debated and controversial issue in performance of Gregorian Chant since its nineteenth-century revival has been the issue of rhythm. The earliest manuscripts with Gregorian music have no staff and so present grave problems of interpretation for pitch and intervals. But they do have a whole series of other marks that in part seem to indicate phrasing and rhythm. Manuscripts after the eleventh century have a staff with a clef to indicate "do" or "fa" and so solve the interval problem, but they seem to have little or anything to indicate phrasing and rhythm.

The modern science of musical semeiotics has attempted to resolve this problem. The Solesmes marks and system(s) allow an interpretation of both pitch and rhythm. What I am about to describe is how Dominicans have dealt with the problem of rhythm since the 1200s. I do not claim that this system is "better" than any of the modern Solesmes methods, including that of Dom Mocquereau, currently in favor with workshops sponsored by the Church Music Association of America. Those who want to sing Dominican music, however, cannot use that method because it depends on the presence of Solesmes marks, and these do not exist in our books.

Principles of Phrasing

As you read these principles, you can consult the sample image of various bars and neumes from the Processional, pasted here:

1. The most important determination of phrasing is the meaning of the Latin text. Dominicans break the text into what might be called longer phrases, and then break the longer phrases into shorter ones. This is the way we punctuate English. A period or semicolon indicates the end of a major phrase; the comma (or implicit unwritten "comma") indicates a "joint" between the shorter clauses.

2. The punctuation marks of Dominican chant are the double bar (A), the single bar (E), and the half bar (D). The quarter bar (C), which moves up and down, is NOT a punctuation mark. The single bar and double bar are very similar: both indicate a full stop and a point at which a full breath may be taken. The difference is that after a double bar, the other choir takes up the new phrase; the same choir continues to sing after the single bar. In small choirs it is common not to alternate between choirs. In that case, both of these bars are treated the same way--although it is common to make the pause after the double bar a bit longer since it sometimes (as in hymns) indicates a new stanza. The half bar indicates a minor break or joint in the phrase. The pause is shorter and a short breath may be taken.

3. The quarter bar is not a rest or a break; its function is similar to the Solesmes dot. It serves ONLY to lengthen the previous note. No breath is taken at a quarter bar. There is no break in the phrase. Back at choir practice in our Western Dominican House of Studies in Oakland during the 1970s, this was called the "don't breath here bar." Some people even cancelled them out with a pencil and put in a dot instead. Although lazy singers often took a breath at quarter bars (and the lengthening of the previous note tended to cover this), it is very bad form and ruins the phrasing. It is better for people to stagger their gasps for breath at any place other than a quarter bar.

4. Although it is not noted in the music by any mark, singers "take off" (slightly accelerate) as they begin a phrase, and "make a landing" (slowly retard) as they end the phrase. They do this in a more delicate way in the minor phrases ended by half bars. This acceleration and retard is so delicate that it should be hardly noticeable, just as the similar change in velocity is hardly noticeable in ordinary speech.

5. Singers hold the syllable(s) that precede the bar, be it a double, single, half, or quarter bar. This is distinct from the gentle progressive retard of the "landing." The proper way to do this is to slow down, hold, and taper off in volume so as to let the note fade away. Avoid, at all costs, staccato stops.

6. In music with one neume per syllable, the number of syllables held is determined by accented (or virtually accented) syllable.

6A. Hold the last syllable before the bar if the word there a monosyllable (e.g., "te").

6B. If the word has two syllables (e.g., "nobis"), or more than two and the penult (second to last) is accented, lengthen the last two notes, but with the ultima (last syllable) a tiny bit longer than the penault. BUT, if the text is metrical (poetic) you only hold the ultima, not the penult, even if it is accented. One text often sung incorrectly (because the singers don't bother to note that it is poetic) is the antiphon on the Nunc Dimittis during much of Lent, the Evigila.  It's a poem; it even rhymes!  So you only hold the ultimas---the one place you hold the penult (on "nobis") has to show this by two puncta! That is because, as poetry, you would not automatically lengthen that note. There are actually quite a few poetic antiphons, e,g., the O Lumen!

6C. If two syllables follow the accented one (e.g, "Dominus"), hold the last syllable only: it is considered "virtually accented."

6D. So, you never lengthen more than two syllables.

6E. Finally, again (see 6B above) if you are singing a hymn, sequence, or other music with a poetic meter, ignore these rules and hold ONLY the last syllable of the line.

7. In melismatic chants, i.e., those with more than one note per syllable, lengthen only the last neume or part of a neume.

7A. If the last syllable of the last word is melismatic and the penult is a simple punctum, you lengthen only the last syllable, even if the penult is accented. Conversely, if the if the last syllable has a punctum and the penult is melismatic, you lengthen only the last syllable, even if the penult is accented.

7B. When the last syllable has or ends with a two-note neume, e.g. a podatus or a clivus (see B above), lengthen both notes of that neume.

7C. When the last syllable has a meume with three or more notes (see F above), lightly retard but hold only the last note in the neume.

7D. In very melismatic music, such as Alleluias and Responsories, there will be a part of the multi-note neume at the end, usually a clivus, podatus, or torculus. This part of the neume is often detached, as is the case in B and F above. Hold that part of the neume only, using rules 7B amd 7C.

8. Strive to sing all chant legato. This is very important to phrasing. At the end of all phrases avoid staccato stops. Let your voice volume taper off to silence. Do not gasp for breath; if necessary, stagger your choir's breathing.

That's all there is to the system. It may seem more complex than having the Solesmes marks to tell you what to do, but, in fact, it is very natural and intuitive. And it seems, from writers like Jerome, that all thirteenth-century singers of chant sang this way, or in a very similar way. So, now you twentieth-century singers can sing the music in the new Roman liturgical books like the Missal, where there are no Solesmes marks!

The Salve Regina

If you do not have the French Dominican C-D Dominican Chant-Dominican Liturgy, which has the example discussed on track 20, you can order it here.

Here is the music of the famous Dominican Salve Regina used all year after Compline, taken from the Antiphonarium. Take a look at it and I will apply the principles of phrasing I explained earlier. If you have the French Dominican recording of this, you might listen along as you read, in it you can hear how the phrasing and the lengthening of ending notes merge into an organic whole. They are also very good on distinguishing the strong retards at the end of major phrases from the lesser ones of the shorter clauses that make them up. To allow you to consult the rules given in the last section, I will reference them in brackets as I comment.

Phrase 1 "Salve Regina, Mater misericordiae.

This is a major phrase with a break in the middle after Regina indicated by the half bar. So the major rhythmic phrase is the whole verse and there are two minor phrases [rules 1, 2]. The take-off begins on Salve and the landing on "ae" of misericordiae. There is a light touch down on "na" of Regina because of the half bar. This kind of subtle articulation is always observed in the rest of the antiphon at the various breaks, so I will not repeat it [rule 4].

There are no "true" double bars in the antiphon Salve Regina until its end because it is traditionally sung by all, not alternately by choirs. That first "double bar," after Salve, does no more than indicate the place to which the cantor sings and at which all come in. There is no lengthening of "ve" since this double bar represents merely that the choir comes in. The clivus on "na" of Regina is lengthened, but not the melisma on "ce" of dulcedo because even though "ce" is accented, it has a compound neume [rule 7A]. The same is true of misericordia, where "cor" is accented, but the second to last note is melismatic: so only the simple neume on "ae" is lengthened [rule 7A].

Phrase 2: "Vita, dulcedo et spes nostra, salve."

This major phrase has joints after Vita and dulcedo, indicated by the half bars. The mark after nostra is a quarter bar--equals a Solesmes dot, not a phrasing mark, so it does not divide a phrase and no breath is to be taken [rule 3]. The final syllables "ta" "do" and "ve" are lengthened, but not any previous notes because the previous notes are all melismatic [rule 7A]. The melismatic neume on "tra" of nostra is lengthened because of the quarter bar, but only on the two notes of the podatus [rule 7B].

Phrase 3: "Ad te clamamus exsules filii Hevae."

This major phrase has a joint after clamamus indicated by the half bar. Notice the two quarter bars. The compound neume on Ad ends in a Ephiphonus, which scants the smaller note. The quarter bar lengthens te but without pausing or breath [rule 6a]. Only "mus" of clamamus and the "vae" of Hevae are lengthened because the previous accented syllables have melismas [rule 7a]. The last two notes of the porrectus on "les" of exsules are lengthened without a pause because of the quarter bar [rule 7B].

Phrase 4: "Ad te . . . valle" contains nothing new. Apply rule 6A to te, 7A to clamamus, 7B to hac, and 7A to valle.

Phrase 5: "Eja ergo advocata nostra."

The phrase is jointed after ergo by the half bar. The odd looking podatus over the "E" of Eja is composed of a regular punctum and the large note of the liquescent Chephalus because the "j" is a semivowel: but this has no affect on phrasing. Notice the liquescent Ephiphonus on "er" of ergo because r is a liquid. Because that Ephiphonus is a compound neume only the "go" is lengthened by the half bar, even though "er" is the accented syllable [rule 7A]. In contrast, nostra is a bi-syllable with simple neumes on each one: so since "nos" is accented, both syllables are lengthened [rule 6B].

Phrase 6: "illos tuos . . . converte."

There is nothing new here except the lengthening of the last note only of the neume on "los" of oculos because it is a three-note neume [rule 7C]. Otherwise, apply 7A on tuos, 7B on misericordes, and 7A on converte.

Phrase 7a: "Et Jesum benedictum fructum ventris tui, nobis . . ."

This is routine: apply 7B on Jesum, 7A on fructum, tui, and nobis. The next part of this phrase has items of interest. Here it is with the rest of the antiphon:

Phrase 7b: "post hoc exilium ostende."

Note the lengthening of both notes of the clivus on "um" of exilium by the quarter bar, even though "si" is the accented syllable [rule 7B]. Notice also the big note (without the small) of a Ephiphonus on "sten" of ostende because "n" is a liquid--but this has no effect on singing. Both "sten" and "de" are lengthened here because they have simple neumes and "sten" is accented [rule 6B].

Phrases 8-9: "O clemens" and "o pia": nothing new. Apply 7B to both of the Os; apply 7A to clemens and pia.

Phrase 10: "o dulcis virgo Maria."

The joint in the phrase is at the half bar, which is actually in the middle of the notes on O. Note the quarter bar along with that half bar in the melisma on O. The quarter bar, as usual, lengthens the last two notes of the first neume on O without a pause or break [rule 7B]. The next neume on O after the quarter bar is lengthened only on its last note by the half bar because it has four notes [rule 7C]. Although it might seem bad form in Benedictine chant, we Dominicans normally take a small pause and a little breath, here at the half bar before attacking the rising notes that complete the melody on O, which are not lengthened because no bar follows. (Those with the French recording will notice that the French Dominicans manage to sing this without taking a breath here--a particular interpretation which may reflect a variant in the Competorium book.) The phrasing joint in this passage comes here in the middle of the run on the O. This is a not uncommon practice in Dominican chant, especially in Responsoria. Perhaps it is a concession to less disciplined non-French friars who, unlike monks, cannot sing through long runs in a single breath.

On dulcis, note the liquescent because of the "l" and that only the podatus is lengthened by the quarter bar [rule 7B]. There is a one note liquescent on "Vir" because of the "r". Finally, note that only the two notes of clivus on "a" of Maria are lengthened not the neume on the previous accented syllable [rule 7B].

Phase 11: the Alleluia.

The antiphon normally ends with Maria, so there is a double bar after it, but in Paschal time the Alleluia is added. Since there is no choir change before the Alleluia, the double bar magically equals a single bar. There are two things to notice in the Alleluia: the lengthening of the two notes of the clivus on "lu" by the quarter bar [rule 7B], with, as usual, no breath or pause before adding the last three notes. Finally, "ia" has a clivus, both notes of which are lengthened, but not the podatus on the accented syllable [rule 7B].