Greetings! I have taken an interest in the theory presented by Jan van Biezen, a short summary of which can be found here: http://www.janvanbiezen.nl/gregorian.html
And here also is a link to his book with a detailed review of the theory: https://www.amazon.com/Rhythm-Meter-Tempo-Gregorian-Chant/dp/1945416009#customerReviews
I study Byzantine chant and do not practice Gregorian chant, but nonetheless take interest in any new "approaches" to interpreting the Gregorian repertoire besides the famous Solesmes approach. Having found out about Jan van Biezen's approach, I tried to search for audio recordings of chant done according to his method, but found none. Do any such recordings exist? And is there any critique of or consensus on this method by other researchers? Thank you in advance!
Replies
<I've seen this long-substitution all over the Alleluias. But it is also in the more standard repertory as well. The opening figure D-F-G-F of Introit Dominus dixit, found also a few other Mode 2 neumatic antiphons, is properly interpreted as grace-short-short-long, 0 + 1 + 1 + 2 time units = 4 time units = 2 beats, an even rhythmic number. In the Offertory Benedicite gentes (Offertoriale Triplex, page 72), that first D is a long in both St. Gall and Laon, and the F-G-F is replaced with a torculus over "qui non amovit".>
While grace notes can be insertions into tunes sung to a rhythmically regular metre, there are other kinds of insertions too. I would not see every insertion as a grace note. The D pitch on non amovit is an interpolation and, if I understand you correctly, you may be improperly interpreting it as an ametrical grace note rather than a straightforward metrical insertion in order to avoid outright repetition of the tune of the word benedictus.
<The rationale is clear to me. The upper second note has both rhythmic weight and melodic importance. As such, the first note may be weightless and unimportant, or it may be rhythmically acceptable (lengthened to a beat) but still unimportant, since, no matter whether it is 0 time-units or 2 time-units, its function is the same, namely to lead into the second higher note.>
Your rationale is also clear to me but that does not make it unarguable. The first note has rhythmic weight in my estimation. This kind of argument about “melodic importance” runs into the danger of saying that the likes of Greek and Coptic chant are unmusical because so many pitches of “melodic importance” are sung after the beat on account of all manner of gracings being sung on the beat.
<So, I think we cannot simplify all grace-, liquescent-, or debilis note interpretation to a single rule. That first note may be important in some cases but not in others. Attention is required to the rhythm, the melody, and how they together fulfill the function of each note. And this was Jan van Biezen's concluding point in his Gregorian chant article.>
I agree but I would add that norms of European singing, especially in living Christian chant traditions, need to be properly appreciated before they are thrown out of the window (on some theoretical pretext) in relation to being a medium through which one can interpret the notations of the likes of Laon 239 and Einsiedeln 121. While all living Mediterranean communal chant is sung to a fairly regular rhythmic metre while being capable of being rhythmically very complex and taking all sorts of liberties with verbal stress, I would give a great deal more weight to the possibility of those features existing in Laon 239 and Einsiedeln 121 than perhaps many others do.
<The contrary example you gave from Alleluia Jubilate Deo highlights the parallel between Laon and St. Gall of a short-long on one hand and a long-long on the other. I think the most reasonable explanation here is that the second note is indeed a long of importance --- it is a higher note to which the melody rises --- while the first note is by default a rhythmless note, i.e. an initium debilis or the like, meant to raise the melody up to the higher note, and that its own duration may be expanded from grace-duration to long-duration. What the first note cannot be is short, because a short is 1 chronos, 1 time unit, which is half of a rhythmic beat. It must therefore be either 0 beats (i.e. a vinnula or initium debilis or however you want to perceive), or else 2 beats, i.e. a long.>
You're expounding ideas that pitches either take a theoretical long value or a theoretical short value or a theoretical zero value. If the chant is to be sung according to that theory, a notated pitch cannot be realised with zero value, and yet it is theoretically bound not to have a duration as large as a short. The only sung solution, if the pitch is not to be long, is for it to be given a duration of intermediate value between zero and a short. The addition of such a duration into a performed metre would mean that the metre couldn’t be sung in a regular metrical rhythm. The likes of Greek and Coptic communal chant can be performed in a regular metrical rhythm, however,. Your description of an intermediate sung value, apparently knocking any regularity of metre off kilter in practice, contrasts with the practices of other traditions of Christian chant.
I have no problem with this idea insofar as it is practicable but it can be argued against as being unnecessary. One can argue that writings such as the Commemoratio brevis are based on a tradition that makes much use of organum; that the example of the antiphon Ego sum via indicates that a single pitch on a syllable is regarded as short, not as long; that such an equalist tradition strongly opposes the practical use of pitches of such intermediate durations as you mention; that the likes of Laon 239 and Einsiedeln 121 represent traditions criticised and superceded by such an equalist tradition; that those traditions not only made use of durations intermediate between zero and short but also of durations intermediate between short and long; that such traditions have their correspondence in the likes of Greek and Coptic chant; that theoretically regular metre is realised in practice in the likes of Greek and Coptic chant; that the traditions that produced Laon 239 and Einsiedeln 121 may also have realised theoretically regular metre in practice.
One can further argue that it is not the norm for communal Christian chant to allow pitches of intermediate metrical values to knock the regularity of the rhythm off kilter such that a regular rhythm is no longer to be heard; that it is the norm to incorporate pitches of intermediate value into a metre that is sung as regular by allowing the duration of such pitches to decrease by proportion the durations of adjacent pitches. Since this is the norm, I would advance the case that one really needs to demonstrate convincingly that the norm does not apply to notations such as those of Laon 239 and Einsiedeln 121. I don’t regard your good self or anyone else as having done that yet to my own satisfaction at least.
<Case 2: Written transmission, i.e. manuscript copying, increases the odds that an otherwise inconsequential vocal effect, such as liquescence, epenthetics, or whatever, be perpetuated. I think this is the chief reason why the 'Jerusalem' example you gave has the seemingly superfluous lower short note in both St. Gall and Laon: its presence in two manuscripts is more likely, in my opinion, owing to manuscript copying than to ingenious musical rationale.>
Musing on your use of the word “inconsequential”, I would again advance the argument that the recall of discrete pitches may be more fundamental to the memory than instances of liquescence. Because of its dating, I assume that a liquescence in Einsiedeln 121 is a reflection of practice rather than a reflection of written record, because whoever wrote out the communion Jerusalem quae aedificatur most likely sang it in choir too.
If the purpose of the document was to confirm pre-existent choral practice, then manuscript copying is not likely to have been the reason for the marking of the liquescence in writing. If the purpose of the document was to confirm choices of options for choral practice, then manuscript copying would again not be likely to be the reason for the marking of the liquescence in writing. If the purpose of the document was to transcribe another document, we would not know whether that transcription was mere and faithful or whether innovations (such as accommodations to the choral tradition of the transcriber) were made.
Musing on your use of the word “superfluous”, I would advance the argument that, because of its dating, Einsiedeln 121 ties into then contemporary sung repertoire which, by its very nature, exhibits “ingenuity”. I would also advance the argument that one can only copy from manuscripts containing what you call a “seemingly superfluous lower short note” because any manuscript copied from would reflect an existing practice which, by its very nature, would exhibit “ingenuity”. I would correspondingly advance the argument that the ultimate explanation for the existence of a first pitch of D on (Jerusa)lem is more likely to be found in oral practice than in any purported scribal innovation that would have been a fiction (with regard to actual practice) at the time of the feature’s introduction into writing.
<Case 1: If you have heard "Ad te levavi" or a similar figure only ever sung with that 'liquescent' lower D between the first two syllables, then you are far less likely to omit the D, however the D is performed, because the D is ingrained in the memory as a part of the melodic shape. The 'liquescing' is just the effect of the note being diminished by the textual articulation. Similarly, with "verbum tuum" in Communion In Salutari.>
I regard the pitch of a melody as likely to change because of the likes of improvisation, poor memory and poor imitation/reduplication. I regard liquescence itself as one improvisational option of many rather than as a causative factor in melodic change. Even if liquescence affects the pitch memory more than some other feature, I would still regard the liquescence on a note as being more likely to be forgotten than the pitch of the note. I have memorised the Sunday introits and I find that my memory of liquescence is much worse than my memory of pitches and I note that you appear to be arguing that the D being “part of the melodic shape” is the cause of any stability in transmission regardless of liquescence (“however the D is performed”) so you appear to be making my last point for me.
The first note G does not appear in the Graduale Novum and I would not regard that first note G as protyptic, just as I would not regard the first note of the Sciant neume as protyptic. The original bones of the melody of the words Ad te levavi may have been D F A A G. If so, an original liquescence on the word ‘ad’ may have been on a C pitch.
A/the tenor, that is, the 'mora' of the last note - which, however small it is in a syllable, is larger in a part and longest in a distinction - is a sign of division in these. And so it is necessary that the melody should be clapped as if with metrical feet and that some notes compared with others should have a duration twice as long or twice as short, or else trembling, that is, a varying tenor, which is sometimes shown to be long by a plain virgula added to a letter.
The Latin words 'mora' and 'tenor' can mean 'any duration of time'. They do not imply length anymore than the question "How long is it?" implies no brevity. Hence, I would paraphrase this paragraph as follows:
Next, the "virgula plana" clause does not apply to the "varium tenorem" phrase. Aurelian really helps to understand this.
The first one is provably a light passing note (or somewhat equivalently, a slide). The second is a flutter, a light neighboring grace note, or however you want to perceive it. In both cases, a small note is inserted which is diminutive ("debilis"), "tremula" (in the sense of "timid"), and "graceful" ("vinnula").
The "virgula plana" clause cannot possibly refer to either the trembling inflection or the trembling repercussion. It cannot refer to the quilisma because the quilisma is never episematic. Nor can it refer to the repercussive grace, because the grace note in a repercussion is never notated. As such, the clause "which, whenever it is long, can be signified with a virgula plana" refers to the "tenor" in general, and not to the "tremula". Re-paraphrased:
The "varium tenorem" can describe either the "tenor" in general or the "tremula", depending on how you read the sentence.
I believe the first reading is correct, wherein "varium tenorem" describes the tenor in general, which may be longer, shorter, or trembling. The accusative of "varium tenorem" does not link it to "tremulam". "Longiorem" and "breviorem" are also accusatives, and also adjectives. "Tremulam" is an adjective. To my knowledge, it was hardly employed as a noun until after Guido's paragraph was commented upon decades later. But most importantly of all, the verb, which typically ends a clause in Latin, is found before "id est, varium tenorem".
Therefore, I believe "varium tenorem" describes the "tenor" in general, and not the "tremula" specifically, and the sentence should be read with a comma or dash between "tremulam" and "habent".
If you subscribe to the second reading, wherein the "tremula" is the "varying tenor", then you must explain how it is varying. "Varius" cannot mean "either short or long", since the tremula, which Aurelian equates to the quilisma and grace note, is neither short, nor long, nor ever marked with an episema. Perhaps "varius" could describe the nature of the note that is modulated by the "tremula" in a repercussion: in practice a repercussed note is three notes --- two notes separated by a graceful inflection ---, but one could argue it to be seen as a single long note that is 'bent' halfway through its production. In that sense the tenor can be said to be "varium". But I really think this argument is a stretch.
Summary:
"I have never thought of liquescence as being particularly connected either with melody transmission or consonant clusters, as you mention in point3."
Case 1: If you have heard "Ad te levavi" or a similar figure only ever sung with that 'liquescent' lower D between the first two syllables, then you are far less likely to omit the D, however the D is performed, because the D is ingrained in the memory as a part of the melodic shape. The 'liquescing' is just the effect of the note being diminished by the textual articulation. Similarly, with "verbum tuum" in Communion In Salutari.
Case 2: Written transmission, i.e. manuscript copying, increases the odds that an otherwise inconsequential vocal effect, such as liquescence, epenthetics, or whatever, be perpetuated. I think this is the chief reason why the 'Jerusalem' example you gave has the seemingly superfluous lower short note in both St. Gall and Laon: its presence in two manuscripts is more likely, in my opinion, owing to manuscript copying than to ingenious musical rationale.
"I would need real contrary evidence to persuade me that it is possible that the first note in those many instances did not receive the primary stress normally assigned to first notes of multiple note groupings on a syllable."
The contrary example you gave from Alleluia Jubilate Deo highlights the parallel between Laon and St. Gall of a short-long on one hand and a long-long on the other. I think the most reasonable explanation here is that the second note is indeed a long of importance --- it is a higher note to which the melody rises --- while the first note is by default a rhythmless note, i.e. an initium debilis or the like, meant to raise the melody up to the higher note, and that its own duration may be expanded from grace-duration to long-duration. What the first note cannot be is short, because a short is 1 chronos, 1 time unit, which is half of a rhythmic beat. It must therefore be either 0 beats (i.e. a vinnula or initium debilis or however you want to perceive), or else 2 beats, i.e. a long.
I've seen this long-substitution all over the Alleluias. But it is also in the more standard repertory as well. The opening figure D-F-G-F of Introit Dominus dixit, found also a few other Mode 2 neumatic antiphons, is properly interpreted as grace-short-short-long, 0 + 1 + 1 + 2 time units = 4 time units = 2 beats, an even rhythmic number. In the Offertory Benedicite gentes (Offertoriale Triplex, page 72), that first D is a long in both St. Gall and Laon, and the F-G-F is replaced with a torculus over "qui non amovit".
The rationale is clear to me. The upper second note has both rhythmic weight and melodic importance. As such, the first note may be weightless and unimportant, or it may be rhythmically acceptable (lengthened to a beat) but still unimportant, since, no matter whether it is 0 time-units or 2 time-units, its function is the same, namely to lead into the second higher note.
So, I think we cannot simplify all grace-, liquescent-, or debilis note interpretation to a single rule. That first note may be important in some cases but not in others. Attention is required to the rhythm, the melody, and how they together fulfill the function of each note. And this was Jan van Biezen's concluding point in his Gregorian chant article.
Point 2a is a good analysis and I would only add that N can produce a schwa as well.
With regard to point 2b, I would note that liquescence is a linguistically determined feature and that the stress of verbal syllables never falls on the consonant but on the subsequent vowel not just in speech but in all European traditional song and traditional ecclesiastical chant that I have heard. As for verbal stress, evidence of normal verbal stress being mismatched under the influence of melo-rhythmic stress needs to be substantial (like the examples we in Erit vobis) and where there is no liquescence mark, there is no evidence of any pitch occluding liquescence that requires marking.
Comparison of sources shows countless instances of single longs being varied by division into two shorts without any indication of any kind of meta tone or liquescence. In Scotland and Ireland, the instrumentalists do this all the time in reels and there is no liquescence on a fiddle or pipe.
I have never thought of liquescence as being particularly connected either with melody transmission or consonant clusters, as you mention in point3. There are too many examples of liquescence that do not connect with either of those features and liquescence does not occur for those reasons in my own tradition either.
The argument you make in point 4 would be unsuccessful if applied to my culture. There are plenty of instances in Scottish music of melo-rhythmic stress on what is only a slur preceding a main note so I could not make such an assumption about melo-rhythmic stress purely on that argument.
Kevin M. Rooney said:
I am grateful to everyone for engaging in public discussion in this way. I think this point is very important to discuss publicly. My own culture makes much use of epenthetic vowels in speech alone and, ever since starting to examine the subject, I have been somewhat bemused by certain opinions I have read about liquescence in Latin ecclesiastical chant.
In order to deal with one issue at a time, I will ignore the question of the emphatic production of an epenthetic vowel in singing, and will also ignore the question of the emphatic production of long sounded connnnnnnnnnsonants, and focus on the matter of liquescence of apparently short duration on sounding consonants only.
It should be superfluous to say that a sounded consonant must carry a pitch and therefore cannot do other than liquesce, whether to a lesser or to a greater duration of time. On a monotone, short liquescence is inevitable on every sounded consonant (no matter how long or short the duration) and on whatever epenthetic vowel one makes). When writing music down, it would be needlessly laborious and pointless to note down every occurrence of such liquescence on a monotone because it is impossible not to do such liquescence on a monotone.
For the most part, it is only therefore worthwhile noting liquescence when the sounded consonant is assigned a pitch that is i) different to the last pitch on the preceding vowel but similar to the first pitch on the following vowel, or ii) different to both the last pitch on the preceding vowel and to the first pitch on the following vowel.
I regard the above as the reasons why, in Montpellier 159, virtually every liquescence mark is accompanied by its own alphabetical pitch indication, and why that pitch is different to the last pitch of the vowel preceding the liquesced consonant.
In the case of letter L of the communion antiphon Jerusalem, quae aedificatur, I correspondingly regard this consonant as having to be liquesced whether or not it is marked in the ancient sources as having to be liquesced. I regard it so for the reasons given above. However, at (Jerus)ale(m), there is no change in pitch between the last note of the preceding vowel and the first note of the following vowel and there is no indication that the consonant has been assigned its own discrete higher or lower intervening pitch. It is therefore pointless to notate any liquescence.
If one takes the view that the first short note on the following vowel E is tantamount to a liquescent mark, that opens the door to every single first short note on a vowel being viewed as tantamount to a liquescent mark. I regard there as being no liquescent marks here because neither of the two conditions that would make noting liquescence worthwhile are seen to apply: we see merely the rhythmic pattern of the second and third syllables of the word extending to the fourth syllable of the word upon which there is a terminal flourish of notes.
In my own singing culture, the first pitch of a syllable is, as a rule, not protyptic. In Scotland, we would happily stress the first pitches here in both Messine and Swiss notations without the word ‘jolt’ entering our minds.
Why, objectively, would one regard one instance of a first short pitch of a vowel as protyptic and as not being like any another instance of a first short pitch? No beat stress (for either the first pitch or the second) is overtly marked in the notation. Why would one decide not to regard the first note as taking the normal stress? Even reference to Vaccia cannot answer this question for me because he explains his own portamento as existing in two forms, the first proptyptic and the second not.
When learning a tune from an instrumentalist, another instrumentalist could, in a given case, so easily confuse a slur with a dotted rhythm because he does not have the sound of a change of verbal syllable to clarify to his ear that the short note takes the rhythmic stress in that case, rather than the following long note.
As evidence, comparative analysis of the ancient sources of course is better than Vaccia and the Scottish song tradition with regard to this question, which is the only area of Jan van Biezen’s analysis that I find weak. While there are very good reasons for regarding the first note of the Sciant neume as ametrical and thus possibly proptypic, it is perhaps harder to make an argument for plain short notes (particularly in the Messine notation) being seen as playing the same role as in the Sciant neume.
The Al(lelu)ia in the Alleluia Jubilate Deo in page 222 of the Graduale Novum is a case in point. It shows Laon 239 as having one short pitch G followed by one long pitch B but both these pitches are long in St Gallen 359. Additionally, it shows Laon 239 as having one short pitch D followed by one potentially long pitch G but both these pitches are again long in St Gallen 359. St Gallen 359 is highly unlikely not to have stressed the first pitches of each syllable.
This is only one of many, many instances of one tradition recording two long notes at the very start of a syllable and another tradition recording them as one short followed by one long. As a Scotsman, with this kind of evidence I would need real contrary evidence to persuade me that it is possible that the first note in those many instances did not receive the primary stress normally assigned to first notes of multiple note groupings on a syllable.
Kevin M. Rooney said:
My own reading of the Commemoratio brevis does not regard it as presenting the tremula as a brief note. I will present the passage and an English translation in order to refer to it clearly.
Tenor uero id est ultimae uocis mora. qui in sillaba quantuluscumque est. amplior in parte. diutissimus uero in distinctione. signum in his diuisionis existit. Sicque opus est. ut quasi metricis pedibus cantilena plaudatur. et aliae uoces ab aliis morulam duplo longiorem uel duplo breuiorem. aut tremulam habeant. id est uarium tenorem. quem longum aliquotiens apposita litterae uirgula plana significat.
A/the hold, that is, a/the delay of a/the last note - which however small it is in a syllable is larger in a part and longest in a distinction - is a sign of division in these, and so it is necessary that a little song should be clapped as if with metrical feet and that some notes should have a duration twice as long or twice as short compared with others, or a/the shake, that is, a /the varying hold, which is sometimes shown to be long by a plain virgula added to a letter.
Guido at least describes the tenor as the (or a) mora ('delay') of the last note. He at least describes the tremula as the (or a) uarium tenorem (varying hold). I don't regard the concept of 'hold or delay of last note' as applicable to the first note of the Sciant neume.
Kevin M. Rooney said:
Jerome F. Weber said:
I am grateful to Solesmes for its efforts, to the early mensuralists for theirs, to Vollaerts for his, and to the semiologists for theirs, but I find Jan van Biezen's metered-rhythm theory to surpass them all in scientific accuracy and promising benefit. Namely because it provides a strikingly elegant solution to nearly all of the problems of 'mensuralism' that ties it into not only the notation, but also the words of the fathers, the customs of the East, and modern musicological knowledge on oral transmission.
Semiology, on the other hand, relies on sola notatio. One cannot interpret the notation without guidance anymore than one can interpret the Bible without guidance.
-
1
-
2
-
3
-
4
-
5
of 7 Next