Trill

When did the trill convention change?

Dale Gedcke wrote (November 29, 2004):
Bradley Lehman wrote:
"We're damned if we play the ornament with two twiddles instead of three, or four instead of three, or omit it altogether, or put on a termination, or speed up or slow down, or start on the upper note or on the main note."
MY QUERY:

I have read in various places that the convention for baroque music was to start the trill on the upper note. In fact J. S. Bach published a guide prescribing that convention for his music (see: http://members.aol.com/kjvisbest/jsb_ornm.htm).

David Hickman, in his Piccolo Trumpet Big Book, advises that starting on the upper note of the trill is strictly prescribed at the cadence (end of a significant phrase or passage). He suggests that the trill can start on either the upper or lower note at other points in Baroque compositions, depending on which sounds better.

In more modern music the convention is to always start the trill on the main note.

In the published prompts for the trills in Haydn's Trumpet Concerto in Eb, the trills with turns start on the main (lower) note. That concerto was composed in the early 1800s.

My question for those more experienced in the historical development of music is simple: "When did the convention change to starting the trill on the lower note? Was it in the 1800s or the 1900s?

Bradley Lehman wrote (November 30, 2004):
Dale Gedcke wrote:
< I have read in various places that the convention for baroque music was to start the trill on the upper note. In fact J. S. Bach published a guide prescribing that convention for his music (see: http://members.aol.com/kjvisbest/jsb_ornm.htm ).>
It changed gradually throughout the whole of the 18th century, and into the 19^th. There was no definite cut-over point (obviously). It's a very complex issue that goes not only into trills and mordents, but also the various types of appoggiaturas from both directions, and the handling of rhythm, and speeds of ornamentation, and Italian/French/German/other style, and improvisation of other ornamentation that doesn't fit into the signed/named symbols, and so much more. It simply can't be summarized into a couple of paragraphs.

It also gets confusing back into the 17th century (i.e. REALLY the Baroque period) and on into the 16th, as some of those trills were probably starting on the main note!

A decent presentation is Frederick Neumann's 600-page book..... :)

< David Hickman, in his Piccolo Trumpet Big Book, advises that starting on the upper note of the trill is strictly prescribed at the cadence (end of a significant phrase or passage). He suggests that the trill can start on either the upper or lower note at other points in Baroque compositions, depending on which sounds better. >
Essentially, yes, but again it's not as simple as all that. Depends how the note above the trill has been approached, and on the position within the meter, and possibly also on some other factors such as avoiding direct fifths or oblique fifths or octaves with other parts....... It also depends how fast the music is going, if there's even time to slip more than two or three notes in there at all. Sometimes in the fastest stuff trills have to be simulated by a single crushed note.

< In more modern music the convention is to always start the trill on the main note. >
Unless the composer or arranger was heavily influenced by 18th century music in the styles underlying the composition, in which case this is reversed. For example, when I write any ornamentation symbols into my compositions, I expect them to adhere to normal 18th century usages; and to perform my music correctly, somebody would have to know about me that I came to the composition process as a specialist performer of that earlier repertoire.

And, I don't expect that such a situation is all that uncommon anymore. If we play Bartok's or Rachmaninoff's or Godowsky's arrangements of Baroque compositions, what's correct: to render the ornaments the way the arranger might have known, or the way the earlier composer might have known? How about Grieg's suite "In Holberg's Time"? Play the ornaments from Holberg's time, or Grieg's? I ran into this same question in an unexpected way, during my oral defense. In one of my projects being defended I'd given a performance of Händel's B-flat variations, the same one that Brahms later arranged for piano as a different set of variations. Pretty much automatically, I rendered all the trills starting on the upper note since that was usual practice around Händel. Somebody on my committee knew the Brahms version and had never played the Händel; and then challenged me on that point (a question that blind-sided me...) because my performance had sounded unexpected, not starting them on the main note as that player would do in the Brahms. So, what's a pianist to do with the Brahms: go figure out how concerned Brahms was with 18th century accuracy, or ignore the fact that this is derivative from a Händel composition, or take some hybrid approach? How do we decide? This is a big question in the area of authenticity. Whose authenticity? According to what, and when?

< My question for those more experienced in the historical development of music is simple: "When did the convention change to starting the trill on the lower note? Was it in the 1800s or the 1900s? >
See above. It goes in both directions, into and back out of the 18th century from both sides....

Doug Cowling wrote (November 30, 2004):
Bradley Lehman wrote:
< It also gets confusing back into the 17^th century (i.e. REALLY the Baroque period) and on into the 16th, as some of those trills were probably starting on the main note! >
Much amusement and frustration can be enjoyed in trying to plot the trills in the alto aria "Awful Pleasing Being Stay" in Händel's oratorio "Joshua". Händel tosses in a few markings and then leaves you to decide where the rest are implied and how they should be appproached, It has them all, cadential and initial trill, ornamental and melodic. Interestingly, Robert King in his period instrument recording plays the score as it is written with no ornamental or rhythmic editing.

Bradley Lehman wrote (November 30, 2004):
[To Doug Cowling] Sorta turns into a puzzle. I like the general guideline: the signed ornaments being an indication to do something musical and tasteful according to experience and training, with melodic and rhythmic flexibility...more than anything pedantic or trying to solve all the "problems" on paper. More important to sound free and loose, musically intense and relaxed at the same time, than to amuse pedants with a presumably immutable solution.

Thomas Braatz wrote (November 30, 2004):
Dale Gedcke wrote:
>>"When did the convention change to starting the trill on the lower note? Was it in the 1800s or the 1900s?<<
Here is a recent summary about this written by Clive Brown: [Grove Music Online, Oxford University Press, 2004, acc. 11/29/04]
>>The elaborate systems of ornament signs developed by 18th-century keyboard players was not widely adopted, even in keyboard music, during the Classical period. For other instruments composers rarely employed anything but 'tr', the mordent sign and various forms of turn sign....The sign 'tr' usually indicated a trill with a number of repetitions of the upper auxiliary, while the mordent sign indicated only one or two repetitions (depending whether it began with the auxiliary); however, each of these signs was sometimes used with the meaning usually applicable to the other. The various forms of turn sign cannot reliably be related to particular melodic and rhythmic patterns; sometimes they too could be synonymous with 'tr', and in manuscript sources the distinction between [various examples given] is often unclear.

During the 19^th century, as composers became concerned to take greater control of their music, they increasingly wrote out ornaments in full. The progression is neatly illustrated by Wagner's turns: up to Lohhe used signs, but in Tristan and his later operas he always incorporated the turns into the notation. Inverted mordents were often indicated either by small notes or in normal notation, and even trills were sometimes fully notated, for instance by Dvoøák (op.106) and Tchaikovsky (opp.64 and 74). Considerable controversy has been generated by the question of how trills in music from the period 1750 to 1900 should begin. Scholarship has clearly shown that, although the upper-note start was never quite as self-evident as advocates such as C.P.E. Bach implied, it was undoubtedly the dominant practice in the mid-18th century. When and where a general preference for a main-note start began to emerge remains uncertain. Moser identified the strongest support for the upper-note start as being in north Germany; he asserted, however, that in Mannheim the trill was to begin from above only if specifically notated, and that C.P.E. Bach's authority was countered by 'the powerful influences which stemmed from the Viennese masters of instrumental music' (Violinschule, iii, 19-20). What evidence Moser may have possessed for this statement, other than received tradition by way of Joachim, remains unclear. Certainly, a considerable number of the trills on the musical clocks from the 1790s containing Haydn's Flötenuhrstücke begin on the main note, but there is no consistency and no connection with Haydn's notation. Arguments for and against Mozart's preference have been advanced, and the matter has been exhaustively examined by Neumann. For Beethoven, too, the evidence is largely circumstantial. In 1828, however, Hummel published his unambiguous opinion that a main-note start should be the norm, and Spohr followed suit a few years later. Baillot offered four different beginnings without recommending the primacy of any. Some 19th-century composers took trouble to indicate the beginnings of trills, particularly to show a start from below, and their manner of doing this was used by Franklin Taylor in 1879 as evidence for their normal practice. It seems probable that among major 19th-century composers Weber, Chopin and Mendelssohn generally favoured an upper-note start. In this as in other aspects of performance, however, dogmatism and rigidity are undoubtedly out of place.<<

Gabriel Jackson wrote (November 30, 2004):
Thomas Braatz wrote:
"In this as in other aspects of performance, however, dogmatism and rigidity are undoubtedly out of place.<<"
Indeed. Would Thomas tell us whether he agrees with this statement, which he quoted with approval (presumably), since his every pronouncement would indicate a devotion to dogmatism and rigidity?

Thomas Braatz wrote (November 30, 2004):
Gabriel Jackson asked me whether I agreed with Clive Brown's statement regarding trills performed in 19^th century musical literature. Brown's statement, taken out of context, reads:
>>In this as in other aspects of performance, however, dogmatism and rigidity are undoubtedly out of place.<<
Gabriel's loaded and deliberately misleading comment was:
>>Indeed. Would Thomas tell us whether he agrees with this statement, which he quoted with approval presumably), since his every pronouncement would indicate a devotion to dogmatism and rigidity?<<
For the past few centuries, there have always been composer/performers who have not followed slavishly the performance standards of their time and place, just as there have also been a few who have meticulously tried to indicate in their scores just how they wanted to have their ornamentation performed. J. S. Bach, after a recent BCML discussion on this matter based upon clear evidence arising out of the Scheibe/Birnbaum controversy, is an example of a composer who knew that almost everyone else who would perform his music did not possess his high standard of good taste in music, hence Bach took great care, as much as it was possible for him, in preserving his musical intentions and sparing them from being diminished by the efforts of performers who believe they understand better (have better musical taste than Bach) how his music should be performed. These are performers who are absolutely opposed to what they term 'a devotion to dogmatism and rigidity' not realizing that adhering closely to Bach's intentions means demonstrating a willingness to learn from a great master whose good taste in performance practice is unparalleled.

Gabriel Jackson wrote (November 30, 2004):
[To Thomas Braatz] This doesn't answer the question, of course. But why should that suprise anyone?

Gabriel Jackson wrote (November 30, 2004):
Thomas Braatz wrote:
"Gabriel's loaded and deliberately misleading comment was:
>>Indeed. Would Thomas tell us whether he agrees with this statement, which he quoted with approval presumably), since his every pronouncement would indicate a devotion to dogmatism and rigidity?<<"
I'm surprised that Thomas, as a pedant, is unable to distinguish between a comment and and a question. As to 'loaded' and 'deliberately misleading' ....how?!!

Dale Gedcke wrote (November 29, 2004):
[To Thomas Braatz]
MY COMMENTS:

In his book, The Piccolo Trumpet Big Book, David R. Hickman talks about adding free ornamentation to Baroque music and then advises:
"Be aware of the national styles of Baroque composition.

French music shows a predilection for shorter and quicker ornamentation such as trills, mordents and turns. The use of too many specific ornaments usually precludes any free ornamenting of the line.

Italian music, on the other hand, requires the most free ornamentation. Quantz states the following concerning the Italian style:

'In music composed in the Italian taste, much is left to the caprice and capacity of the player. ...... One cannot deny that for a piece to make its full effect in Italian music, the performers contribute as much as the composers.'

Although German composers manifest a strong influence from both the Italian and French schools, they wrote out most of their ornaments. Because German music tends to be highly contrapuntal, free ornamentation would often disrupt the contrapuntal flow."

That analysis from Hickman seems consistent with J. S. Bach being more specific about what ornamentation he wanted in his compositions.

Dale Gedcke wrote (November 29, 2004):
Here is the specific advice that David Hickman gives trumpet players concerning trills in Baroque music. The quotation is from his book, The Piccolo Trumpet Big Book (highly recommended for anyone considering adding a piccolo trumpet to their portfolio).

"The trill has two main functions in Baroque music. It can be used to intensify harmonic activity or for melodic decoration. In the former function it is located at a cadence and is known as a cadential trill.

The pertinent questions regarding correct performance of the trill are which note to begin on, and how to end the trill. These are contingent upon the context in which the trill is performed, and which of its two functions the trill is fulfilling. Each of these cases will be considered.

CADENTIAL TRILL: This is the most obligatory ornament in Baroque music. Its use became so standard that often composers did not bother to indicate it with a symbol. The cadential trill has three parts:

1. The PREPARATION is the upper note. The cadential trill always begins on this note. It occurs on the beat and should be slightly louder and prolonged.

2. The TRILL immediately follows the preparation. The speed at which it is carried out should tastefully suit the piece. (Allegro trills are immediately fast, whereas Grave trills begin slowly and accelerate.)

3. The TERMINATION of a cadential trill can be carried out in two ways: With an anticipatory note [example deleted], or with a turned end (most common in late Baroque, Rocco and Classical works) [example deleted].

The three parts of the trill should flow together as naturally as possible. The harmonic function (tension release) is achieved because of the appoggiatura effect that is created by beginning the trill on the upper note.

THE MELODIC TRILL: Trills that occur at places other than at cadence points are called melodic or embellishing trills. The treatment of trills is not rigidly defined as with the cadential trill. It may begin on the upper or the lower note, whichever better suits the line. Melodic trills may use a terminating turn or anticipatory note, although more often than not the trill merely stops. [example from J.S. Bach's B Minor Mass deleted]"

ADDING MY OWN COMMENTS

I have two different scores of the Trumpet Voluntary by Jeremiah Clarke (1673 - 1707) where the trills are written out, one arranged by the famous trumpet player of the 1950s, Roger Voisin, (1962 copyright), and another arranged by Carl Strommen (copyright 1989).

The former advises playing both the cadential and melodic trills beginning on the upper note and ending with a turn. The latter advises melodic trills beginning on the upper note and ending with a turn. But, it offers several alternatives on the speed of the trill and a possible longer note just before the terminating turn. The cadential trills begin on the upper note and end with an anticipatory note. From looking a few other scores of the Trumpet Voluntary, where the details of the trill are not written out, the Strommen version seems more consistent with the original score (which does spell out the terminating turns for the melodic trills and anticipatory notes ending the cadential trills). Also, most recent recording artists seem to follow the style specified by Strommen on this particular piece.

What this demonstrates is that the artist has final control over how the ornamentation is actually performed. The artist must have the skill and training necessary to exhibit good taste. The listener, of course, must decide for himself, whether or not he likes the style of the performance.

Bradley Lehman wrote (November 30, 2004):
< What this demonstrates is that the artist has final control over how the ornamentation is actually performed. The artist must have the skill and training necessary to exhibit good taste. The listener, of course, must decide for himself, whether or not he likes the style of the performance. >
Good points. And, if the listener as critic doesn't like the style of the performance, such a critic is not in the position to assert that the performer is wrong or has bad taste; but only that the contemptuous critic himself doesn't fancy it (and/or understand it).

=====

Some bits from Geminiani's "Treatise of Good Taste in the Art of Musick" (1749, but reflecting his own conservative preferences back to those of his teacher, Corelli, d1713):

"Playing in good taste doth not consist of frequent Passages, but in expressing with Strength and Delicacy the Intention of the Composer. This Expression is what every one should endeavour to acquire, and it may be easily obtained by any Person, who is not too fond of his own Opinion, and doth not obstinately resist the Force of true Evidence. I would not however have it supposed that I deny the powerful Effects of a good Ear; as I have found in several Instances how great its Force is; I only assert that certain Rules of Art are necessary for a moderate Genius, and may improve and perfect a good one. (...)"

The rest of the book then goes on to illustrate a bunch of specific techniques, carefully notated and explained. Good taste is a quality that a good performer can apply objectively to the performance of a piece of music, bringing out its beauty by going beyond the page's notation in specific ways. (Melodic and rhythmic ornamentation, and the addition of crushed dissonant tones.)

Indeed: the craftsmanship of a good performance consists in large part of following such specific advice above one's own Opinion, because the techniques have been proven empirically to move people. (Not just in Geminiani's own time, but now: when good performers do those specific things, whether it's 18^th century music or something else, people are moved.) And that's something that Quantz also wrote (1752): to judge the quality of a performance and a composition, closely observe the reactions of an audience that consists of both connoisseurs and ordinary folks.

Geminiani's melodic advice is the kind of stuff practised by Bing Crosby, Barbra Streisand, Petula Clark, Linda Ronstadt, and other popular 20^th century singers (although they probably didn't get it from reading Geminiani). Why are these singers popular? Because they move people, with performances that are beautiful. These techniques, objectively, when used well in service of the music, move people! Geminiani said so explicitly, and he is right. What is a correct way to perform music, tastefully? With a delivery that moves people.

Geminiani again: "Lastly, as the chief End I have in view, is to contribute as far as my Abilities will permit, to the Perfection of an Art that I love, and to rescue the Character of Musician from the Disgrace and Contempt which the Follies of ignorant Pretenders have brought upon it, I hope no acknowledg'd Master will lend his Countenance to the Misconstruction which those Pretenders may think their Interest to pass upon it."

By request

Douglas Cowling wrote (June 7, 2008):
<>
Somebody please start an argument about whether trills should always begin on the upper note.

Neil Mason wrote (June 8, 2008):
[To Douglas Cowling] Trills should always finish on the lower note.

Neil Halliday wrote (June 8, 2008):
Neil Mason wrote:
> Trills should always finish on the lower note.<
I was wondering how to reply to Doug's heartfelt request.

Thanks for the answer (and the chuckle).

Joost wrote (June 8, 2008):
[To Neil Mason] Maybe it is more correct to say that trills always shoud finish on the main note (since a mordent, which is some kind of trill as well, makes use of the main note and the note below it).

Cara Emily Thornton wrote (June 8, 2008):
[To Neil Mason] Finally something we can agree about ;) I guess the jury is still out about whether we begin on the upper or lower note?

Neil Mason wrote (June 8, 2008):
Cara Emily Thornton wrote:
< Finally something we can agree about ;) >
Ah, it appears not - what a hoot.

Ed Myskowski wrote (June 8, 2008):
Cara T wrote:
>Finally something we can agree about ;) I guess the jury is still out about whether we begin on the upper or lower note?<
in response to Neil M.
> Trills should always finish on the lower note.<
I interpreted Neil as intentionally, humorously, non-responsive to Dougs request (to discuss which note should a trill begin on?) Even if by accident, I find it amusing.

Ed Myskowski wrote (June 8, 2008):
<So what about a trill 'ex nihilo' which begins a melody rather than forming part of its conclusion?>
Trills? Nothing but <molto vibrato> to my ears, to be eschewed at all times by HIPsters.

N.B. - HIP is an acronym, not to be confused with a SHOUT.

Another question on trills

Thérèse Hanquet wrote (June 8, 2008):
Doug's suggestion reminded me a question I had about trills in Bach's scores.

There are here and then trills in choruses of the cantatas. An example fresh from my memory is BWV 75/1. Each time, we (in the choir) encounter a trill, we ask the conductor what we are expected to do. Sometimes we are told not to perform it, sometimes the conductor suggests that only those feeling at ease with it perform it. There are also discussions about the length / speed of the trill(s).

It is indeed quite difficult for many singers to perform a trill all perfectly together. Even in duets, when trills fall together on the two vocal parts (I do not know if this happens often, but I have encountered it for example in Pergolesi's Stabat Mater), it takes agreement between the singers and practice to achieve a good sounding result. On the other hand, if the trills are not together it can sound awful. I have also taken part in a choir performance of the same Stabat Mater, and there our conductor told us to leave out all trills in the parts sung by the choir.

Thus... my question was: do trills in choruses not give some evidence that the concerned cantata (I underline) was to be performed by OVPP? or maybe TVPP (two/three voices per part)? Or is there evidence of trills satisfactory performed by large choirs?

Another element that struck me is that trills in choruses generally afone voice at a time, even in parts where several voices sing the same text (example: measure 21 of BWV 75/1 - only the alti have a trill). This may be also a hint of the difficulty of performing simultaneous trills. Or is there an artistic reason?

Thanks for your insight

Douglas Cowling wrote (June 8, 2008):
Thérèse Hanquet wrote:
< Thus... my question was: do trills in choruses not give some evidence that the concerned cantata (I underline) was to be performed by OVPP? or maybe TVPP (two/three voices per part)? Or is there evidence of trills satisfactory performed by large choirs? >
This is certainly a question that continues to nag at me. We know that Handel's works were sung MVPP and contain ornaments. The most famous is the trill in the Hallellujah Chous on "And he shall reign for ever and ev-[trill]-er" when it was reused as the finale of the "Foundling Hospital Anthem".

On the other hand, trills and ornaments were daily fare for 18^th century singers and didn't have to be learned in the way modern musicians have to make a distinction between French and German Baroque trills, not to mention Classical and Romantic ornaments.

In my expereince, most choral conductors who are not Baroque specialists ignore all choral ornaments. This is particularly unfortunate in French Baroque works by composers such as Delalande which are full of embellishments.

Some conductors ask the singers to add just the openng apoggiatura of the trill or the appgiatura and final anticipatory note. This at least preserves the composer's harmony. One clever conductor asked the choir to sing the appoggiatura and had one or two of the singers execute the trill. This works rather well if it's a small group, say 16-24.

Off this evening to hear Cantata BWV 26, "Ach Wie Flüchtig" sung as part of a Bach Vespers. My son, who sings countertenor, says that the guest Russian conductor is really putting the "flüchtig" in the opening chorus. I have to confess that I arranged orchestral parts for the congregational chorales in the service, although there is little historical justification -- it's fun to sing with an orchestra.

More information at: http://www.theredeemer.ca/pages/news.html

Thérèse Hanquet wrote (June 8, 2008):
[To Douglas Cowling] Thanks Doug, I wish you a good concert. Yes it is fun to sing with an orchestra!

We are precisely going to rehearse Haendel's Halleluyah tomorrow night, for a wedding mass right at the summer solstice. I will pay special attention to these trills... All the rest of the music will be by Bach, notably the opening chorus of BWV 193 in the beginning and that of BWV 120 at the end... A rather exultant wedding! The bride sings regularly as soloist with us, this may explain that...

Ed Myskowski wrote (June 8, 2008):
I previously wrote (Re: By request):
>Trills? Nothing but <molto vibrato> to my ears, to be eschewed at all times by HIPsters.<
No sooner did I hit send, and then open a book (the incorrect sequence, not the first time for me, by a longshot), but I realized this is not so bizarre, and might not be recognized as an attempt at humor, (in the spirit of the request?). Whatever.

In particular, I note the very old, historic origin of the trill, and its not exactly precise distinction from tremolo and vibrato. Leaving the modern HIPsters without an H to stand on?

I also note the favored English terminology: <shake>. Could the effect have its ultimate origin in the natural tendency (?) of the human voice to tremble, with fear?

To the vocalists: can you get close to a trill with <molto vibrato>, or can I retain that as a joke.

Some wag (Shakespeare? I will get the book out, as soon as I hit send) wrote, to the effect: <Humor is nothing but the truth writ large.>

Thérèse Hanquet wrote (June 8, 2008):
Ed Myskowski wrote:
< [...]
To the vocalists: can you get close to a trill with <molto vibrato>, or can I retain that as a joke. >
You should have a very "well tempered" vibrato then, which makes just one tone / one half tone according to the key and to the starting / finishing notes... but who knows...

Cara Emily Thornton wrote (June 8, 2008):
Ed Myskowski wrote:
< In particular, I note the very old, historic origin of the trill, and its not exactly precise distinction from tremolo and vibrato. Leaving the modern HIPsters without an H to stand on? >
Umm, I think the distinction is very precise and is reflected in Baroque treatises containing 'written out' forms of each ornament.

< I also note the favored English terminology: <shake>. Could the effect have its ultimate origin in the natural tendency (?) of the human voice to tremble, with fear? >
To me, the word shake has the connotation of shortness when applied to a trill. In other words, that the note being trilled is of short time value.

< To the vocalists: can you get close to a trill with <molto vibrato>, or can I retain that as a joke. >
A joke only, please. A proper trill contains two and only two different, rapidly alternating notes, one of which is notated in the score, the other of which is the note above it. The two notes in question can be either a half step or a whole step apart, depending on the context. And yes, to properly execute a trill, you should be able to hear the difference (half-step vs. whole-step).

Neil Mason wrote (June 9, 2008):
Ed Myskowski wrote:
<< To the vocalists: can you get close to a trill with <molto vibrato>, or can I retain that as a joke. >>
In all seriousness, when I try to teach vibrato to one of my singing students, I sometimes use a trill as a basis to start with.

In both vibrato and trills, the singer needs to feel a certain "abandonment" of control, in other words freedom.

Ed Myskowski wrote (June 9, 2008):
Cara T. wrote, in response to my post:

EM:
<< To the vocalists: can you get close to a trill with <molto vibrato>, or can I retain that as a joke. >>
CT::
>A joke only, please. A proper trill contains two and only two different, rapidly alternating notes, one of which is notated in the score, the other of which is the note above it. The two notes in question can be either a half step or a whole step apart, depending on the context. And yes, to properly execute a trill, you should be able to hear the difference (half-step vs. whole-step).<
EM:
Note my special effort for clarity of the thread (EM and CT clearly indicated). Supplemental help available, on request.

I believe Cara answered the question I posed (with intended levity on my part). Vibrato can never be as wide as a half step (twelve tone step?), anything less is vibrato.

There are singers I will go back to for a second listen, but I expect Cara is exactly right. I do not think anyone has ever performed <vibrato> approaching a twelve tone step. Additional thoughts invited. I have Yma Sumac in the back of my mind.

As a HIPster might say, either sing the note, or sing a trill. Should the trill begin on the half, or whole, tone above the written note?

Q.E.D. No, wait, that means something else.

<Up and down, I will lead them up and down>. Or down and up. That is Shakespeare (via Puck), no book required.

Cara Emily Thornton wrote (June 9, 2008):
Neil Mason wrote:
< In all seriousness, when I try to teach vibrato to one of my singing students, I sometimes use a trill as a basis to start with.
In both vibrato and trills, the singer needs to feel a certain "abandonment" of control, in other words freedom. >
I am astounded when you speak of teaching vibrato to singers. I've never met a singer yet who didn't do it more or less naturally. The problem is that they often do too much. And especially in the case of Baroque music... Often folks who make the transition to early music have to unlearn the habit of automatic vibrato (just as in the case of string players and no doubt others as well who make the same transition).

Douglas Cowling wrote (June 9, 2008):
Cara Emily Thornton wrote:
< I am astounded when you speak of teaching vibrato to singers. I've never met a singer yet who didn't doit more or less naturally. The problem is that they often do too much. >
In my experience, the greatest singers treat vibrato as an ornament which they can control and apply. There are few more exciting moments than in a Rossini aria when a soprano finishes a coloratura run focussed as head tone and then begins a sustained note and applies vibrato allowing the size and pulse of the ornament to grow then pull back. Alas, too many singers do not have the breath control and accuracy of pitch to execute vibrato. The natural vinrato at 20 will become a wobble at 40.

The dominance of head tone, especially among sopranos, in much modern Baroque practice has meant the loss of an authentic vibrato. In an aria such as "Ebarme Dich", that is a major interpretative loss.

James Atkins Pritchard wrote (June 10, 2008):
Douglas Cowling wrote:
< The dominance of head tone, especially among sopranos, in much modern Baroque practice has meant the loss of an authentic vibrato. In an aria such as "Ebarme Dich", that is a major interpretative loss. >
Which is why I find myself still turning to Christa Ludwig...

Cara Emily Thornton wrote (June 10, 2008):
Douglas Cowling wrote:
< In my experience, the greatest singers treat vibrato as an ornament which they can control and apply. >
This is precisely what I am speaking of. In Baroque performance practice, at least from what my friends who read the treatises and such tell me, vibrato exists but is an ornament - not to be automatically applied. It is rare to
find a singer (and probably even rarer to find a string player) who does only 'modern' (i.e. later than Baroque) music AND treats vibrato as an ornament.

< Alas, too many singers do not have the breath control and accuracy of pitch to execute vibrato. The natural vinrato at 20 will become a wobble at 40. >
I've heard some singers with pretty wide vibrato in my time - for some reason they seem to be female altos for the most part. But they at least kept on pitch, even if they oscillated pretty widely around it... But that's a rarity. I mean, is someone really 'a singer' if they don't have breath control and pitch accuracy?

< The dominance of head tone, especially among sopranos, in much modern Baroque practice has meant the loss of an authentic vibrato. >
I have more often heard of what you are talking about here as something to the effect of 'forward placement'. But yeah, it does amount to cutting off the 'chesty' or 'dark' element of one's voice. On the other hand, I think it is possible to leave the 'chesty' or 'dark' element in the mix and still have little or no vibrato. And on the other hand, those Rossinian sopranos whom I deleted have head voice galore, but I never heard anyone accuse them of having too little vibrato...

My two cents for the night.

Neil Mason wrote (June 10, 2008):
Cara Emily Thornton wrote:
< I am astounded when you speak of teaching vibrato to singers. I've never met a singer yet who didn't do it more or less naturally. The problem is that they often do too much. And especially in the case of Baroque music... Often folks who make the transition to early music have to unlearn the habit of automatic vibrato (just as in the case of string players and no doubt others as well who make the same transition). >
In an ideal world vibrato would come to singers naturally, and for most lucky people it does.

However this is not true for everybody. Yes, it is true that some people have too much (or perhaps too wide) vibrato.

You mention string players. Some weeks ago I was at a master class given by Neil Semer of New York. He was asked "do you teach singers vibrato or does it just happen?", pretty much what I was talking about. His answer was that string players don't just let it happen!

There is little consensus about the amount of vibrato that is "right" for early music, except to say (as I think you do) that many opera singers have too much for early music. But IMHO it is wrong to have none. Even Ruth Holton uses it sparingly, and nobody could complain about her clarity.

This is an interesting topic. The speed of (natural) vibrato seems to be changing over time, and it is now much slower than about fifty years ago. I have no idea of the reason.

Douglas Cowling wrote (June 10, 2008):
Neil Mason wrote:
< This is an interesting topic. The speed of (natural) vibrato seems to be changing over time, and it is now much slower than about fifty years ago. I have no idea of the reason. >
I actuallly had this very debate with a voice specialist during the intermission at a recent performance of the "Barber of Seville" in Toronto. My wife and I didn't like the tenor because he had what we thought was a "small" voice and very unlike current stars like Roberto Alagna who has a huge instrument. Our friend argued that this tenor had a scale of voice which was much more authentic for the period. We had to agree that his coloratura was crystal clear and that he had a real embellishing vibrato unlike the wide bluster that passes for 'expressive' interpretation in other 'can belto" singers. I suspect that if we reach even further back in time to Bach, it becomes even more difficult to reconstruct vocal technique in his cantatas.

Ed Myskowski wrote (June 10, 2008):
Cara T. wrote
>I've heard some singers with pretty wide vibrato in my time - for some reason they seem to be female altos for the most part. But they at least kept on pitch, even if they oscillated pretty widely around it...<
Isnt this the issue in much of the BCML discussion on the topic? Does <on pitch> mean that the vibrato is centered on the correct pitch, or simply that the correct pitch is somewhere within the vibrato range.

I presume your statement means <centered on pitch>, but can you confirm that?

Cara Emily Thornton wrote (June 10, 2008):
[To Ed Myskowski] Yes, that is what I meant. If it's only 'somewhere in the range', then that is indeed a wobble...

Cara Emily Thornton wrote (June 10, 2008):
Cara Emily Thornton wrote:
<< Often folks who make the transition to early music have to unlearn the habit of automatic vibrato (just as in the case of string players and no doubt others as well who make the same transition). >>
Neil Mason wrote:
< You mention string players. Some weeks ago I was at a master class given by Neil Semer of New York. He was asked "do you teach singers vibrato or does it just happen?", pretty much what I was talking about. His answer was that string players don't just let it happen! >
Absolutely not. It is a real rarity for someone to have a natural vibrato on a stringed instrument. Although I recall reading the statement of someone from the old school, I think the guy was from Romania, whose teacher really did subscribe to the idea that it has to be natural and refused to 'teach vibrato'. But nowadays, in general, at a certain point, God knows how this point is determined, the teacher decides it's time for you to 'learn vibrato', and gives you some exercises to do...

This is the thing, however: I tend to think that for stringed instruments, ideally, the vibrato ought to be in some way a response to what you are feeling on the fingerboard. Which means that there is just no basis for the vibrato until your intonation is absolutely crystal clear (otherwise you'll be feeling dissonance on the fingerboard - at least for those lucky few whose fingers are sensitive enough to tell the difference, which I understand is a real rarity).

But if that is how you approach vibrato, it is going to naturally lead to using it as an ornament. Except maybe if the relationship between your fingers and the fingerboard is just so spot on right from the beginning that you have that basis for a vibrato right from the beginning - which I understand is kind of a must even for Baroque music if you are playing on metal strings (i.e. a modern instrument), to neutralize that... metallic quality, but is a non-issue if you are playing on 'on guts' ('na jelitach', as we say in the Polish early music community).

The matter is more complicated when you are dealing with a singer, in that there is no apparent 'fingerboard', much less fingers, in the mechanism. Although I could swear that when I've got my placement just right, I can feel fingers dropping on a fingerboard as I move up the scale, and picking up as I come down... But never mind. I think a lot of the problem with modern singers (or indeed, other modern musicians, now that I think of it) is that instead of vibrato being a response to that perfect placement, it's used for quite the opposite purpose: to hide the fact of one's not-quite-perfect placement.

< This is an interesting topic. The speed of (natural) vibrato seems to be changing over time, and it is now much slower than about fifty years ago. I have no idea of the reason. >
I personally would bet that it has to do with the increased sound volume required over time. The tendency is for bigger and bigger voices, which I suppose is in turn caused by bigger and bigger halls, and I recall a discussion some time back on this list to the effect that as the volume increases, the vibrato has to increase too. It's true on the violin as well, actually, that you increase the vibrato as the volume goes up. I guess the point is that with greater amplitude of sound waves proper, there has to be an increased vibrato to get the intended effect.

Another thing I notice about modern singers nowadays - although you only really notice it if you're hearing the performance live, it's not so noticeable if you are hearing a recording - is that they seem to be fighting with the hall, to project their voice (Doug Cowling's amusing expression 'can belto' comes to mind here, at any rate I get a stiff neck and even a headache listening to it). I suspect that if they were treating the hall as part of their instrument and working with it instead of against it, the effect would be very different.

Indeed, I wonder if there has been some sort of change in philosophy over time, whether perhaps way back when, there was more of a sense of working with the space instead of against it. I've never heard of anything like that being written about in the treatises, it may be there is just no documentation to lead us to a conclusion in this matter, but it is interesting to speculate anyway...

My two cents for now, my client is waiting for the last bit of that translation job...

Neil Mason wrote (June 10, 2008):
Bottom posting.

Neil Mason wrote:
<< This is an interesting topic. The speed of (natural) vibrato seems to be changing over time, and it is now much slower than about fifty years ago. I have no idea of the reason. >>
Douglas Cowling wrote:
< I actuallly had this very debate with a voice specialist during the intermission at a recent performance of the "Barber of Seville" in Toronto. My wife and I didn't like the tenor because he had what we thought was a "small" voice and very unlike current stars like Roberto Alagna who has a huge instrument. Our friend argued that this tenor had a scale of voice which was much more authentic for the period. We had to agree that his coloratura was crystal clear and that he had a real embellishing vibrato unlike the wide bluster that passes for 'expressive' interpretation in other 'can belto" singers. I suspect that if we reach even further back in time to Bach, it becomes even more difficult to reconstruct vocal technique in his cantatas. >
I'm not sure what the link is between that and the speed of vibrato, but never mind!

I believe your friend is most probably correct about Rossini tenors. I don't have the reference handy, but Rossini himself complained about the "new-fangled" or some similar description of the fashion of loud high notes by tenors singing his operas. He wondered why they no longer sang in "falsetto". His comments still lead to debate about what he meant by falsetto.

Of course all this in practice is influenced by the size of orchestra employed and the size of the venue. Falsetto would never work in the Met but would in Drottningholm. The connection with vibrato (which is where I came in) is that opera singers sing with vibrato because this promotes the singer's formant (about 2800Hz) which allows the singer to be audible through a large orchestra.

So, I would agree that this is of itself not relevant to Bach cantatas, except that some singers don't know how to "switch this off". That's why Pavarotti rarely (perhaps never) sang Bach, or Rossini.

But less vibrato does not mean IMO no vibrato.

John Pike wrote (June 10, 2008):
Developing good vibrato on the violin is certainly not something that comes naturally at all. It is also very important to keep it in check all the time, not just for baroque music, but across the repertoire. Too much vibrato, or too big a movement, not only leads to intonation problems; it also sounds too "hectic" and just unpleasant. I do this now almost without thinking, but sometimes I have to make a conscious effort to keep it right down. I often start a note without any vibrato at all and then start just a very slight movement later in the note.

Cara Emily Thornton wrote (June 10, 2008):
[To John Pike] Oh, those modern types would be horrified to hear about that approach to vibrato :) And it's true, it doesn't work as well with the Romantic repertoire - it sounds so... Baroque!

Cara Emily Thornton wrote (June 10, 2008):
Comments interspersed below, unnecessary items snipped out.

Neil Mason wrote:
< I believe your friend is most probably correct about Rossini tenors. I don't have the reference handy, but Rossini himself complained about the "new-fangled" or some similar description of the fashion of loud high notes by tenors singing his operas. He wondered why they no longer sang in "falsetto". His comments still lead to debate about what he meant by falsetto. >
I could make an educated guess that what is meant is a type of vocal production more similar to that of a male alto, instead of doing everything way down in the chest, which becomes artificial somewhere not too far above middle C (as in, the one in the middle of the piano keyboard, not the transposed one in tenor parts), which is why my old voice teacher in the States (a tenor himself) told me it takes so long to train a tenor properly - twice as long as a soprano.

< Of course all this in practice is influenced by the size of orchestra employed and the size of the venue. Falsetto would never work in the Met but would in Drottningholm. The connection with vibrato (which is where I came in) is that opera singers sing with vibrato because this promotes the singer's formant (about 2800Hz) which allows the singer to be audible through a large orchestra. >
This concept of formant sounds very interesting. Could I ask for an explanation?

Continue of this part of the discussion, see: Vibrato - Part 3 [General Topics]

Neil Mason wrote (June 11, 2008):
Cara Emily Thornton wrote:
< This concept of formant sounds very interesting. Could I ask for an explanation? >
Sure. I'll do my best.

Basically a formant is a strong partial, whose pitch does not vary with the fundamental.

You may have noted that I mentioned that the "singer's formant" is about 2800 Hz. (It does vary from singer to singer, form about 2600Hz to about 3000Hz - interestingly, this is a frequency range that is the best our ears can hear). This singer's formant is the same (in frequency) throughout a singer's range, although it does vary in intensity, and is normally more prominent on higher notes. In tenors we identify it as "squillo".

This is an extremely useful technique for singing with a full-sized orchestra in the romantic repertoire.

It's not really needed for a Bach-sized orchestra.

I have also noted your comments on Rossini. I do not myself believe he meant what we call "falsetto", in which the full length of the vocal folds is not involved in oscillation, but only a proportion of the length. I think he meant a "thin-fold" configuration, in which only part of the thickness comes together in the closed phase of oscillation. But of course that's just my conjecture.

Douglas Cowling wrote (June 11, 2008):
Neil Mason wrote:
< This is an extremely useful technique for singing with a full-sized orchestra in the romantic repertoire.
It's not really for a Bach-sized orchestra >
Which is why the whole vocal production question is changed completely if the orchestra consists of single players. Voices simply don't have to resort to the Romantic techniques necessary to cut through a symphony orchestra. I read a very interesting interview with a noted tenor who specializes in Bach (the name escapes me for the moment). The interviewer asked him how he found the stamina to sing in OVPP performances of large scale works like the Mass in B Minor (BWV 232) or even the Magnificat (BWV 243). He said that he found it easier to sing the tenor choir part solo than to be a section lead in a large choir and then shift to a "solo" voice for the arias.

Cara Emily Thornton wrote (June 11, 2008):
Comments below

Neil Mason wrote:
< Basically a formant is a strong partial, whose pitch does not vary with the fundamental.
You may have noted that I mentioned that the "singer's formant" is about 2800 Hz. (It does vary from singer to singer, form about 2600Hz to about 3000Hz - interestingly, this is a frequency range that is the best our ears can hear). This singer's formant is the same (in frequency) throughout a singer's range, although it does vary in intensity, and is normally more prominent on higher notes. In tenors we identify it as "squillo". >
What would be the physiology or physics behind such a tone, that does not vary with the fundamental? Could it vary with something else, e.g. the manner of vocal production? I lack the theoretical background to describe this in physiological terms, but what I have in mind here is that if someone is capable of producing tone qualities ranging from female soprano to male alto or perhaps even tenor on a good day, would the formant remain the same throughout all of that?

< This is an extremely useful technique for singing with a full-sized orchestra in the romantic repertoire.
It's not really needed for a Bach-sized orchestra. >
But is the formant still there anyway? So the question is simply whether we are subjecting it to enhancement?

< I have also noted your comments on Rossini. I do not myself believe he meant what we call "falsetto", in which the full length of the vocal folds is not involved in oscillation, but only a proportion of the length. I think he meant a "thin-fold" configuration, in which only part of the thickness comes together in the closed phase of oscillation. But of course that's just my conjecture. >
That sounds like a reasonable physiological explanation or at least conjecture for what I was talking about in terms of 'male alto' tone quality.