On 14 Feb 2006 at 9:27, Dennis Bathory-Kitsz wrote: > At 04:42 PM 2/13/06 -0500, David W. Fenton wrote: > >I'm not certain I understand. If your original tempo is ca. q=80 and > >your new tempo is ca. h=80, then would q=84 in the first section and > >h=78 be Ok for the new meter? Or do you mean that once you've chosen > >q=84, the half has to remain 84? How do you indicate flexible initial > > tempo but strict proportion? > > It's hard to anticipate how a performer would interpret markings. > (Realize that I'm not one of those who specifies tempo in painful > detail anymore, but I do understand the kind of scores that would.) > When it comes to matters which might be misinterpreted, composers seem > to go for redundancy and comment, even in those fat 19th century > scores. A parenthetical phrase such as "overall tempo is flexible, but > keep rhythmic proportion exact" would do.
Well, one problem: you haven't indicated the rhythmic proportion, except indirectly -- the performer would have to remember the original tempo marking in order to figure out what note value is proportional. But no explanatory note is needed with ca. q=80 at the head of the score, and <-q=h-> at the time change. Right? Is there any ambiguity there at all? > For me it's a matter of disambiguation. If it's important, mark it. > That goes for bowings, conducting beats, etc. Some composers clutter > up their scores with hundreds of markings. I've engraved some of > those, and where they matter, they produce an amazing result. But > arbitrary over-marking is different. But your example explanatory note seems to me to not clear up all the questions, whereas my solution is unambiguous (so far as I can tell). You could also use "l'istesso tempo (ca. h=80)", and that should be clear, too, seems to me. Simply stating the new metronome marking is going to result in the same thing (assuming the interpretation that the same ca. 80 tempo is used for both), but it requires more interpretation on the part of the performer to figure it out. I'm all for notation and engraving that leaves no questions in the mind of the performer. > >My point is that proportions can handle all the permutations very > >clearly, whereas metronome marks require increasing verbosity to > >express the same thing. > > The only reason I don't assume you're right is because the question > came from an accomplished musician on this list. If there is a > question, then there's an ambiguity that needs to be clarified. But wasn't the ambiguity over the order of the proportional values? Is it not the case that the arrows and having the equal sign centered over the bar line would completely disambiguate that problem? What other interpretation could <-q=h-> have? > (This also relates to the dangerous 'swing' question and its stylistic > suggestion that may get lost or changed in the mists of history. The > 'swing' components are hardly alike in Joplin, Ellington, Goodman, > Parker, Hurt, Monk, Ayler, Coltrane, and Ware.) I take it you mean that there are many different "swings," not just one. Seems to me that even within the style of a single performer, there are lots of different "swings," depending on the musical context (8th-note swing can be different from quarter-note, etc.). > [...] > >> >It seems to me that <-q=h-> is completely unambiguous and easy to > >> >understand. > >> > >> Or not, unless it's explained. :) > > > >Who wouldn't understand it? Seriously -- what trained musician would > >misinterpret it? > > Arrows suggest direction and change. They might be interpreted as > moving from one to the other rather than changing abruptly. If a > properly positioned q=h and <-q=h=> are functionally equivalent, what > exactly to the arrows show? Maybe it's just me, but I can't read it > instantly. They make it clear that which value is which, that the first refers to the previous section and the second to the next section. Without the arrows, there is still the question whether you're saying "new h=old q" or "new q=old h." Personally, I never know which way to interpret it, unless one interpretation is nonsensical. > >(it also leaves open the ability to specify a loose proportion by > >using the = that is squiggly, i.e., two ~ atop each other) > > Yes, indeed. > > >I'm not saying metronome markings should be ignored. I'm just raising > > the question of how helpful they are for establishing proportions > >between two meters. If there's no actual proportion, then a metronome > > mark is going to have to do (absent some tempo marking that does the > > job), and that's fine. > > I agree. My concern is that notational shorthand is not always a good > thing. Further explanations help. Explanations might not facilitate a > first reading, but they should at least clarify a meaning to a > conductor or a performer who will spend more than a casual reading on > a piece. While I'm not against explanatory notes, I think one should avoid them if there's some succinct method of conveying the information. I'm arguing that my way is more succinct than yours, and that this is an objective good. > >But I just can't see how q=80 at the beginning and h=80 at the change > > of meter is going to be more easily perceived by the performer than > >q=80 at the beginning and <-q=h-> at the time change. [...] > > You may well be right. We share (or shared, since I haven't done it > for at least 15 years) considerable experience in early music. There > was a great mystery in these proportions and rhythms in HAM vs. when > those 'new' Oxford editions were first appearing, for example. The > interpretation in various editions was sometimes wildly different. I > have a hideously expensive Minne/Meistergesang Troubador/Trouvere > edition ca. 1970 (expensive when I was younger, that is, and sadly > lost in one of my moves) which differs so radically in rhythm from the > Appel presentation ca. 1950 that it might be a different piece > entirely (Be m'an perdut). Not being an early music scholar, I > couldn't put together my own. The scholars didn't know, either -- they were just making it up. The Mendel articles on tempo and proportion from the 40s show exactly how far wrong things can go when you start out with misinterpretations of th esources. In performance, I ignore editorial proportion suggestions in early music, unless they make musical sense. And I don't really believe in maintaining strict tempo proportions unless it works musically. I also prefer editions that stick as close as practicable to the original note values, because the longer note values imply certain things about harmonic rhythm and figuration that are lost (or simply misrepresented) when the note values are shortened. Obviously, this is only practical for certain original values. I wouldn't find a 4/1 modern edition particularly easy to read, for instance, even if it is a literal transcription of the original note values. > Admittedly, evidence from 500 years past is scant. But imagine if this > same interpretive dilemma faces future performers trying to get a real > sense of what composers wanted in their dim past. Despite our > propensity for playing something contemporary once and then abandoning > it, that's not necessarily what the future will be like. If what we > write is important, we have to clarify our meaning. Well, it seems to me that the approach I've suggested avoids any ambiguity, as long as the future musicians understand the metronome markings and the arrows. Perhaps the "l'istesso tempo (ca. h=80)" would be superior after all. > >To me, it's like the difference between piano roll notation and > >traditional notation. The former has the virtue of specificity but > >lacks the clarity and flexibility of interpretation of the latter. > > I don't agree, but then I can read piano roll notation. (I had to -- I > was teaching a composition student who had learned to read it as his > first notation, and could make absolutely no sense out of the > irregular steps in staff notation! And it was amazing how much clearer > his vertical relationships became for me without the intervening > abstraction of altered pitches (vs. absolute pitches).) Well, OK, perhaps it's OK for pitch, since it's very graphical, but for rhythm, it's not terribly helpful. One of the things I teach in the pre-theory class I've taught for so many years is that there's a good reason for writing 4 eighth notes with a stacatto dot instead of 16th note, 16th rest, etc. It's because the notation is *more* precise in capturing the range of allowable performance alternatives, while being less precise about exactly what should be played. If the composer does not want to imply any such interpretive freedom, she'd use the 16th notes and rests. But if one wants that kind of freedom from the interpreter, the stacatto version is more exact in conveying what you're after, because it encompasses the full range of possible interpretations. To me, piano roll notation records a performance, but not a score, because it is exact without any way for the interpreter to know where the parameters of interpretive freedom may be. Conventional notation makes notating certain things every difficult (everything that doesn't fit into the traditional musical system that produced the notational system), but for the things it was designed for, I'd say it's pretty good at striking the balance between specificity and intentional variability. > Let me veer off a little here, because it's not always an issue of > specified tempo vs. interpretation. This is not about piano roll > notation either, but about the specificity of beat and how performers > can be confounded by even a simple one. I had the experience recently. > For background, think of Beethoven's "Coriolanus" overture (I think > that's the one) where the tempo remains precise and the note values > and frequency of rests change to create the illusion of rallentando at > the end. > > My string quartet from last year was written for a group in Ghent that > frequently performed newer music. The tempo was to be held to fairly > tightly, with much of the bending of tempo and rhythms written into > the music -- in particular because much of my music doesn't have much > interest in barlines, but performers do. So I give them barlines in > plain old 4/4, but they are only positional guideposts. (The quartet > that was supposed to play it didn't; their leader was sick. The > replacement quartet, made up from players from four countries, was at > a loss to understand how to play it -- it wasn't hard, but adding > their string-quartetty romanticism to it caused, shall we say, a loss > of stability in performance. To be kind.) Incompetent performers are to blame for this, not the notational system. Your explanation above should have been perfectly sufficient. Brahms wrote out a ritard at the end of the Edward Ballade (if I'm remembering the title correctly), and it's pretty clear without there being any explanatory note that the player doesn't need to layer an additional ritard on top of that. But some musicians are insensitive to these kinds of things, and there's nothing that can be done about that, except to either beat them over the head to do what they should, or find better players. > So I pose this question to you and anybody who's still reading. :) > Given a situation where a composer wants a Coriolanus-esque change of > temporal expression, but throughout a piece, what's the best route? Seems to me that whatever you notate, just say "senza ritard" and have a line over the period that this directive extends. Or "in strict tempo" or some other equivalent instruction. > 1. No barlines (so as not to suggest anything that isn't true) save for > dotted barlines in very occasional places where the material is > simultaneous. The note lengths determinine the timing. That might be the best solution for particular musical content, but not for all. It would depend on the passage, I'd think. > 2. Same as #1, > but with arbitrary regular barlines as counting signposts, and the > note values reworked to fall properly within the pseudo-measures. Perhaps Mensurstrich? Again, I think it would depend on the passage. If the first solution fit naturally with the musical gestures, I don't see why barlines should be necessary, unless there are problems of coordination between parts that are quite independent. > 3. Same as #2, but with frequent meter changes to accommodate the note > values as much as possible. If there really are meter changes with the associated shifting accent patterns, then that would be the best solution. But if you don't want the implied metrical accent patterns, then I'd say that adding in the time changes would just be making a whole new set of problems. > 4. Barlines with ongoing "rit." and > "accel." markings along with tempo changes every couple of bars. In a passage where the rhythmic notation is conveying the temporal alterations, this seems unnecessary to me, and would require a complete rewriting (not just transcribing into a different metrical framework). And would result in something completely different. The difficult question is whether this would be more likely to get what you're going for from particular performers. THe hard part of this is that I expect that different performers would respond differently to the different versions, with some doing better with #4, others with #1, depending on their experience with these kinds of passage in contemporary music. > 5. A > combination of 3 and 4. > 6. Another option? I can't really answer in the abstract. It seems to me that the answer depends on two factors: 1. the performers and their backgrounds, AND 2. the exact musical content. > It's not a question of leaving the interpretation up to the > performers. Something specific needs to be indicated, or the piece > can't be understood. A new piece is a new piece. It doesn't belong to > a tradition yet (and where it might, it wouldn't need notation to > explain that). And this isn't 18th century music, and not even 19th > century music that *was* frequently littered with changes. I'd tend to go with whatever feels most pure and uncompromised to *you*. I might then offer an alternate transcription in one of the other modes if the performers are not able to manage the pure version. > When I've done #1 above, performers have complained that it's > impossible to know where they were. They refused to read the score > (even with only a few parts) and so were stuck reading from separate > parts without 'signposts'. . . . Well, then they were idiots. > . . . I have often had to make performance scores > with arbitrary barlines as in #2 -- but the failure was that the > performers then *counted* and misinterpreted tied notes and > irregular-appearing rhythms as syncopations and got the emphases all > wrong (think of bad editions of Ockeghem, for example, without the > barlines between staves ... what's the term for that? I forget. > Mensurstreiche or something that's escaped my fading mind). Again, I'd go with whatever notation you believe best conveys the music, then deal with performers on a case-by-case basis. Reading from score is a requirement in cases like this, and not at all unusual in modern repertory. > I've given up doing #3 because I frequently write for professional > musicians with limited new music experience -- and counting is the > primary failure I've seen. I've never done #4 because it just seemed, > well, excessive and not likely to represent the linear flow accurately > at all. > > So this is a kind of expanded question about these tempo issues. > Onward... Many composers throughout history have written music that performers couldn't manage, and have been forced to simplify or provide alternate versions in order to get it performed. In your position, I think I'd try to make them better performers, insofar as that's possible, by attempting to stretch them beyond what they're accustomed to doing. This would require a pedagogical flair that perhaps you lack (I doubt you lack it, though), and also more time than may be available. I don't know what the solution is, but I'd start from the ideal and compromise only when forced to do so by circumstances. > >I would also argue that you *can't* have specificity of dynamics and > >tempo, because all performance situations are different. Environment > >forces changes to dynamics and tempos while not altering pitch, > >rhythm and instrumentation -- you likely wouldn't want the same tempo > > in a dry hall as in a hall with a 5-second acoustic. You wouldn't > >use the same dynamics and balances in a hall that favored high > >frequencies. Thus, dynamics and tempo must by definition be variable, > >even if there is a definite conception in the mind of the composer. > >This is very different from the other parameters of music > >composition. > > It's sometimes thought of as different, but we've come a long was in > music since the days when its basics were just "melody, harmony, > rhythm". Environmental forces may play into it, but so do they in > relativee harmonic content of orchestration, in a relative pitch base > (even in Baroque music, and wildly so), etc. While these elements have > been increasingly standardized (or were, before the original > performance movement), the idea that tempo and dynamics cannot be > specified well has been an intractable and I think faulty assumption. I don't see how dynamics could be specified any more specifically. In decibels? Measured from where? The audience? The conductor? The player? What other methods of notating dynamics would there be? > My assumption is that 99% of listeners will hear my music on a > recording, not in a live concert. That considerably alters what a > composer can reasonably expect in terms of tempo, dynamics, balance, > etc. The failings of a live space do not have to be replicated on a > recording -- even if it is a recording of the same concert in the same > live space concertgoers heard. So these elements can be specified with > the assumption that 99% of listeners will hear them as written, and 1% > may not because of extra-compositional factors. Well, if you're talking about composing for recordings, then you're in Glenn Gould territory, and so far as I'm concerned, you've left the part of the realm of music that interests me. Music is a social act, and music performance benefits from the interaction of musicians. If what you want is control of the final recording, then put a mike on every performer and use a multi-track recording studio and set the dynamics and balances in the recording studiot. Of course, that won't really work, either, because you aren't necessarily going to have gotten the right sounds from the original performers -- an oboe played loud doesn't have the same tone quality as an oboe played soft. All intruments vary their tone according to volume. Secondly, dynamics generally indicate more than just volume, because changes of dynamics often go hand-in-hand with changes of mood and affect. So, you can't just layer on the volume by adjusting sliders on the mixing board, because you need some part of the dynamic profile of the piece to be incorporated by the performers in manipulating their instruments. > In other words, there's no reason for me as a composer to be chained > to an outdated assumption. Well, I've suggested to you before that you seem ill-suited to composing for human beings. As synthesizers become better and better, perhaps you can avoid all the problems of dealing with fallible human beings entirely. But for me, dynamics are inextricably tied up with the act of playing an instrument, not a parameter independent of all other aspects of the instrument's sound or the performer's interpretation. On the other hand, if all you need to do is fine tune a performance that basically gets the dynamics right, and all you're wanting to do is clean up the dynamics and balances that didn't come out right (e.g., conductor didn't balance the group right, or the hall caused certain instruments to be covered), then I have no objection to using the recording studio to bring the imperfect recording closer to the ideal in your head. But that wasn't what I took you to be describing. > >Further, dynamics and tempo are not something that are controlled > >with mechanisms that have fixed intervals. That is, rhythms have only > > one correct value, pitches have only one correct frequency (within a > > context -- a B natural in a G chord may be a different frequency > >than a B natural in an e minor chord; but within any particular G > >chord there is really only one correct frequency for the B, given an > >agreed- upon temperament), and our instruments are mostly designed to > >give us these precise pitches and rhythms. > > From a historical perspective, yes. But tempo is surely controllable if > the expectation is that the marking will be respected. . . . With a click track, yes. Without it, there's always going to be variation. > . . . There's > fluctuation in pitch and harmony and rhythm, and I see an equivalent > fluctuation of tempo as acceptable. Too much so, and you have changed > the style of the composition. Well, I was thinking at a simpler level (assuming reasonably tempo variation within a piece of music), about setting tempos according to metronome markings in a score. It's never going to be exact because humans can't remember tempos as precisely as the metronome markings can specify them, and because human perception is subject to distortion by context and environment. > As for dynamics, this issue has been developing over the past > half-century. Numerous dynamic systems have been suggested to replace > the 11 dynamics from ppppp to ffff. . .. Well, let me say that anything other than ppp to fff is ridiculous -- there's no possible way to actually make such gradations in any meaningful sense. Anything beyond that range is voodoo notation, in my opinion. > . . . None have solved the problem > because 11 dynamic levels is really a pretty good outline -- *not* > good enough for detailed presentation in the final 'mix' (whether live > or on recording), but a good outline. I'm not going to take up the > fight on this one, but clear levels of relative balance exist in this > composer's mind (and the recording engineer's mind, especially the > very good ones). How these balances are to be indicated accurately on > the page is for the next generation to deal with. And they might only > deal with it in the mix, since recording remains how most people hear > music. I don't think human perception is precise enough to allow indications for performers that would have any unified result. It might be one thing for a solo piece, but there's no objective way for two or more performers to agree on the exact loudness of ppp in any particular context (let alone in the abstract). This seems to me to be a feature rather than a bug. Music is all about relationships, how one musical artifact can change its meaning depending on the context in which it is heard. Likewise, a pp in one context might be the exact decibel level of another context's mp. Music is about shaping perception and experience over time, and thus everything that happens is relative to what has gone before, not to some abstract outside reference point. For me, that is fundamental to the way music works, and anything that tries to abstract parameters like dynamics out of their context is going to approach automated music making. I've nothing against automated music making where it can convey the full content of the music involved. But if one is notating for some form of automated music making, one notates completely differently than if one is notating for live performance. > >With dynamics, we don't have the ability to set the dynamics in > >absolute terms, nor tempos (without some outside reference such as a > >metronome). So, I just don't see this as a problem -- it's a feature > >of human beings making music, and to me is a *good* thing, as it's > >the source of variety in performance. > > I agree with the latter, but not without caveat. It seems to me that > raising the level of respect, if you will, for the good judgment of > the composer in matters of tempo and dynamics is important. . . . This is how these discussion with you typically run. I make a point that you agree with but you provide a caveat that berates bad performers. We agree that bad performers are bad performers. We agree that performers should attempt, to the best of their abilities, to play what is in the score, and not fancifully add things that are not there, nor arbitrarily ignore things that are there. I find Ivo Pogorelich's Chopin infuriating for just this reason -- he willfully ignores the dynamic markings and tempo indications in order to apply his willful "interpretations." I think that's wrong in Chopin, and it's wrong in Bathory-Kitsz. But that's not a notational question. It's a lament about the state of modern performance. I could go on and on about the sad lack of musicality and musical intelligence in the young conservatory graduates I coached in my two stints on the faculty of the California Music Festival. They lacked so much in basic training about how to approach the music as well as lacking basic performance skills. They were, in my opinion, very ill-served by their teachers, especially given how easy it was to get them to do things right. > . . . Performers > continue to roll over these indications as if they were simply hints > of a thought of a suggestion of a possibility. The move to specify > tempo (as in indications like 'q=62.2') and rhythms and tempi (again, > ref. the new complexity composers) brings with it an awareness that > such markings *do* matter. With experience and persistence, they will. Well, I move on when I see an indication like that. It can't possibly be done by a live musician, so is what I earlier classified as "voodoo" notation -- i.e., writing down something you know can't possibly be done exactly in the hopes that the mere specificity of it will produce the desired effect. > [...insightful comments snipped...] > > >I think your valid objections are to willful misinterpretation of > >scores with clear indications, rather than to small variations like > >those described above. I'm with you on those objections -- they are > >the sign of bad or indifferent musicians, and that's something any > >composer has every right to protest. > > I see 'willful misinterpretation' as any deliberate and significant > variation from clear markings on a score. . . . Assuming, of course, that the markings are not musically nonsensical. > . . . Times have changed, it seems > to me, and 'q=72' does not mean the same as 'q=ca. 72'. It's incumbent > on the performers to understand the intent of the composer well enough > in taking on a piece to determine if the composer meant the former or > the latter. Well, even q=72 is going to be performed with certain variability, unless the performer keeps a metronome on the music stand at the performance. To me, q=72 is absolutely equivalent to q=ca. 72, as it must be. "Circa" only really has any definite meaning if a range of possibilities is provided. Otherwise, you're not really adding any meaning, since you haven't said what the acceptable range of possibilities is. And I also would say that with a particular group of performers, the same music at q=74 may sound better than if that group played it at exactly q=72. Tempo, like pitch and dynamics, is often about relationships, and tempo is not an abstract, independent parameter laid on top of music. It affects everything about the performance, including dynamics, agogics, accentuation, articulation and everything else. If slavish devotion to the indicated metronome marking means that the performers miss all the other aspects of the performance, and adjusting slightly to q=74 allows them to get the other aspects in place, why would any composer want them to treat the q=72 as the only sacrosanct parameter of the performance? Getting the music off the page is a complex interaction of a web of parameters, each slightly altering the specifics of the other, and what one wants in a performance, I'd think, is an overall gestalt that conveys the content of the music, which is, one would hope, somehow conveyed by the accumulation of detailed and specific performance instructions. A good performance may not necessarily reflect every performance indication exactly, because some parameters will cause adjustments to others. But since it's all in the service of conveying the overall musical content, it should still work, as long as the performers have grasped what that content is intended to be. The problem with the performers, it seems to me, is not so much the failure to honor q=72 (or any other notational detail) literally, but with the failure to grasp the musical content and meaning of the work as a whole. This is one of the unavoidable problems of new music, in that the performers don't necessarily bring any background and experience that will help them perceive the content/meaning of the music from the score alone. Slavish devotion to the text is a really good place to start, but won't get you all the way to a performance that gets the music off the page. > Also, I have a stricter standard of performance, particularly in the > matter of tempo. I'm far from alone. In post-performance quiet, when > the exuberance has faded, I've talked to composer after composer who > was disappointed that these very basics in the score were not properly > attended to before the performers veered off into 'interpretation' > mode. As the composer, are you sure you've conceived of the tempos correctly? Do you play the instruments you're writing for with sufficient technical knowledge to say if your tempos were well- chosen? Secondly, my experience is that well-written music falls into its correct tempo by itself, and stays there. Perhaps there are aspects of the score that don't work well for the stated desired tempo, and this has caused the performers to choose some other tempo. Third, performances, especially first performances, bring with them a lot of psychological effects on the performers that can skew perceptions. Perhaps the performers thought they were choosing the indicated tempo and performance nerves caused problems with their time sense. It's certainly happened to me. And don't underestimate the issue of performing a piece multiple times. I know from experience that a piece changes greatly once you've performed it once, twice, three times. It takes performance before an audience to temper an interpretation, seems to me, so any first performance has to be taken as a first effort. > >But trying to overspecify beyond the resolution of human perception > >is always going to fail to a certain degree, and that's what I'd > >object to as over-specific. > > Our level of perception rises to the challenge. Just think of Ives, > now a century old, or Stockhausen of 50 years ago, or Eckardt or > Ferneyhough today. We can get it if the performers can get it. Does anyone get Boulez in the same way Boulez himself does? In any event, we don't have the performers of 50 years from now to play our pieces. That's not to say we should never write for those future performers, only that we shouldn't be surprised when today's dedicated performers of goodwill fail to reach our ideals. -- David W. Fenton http://dfenton.com David Fenton Associates http://dfenton.com/DFA/ _______________________________________________ Finale mailing list [email protected] http://lists.shsu.edu/mailman/listinfo/finale
