Re: measure for measure

[Joshua Horowitz <horowitz...>]

The music which I transcribed for the Sephardic Songbook is entirely vocal.
That means that its rhythms are subordinate to the text. But this doesn't
mean that they lack regularity which can be translated into a notation using
measures.

Many of the compound meters (i.e. meters in which there is a combination of
2's and 3's, like in the 9/8 meter: Daka Daka Daka Dakata / Daka Daka Daka
Dakata, whereby the first 3 Dakas are 2 beat units, and the last Dakata is a
3 beat unit) are actually isometric. That means that the amount and
placement of the 2 and 3 part units remain intact during the course of the
section. Isometric meters can indeed be placed into a measure corset, but
the problem which occasionally arises in the Sephardic songs is that even an
isometric meter can be broken when the text in a verse demands this or the
performer decides to get fancy and stretch out a note or two.

Vocal folk music in many traditions relies on poetic meter for its
construction. So the texts will have a certain number of poetic "feet."
Poetic feet (consisting of one accented syllable plus one or two unaccented
syllables), are not necessarily synonymous with fixed musical time values
like one eighth note and one quarter note, but are proportioned in terms of
stress. So a word like "YESterday" would be dactylic, with one stressed
syllable and two unstressed ones (YES-ter-day). The simplest reading of this
word would indeed be in 3/4 meter, like a waltz, as YES-ter-day (ONE two
three).

A poet, however, when reading this, may stress the YES syllable by drawing
it out for 2 beats, in which case the meter would be 4/4
(YE-ES-ter-day). This poetic flexibility, common to reading poetry but less
common in the modern world on non-improvised music is still found in the
renditions of the old singers of Sephardic vocal songs, making possible
versions which occasionally break the MUSICAL meter but generally stay
within the POETIC one. Usually this is done only punctually, so it is not as
though there is a wild discrepency there. Keeping this flexibility of the
performer in mind, I tried to explain its effect on the music in the
introduction of the Sephardic Songbook, which I'll quote here (soon members
of this list will have the entire intro of the book if we keep this up)...

" 1) There is usually no ³fixed² melody of a song as you would find it in,
say, western classical music, but rather an abstract, yet traceable notion
of the melody, recognizable in all its variants, but always in motion. The
extent and style of these variants differs necessarily from performer to
performer. Therefore, a rendition of a ballad by someone trained extensively
in classical Turkish music and/or cantorial singing might result in a more
highly ornamented and elongated form than a rendition by an untrained
singer. Strikingly different variants may be found of each and every tune,
even affecting such basic qualities as meter. For example, the song Bre
Sarica, commonly known in duple meter, is found here in a rare 7/8 meter
version; and the melody of Los caminos de Sirkeci, often found in Balkan
countries in 7/8, is here found in 6/8 meter.

2) Most Turkish and Balkan performance practices tend to allow improvised
variations from verse to verse. The extent to which a performer chooses to
stretch out melodies melismatically and incorporate effects and nuances into
a song is a matter of taste, style, preference and musical background. As
such, it feels almost a shame to present their performances as truncated
abstractions of a melody that would rarely be performed more than once in
that particular way or form.

3) A song containing several verses will generally need to accommodate the
text of each verse with  corresponding melodic variance. Therefore, syllabic
differences from verse to verse render impossible a unified melody which is
utilizable in unchanged form for each verse..."

In the introduction, I tried to detail as thoroughly as possible how and why
I came to the solutions I did. In presenting editions of transcribed music
such as this, you have to be crystal clear about your intentions; we were
making a songbook to be used and reused (of course to be abused is also a
possibility) by the general public. I write about this in the intro as well.

A musicological transcription will be quite different from a songbook
transcription. Bartok's transcriptions - as wonderful as they are - are not
intended for Joe and Joanna Shmo, but served as a basis for his composed
music and allowed him to study the mechanics of the music he was using. So
you find an extremely exacting orthography in his transcriptions, with
notated  changes in tempo which are as small as a few metronome markings,
ornaments scribbled down to the last tea-drop and sometimes larger metric
shifts which do not necessarily correspond to the informant's "intention",
but which occur "in-situ" and are therefore taken as biblical writ by
Bartok. You can almost "hear" the performance of the informant alone without
the recorded music if your reading skills are that good (mine are not).
Actually this is not really true, as Judith hit it on the button when she
wrote that you cannot get the full idea of music through notation alone. As
obvious as this may seem to be, you can go to any conservatory anywhere in
on this planet and be totalitarily confronted with the tyranny of the
written score.

Many aspects of musical performance are left out of a transcription, one of
the most central being timbre. Timbre is the tone quality of an insrument or
voice, and if you were to notate this, you would have to find autonomous
symbols for each aspect of tone production. Can you imagine reading a symbol
which expresses the following:

a half-pinched glottis with the tongue raised to touch but not press against
the soft palate.

A score with that level of precision would be unreadable (There actually is
a modern piece which has a kaleidosopic array of vocal gymnastics which does
this, called Glossolalis by Dieter Schnabel).

But lets be content when we can relay any aspects of culture, even when our
methods for preservation pale in comparison to the original and risk
diluting its essence if taken at word.

Of course, at the end of the day, writing about music is like describing a
lunch. If you know that, then the foregoing discussion is irrelevant. Josh


>From: Judith R Cohen <judithc (at) YorkU(dot)CA>
>To: World music from a Jewish slant <jewish-music (at) shamash(dot)org>
>Subject: measure for measure
>Date: Sun, Jan 13, 2002, 6:05 AM
>

> hi, I'll make some  more detailed comments about JOsh's excelllent
> songbook later - just off-hand, I 've never seen any problem whatsoever
> with dividing a measure between lines, especially if the musical and/or
> text phrasing calls for it. Measures are a relatively new aspect in the
> history of western music notation, they have no intrinsic value or
> meaning except as a visual aid in marking off the units chosen as those
> best describing as the movement of a piece of music - the layout should
> serve the music and not vice versa.  Also, measures, and notation
> altogether, have a somewhat different function in composed music from
> that of oral tradition. If one is composing something new, to be read
> and performed by others who are not likely to ever hear oneself singing
> or playing it before they do, i.e. who are using this notation as their
> only clue to what the music sounds like, then the bar lines and other
> aspects of notation have definite functions, and must be carefully and
> deliberately chosen.
>
> In ethnomusicological transcriptions, one can (and often does) make a
> major case for, unless there is a clearly delineated metrical pattern,
> simply not using bar lines. Personally, that is my approach and that of
> many colleagues, but others prefer to use barlines and indicate shifting
> patterns by changing time signatures several times within a song. There
> are positive and negative sides to both approaches.
>
> (Most software music writing programmes appear to have been designed
> especially to thwart this kind of flexibility (and most other kinds)
> without elaborate circumvention inventions.)
>
> But when it is a question of bending into the straitjacket of western
> musical notation, a piece from an oral tradition often not meant for
> notation, then it is a different story. A major part of part of
> ethnomusicology has traditionally been to do just this - figure out how
> best to pin an elusive piece of aural and oral tradition to a piece of
> paper so that someone can then read it and reproduce it as closely as
> possible to the original.Often the choice of time signature is somewhat
> arbitrary - asymmetrical metres were not originally thought of as, say,
> 2+2+3; that's how people worked out how to write down what was being
> done. Then a convention developed for calling it, say 7/8 for slower
> tempi and 7/16 to indicate faster tempi, but it's just that - a useful
> convention, to be used and bent as a tool. Until I learned to DANCE
> these metres, it was very hard to play them or sing them properly years
> ago, but once I had them in my feet and rest of my body, playing and
> singing them was automatic.
>
> This is generally impossible.The most unbelieveably precise
> transcriptions of, say, complex polyrhythmic African drumming, have been
> produced, or, perhaps closer to this lists' immediate concerns, Bartok's
> exquisite transcriptions of eastern european and Turkish music. But
> music is elusive, and one could not reproduce the sounds from even these
> masterworks (no irony here, I mean it) of transcriptions unless they had
> actually heard the sounds first and knew how to interpret them.Even
> simple tunes, unless one knows the style , especially the vocal timbre,
> will not sound the same. A piece of music (as most of you know) is MUCH
> more than the pitches and time values of the notes . I shudder to
> remember the various forms of musical castration I've heard in schools
> where music teachers blithely pick up the latest curricular virtuous
> multi-cultural songbook and accurately teach students the pitches and
> time values of a Caribbean or African or Korean song, using a perfectly
> innocuous transcription, and transform it into a bland, sugary piece of
> nothing for the school concert... but I digress.
>
> That's why I left medieval music as my main focus - got too frustrated
> at knowing I'd never hear the sounds - the vocal timbre, the
> inflections, the way of matching words to music - but especially HOW the
> singers sang - all that stuff which notation can't capture.
>
> So this is NOT a diatribe against notation, which I use constantly, and
> is an essential tool. Just a longer response than I expected to make
> about dividing measures! I know most of the songs in Josh's book, and in
> some cases have actually collected them myself as well, from Salonica
> Sephardim in Salonica itself or Israel or North America, and from my
> cursory (so far) look, these are very fine transcriptions indeed of a
> complex and beautiful repertoire.
>
>  But yes, it would be ideal to have the CD in the sort of package which
> it has become, fortunately, common to produce now that it's cheaper to
> do so than it used to be, because again, no matter how good the
> transcriptions are,
>
> there's simply no way to know how a song is sung without hearing someone
> sing it.
>
>

---------------------- jewish-music (at) shamash(dot)org ---------------------+