What comes after 4-part writing?

[fwe3k]

I've written a bunch of pieces with more than 4 voices, but always by ear. I studied up through traditional 4-part writing in college, but no further. Is there a body of theory about how to expand beyond 4 voices? Let me know if I need to explain myself better, and thanks in advance to anyone who answers.

[Mr Arkadin]

Five-part perhaps?



Sorry couldn't resist. However i think the thing to do is to divide the the choirs into multiples of three- to five-parts. An extreme example would be Thomas Tallis' amazing Spem in alium which is written as a 40-part motet. He has a series of eight five-part 'sub-choirs' which in the church setting would be placed round the church space giving an early surround effect.

Mind you, even dividing in this way is complex as each 'sub-choir' has to relate to the next. The key with the Tallis piece i think is that its harmonic structure is actually quite simple, with the lines seeming the move quite freely over the top. Also reducing the choirs to solo or two-part in places adds dynamics and simplifies the writing.

As to whether there is a method or name for this i don't know.

[Ogg Vorbis]

In traditional arranging/orchestration/part writing, once you've got four-part writing down, extending the number of voices is pretty easy. The reasons I thought of as to why you'd do so is to:

1) Reinforce the parts through doubling (either at some octave or unison, etc.)
2) To manage variation in "color" (like cueing in an additional instrument to take over the melody or a counter melody) while the previous instrument "lays out"
3) Add even more melodic, rhythmic or harmonic "layers" to the texture
4) To accomplish spatial effects like the ones described above.

I think these principles hold true in pop/rock/jazz/country/hip-hop/bluegrass/showtunes/etc. as well.

But I can't think of any principles of theory that govern cases of 5+ voices that would be different than, say, 3 except that when doubling in triads, you should tend to double the root or the fifth before thinking about doubling the third. In extended triadic chords like 7ths, 9ths, etc., you should still think about gravitating toward doubling the root, fifth and not so much the 7th or 9th (unless these are a part of your melodic elements that you are reinforcing through doubling).

There are principles of orchestration and arranging that come into play. In general, keep the voicing with more open intervals in the lower range, with the closer intervals in the higher range. This is not a hard and fast rule but it's a general one that keeps things a bit more "clear."

Basically I think that you have to be careful not to allow the texture to get so complex that the parts compete for attention in an undesirable way, or cover up your main idea that's going on. Your mix should not get muddy or too thick in the lower ranges (unless you are doing that on purpose for effect). If you need to have busy elements going on in the same register, you can try panning them left and right, or arrange them physically far from each other to keep them from becoming gloppy.

I've heard orchestrators and arrangers talk about the "rule of three" which states that even in the largest ensembles (Mahler's Ninth, Rite of Spring) there are usually (NOT ALWAYS) no more than three major elements going on at any one time. Each part will be contributing its sound in some way to one of these three elements. This applies to pop/rock as well, even though you have lead vocal, backing vocals, keys, guitars, bass, perc., drums.

It's kind of funny because I've noticed the same tendencies I have with arranging also work against me in my cooking! In other words, my tendency is to add, add, add, spice, aroma, salt, sweet, etc. until it knocks the shit out of you. So I am learning to leave "openings" in my arrangements (and sauces).

[MoreDread]

http://en.wikipedia.org/wiki/Micropolyphony
http://en.wikipedia.org/wiki/Atmosph%C3%A8res

[llatham]

I've written a bunch of pieces with more than 4 voices, but always by ear. I studied up through traditional 4-part writing in college, but no further. Is there a body of theory about how to expand beyond 4 voices? Let me know if I need to explain myself better, and thanks in advance to anyone who answers.

Actually, I'm going to say 3-part writing, 2-part writing, and 1-part writing!

I'm not being silly, but serious.

As a theory professor at a university, we typically approach part-writing in the "traditional" style as 4-part.

Very little about 3-part is discussed, and 5+ part only for particular examples.

Unless you took a Counterpoint class in addition to your Harmony (or Theory) Course, you probably didn't do nearly as much work in 3-part as 4-part, and far less in 2-part, and probably little 5-part.

Furthermore, good melody writing (1-part, or monophony) is hardly addressed at all. Usually, melody in harmony studies is either a Chorale Tune, or, it's a result of the part-writing. In Counterpoint, it's either the Cantus Firmus, or a result of contrapuntal writing!

If you feel pretty confident in 4 parts, don't fall into the "more is better" trap.

There are certainly some 5+ part works/styles - Monteverdi's 5-Part madrigals come to mind, as do a lot of 6-part Ricercare, etc.

But there is a LOT of music in only 2 or 3 parts.

The interesting thing to me is that there are two styles - one style is contrapuntal, where intervallic relationships are most important, and the other is like "4-part with a missing voice" - which leads to the harmonic progressions of traditional 4-part writing, but having to imply full chords with fewer notes.

The latter is typically a *serious* deficiency in most student's studies (or most graduate's skills), but there's a huge body of 3-part music - Trios, 3-Part Inventions, Fugues, etc. etc. etc.

It takes a lot more skill to make full sounding harmony or imply logical progressions with only 2 or 3 parts!

HTH,
Steve

[Mister Natural]

At my school - 4 part was followed by 2 semesters of counterpoint. Don't know what was/is taught anywhere else

[jancivil]

...

There are principles of orchestration and arranging that come into play. In general, keep the voicing with more open intervals in the lower range, with the closer intervals in the higher range. This is not a hard and fast rule but it's a general one that keeps things a bit more "clear."

I've heard orchestrators and arrangers talk about the "rule of three" which states that even in the largest ensembles (Mahler's Ninth, Rite of Spring) there are usually (NOT ALWAYS) no more than three major elements going on at any one time. Each part will be contributing its sound in some way to one of these three elements. This applies to pop/rock as well, even though you have lead vocal, backing vocals, keys, guitars, bass, perc., drums.

It's kind of funny because I've noticed the same tendencies I have with arranging also work against me in my cooking! In other words, my tendency is to add, add, add, spice, aroma, salt, sweet, etc. until it knocks the shit out of you. So I am learning to leave "openings" in my arrangements (and sauces).

ok, good stuff!
but then, you got this in here, which rather ignores context...

But I can't think of any principles ... except that when doubling in triads, you should tend to double the root or the fifth before thinking about doubling the third. In extended triadic chords like 7ths, 9ths, etc., you should still think about gravitating toward doubling the root, fifth and not so much the 7th or 9th (unless these are a part of your melodic elements that you are reinforcing through doubling).

WHY? Does this not assume some things not in evidence?
to wit:
in a lot of extended harmony, you might not even include the fifth. the root might be present enough when underplayed, or even absent per se. the inclusion of every note of the whole build of a harmony, once you get into the sort of practices which use such sonorities, I would bet real money is statistically more rare than frequent.

I have noticed something in a whole lot of, I mean all the time in classical music, which does what you say about this doubling the root fifth, it's too thick by far without being very resonant, it's gross.
now, someone else might think that's the swell sound they really want to hear, but it's not going to enhance every kind of, or even most kinds of musics.

[Ogg Vorbis]

in a lot of extended harmony, you might not even include the fifth. the root might be present enough when underplayed, or even absent per se. the inclusion of every note of the whole build of a harmony, once you get into the sort of practices which use such sonorities, I would bet real money is statistically more rare than frequent.

I have noticed something in a whole lot of, I mean all the time in classical music, which does what you say about this doubling the root fifth, it's too thick by far without being very resonant, it's gross.
now, someone else might think that's the swell sound they really want to hear, but it's not going to enhance every kind of, or even most kinds of musics.

Jan, I don't doubt you are right. My point was not that one OUGHT to double chord members, but in cases WHEN you are doubling, the general idea is that you should leave doubling the third of the chord as the last considered option. The root (in unis. or at some octave) is normally the most "likely candidate" for doubling, then the fifth.

This guideline is something I learned in traditional SATB writing, and you can see it farily well adhered to in Bach chorales. This general principle can be extended into parts involving more than four parts.

In truth, the context is going to determine what REALLY happens. You see the third being doubled at the octave a lot, especially where it's a melodic component.

And you are right, if you double the fifth extensively you're going to get a very "thick," as you say, sound. So, I believe it's better to double it at a higher octave (if you're going to do it at all).

But one must confront the issue of doubling when you consider tutti writing in the orchestra, right? I mean, with only three different note names in a triad, you must decide how dbs, vcs, vlas, vns (1 and 2), perc., harp, hns, tubas, tbs, trps, bns, cls, obs, fls, pic, etc, are going to play that triad. Even if it's some extended tertian sonority, there's inevitably going to be doubling going on somewhere.

The thing is...I recognize that no rule can be laid down in music. As soon as someone lays down a rule, someone else can discover instances in the literature that violate that rule. But for me, it's been nice to understand the guidelines that I learned in SATB writing so many years ago, and extend those principles into my thinking when writing for more than four parts. It doesn't mean that I always follow those principles, it's that I'm a bit more CONSCIOUS of what's going on. Awareness is power, do you agree?

Then there's the matter of DESIRING to write "thick" and "gross" textures... we know what we must do then, don't we?

[jancivil]

in a lot of extended harmony, you might not even include the fifth. the root might be present enough when underplayed, or even absent per se. the inclusion of every note of the whole build of a harmony, once you get into the sort of practices which use such sonorities, I would bet real money is statistically more rare than frequent.

I have noticed something in a whole lot of, I mean all the time in classical music, which does what you say about this doubling the root fifth, it's too thick by far without being very resonant, it's gross.
now, someone else might think that's the swell sound they really want to hear, but it's not going to enhance every kind of, or even most kinds of musics.

Jan, I don't doubt you are right. My point was not that one OUGHT to double chord members, but in cases WHEN you are doubling, the general idea is that you should leave doubling the third of the chord as the last considered option. The root (in unis. or at some octave) is normally the most "likely candidate" for doubling, then the fifth.

This guideline is something I learned in traditional SATB writing, and you can see it farily well adhered to in Bach chorales. This general principle can be extended into parts involving more than four parts.

In truth, the context is going to determine what REALLY happens. You see the third being doubled at the octave a lot, especially where it's a melodic component.

And you are right, if you double the fifth extensively you're going to get a very "thick," as you say, sound. So, I believe it's better to double it at a higher octave (if you're going to do it at all).

But one must confront the issue of doubling when you consider tutti writing in the orchestra, right? I mean, with only three different note names in a triad, you must decide how dbs, vcs, vlas, vns (1 and 2), perc., harp, hns, tubas, tbs, trps, bns, cls, obs, fls, pic, etc, are going to play that triad. Even if it's some extended tertian sonority, there's inevitably going to be doubling going on somewhere.

The thing is...I recognize that no rule can be laid down in music. As soon as someone lays down a rule, someone else can discover instances in the literature that violate that rule. But for me, it's been nice to understand the guidelines that I learned in SATB writing so many years ago, and extend those principles into my thinking when writing for more than four parts. It doesn't mean that I always follow those principles, it's that I'm a bit more CONSCIOUS of what's going on. Awareness is power, do you agree?

Then there's the matter of DESIRING to write "thick" and "gross" textures... we know what we must do then, don't we?

well, I shoulda said 'turgid' rather than 'thick'... maybe.
you're right, double higher; if you're going to obey principles, observe acoustics and follow that*. (that sound... it seems like every time I pass the 'arts' channel on cable here, which I think is some deal outta germany, I get that ugly-ass voicing I mentioned, which is all double in the mid and lower mids. and, this is all 'masterpiece' sort of music, but I don't buy into it at all. but, I don't go for that music by and large.)

*: hence> a fifth is already generated by a 'fundamental'; an octave doubling is going to reinforce this.
a ninth chord tends in most textures to give a sort of 'undertone' or difference tone and produce that fifth in its absence as well. this is why, look at Ravel, or jazz, even that master of turgidity Wagner and you'll see, not a lot of the fifth of a chord where the ninth is present.
{BTW, I'm only 'correcting' you there for the benefit of perhaps the less-experienced proponent, like.}

the third, being another partial higher, might be doubled without harshing your buzz as much, but of course we agree that orchestral texture, and style does govern.

anyway, I don't think there IS a lot 'beyond 4-part' in terms of chordal thinking.

I do agree with llatham, that in a sort of less = more sense, that, eg., 2-part and just going for melody can be beyond 4-part.