[CAUT] "key coloration"

[Fred Sturm]

On Mar 4, 2009, at 10:13 AM, Fred Sturm wrote:

> If we are really talking about historical practice, and we set aside  
> the fantasies about 19th century key coloration (that Jorgensen  
> promoted - and which have very scant historical basis

	In the interest of trying to establish clarity, I need to explain  
what I meant by the above. I don't want to be misunderstood as denying  
the "idea of key characteristics," for which there is plenty of  
evidence. The evidence is contradictory, but it is there.  
Interestingly, Rameau and Rousseau disputed along these lines in a way  
similar to arguments that were made on this list very recently. Rameau  
argued for coloration based on harmonic progression (advocating ET),  
and Rousseau argued for coloration enhanced by tuning (arguing for  
"French Ordinaire" which is a fairly extreme circulating temperament  
with one to three pure M3s and some wide 5ths, meaning the wide M3s  
are substantially wider than Pythagorean - and the sizes of M2s and  
m2s vary enough that scales sound quite different from one another).  
Rameau argued that the normal tuning of his time resulted in a lot of  
"out of tune sounding" intervals, which didn't enhance the music, and  
that ET was a better compromise. (Based on my own ears' response to  
French Ordinaire, I'm with Rameau, though I try to have an open mind  
about it - the distant keys are "too distant").
	In any case, leaving that digression aside, what I was referring to  
as " the fantasies about 19th century key coloration" was Jorgensen's  
notion, which he emphasized repeatedly, that 19th century tuners  
"heard differently and tuned differently." Here is a typical quote  
from Jorgensen:
“Nineteenth-century tuning by ear was a highly developed art based on  
aesthetic judgments for every tone, and test chords were used more  
than test intervals. By contrast, twentieth-century tuning is a  
mathematical skill."
	The implication, reinforced by other statements he made, is that 19th  
century tuners had these "key colors" in their ears, and used them as  
the basis for making tuning decisions, and as a method for achieving  
them. Let's look at that notion with a practical mind. Suppose you  
reverse his statement and say that 20th century tuners had ET sounds  
in their ears, and that they used the color of ET chords to create  
their tunings, adjusting each one until all sounded "just so." Imagine  
somebody actually trying to tune that way. In my imagination, that  
somebody is going to be chasing his tail for hours and never arriving  
at the desired result.
	It is very clear to me (and from all the written evidence) that  
tuners in the 19th century actually tuned by the same means as tuners  
of the 18th, 17th, and 20th: they divided the octave using 5ths and  
checked their results using M3s (and, yes, in the 19th century,  
instructions call for listening to triads as well as bare M3s). In  
order to get a pattern of varying sizes of M3s, you need to understand  
how the cycle of 5ths interacts with the M3s. And it is  
counterintuitive. Dirty 5ths produce cleaner M3s. And very subtle  
differences in 5th sizes add up fast to make fairly large differences  
in M3s, as all of us know (all of us who have tuned aurally). Trying  
to tune by adjusting M3 sizes individually is a recipe for disaster.
	So, while I am willing to believe that there were fine tuners in the  
19th century who might have adjusted their "quasi-ET" tunings to favor  
the white keys (though I doubt very many were that adept), I am quite  
convinced that they did so by adjusting and balancing 5th sizes, not  
by listening to "key coloration of chords." It is possible that they  
used "key coloration of chords" as a "proof" of their tuning. But they  
needed a procedure, a template to come up with a pattern, and there is  
no surviving written evidence pointing to such procedures, other than  
assuming that one would use standard WT or MT procedures (meaning  
patterns of 5th sizes) and moderate them by using different 5th sizes  
- subtly different 5th sizes, if you are to come up with a refined  
result. That's a lot of assuming.
	This is separate from the notion that some tuners actually continued  
to follow WT and MT procedures through the 19th and into the 20th  
century without aspiring to "quasi ET." There is sporadic evidence of  
that, especially in England, Italy and the US. But that is very much  
an exception, and focused more on the organ than stringed instruments.
Regards,
Fred Sturm
University of New Mexico
fssturm at unm.edu