Re: Tools for harmonic analysis (Riemann style)

[Urs Liska]

Hi Lukas,

Am 26.04.19 um 14:45 schrieb Lukas-Fabian Moser:
> Hi Urs,
>
> > finally I'm in need of symbols for harmonic analysis (in about the 
> > Riemann flavour).
>
> Just to provide some context to this request (of course, I assume you 
> are perfectly aware of all this): 


It's somewhat painful to admit but I have to some degree lost track of 
much of the discussion of the last 20 years, and when I actively worked 
with that kind of written harmonic analysis (while at the 
Musikhochschule) I didn't have much of a scholarly mind-set. So in fact. 
I was *not* aware of all this...


> The style of harmonic analysis is quite different from any of the many 
> flavours Hugo Riemann proposed originally. Rather, in incorporates 
> fundamental changes established by later German-speaking authors, most 
> importantly Reger, Grabner and Maler. (To name but a few: The 
> abandoning of Riemann's dualistic approach to minor keys, the 
> upper-/lowercase distinction for major/minor sonorities, the 
> designation of inversions by giving the bass note relative to the 
> root, the notion of 'Gegenklang' instead of Riemanns 
> 'Leittonwechselklang' etc.)
>
> But what's important is that the exact style in used in Urs's example 
> became quasi-standard in German-speaking countries in the second half 
> of the 20th century, and apparently it still is so, at least for 
> highschool/undergraduate level. 


As far as I've heard in recent years, there have been significant 
developments in that area over the last two decades, and it seems in 
current music theory (in the silly German distinction between "music 
theory" and "musicology") that "Riemann"-style analysis is mostly being 
considered pretty much outdated and too schematic to reflect actual 
musical reality.


> (I'm often amazed in my music theory classes at German/Austrian music 
> universities when some alternative descriptions of sonorities such as 
> vii°6, ii65 or a French Sixth are quite often only very reluctantly 
> accepted by the German-speaking students who state they learned the 
> "real" explanation, i.e. Maler-style functional analysis, at school.)


Oh, yes, the "real" understanding ROFL.


> To wit, Urs's request is probably shared by lots of users from 
> Germany/Austria. (I, for one, never needed them in Lilypond until this 
> year, at which occasion I used the LSR snippet that Harm already 
> pointed you to).
>
> I'd very much be interested in discussing questions of design, 
> interface, variants-to-be-supported etc. for this. 


I was first surprised and then not at all to see that the LSR snippet 
was created by Klaus Blum - who has also provided amazing stuff with 
boxes and arrows, now available in the openLilyLib 'analysis' package 
(https://github.com/openlilylib/analysis) - which is also the place a 
more comprehensive solution for harmonic analysis.

This should indeed include much more than the exact style I am currently 
looking for, although creating a package that provides multiple basic 
'styles' and substantial configuration options seems pretty challenging. 
But I'd say it would warrant a kind of "working group" because I think 
together with Klaus' existing analysis tools and my 'scholarly' package 
we have the foundation to provide pretty much a killer toolkit for 
academic purposes, and I'd be more than happy to increase its coverage 
and consequentially (hopefully) impact.


> One might probably learn a lot from David Nalesnik's excellent roman 
> numeral analysis tool 
> (https://github.com/davidnalesnik/lilypond-roman-numeral-tool). I 
> imagine that matters will probably be more complicated than with roman 
> numerals, but at least for these, David implemented an ingenious 
> syntax and layout engine.


I'll have a look at that too. Roman numeral analysis should of course 
also be merged into a comprehensive package.


> And @Malte, am I wrong if I remember that I saw some work on 
> functional analysis symbols done by you?
>
> > > I would go for it by creating a lyrics-like context for the 
> > > horizontal and vertical alignment and produce some functions that 
> > > would create combined markups or stencils for the symbols. 
> > > Additional challenges would be to define practical ways to use 
> > > something like stanzas for modulations and boxes around that (a 
> > > sample image attached).
> Stanzas are ideal for indicating key regions (I use them regularly for 
> roman numeral analysis). As for the boxes, this probably needs some 
> new kind of interface.

The "frames" module in the "analysis" package already provides something 
that should be usable for that (see attached image).

One thing I'm a bit worried about is that when changing the harmonic 
region I would prefer having a solution that works like \change Staff = 
"" (well, ideally it should not depend on an existing context but 
*create* it on the fly) so I don't have to provide skips in the second 
layer/stanza.

> > I might also be willing to use LaTeX for the creation of the symbols 
> > because I could then combine efforts for a standalone LaTeX package 
> > to produce the symbols in continuous text too.
>
> I'd very much argue for a LilyPond-only solution in order to rely as 
> little as possible on a specific toolchain.


Good point. However, based on discussion of the interface and required 
features a *separate* LaTeX package would be desirable.

Best
Urs

> Lukas

frames.png
Description: PNG image