If you saw my previous posting about a quiet but emotionally powerful tune in Mahler’s Das Lied von der Erde, you’ll recall that I included a scanned music example that showed two divergent attempts — one by Schoenberg, the other by Dahlhaus — to analyse the fascinatingly irregular melodic structure. I mentioned that Schoenberg wrote about the tune in his famous 1947 essay ‘Brahms the Progressive’ — but, as it happens, I didn’t show the example I was looking at (in the 1975 edition of the collection Style and Idea). All I did was scan Dahlhaus’s printed example, and use a graphics programme to add to it an easily comparable representation of Schoenberg’s analysis.
I also included a parenthetical note that will have given you a clue to the way my mind was working:
One needs to be a little cautious here, as Schoenberg’s essays have been printed and reprinted with shockingly wrong things done to some of the music examples: I don’t totally believe any example until I’ve compared all the existing sources — and in this case, I haven’t.
Well, I hadn’t — but now I have: after writing that I popped off to my local university library and did some digging about. Let me tell you what I found.
To start with, here’s the example I uploaded last time. The bracketed numbers below the staves are Dahlhaus’s originals; the bracketed letters above the staves were added by me to show how Schoenberg divides up the tune in his analysis. (I used letters to indicate where Schoenberg’s units start; he himself used horizontal brackets to show their extent):The reason I used Dahlhaus’s example is partly that its layout made it easy for me to add Schoenberg’s parsing above it — and partly that just by looking at the example in Style and Idea I could tell there was something wrong with it.
Here’s what I was looking at: Now, there are a few things to be said about this example before we go further. First, you’ll see that Schoenberg has put in an F major key signature that clarifies this single-line extract for the reader. (Mahler’s actual score looks like Dahlhaus’s, with no key signature at this point). Secondly, you’ll notice what I didn’t try copy literally: the long horizontal brackets that Schoenberg uses to parse the tune clearly into smaller units. Thirdly, you’ll see that this version also leaves out quite a bit of dynamic and articulatory notation that Dahlhaus actually includes — which is fair enough: Schoenberg’s point here is to highlight the irregularity of the construction, not to cram in all the performance detail that’s in the score. At the same time, though, the omission of three trill signs and a wiggly mordent is a bit more questionable: surely they are a part of the melodic and motivic structure? Fourthly — and here’s the mistake — if you know how tadpoles work, you’ll see that the first tadpole in the penultimate bar is a C in Schoenberg’s example, while the corresponding tadpole in Dahlhaus’s version is a B flat — as it ought to be, according to Mahler’s score. Yes, I know: don’t go mad over a misprint; these things happen; surely you’ve done it yourself; etc.
But the wrong note wasn’t the only thing that made me suspicious. To my eye, it also looked jolly weird that Schoenberg had apparently written in phrase lengths as numbers of bars (or ‘measures’) only for the last two of his units (‘1 ms.’; ‘2 ms.’) — leaving the other three officially uncounted. So I went and dug out the much smaller 1950 edition of the collection — in which I found the following:
Yes, this is all getting a bit nerdy; but it’s important, so don’t stop reading. First, what you see is that here too we only get numbered phrase-lengths for the last two units. Secondly, you’ll note that the penultimate bar of this version contains the same wrong note. Clearly, then, the re-copied example that appeared in the 1975 edition was a completely mechanical re-copying of something — all the way down to reproducing an erroneous C natural, which anyone who actually read the example and knew Mahler’s tune would instantly have detected. All of which only made me want to see what it was that the old 1950 version was based on…
So, I went to the fantastic online resource that is the digital archive of the Arnold Schönberg [sic] Center in Vienna, and looked up the scan of Schoenberg’s original musical example. What I found was the following (I’ve smudged out the non-relevant parts of what is a very crowded page; click the example to get a larger image):
The first thing to say here is that not all the notes are drawn or placed quite as cleanly as you would want them to be: an indication, perhaps, that this page was produced for Schoenberg’s very late ‘full reformulation’ of what was originally a lecture from 1933. (When I asked Lawrence Schoenberg about his father’s eyesight, he told me that the elderly Schoenberg was apparently not able to focus on smaller items — and reminded me about his need for special, large-stave manuscript paper in his final years.) All the same, see the penultimate bar: Schoenberg has very obviously written a correct B flat, and not an erroneous C. No less striking, though, is what we see below and above the staves. Yes, Schoenberg has actually written numbers for the lengths of all his parsed phrases: what we get is ‘3 meas. … 3 meas … 2 meas … 1 meas … 2 meas’. In other words, there’s no reason at all why Schoenberg’s example shouldn’t have appeared in print with (i) all the notes right, and (ii) all the phrase lengths counted. Nor, in fact, is there any earthly reason why a little editorial attention couldn’t have fixed up the trill and mordent signs that he didn’t put in. If you want to see the entire example looking as it really ought to have appeared in print, right from the start, here’s my digitally corrected version :
Doesn’t that look a whole lot better…?
Now, I get endless stick from people I know — very good friends, as well as colleagues — who think that I’m too critical of musicology and music publishing and the whole arts-and-humanities racket; but tell me: isn’t this sort of thing — you know: the ‘copying stuff out’ bit — supposed to be the easy part? Aren’t errors and omissions of this kind the things that are meant to get fixed at the proof stage, at the very latest? And remember: this example didn’t just go wrong in the Style and Idea of 1950, but was then, decades later, mindlessly re-copied for the 1975 version — and was still being printed that way at least nine years after that in the copy I have, which is the 1984 ‘First Faber Paperback edition with revisions’. This messed-up example may not be a ‘big deal’ or a ‘game-changer’ — but it’s ragged; it’s inept; it’s totally unnecessary; and it isn’t fair to Schoenberg or his analysis. What’s more, it’s far from being the only such example in Style and Idea (I’ve published corrections of three others — and I have more still to discuss). Finally, it all goes to create and reinforce the impression that no-one actually gives a damn. I mean, it’s almost as if … oh, yes: wasn’t I talking about ‘functional extinction’…?
If you enjoyed this posting, remember that I am a regular contributor and columnist for the UK magazine Musical Opinion. The magazine’s website can be found here; to see its Twitter feed, click here; to see its Facebook page, click here. To subscribe to Musical Opinion, click here.