(archived from this ihm thread)

Part I:

What I wanted to add to this thread was how to judge an interpretation of the score.  This is an incredibly complex issue but there are I think several basic rules of thumb that can be applied. First off, to get the most common objection out of the way, if you like a particular recording as a piece of music then that is great, no-one is denying you that.  On this issue I think there is no discussion to be had.  The question is though, is a particular recording an actual realization of the score, or more influenced by the score?

The key to interpreting the score is that you build up a consistent vocabulary for the symbols of the score. This vocabulary needs to be flexible as the symbols rarely stand on their own: they are constant intersected, amended, interrupted or overlaid with other symbols. This is an important issue that I think leads to failures in many interpretations, but I’ll get back to that in a bit.  The consistency issue is equally important, if you do not consistently interpret the symbols then you aren’t really playing this as a score. Let me give an example; if you created a version of Beethoven’s Ninth Symphony that played all the notes, used the exact orchestration, followed all of the rules of the score, except that you played it at a tempo such that it lasted nine hours, this would no longer be Beethoven’s Ninth Symphony.  It would be at best “inspired by” or a gloss on it. Why? Because Beethoven gave explicit directions on the tempo for the score. Now there is a lot of debate on that and realizations of the score can vary in times by 15 or more minutes.  But these I think are all valid, they fall within the parameters of the score.

So am I saying that doing a short version of Treatise is a priori invalid? No, but to do a short version of the score requires a huge amount of work. Cardew’s composition Volo Solo is apparently a transcribed version of the entire scored to be performed by a piano virtuoso which can be done in about 10 minutes. But this is meticulously scored (and it is rather difficult to see exactly how it is transcribed from the score. Tilbury (for whom it was composed) goes into this in Cornelius Cardew: A Life Unfinished) and contains thousands of notes.  So this is one solution to the issue of consistency and flexibility: rigorous transcription. In essence you are compressing the score into a smaller more compact version, a translation of all of this syntactic content into another system.

Beyond that level of transcription what else do we find in interpretations of the score? When you are not converting the symbols directly into traditional notation you tend to utilize varying degree of abstraction. That is you associate techniques, events, or sounds to the symbols. You then need to take into account the various permutations of the symbols, the spaces between them, the center line and the numbers. When sight reading from the score this is pretty challenging and the pages themselves contain varying degrees of density. In these sort of interpretations it takes much longer to work though each page, some pages have a lot of content to work through others have very little. Though the relationship between density and tempo is not precise, so pages with very little syntactic content can feel very slow while others dense with information feel fast.  When sight reading the score, the amount of time it would take is certainly highly variable, but I think that on average you’d be pretty hard pressed to fully interpret a page in less then five minutes.  Maybe a few pages would be less but many many more would require more time.  Even if you did spend one minute per page, which would really be a gloss over most of the symbols, you would end up with a three hour+ performance.  So when you are presented with a short performance there are basically five options:

  1. It is fully transcribed to traditional notation and is played at a very high tempo with a high density of notes
  2. The score is being broken up among the players such that for each page each player is playing only a small subset of the information there.
  3. The score is being played layered: that is you each divide the score up among each player so that each player is only doing about 20 pages)
  4. A subset of the score is being performed
  5. It is a gloss of the score

Now lets specifically consider the Hat[now]ART version of the piece. I think its fair to rule out option (1) – it is too sedate, too ambient to be like a 120 minute Volo Solo. (2) and (3) are possible though of course this is impossible to know.  Ideally any performance of the score should include information on the pages played and some description of the strategy adopted. I would guess that it isn’t a layered version of the score – it again doesn’t feel dense enough. (2) is certainly possible and I think would be the most obvious to someone who reads up on the available resources on the score:

“Performance advice. Divide the musicians into those involved in dot events (percussionists and pianists?) and those involved in line events. Dot events to be exclusively soft.” -Cornelius Cardew,Treatise Handbook(1)

However in this fashion to really treat it as a score, you’d basically have long rests between a lot of the events that you are playing. So it’d still come down to spending less then a minute a page, which even if you were only making a few amount of sounds would still come out to all the symbols in the score being played within this time.  Again this take just isn’t that information rich, or quick in tempo. Thus it seems fair to say that this performance is a gloss, a “take” on the score. And there is nothing wrong with that, except that they say this is the “World premier complete performance” on the set (which rules out (4) of course). You can still like the music here, enjoy it for what it is, but it isn’t Treatise anymore then a nine hour version of the Ninth Symphony isn’t Beethoven

This post has gone on long enough, but I think it establishes a set of criteria for examining realizations of the score and demonstrates how to apply it. If there is further interest, I’ll add examinations of a couple of other versions here over the next couple of days.

Part II:

The most successful (and common) strategy of interpretation of Treatise is to play a subset of the score.  This allows you to devote the time that each page requires.  When playing as a group without any sort of strategy of synchronization these types of interpretations necessarily become layered versions of the score. I for one am a fan of that type of layering and this was something well embraced by the 60s Experimental Composers (c.f.  all of the simultaneous Cage performances) including Cardew. The entire score could be played through in the fashion, playing a page or series of pages in a number of sessions. We began this process in the Seattle Improv Meeting and played through 44 pages in this fashion (these can be downloaded here). This attempt  to play through the whole score in sequence was stopped at that point as the interpretation was suffering in other ways.  This is a good example of one of the other failures in interpretation: using the score for structured improvisation.

“The score must govern the music. It must have authority, and not merely be an arbitrary jumping-off point for improvisation.” -Cornelius Cardew,Treatise Handbook(1)

Use of the score for structured improvisation again becomes a gloss of the score; one is not interpreting the symbols of the score consistently or rigorously. Again you can make good music this way and it is not an invaluable exercise. In fact I quite like structured improvisation, I think that it often adds depth to improv, just enough structure applied to give form and to curtail certain impulses.  However it is not following a score and as is quite clear with Cardew’s own intentions (“It must have authority”) this is not a rigorous realization.  I should point out that in the case of the SIM this was just how the group developed. In our three years of working with the score, we evolved from a tentative learning process, to a rigorous consistent approach and then it kind of slackened off into this more loose approach. See some of the recordings from the second half of 2005 to the first couple of months of 2006 for some of the more consistent realizations (my personal favorite from this period is Pages 146-148)

Other versions that most likely fall into the gloss of Structured Improvisation would include Formanex Treatise- Cornelius Cardew(Fibrr).  While they don’t explicitly say which pages they are playing on this recording (as opposed to their other recording) overall the music here is too overtly ambient to fit much of the score. However it is possible that this is the last three pages of the score in which this sort of interpretation would be valid. Again this is a case where it is frustrating that they don’t tell you the page(s) that are played. If you contrast this though with the other Formanex recording, Treatise-Live at Extrapool (also on Fibrr) it is markedly different, that one has the spaciousness and spikiness that more strict realizations tend to feature. I would also guess that the performance of pages 21 & 22 On the hat[now]ART release Material also leans this way. At least by some of the members (which is worth noting, there can certainly be degrees of rigor in any group performance of it) especially the vocalist who seems to be free riffing well past any material available from the score. This sort of free improvisation on the score I think is particularly egregious as this is adding semantic content which is not contained within the score.

Part III:

Quax Ensemble plays Treatise

Cornelus Cardew Treatise (Prague version 1967) (Mode)

The most recent Treatise recording out there came out just a couple of weeks ago on Mode.  This release is certainly of much historical interest, the QUaX Ensemble being one of the earlier groups to work with the piece. During the course of the development of the piece a number of musicians worked with it (including the members of AMM, in various combinations and of course AMM itself) and one of these was Petr Kotik. He was pretty young at the time, (early-mid twenties) and had was still in conservatory when he met Cardew and developed a relationship with him. He was able to get a number of pages (alas and annoyingly the liner notes do not specify these) prior to the scores publication in 1967 and put together the QUaX Ensemble to play it among other scores.  They played from these pages frequent but and eventually put together a 2 hour version of the pages they had which they performed once on October 15th 1967. That of course is this recording. He quite clearly states in the liner notes that he received a copy of the complete score in Buffalo NY in 1969, two years after this performance. So this two hour version is clearly a subset of the score, thus eliminating the issues that are rife with short complete versions.

“There is much to admire in this 1967 version of Treatise by the QUaX Ensemble from Prague: the feeling of spontaneity, its uninhibitedness, the rough-hewn sounds, the accidental, the half-intended, the blurred.” -John Tilbury(3)

The liner notes include a page from John Tilbury which tellingly is half devoted to quotes from the Treatise Handbook. The above quote from him, which is rather amusingly used in the Mode PR, I think sums it up perfectly: There is much to admire yes, but there is also much that is not so admirable.

First off they clearly did spend the time developing a consistent take on the pages they had. The liner notes includes reproductions of two of Kotik’s pages which have numerous annotations on them. These include notes on what to play for certain symbols but are actually mostly devoted to timings. Most interestingly the notes mostly resolve around who is to play for a given symbol.  For instance on the back cover there is an except of a page where the same symbol (black filled circles) are notated “Kotik”. Other notes seem to be either instruments or perhaps performance techniques that I can’t decipher (anyone who can, do let me know). These notes are pretty revealing in their take on the score, which to me seems like a fairly typical case of classical musicians trying to improvise.  It in fact reminds me a lot of the workshop and performance I did with Vancouver New Music (that Joda references (in the ihm thread) and which I wrote up at length here: VNM Treatise report) where I’d say most of the ensemble never really managed to work with the score in and of itself, they always were using it as a springboard.

In the case of this version, they seem to have a concern that I’ve encountered among virtually everyone I’ve seen play the score: playing together. It seems for classical and jazz musicians that the concept of everyone working through it at their own pace is a difficult concept.  This was constantly raised in the VNM group and also something that plagued the first half dozen or so sessions with the Seattle Improv Meeting (at least among certain people in both groups). Even a more recent performance that I did with a dedicated graphic score group (EyeMusic) under the aegis of Keith Rowe this issue was also raised. In this case since they seemed also to be working with the score in parts (that is assigning symbols to various performers) this seems a much more traditional approach.  Of course Cardew definitely worked with the score in this way especially before AMM and I wouldn’t be surprised if he recommended this approach to Kotik.

Based on the annotated pages that we have in the liner notes it appears that their working out of the score is more an assignment of who plays what, not much of an indication of what they play. In other words this approach is structured improvisation. Listening to this before I closely examined those score excerpts there were a number of passages I found troubling that are explained somewhat by this approach. These passages involve overtly melodic material from the saxophone (Pavel Kondelik) and kind of piano jazz breakdown complete with vocals (Vaclav Zahradnik). Now one might think that one could assign melodic content to the symbols as long as one is consistent and in theory you could. But as I pointed out in my earlier post, any symbolic association must take into account the various fragments, interjects, incomplete symbols, overlapping symbols and so in. With very few exceptions no consistent melodic line would survive that for long. Thus the multiple long melodic sax passages seem to me outside of a strict reading of the score, but would make sense if you just were using the symbols as sigils for when a particular musician was going to perform.

Apart from this melodic content there are a lot of great sounds in this recording and lots of space. Apart from the center line, there are many pages with long gaps between symbols and thus any performance should have these gaps (unless someone is focusing on the center line, but even in this case the other musicians should respect the spaces). There is even a nice background radio grab for a bit, giving it a bit of a Cage or AMM feel.  Much of the sounds are generated on traditional instruments with extreme extended techniques, using many of the sounds associated with avant-garde composition (ala Lachenmann) or post EFI improvisation. Additionally there are various ambient sounds, passing traffic, shifting chairs and the like that places this within a space and I think add a lot to the overall environment. Long stretches of this performance is fantastic in my opinion, though there often is something coming in that one may not like.  For me it really is the melodic content and even worse the semantic content from the singing that mar the performance.

Cardew on a number of occasions expressed his dissatisfaction with classical musicians performance of this piece. Too hard for them to break out of their routines and notions of performance. Impossible for them to capture the right balance between the spontaneous and the structured. This recording I think is somewhat exemplary of that. The fact that the performers were young and still students probably gives this the life and drama that it does have. It still I think would have somewhat dissatisfied Cardew in that it doesn’t go all the way to where he wanted to go w/r/t performer involvement but probably wouldn’t be the total disappoint of some of the performances by highly trained and rigid musicians. As Tilbury says there is much to like here, but this is not I think a wholly successful interpretation of the score. It is definitely recommended though, it is an important piece in the history of the score a history that is quite lacking in the early performances.

Part IV:

Getting back to analysis of performances of the score there is two important notions that I have thus far neglected to mention. This was deliberate as I wanted to approach these recordings in terms of their consistency and rigor. These notions basically subvert that to some degree but in general they can’t be used to excuse those primary notions.  These issues are numbers in the score and the notion of a perverse reading of the score.  The numbers are inescapable in the the score so lets take a look at those first.

As with everything else in the score there isn’t any instructions on how to handle the numbers. However in the Treatise Handbook he says thus:

“The numbers are included at the pauses for the reason that: any act or facet of the conception or composition of the score may have relevance for interpretation.[…] It is the fact that there were 34 blank spaces before the first sign put in an appearance” – Cornelius Cardew, Treatise Handbook (2, p. 251)

This singular note on the numbers from Cardew doesn’t really give much info on how to interpret them.  However it has come to be interpreted as a repetition of an event for the number of times of the numeric value. As there isn’t a specific symbol tied to the event the event is treated as outside of the score, that is it is independent of your consistent interpretation. So theoretically a performer could say play a chord 34 time, or make 34 sounds or what have you.  So does the potential arbitrariness of the numbers make possible the overtly melodic sax in the QUaX version discussed earlier? Probably not. In general the numbers aren’t too large after the initial 34 and there are multiple long sax segments in that version. So I think my previous read of it still stands, but it is worth taking the numbers into account when you listen to a version where something may seem out of place (for those interested Tilbury includes a breakdown of the number distribution in the footnotes to the Treatise chapter in his Cardew bio, on page 278 No. 16).

The notion of Perverse Readings of the score is a lot more problematic and there isn’t a lot of information to go on. From Tilbury’s bio:

“…John White’s precedent for ‘perverse readings’ by reading ascending lines as descending intervals” – John Tilbury (2, p. 251)

The first time I encountered this notion though was for the AMM+Formanex performance of Treatise at the Musique Action Festival, Nancy, France in June 2002. At this performance AMM performed several pages of Treatise with  John White who once again did his trademark Perverse Reading in this case interjecting banal samples into the performance. (Interestingly enough while looking for data on this, I found this tiny bit of video from this event here).

“AMM were extremely spartan (wonderfully so) while White, on some kind of sampling device, intruded with all manner of awful-sounding blurts, cheezy synth tones, sheep baa-ing, etc. It was very annoying and I found myself vainly attempting to mentally tune him out.” -Brian Olewnick

Now considering that Perverse Readings were par for the course, occurring from the very beginning of the scores history, can any out of the ordinary performance be considered “perverse”?  I tend to think not, I think that the perverse readings are still consistent and rigorous, but in such a way that may be contrary, controversial and against expectations.  Brian later asked Keith Rowe about that aspect of the performance and he went on to say:

“He felt that ‘Treatise’ performances had a tendency to get overly somber and reverential and wanted to inject a “low” element that he thought of as the sonic equivalent to US rags like Weekly World News, to introduce “transgressive” sounds that “simply aren’t done” at these types of events.” -Brian Olewnick

And this to me I think is the essence of the perverse read, not falling into a free for all, do as you wish sort of situation but subverting received notions. But I do find it a troublesome notion and this would certainly be an interesting point to have clarified.


1) Cornelius Cardew (1936-1981) A Reader (Copula, 2008)
2) Cornelius Cardew: A Life Unfinished, John Tilbury (Copula, 2008)
3) Liner notes from Cornelus Cardew Treatise (Prague version 1967) (Mode)