On composing Tom Phillips' Irma
In February and March of 1977, for Brian Eno’s Obscure Records, I made a version of Irma. The following notes on the piece arise out of that involvement and try to show how the piece can be made into a performance state.
Irma is a curious score – it is printed on a single sheet 50cms x 50cms. The notation consists of fragments from Tom’s continuing treatment of the victorian novel by W. H. Mallock, which he calls A Humument, and utilises those short verbal fragments that refer to either ‘‘libretto’’, ‘‘decor and mise-en-scène’’ or ‘‘sounds’’. These 3 categories are arranged in separate sections on the square sheet with, at the bottom, a line of stave notation. At first sight it looks like a piece of indeterminate music – clearly there has to be some preparatory work done before it is performable and no-one would venture to perform directly from the score - but if it is approached in this spirit, like realising a piece by John Cage or Morton Feldman written during the 1950’s, the sounding results are either largely of a documentary
interest, or rely entirely on the gifted performer to make into a coherent sounding whole. True, one could say the same thing for a piece by Cage, such as Variations I, but there the
performer is given a number of precise parameters of sound within which he should work, whereas Irma needs to be re-composed rather than realised.
If the distinction between ‘‘composing’’ and ‘‘realising’’ is overlooked, and if only the materials present in the notation are used, then the result is likely to be impoverished and it is clear that, looked at in isolation as a self-contained work, the score is notationally very thin. So one either produces an impoverished piece of sounding music, or one takes the responsibility to look further. Tom does not say explicitly that one must go beyond Irma into the rest of his work, but he does say that one has to go outside the piece. On the score he writes: ‘‘Perhaps to treat the indications here given as if they were the only surviving fragments of an ancient opera, or fragments of eye and ear witnesses’ accounts of such, and given no knowledge of performance tradition of the time, to reconstruct a hypothetical whole which would accommodate them economically, would be an appropriate basis of approach to a production.’’ So, try to put it back together and try to fill in all the gaps between these fragments. This approach, which, incidentally coincides with an interest in such procedures within my own work, seems to be the most suitable. If the ‘‘composer’’ uses the sorts of methods that Tom evidently uses in producing pictures, in making A Humument (of which Irma is a part), and if he uses the notations of Irma as clues to lead him into whatever area seems likely to yield rich results, then a much more satisfactory outcome is likely – satisfactory both in terms of the quality of the sounding material and in terms of consistency with the rest of Tom’s oeuvre.
There are clearly many ways in which the various fragments of verbal notation can be used. One fruitful way was to take each of the fragments as the notes of, say, a critic at the only performance of the work (in a hypothetical past), perhaps jotted down on the back of an envelope (then torn into fragments in a rage, or through frustration at some element in the piece? Make the piece inadequate in some way?!). These elements, then, would have been memorable for some reason or other, or used as an aide-memoire to recall something else (even something outside the work). The elements could have occurred at evenly-spaced intervals throughout the performance, they may have all been featured in some way (loudly, as solos), they may have been the worst parts (being retained to damn the piece in a subsequent review, since lost or never written – the composer got wind of the review and murdered the critic, retaining the fragments as the only link with the crime. . .). On the other hand they could be used as discrete musical units quite separate from the main body of the work, which has to be looked for elsewhere. Whatever solution, or combination of solutions, is found it is evident that the composer and librettist are more or less obliged to move outside the work itself i.e. outside the printed score. (One of the original ideas I had, which was not very practical, was to see if I could use another opera called Irma. A possibility was one written by H. J. Banawitz first performed in 1885, which would have had the right period in terms of the connection with the W. H. Mallock original. This seems to have had few performances, perhaps only one, and seems to have disappeared. I thought of looking for the manuscript, treating it in the same way as Tom had treated the Mallock novel, and making a sort of ‘‘musical Humument’’ out of it. While that seemed to have some intellectual sympathy with Tom's work, it might not have sounded anything like an opera, and it did seem to me that one of the notions of Irma is that it is conventional to some degree. Indeed, while much of Tom’s musical work lies within the field of experimental music and graphic notation, his musical taste is conservative, and the greater part of the musical references in the main body of his work are to past, and historically respectable, composers like Brahms, Mozart, Fux, Scarlatti and so on.)
The sources that were used, then, in making the piece apart from the score itself involved the following. I obtained the volumes of A Humument and noted all connection with music, with the role of Irma, and with the possible narrative; I looked at all the prints of Ein Deutsches Requiem after Brahms, which use elements from the Humument and refer directly to a musical work; I went through the catalogue of Tom’s work (Works. Texts to 1974); I went through Trailer, which uses the Humument, in fact a spin off from the main work; I went through all the other pieces of music that he has written to see if they could be used in any way; and I checked as many paintings/prints that I could which had a direct or indirect connection with either A Humument, Irma or with music. The painting The Quest for Irma (1973) which shows her in back view looking out to sea gave much information. It is the only portrait of her and she appears even here as anonymous, or rather, faceless. It gives a marine setting for the work (though since at least two pieces of music that I have written deal, to some degree, with marine incidents it might be argued that I might have been better off avoiding such a reference, but it is very blatant). She is looking out to sea from the Dorset coast and this attitude seems to be characteristic of her behaviour: ‘‘I tell you. . . that’s Irma herself. . . watching the waves fall. . . repeating certain sorts of verse. . .’’ So here we have an elusive heroine, obsessively watching the sea off the Dorset coast, given to repetition. Further checks within the Humument revealed a spate of marine references: ‘‘boat of dreams. . . lost on rocks’’; ‘‘the sad horizon of sea, hours she spent with her sadness on the beach’’; ‘‘see, see, the things. . . the things from the changed sea’’; ‘‘a cruise in an opium clipper’’; ‘‘marine engines and boilers’’; ‘‘ten years’ travel and sport in foreign lands’’; ‘‘a certain light flashed. . . among the eastern clouds’’; ‘‘sinking lights. . .’’ and so on. On the other hand, she is not in mourning since she wears a bright red dress.
One page of A Humument is almost a summary of the feeling of Irma and is certainly one that I tried to emulate. ‘‘. . . The whole history of it is so vague. . . eagerly, gradually the words that I heard I put aside as an opera, an insufficient one; still organ for what – me, me. . . I can’t quite tell. hardly books. . . it was the libretto of the music, of the music – I can’t tell. . . I can’t tell - but all was for the same thing to capture in drawing, and to express in music, thought and study. . . the loss. . . the least important. . . moon I myself am myself in search of an object for love? way? Yes and no – enter myself. . . associating me and me. It made me within me some mystery. . .’’
Other pages give more precise information about particular sounds, rhythms, timbres and so on. The instrumentation was, to a large extent, governed by the references to musical instruments that I found in all these sources. ‘‘Tube’’ suggested tuba. The piano is mentioned
many times, especially in connection with John Tilbury. The gong is specified – ‘‘suddenly a gong in series’’ – which also gave me the whole of a short percussion interlude between the second and third sections of the work. Strings were suggested by a phrase ‘‘the history viola’’ occurring in A Humument and this gave me a reason to feature the viola in some way, in fact using it in unison with the female voice, identifying the viola with the title of the opera. The fact of having strings is such a convention of normal orchestral scoring that it would really have needed a positive clue to the contrary to have excluded them, bearing in mind the relationship of the piece to musical convention.
I used the tuned percussion, and especially metallic instruments, from certain onomatopaeic syllables, like ‘‘ting’’, ‘‘ping’’, ‘‘ding’’ which I had originally considered using as a chorus of instrumental imitations, but decided ultimately to use the instruments themselves.
Two of the prints from Ein Deutsches Requiem after Brahms gave me a great deal of material for the second section of the opera, a slow duet between the two main characters. Print number 5 shows a number of parasols, both closed and open, and has the legend ‘‘. . . a sound was given up’’ taken from A Humument. That particular picture suggested itself since there is, within both the score of Irma and within the published Humument a fragment which reads ‘‘the first parasol sound’’, with the addition, in Irma, of ‘‘f, f’’ indicating loud. From the text of the Requiem printed on the picture, I could find the precise section of music in the
Brahms original which consists of a solo for trombone (in the score I use baritone horn for its greater flexibility and ease of pitching, but it uses the same range, and has the added advantage of resembling the French Horn, an instrument more closely associated with noble operatic melodies.). The ‘‘parasol sound’’, then, indicated that I should use that particular instrument. What it plays came from another source, from the score of Irma. which gives ‘‘quiet, high, intonation divine. . .’’ and ‘‘. . . drops the tone . . . various phrases. . .’’ all of which enabled me to have that particular instrument playing, with "divine intonation’’, a long melodic line consisting of a descending stepwise chromatic scale from top E down an octave, but very elongated. The other use of the Requiem was for the other half of the slow section, and used the following print, number 6, which refers to a sequence of rather chromatic chords in the original which I used as fragments, like the Irma score, inserting chords of my own between groups of those by Brahms to make a new sequence. So the whole of the second section uses references to the Brahms Requiem – in the first half to the harmonic content (vertical), in the second half to the melodic line (horizontal).
The last section of the piece, a chorus ‘‘Love is help, mate’’ uses a page of A Humument that is dedicated to Morton Feldman, though the actual results bear no relation to Feldman's music as such. What I did with that page was to look through some of Feldman's music to see if there was anything in it that was consistent with the way that I was approaching the score of Irma. It occurred to me to use a vocal piece for something that would be vocal within Irma and since Tom had dedicated another page to Christian Wolff – in fact a page of Trailer – and since Wolff and Feldman were close associates with Cage in the 1950’s, I used a piece called Christian Wolff in Cambridge (in spite of the fact that Tom had attended Oxford, and the Cambridge here refers to Harvard). This is a wordless choral piece which is hummed – and I used a lot of humming in the score, often as a means of separating discrete images – and consists of mildly dissonant chords. There were, however, one or two more consonant ones and I omitted those which sounded like ‘‘modern music’’, and so was left with one or two chords that I used, along with others interjected to produce a smooth flow, as the accompaniment to the melody of ‘‘Love is help mate’’. The addition of other chords was necessary because of the static quality of Feldman’s piece in which each chord is an isolated entity, and this mirrored what I was doing, on a larger scale, with the whole of Irma; taking isolated fragments and finding ways of reassembling them into a continuous whole. It could be said that I was doing to Feldman what Tom had done to Mallock since each of us extracted from a body of material what was needed for a particular circumstance, though my extraction was a good deal more cursory.
The melody that this accompanies comes from a number of sources. One of these is the stave notation and references to specific notes on Irma itself – about 60% of the notes in the melody – the rest being added by myself. One of the ideas for this came from Eric Sams’ researches into the ciphers in Schumann's music, and in particular from the fact that he originally found a clue to the cipher by finding 5-note melodic phrases in which the 1st, 3rd and 5th notes were C-A-A (Schumann’s wife was called CLARA) and this gave the possibility of finding what L and R became in the musical code, and thence other possible letters. Using this notion, using the notes given by Irma, and inserting between them other notes, the melodic lines are composed by myself but taking as a starting point the notation of Irma. The stave notation at the bottom of the score I found more usable in this way, and also as bass-lines, in transposition, rather than as originally given.
There are, obviously, some very direct references in the score, and it is the presence of these that ensure a very eclectic result: references like ‘‘the Ring’’, ‘‘the Emperor’’, ‘‘the International’’. The first of these, allied to a notation that refers to many ‘‘s’s’’ (German for E flat) suggested the opening of the Rheingold. The second, ‘‘Emperor’’, could have been a number of references – the ‘‘Emperor’’ Waltz (Strauss), the ‘‘Emperor’’ Concerto (Beethoven), the ‘‘Emperor’’ Quartet (Haydn) and so on. In the event I used the last two, and toyed with the idea of using the source for Haydn’s ‘‘Emperor’’ Quartet viz. his ‘‘Emperor Hymn’’ which became the Austrian national anthem, and which was, in its turn taken from a Croatian folk tune. I considered omitting all the musical references and only using the words of this latter ‘‘Vjutro rano se ja stanem Mal pred zorom’’ – and relished the fact that I would have been injecting something with precise semantic value, though one which I did not understand, but in the end omitted it for reasons of pronunciation difficulty. With ‘‘the International’’, I was delighted that it was misspelt (Internationale) and this made of it a lipogram (like the Ellery Queen story that omits the letter ‘‘t’’) and so I quoted the music leaving out the note ‘‘e’’. I had also considered the idea of the lipogram in another context. The original of A Humument is the Victorian novel A Human Document which leaves behind the letters AN DOC, and this gives a lipogrammatic anagram of NO ADC, that is, to avoid the notes A D and C in the piece as a whole. This seemed to be excessive, however, since it would have effectively ruled out one of the two vowels available in musical cryptography, and they are not easy to come by.

Gavin Bryars
(extract from an article written for Margins magazine, U.S.A.)

I tell you
that’s Irma herself
watching the waves fall
steep from the sun
which come at any hour
repeating certain sorts of verse

I tell you
that’s Irma herself
the morning tide
the summer sea beyond.
Morning’s murmur
like butterflies
building echoes
(Once more in tones of love.)

In another moment
Rapidly surrounded by
Sinking lights of lamps
and coloured candelabra
paper mirrors
giving rise to strange reflections
Irma you will be mine

A certain light flashed
among the Eastern clouds.
I forsee it all in the eyes
I know
Irma you will be mine.

Love is a fool
A surface longing
A mask
The hat of happiness
Love is help mate