Music sounds. This is not to imply that it in any sense commences, with uplift baton, ha-one, two, ha-one two three four for what actually happens is that the audience here, hawking and spitting impatiently in the darkened hillsided seasided cwm of buttockbreakingly hard seating, suffering, waiting sweating to see they know not what, suddenly start paying attention to the music which has actually been going on for some time now, creeping up into their collective consciousness from unimaginably profound depths; for it is the canon per tonos from the "Musical Offering" by J.S. Bach, a circular canon which transposes itself one tone higher with each repetition. Now since we are already reasonably up into instrumental ranges, no pitch below about 60cps., you can only guess at where we must have started out; but wait- who said anything about starting? This is a perpetual transposing canon and it's a perpetual transposing performance we've latched onto, into, here; it must've been really low a while ago then, not just off the left hand end of the piano keyboard but really low, frequencies below 0 cps... Now, it takes as long for some of these frequencies to complete a whole waveform as it does for several notes to be played at the stage in the performance that we're at now, so it stands to reason that the very young Canon must have been going extremely slowly so that the relatively shorter durations in it would have had time to be defined. From this we can draw another crucial conclusion. Since it now seems to be ticking over quite nicely at around crotchet (quarter note) =c.72, the whole Canon must be undergoing an acceleration. In other words, as the frequencies defining the individual pitch classes get larger, the duration of the notes actually generated by those frequencies are getting smaller; and it follows that repetitions of the Basic Canonic Unit (BCU, sometimes called the transposer) will come closer and closer together, until they enter the audio range at a value of around 20 repetitions per second. This was more or less predicted as early as 1968 by Stockhausen, who pointed out the structural importance of the relationship between duration and pitch in his theoretical papers and in Kontakte for tape with piano and percussion. But we can now begin to see the possibility of a theory which shows his more general formal predictions to have been accurate also, for the result of audio range BCU repetitions towards the end of the Canon's life would be a simple glissando. This, the Big Slide theory, however elegant as a theoretical solution, nevertheless fails to satisfy the requirements of the General Theory of Stylistic Consistency, which does not allow glissandi in Baroque music except as very short range articulative gestures, the portamenti in vocal or string lines. Leaving for the moment the fully developed String Theory (which sees portamenti as very old canons left over from the sketching stage of composition) we need to look for a theory of what happens at the end of a perpetual canon which does take into account GTSC. I should like to propose that the main source of our difficulty in understanding the Canon comes from our too great preoccupation with beginnings and endings. The theory that as BCU repetitions enter the audio range they produce a glissando (which is not allowed in any case by GTSC) fails to take into account the fundamental fact that this is, after all, a perpetual canon, so that there cannot be a glissando later simply because there was not one earlier! If there were, the Canon would not be perpetual- it would change into the Glissando (Big Slide.) BS theorists tried to get round this by saying that the Glissando is a singularity; but this is dodging the issue, and still does not attempt to describe the early sub-audio life of the Canon, except by hinting vaguely at a second singularity. Another objection is that the Glissando would have to be infinite, as there is no reason why it should ever stop. But this is impossible, because the instruments playing the canon are only capable of vibrating a finite number of times before they fall to pieces (Disintegration) or the players die of old age (Dying of Old Age). And yet it was while pondering this problem of perpetuality in preparation for a seminar in Leipzig with some of my postgraduate students that the breakthrough came. We proposed that as the BCU repetitions reached the top of the audio range they did not continue independently until they re-entered as the Glissando but instead began to function as the harmonic spectra (upper partials) of fundamental pitches in other performances of the Canon. Experimental confirmation of this was achieved in Cambridge only days after the Leipzig seminar in what has already come to be regarded as a classic demonstration: copies of the Canon were distributed to organ scholars in several colleges, and performances were given while my postgraduates bicycled rapidly in computer generated random motion between them. In itself this was a major achievement, and I expressed at the time my thanks to the Police and my condolences to the relatives of Marvin of Clare, a great brain sacrificed in the cause of the advancement of science. One fact emerged clearly. More than one performance of the Canon clearly existed or could be caused to exist, and it was this realisation which led to our solution of the problem of the very young, entirely sub-audio Canon. Since older, higher states of the Canon always implied and indeed were identical with the harmonics of younger, lower states, we finally suggested that what had been thought of hitherto as some kind of sub-audio "beginning" was in fact no more than the implication of presently existing, audio-range states. So the Canon was perpetually created out of its own audio-range implications, and could therefore be regarded as a finite musical space without boundaries. A subsequent experiment in Cambridge, this time using free beer and lunches instead of copies of the Canon, revealed that most of the organ scholars involved had already performed the Canon prior to our original experiment, but at different times. Now the final difficulty dropped away, for we had not been able to understand how, if there were no boundaries of beginning or end, it was possible to start playing the Canon; or for that matter, how once having started, you could ever stop- yet here were our organists, mopping up the Greene King, evidently not playing the Canon, indeed, in one or two cases, incapable by now of any non-random motion. Furthermore, when we brought out copies of the Canon, we discovered that it had properties of spin. That is to say, when you rotate the score through 180º, it looks quite different. Only when rotated through 360º does it look the same. But it is impossible to say at this stage what this means because of the uncertainty principle: in simple terms, when you know why it seems important to rotate the score you don't know how many pints you've had; and when you do know how many pints you've had you're not drunk enough to want to sit round a table in a pub with a gaggle of half-pissed organ scholars but I digress. Perhaps the most radical implication of our researches is explored fully in a paper I presented in 1988 to a seminar of reactionary philistines distinguished musicologists in North Oxford. At last I felt I had achieved an understanding of how the Canon moves in non-bounded musical space-time, a theory of fully unified Canonic kinetics (FUCK). But I was unprepared for the hostile response from my distinguished audience when I pointed out to them that since the Canon always generates all of its possible performances it must have created performances prior to its "first" performance, which means that in the fully unified theory there is no place for J.S.Bach as there was nothing for him to do- the canon was already in existence and the concept of a composerly creator clearly unnecessary.

Leighton Buzzard: (seated at the Metzler organ in the University Church of St Mary the Virgin, Oxford, playing the circular canon at the fifth between the two lower parts, modulating key by key, varied theme at the upper part, not by J.S.Bach, hawking and spitting the while into the pedals:) Ascendenteque modulatione ascendat membrum virile!