12 December 2011
Professor Christopher Hogwood
This is the last lecture of this year and it is the equivalent of a trip to the zoo. I shall be parading you past a number of cages, explaining the slight differences between species to you. It shall be a sort of “show and tell” of certain techniques in musical composition that lie behind quite a number of masterpieces.
The two devices that we want to parade past are very simple. One is the round, or the canon. Everybody has sung “Three Blind Mice” at some point, so you know what I mean by the round – somebody starts, and then the next voice follows at a fixed interval and so on. The other device, which is curiously very often connected with the round, is supporting harmonies, a bass line, repeating itself, underlying the piece - sometimes very simply, sometimes quite complicated.
I’m dealing with the two together for various reasons, one being that the earliest example we have of notated polyphonic music in this country actually employs both these devices. It is not seasonal to sing “Summer is coming in”/”Sumer is icumen in”, but it demonstrates these devices. You will notice that it is held up by an ostinato, which says “Sing cuckoo, sing cuckoo, sing cuckoo”, and never stops; over the top, you hear, “Sumer is icumen in” by one voice followed by another followed by another.
That is the first problem with rounds and canons: how to stop them! You might have wondered if it was ever going to come to an end, and in fact, there is no notated end – it could go on perpetually.
A second obvious audible problem is that it never moves anywhere harmonically. Because of that repeating bass, it can never get away from that key. This can be used advantageously, but it is certainly a defining form of a round bass that it sticks to one tonality, just as the canon is defined by the fact that the second voice has to follow the first exactly and the third has to follow and so on. So, once you have heard one and remembered that, you recognise it each time it comes up. This is the idea of how you get your bearings in music by hearing something that has been made familiar by the composer. Either the repeating bass or the repetition of the canonic part or both together make things stick readily in the mind.
Here is exactly the same technique used in a piece that has become very popular, more or less at the same time as Vivaldi’s Four Seasons became popular: the Pachelbel Canon. It was first printed, curiously, in 1941, and recorded shortly afterwards, and now of course is at the top of most people’s hit lists. You can hear the same techniques as “Sumer is icumen in” in use, except the initial bass, the series of harmonies, employs slightly more notes than the two used in the first piece you heard. But the canon is an exact canon: first one violin and then a second violin and then a third violin. They start playing just in scale-wise, then the rhythm heightens a little bit, they begin to move apparently independent of each other, but everything harmonises over this repeating bass pattern and you never once leave the home key.
There is a certain compulsion to nod off - not least for the people who are playing the surrounding harmony, which does not change at all! But there is one improvement on “Sumer is icumen in”, in that Pachelbel did fix an ending for it, so it does not go on forever. He arranges all the rhythmic intensity in the centre of the piece, so you begin slowly, gradually wind up, and then gradually wind down – there is a shape to it. Even so, canon developed just like that would not be a great talking point, certainly not for a whole hour. One has to devise other conditions in which the canon can fit, sometimes almost unnoticed, as part of the fabric of the music.
Some of the most ingenious canons are to be found in Bach’s Goldberg Variations. You will remember that Bach wrote these for Goldberg to play to his aristocratic employer - not, as programme notes often tell you, in order to send him to sleep like the Pachelbel Canon, but in fact to entertain him because he suffered from insomnia and wanted something to occupy his mind in the hours when he was awake. The arrangement of the set of 30 variations that Bach devised is quite ingenious, in that every third variation he made as a canon: that is, although the harmonic scheme of each variation is taken directly from the aria at the beginning of the piece – this is a set of variations based on harmony, not on melody – the aria that starts the Goldberg gives you the ground plan of every movement thereafter. It is what the upper parts do, because they are newly invented for every variation, and while some are single line, some are slow, some are fast, some are toccatas, some are dance forms, some are jibs, the canons, mostly held in the right hand, are arranged with great ingenuity, because the first canon is written at the unison. That is exactly like “Three Blind Mice” - the second voice comes in singing “Three Blind Mice” at the same pitch as the person before. That works for one canon. For the next canon that comes up, three variations later, Bach has just shifted the interval, so the next voice comes in one note higher. The interval gets bigger and bigger, so although the parts exactly follow each other, they are just bodily moved one tone higher, two notes higher, three notes higher, and so on.
By the time we come to the fifth canon, Bach thinks that people know he is going to start here, and the entry will come there, and it will go along in parallel. So, instead, he puts a mirror to the canon, as it were: it is a reflected image. The second voice is the opposite way round.
The construction, although it is extremely technically ingenious, does have that redeeming feature of all masterpieces – it sounds completely natural, and the same goes for all the canonic variations in the Goldbergs.
Bach was a master of this. We tend to think of these techniques as being the purview of the Baroque period in particular, when technical counterpoint was very much prized, but it is not exclusive.
Amongst nineteenth century composers, I guess it is Brahms who most admired earlier techniques and styles. He made a great collection of earlier music, and he did a lot of editing of early music. You almost feel he believed he was born into the wrong century – he would have loved to have been a Baroque composer. Amongst his collected snippets of music were examples of ingenious canons from the very earliest Renaissance, right through Baroque and Classical times. He was obviously fascinated by what could be done with this sort of technique, without sort of announcing it to the public. Just to add to his own collection, he composed, for ladies’ choir, a set of canons which, unless you are told they are canons, are not immediately recognisable as such. It sounds like a freely composed piece of music, but in fact, it is strict canon all the way through. Here is one of the set – he in fact wrote a dozen pieces, just to show that canon could be used, as it were, unobtrusively.
They are all very ingeniously constructed. But, as you can hear, they have the same problem that “Sumer is Icumen in” had and Pachelbel had: they cannot get very far away from the starting key. As a result, you are limited by the range, the canon and its answer, and so, essentially, each example of canonic writing will be quite short.
That was written about 1860, 1859, or so. Just slightly earlier than this, there was a very ingenious canon in the last movement of Cesar Franck’s Violin Sonata. This was a sonata that was a long time coming. Wagner’s wife, Cosima, wanted a violin sonata from Franck. He never got round to it, and in the end, he was pushed into it by the violinist Ysaye, who asked him to write a violin sonata as a wedding present. The wedding date therefore gave Franck a fixed deadline. He did manage it. The premiere took place in an art gallery in Brussels, and somebody had forgotten to turn the lights on, so the reputation of the first performance was that Ysaye ended up playing most of it from memory.
The last movement is led by the piano, followed exactly by the violin. It does manage to convey, occasionally, the feeling that it can escape from these bounds, but it is quite a long chunk of strictly canonic writing before the two instruments become freed from each other and go their independent ways.
Canon goes on being used. If you listen to Bartok’s Concerto for Orchestra, there is a wonderful canon in the third movement, where the winds follow each other for an enormous length of time, clashing occasionally, but with the second part copying the first exactly. It is an inescapable part of a composer’s technique and it is one of the early lessons you learn in writing counterpart: how to write a two-part, three-part, four-part canon.
You also have a series from the Renaissance of puzzle canons. If anybody is interested in cryptography and crossword puzzles, these are available for solving still. They would write the canon out, but without indicating what sort of canon it was. That was up to you to discover. Is it a simple “Three Blind Mice” canon, or does the next voice comes in at a different interval, or is it possible that it works with the second and third voices coming in upside down, or does the second voice come in singing notes twice as long as the first voice (i.e. the slow motion version of the canonic thing). We are sure that these canons worked because the composers assure us that they are canons, but there is quite a collection of still unsolved canons, where even clever computer technology has not managed to find which of the thousands or millions of variants is the actual solution that would make the canon work and whether it is a canon in two-parts or three-parts or four-parts or what have you.
Let’s now leave the canon on one side for a moment and go back to the other features of the Pachelbel and the “Sumer is icumen in”: the repeating bass pattern. This is another very easy way for an audience to immediately acquire solidarity and assurance. You hear the chaconne or the ground bass, the obstinate ostinato bass, never moving from a series of notes or, usually, a series of harmonies attached to these notes. That gives you the ground plan of many pieces of music, from Baroque, from Classical, even the present day; composers writing passacaglias, chaconnes, ground basses. It is another technical challenge, but half the challenge is to make your audience not so aware of the fact that there is a fixed patterning on the bass line.
It is also a very useful operatic device. Many Baroque operas would end with a chaconne or a canon. After much mobility in your evening in the opera house, it was nice to feel you had come home at the end to something that was absolutely stable. When you had heard this bass pattern repeated three, four, five times, you knew everything had been resolved and you were on safe ground.
One of the very early experiments with this sort of form is Monteverdi. I thought it would be nice to listen to part of the Zefiro Torna, one of the madrigals he wrote, over a very simple syncopated bass. It only has four notes in it – it is just a two-bar pattern. It repeats 100 times at least! Over it comes this wonderful poem of the soft south wind blowing, and the sound of the lovers’ voices. It is set for two tenors, who, in a way, imitate each other. This is an example not of strict canon but of something that is canonic sometimes, and these voices sing of the sounds being replicated. The echo comes back from the mountains, and so the second voice imitates the first voice – it is almost a canonic effect. It is a lovely example of ‘picture painting’ because the sound goes up to the top of the mountains and down to the bottom of the valleys and Monteverdi follows all this. The gentle swirling and the coloratura of the parts is the main stage, and are as lulled as you are by the insistence of the Pachelbel canon. You sink away into it, feeling that everything is very secure and the wind is blowing and everything is fine in the world. Then, suddenly, it stops, and the whole thing just shifts up a tone, and you realise this is a very shocking moment. You have been lulled into one set of harmonies, and the text says: “But I am not in love – I’m wandering alone in the forest.” The joyful benefit of having two tenors is that not only can Monteverdi trick you once with this thing – the second tenor can then do the same thing, and you get this miraculous move of tonal centre, as you shift into a completely new territory, a new key. The whole mechanism of reassurance has gone. It is a beautiful picture.
It is a wonderful shock! I forgot to mention, there is that little false alarm that you hear in the middle, where it seems to stop, and then goes on with the endless ground bass. That was nothing to do with Monteverdi – that was just the length of the 78 recording, where they had to stop and turn it over, and they were not able to patch it so that it sounds as though it was completely continuous.
The repetitiveness of the bass is not mirrored by what the voices do over the top, and I think the skill of the greatest contrapuntalists in the Baroque time - Monteverdi being one, Purcell being another - is that they hardly make people aware of the necessity to fit to the repeating bass, something that Pachelbel hardly tried to do in his canon. Purcell wrote in the little Playford guide to writing music that the business of writing on a ground bass and writing in canonic form was “an easy thing”. Purcell is constantly trying to give you the feeling that this is a work of several movements, in distinct meters, some slow, some fast, and even going into a variety of keys. This is a piece with three violins, like the Pachelbel, and the repeated bass with the same number of notes as the Pachelbel canon, but like Pachelbel, nearly all the top parts are in some way derived from each other. It is not a canon from the very beginning, but it is canonic, and it shows Purcell’s skill in mixing the two together. Sometimes there are three voices, all saying the same thing; sometimes it is block harmony; sometimes it is two against one. Never are you aware of the constant measure of that bass line, which could be splitting it into very obvious chapters if you were not careful.
You get some idea from that, I think, of the different techniques Purcell has for avoiding the routine. Also, the ingenuity of having the unadorned bass line not showing at the beginning, but in the middle: suddenly, everybody stops, and you just hear the continuo, the lute and the bass, playing along. He reserves that, in case you had not noticed it was a ground bass. Up to that point, there is nothing but the ground bass, and then you realise that everything you have been listening to so far, which you thought was a multi-section Baroque sonata maybe, was in fact over this inflexible ostinato.
Because of techniques like that I think that, when we come to Dido and Aeneas, the farewell of Dido over a chromatic ground bass never feels as though that repeating pattern is in any way getting in the way of the distribution of words and music above. When you only have one voice, of course, there is no question of canon, so this is purely an example of how to exploit the harmonies possible from a chromatic bass. Most of these chaconne basses had certain connotations: you have quite a jolly one, like the Monteverdi; you have quite a normal hymn tune sequence of harmonies, like that last Purcell. But you have these descending chromatics, going down a fourth, from the tonic to the dominant, in the scale, hitting each chromatic, which was the hallmark of a lament. There are many uses for this chaconne bass, and everybody knows from the minute it starts, even before Dido opens her mouth, you know that this is the end – she is lamenting. This is the traditional shorthand for “end of opera, end of heroine”.
However, notice how Purcell ingeniously uses, not only the music but even the words to get you over this bump where the bass has finished and comes back to the beginning again - the repetitiousness. He does it not only by casting the musical line as a bridge over these joints, but also just by using the words – “when I am laid in earth, may my wrongs create not trouble”. The unfinished sentence, “may my wrongs create…”, is very neatly placed at the end of the chromatic pattern, so you know that the sense is not going to stop, and you follow the grammar as well as the music; you are not at all aware of the fact that the bass line is repeating on itself. That is one ingenuity.
A second ingenuity is the simplicity of a line like “Remember me”, on a single note, which is very beautiful. Unlike Monteverdi, who uses a good trick twice in Zefiro, with the two tenors, Purcell does not repeat his “Remember me”. The second time it comes round, you get a top note. It is the only high note in the whole piece, and it marks a point. Again, you realise when you hear that top note that you have no idea whereabouts you are in the sequence of repeating bass lines. When Dido gets to the end, the lament base does not finish. It goes on, with some of the most extraordinary harmonies. If you think that a set pattern of bass notes requires a predictable series of harmonies, or that you might have heard every possible harmonisation in the course of the lament, then you have to think again; just at the end, before it moves into the final chorus, Purcell gives you all the other harmonies he could have used, as an epilogue.
It is a moot point whether you remember “Remember me” for its bass line or for its harmonies or for its single melodic line but, in general, with the later chaconnes and passacaglias, one remembers them for their sequences of harmony. You read the progressions as harmonic progressions rather than a bass line progression.
I shall now play you two tiny epilogue pieces to show you that in practice: one is an advertisement for the last lecture we shall have this season, when we shall be dealing with Bach’s unaccompanied violin music, and unaccompanied violin music in general. We shall be talking in detail about the Bach chaconne for solo violin, which begins with a series of harmonies that are very memorable.
The ultimate example of reducing the interest of the bass and increasing the interest of the harmony is this last little sample, which is the last movement of Brahms’ Fourth Symphony. It is a passacaglia, very much taken from models that Brahms had collected from Bach’s cantatas. You will notice, in the introductory chords that the orchestra gives out, how the bass stays exactly the same. There are seven different harmonies, and the bass is exactly the same until the very last: so it is tonic, tonic, tonic, tonic, tonic, tonic, dominant, and back to the tonic again. You would have thought that it is a rather unprepossessing sort of bass on which to model quite a substantial movement, but almost immediately, he takes the implications of the harmony as the reiterated ostinato, not the fact that it was a monotonic bass.
That really concludes my little zoo tour, but there are many more animals that could be presented in this guise. I think it is interesting that the very first piece, “Sumer is icumen in”, gave us most of the material from which we can extrapolate these other devices, be it Monteverdi, Brahms, Purcell, Handel, Cesar Franck or any of the others.
© Professor Christopher Hogwood 2011