A few years ago, BBC Radio 3 broadcast 'A Bach Christmas', playing all the works of old Sebastian, and having enjoyed a feast of counterpoint I thought I might surprise the family with a little canon based on Bach tunes. They're not really into classical music (to put it mildly), but as I recorded it using the 'Bells' sound on our electric organ, it sounded very festive - like having our own cathedral to ring in the new year. It grabbed their attention for at least a couple of minutes. Yes, that was one wild new year!
I love trying to write canons but, no, I'm not going to share them with you. Instead, I want to try to give a short overview of what canons are.
A round is a piece of music where one voice (or instrument) begins the tune and two or more other voices (or instruments) join it later, singing the same tune, note for note, overlapping the first voice. The result is that different phrases of the tune are sung simultaneously. The result should sound harmonious.
This definition (my own) of a round can serve just as well as the definition of a canon, as the round is only the simplest form of canon, known as a canon at the unison - that is when the second voice (and then the third voice and the fourth voice, etc) begins the tune on exactly the same note as the first voice. A canon at the octave is very nearly the same thing (and will sound pretty much identical to the listener, especially when a mixed choir is singing), except that the second voice begins...and I think you might guess where I'm going with this one...an octave higher. All my attempts so far have been canons at the unison or octave.
The first-ever (known) piece of counterpoint was a canon at the unison:
Sumer Is Icumen In (Summer is here!), dating from around 1260. Even if you don't read music, comparing the shapes of each of the lines in the score below should give you a good idea of how a canon at the unison looks on the page, and perhaps how it might work in practice:
With Sumer Is Icumen In, it's not very difficult to hear the process in action. Even the accompanying figure (which you can see beginning in the tenor part at the end of printed score) doesn't distract the ear.
That's not always the case. A well-known later example of a canon at the octave is the eight piece,
God grant with grace, in Tudor composer Thomas Tallis's extremely beautiful
Nine Psalm Tunes for Archbishop Parker's Psalter, a piece known as
Tallis's Canon, or as the hymn
Glory to thee, my God, this night. Here the soprano follows four beats behind the tenor and begins an octave higher. The other two voices do their own thing, though occasionally echoing parts of the canon. Whether you listen from 6.10 in
this video or 6.07 in
this one or 6.08 in
this one, can
you hear the canon being sung between the tenor and soprano? I'm afraid I just seem unable to hear the piece as anything other than a soprano tune accompanied by three other voices providing harmony. For me, the tenor's tune gets lost in the lines and harmony of his two non-soprano companions. I can see the canon in
the score but can't hear it in performance. Possibly if some performances brought out the tenor part out more (got Placido Domingo to sing it maybe!) then I might hear the process of Tallis's canon in action, but I'm not so sure about that.
You can have canons at any interval - at the second, third, fourth, fifth, sixth, seventh, etc.
Pachebel's Canon, if you were wondering, is a canon at the unison, with a ground bass for accompaniment:
Canons are fun for composers. They are a bit like doing crossword puzzles or playing chess. They take a great deal of ingenuity to pull off. Many composers want us listeners to share their pleasure in the ingenuity of their canons. They want us to hear what they're up to. We listeners can indeed find it highly satisfying to follow the different voices as they imitate each other. However, some composers are far less concerned that the listener hears the canonic process. They might even wish us not to notice the intellectual scaffolding beneath their music. Or they might care only to please themselves, or connoisseurs of scores, or performers. The anonymous composer of
Sumer Is Icumen In clearly fell in the first camp, and I suspect Tallis fell in the second camp.
Where does Webern fall with his
Dormi Jesu from the
5 Canons on Latin Texts,
Op.16? Well, he's not exactly hiding the fact that his pieces are all canons, having chosen that title for the set. This particular tiny little piece for soprano and clarinet, however, is an example of a type of canon that is rather hard for the listener to follow as a canon -
canon by inversion. Here any rising interval in the first voice becomes a falling interval in the second voice - an instance of contrary motion. It's easiest heard at the very start of
Dormi Jesu, where the first four notes played by the clarinet rise through three intervals and the soprano then enters with four notes falling through the same three intervals, upside down. Thereafter I suspect you will find it harder to follow (as a listener), given the wide leaps in the two lines and the fact that you will probably hear the two parts as being completely different melodies.
Even harder not to hear as two completely different melodies is the
retrograde canon, or
crab-canon, or
cancrizans. Here the second voice imitates the first voice by simultaneously singing/playing the melody backwards. The ear is, I would say, incapable of hearing the second voice's part as a new tune altogether, unrelated to the original tune - except to anyone reading a score. With a crab-canon, the composer is satisfying himself and score-reading scholars, but not his listeners - at least as regards the satisfaction given by following canons. Obviously, the piece can still delight us even if we don't know that there are canons at work. (That really
is obvious!)
This truly excellent YouTube video gives a very clear demonstration of the cancrizans at work in a movement from Bach's
Musical Offering.
And there's more. Stravinsky wrote a deeply expressive
Double Canon, where two canons on different themes proceed simultaneously. You can have
triple canons, or as many canons going on as you like. The great American-born maverick Conlon Nancarrow had twelve of them going on in his jaw-dropping
Study No.37 for player piano! Plus, as the Nancarrow example shows, canons can be played at different speeds. There are
prolation canons where the tune is, indeed, played at different speeds in different voices. A beautiful recent example is Arvo Pärt's
Cantus In Memoriam Benjamin Britten. where an A minor scale falls repeatedly against itself at various speeds
.
Canons have been with us since around 1260 then and continued to flourish for the rest of the Medieval period. They were very popular with the Renaissance and the Baroque. After a very short dip they rose again with the Classical era. Haydn, Mozart and Beethoven all wrote them. Many were jokes, such as Beethoven's delightful
Ta ta ta, lieber Mälzel (the tune of which might well remind you of
this). One of Beethoven's most wonderful creations, the quartet
Mir ist so wunderbar from the opera
Fidelio, is also a canon.
YouTube has huge numbers of these fantastic little chips from Beethoven's workshop
here. Canons lasted even through the Romantic era, where they tended to become far less strict and were usually accompanied examples, such as this gem from Schumann, his
Study No.4 from Op.56. Brahms was their greatest Romantic exponent, using canons in many of his works and writing a fair few stand-alone ones, such as this magical specimen,
Einförmig is der Liebe Gram, based on Schubert's
Der Leiermann from
Die Winterreise. With the coming of the Twentieth Century, canons of the stricter and unaccompanied kind came back with a vengeance. All manner of composers relished them, from
Schoenberg (you should enjoy these!) to Stravinsky (see above), from Pärt (also see above) to Ligeti (not yet on
YouTube).
In other words, they've always been loved by composers and are one of the most durable musical forms. Long may that continue to be so!