Alright, buckle up, bandies, because today we’re going to go all theoretical on your butts.
Harmonies. They’re this complicated thing that maybe you can do by ear if you’ve been singing and playing for a while, but most people people who haven’t done a 4-year music degree don’t really understand how it actually works. They just know that some notes sound good. Some sound bad. Some clash more than others. But they have no idea how to really write some tight harmonies for their song, and therefore, do without. Not because their song couldn’t use it, but because they haven’t bothered how to learn to write a solid harmony line.
(If you haven’t read it yet, I’d recommend reading our article on Melody Writing here first, before reading on.)
Great, now that you’ve done that, let’s go from there. So, as you now know, a good melody is made up of at least 80% chord-notes and the rest are other embellishment notes, because the best (most stable/resolved) melody notes are the chord/arpeggio notes.
Let’s begin with an example. Something basic. Something easy. Something everybody knows.
Great. Now, we’re going to break that down into the notes and chords. If you can’t read the music notes, that’s okay, I’m going to put them underneath for you. The individual note-names go underneath and the chord goes above.
As you can see, along the top, we only have two chords. It just alternates between the C-major-chord and the G-major-chord again and again, C and G, C and G, at specific points through the whole song.
Now, the notes in a C-chord are: C, E and G. (It’s always the 1, 3, and 5 from whichever starting point in whichever key/scale you’re in- C D E F G A B C.)
The notes in a G-chord therefore are: G, B, and D. (G A B C D E F G.)
Because we’re in the key of C, we use the C-major scale (C D E F G A B C) as the basis for all our notes and chords, meaning that there are no sharps or flats to worry about, only white notes on the piano.
So now, if you go along our melody there, you’ll notice that every single note in this melody matches the chord above it. There are no non-arpeggio notes. The C- and E-notes are always part of a C-chord, and the D-notes are always part of a G-chord. Nice and easy. So let’s talk about how we might harmonize this melody.
For starters, you need to know something about intervals. An interval (in musical terms) is the distance between two pitches. In example, if I played both the C and E notes together (the 1 and 3 of the C-major scale), the interval is called a Major 3rd. So the chord (1,3, and 5) is essentially just two 3rds stacked on top of one another. And if you stack 3rds on top of one another on the stave/staff, you find that they stack up really nicely, all line-notes or space-notes.
So, because chords are mostly just made of stacked thirds, the most default, most basic harmony that most people intuitively try first is to just sing a 3rd above the melody. You might recognize this distinctive sound from a lot of pop music and especially pop-punk music, in which a particular melody line might repeat and a second voice will come in with a higher part that seems to go pretty naturally along over the melody.
So, let’s look at that with Hot Cross Buns:
Our chords haven’t changed. Our melody hasn’t changed. It’s still there. We just have 3rds stacked on top of every note. Now, you’ll notice that I’ve specifically marked the F-notes. They aren’t part of the G-chord. However, they basically function in every instance as a passing note between the E- and G-notes.One alternative option, however, is to say that the F-notes are not passing notes at all, but 7th-notes. If we go back to our 1, 3 and 5, starting on G in the key of C-major, we have G A B C D E F G. If we were then to stack just one more 3rd on top, we’d have G, B, D and F (G A B C D E F G), a G7-chord.
Now, we’ve already stated that the nicest, tightest harmonies start with 3rds. So, unless we’re going for something intentionally a little clashy, we don’t really want to have our two voices singing the F- and G-notes right next to each other at the same time. But so far, we don’t need to worry about that at all.
The only potential problem with this harmony line now is that it’s sitting on top of the melody, which actually makes the melody harder to hear. If we were doing traditional 4-part church choir-style harmonies, then usually, this style of 3rd-above harmony line is called a tenor line. The soprano would sing the melody up high, and the tenor would sing something like a 3rd above that but in his own vocal range, generally one octave (8ve) lower. Meaning that it would really look something more like this:
You see? We’ve taken that initial melody line, and thrown it up an octave higher, making our 3rds into 6ths. While a 6th harmony isn’t necessarily as tight as a 3rd, because of the greater space in between the notes, it does have its own particular resonance and is the next best default interval to use when just harmonizing two notes.
But you don’t need to necessarily just follow the melody line to make a great harmony line. The best harmony lines are built the exact same way as melody lines, using chord/arpeggio notes (and wherever possible, 3rds and 6ths). For instance, both the C-chord and the G-chord share a note in common: G. So with this particular song, you could create a harmony line with just a single note.
Again, not a single non-chord note. Even though that harmony might seem like a boring one, it completely works with the chords. This particular style of harmony, sitting underneath the melody like that, is generally called an alto line, because the alto voice sits just underneath the soprano voice. But as you probably can tell just by looking, most of the intervals now are 4ths and 5ths, meaning that this particular harmony, while perfectly serviceable, and even appealing in a stairwell or large church hall, won’t get us a particularly tight-knit sound on a recording. But luckily for us, we have a better, closer option for an alto line.
Yes, they’re singing the same note some of the time, but overall, and especially if we have more than just two voices, this is going to be the superior alto harmony, whether up or down the octave.
Now, the remaining part we haven’t really talked about yet is the bass-line. The bass line is pretty simple, it’s usually just the 1st note of each chord, maybe with some passing notes in between or something. With most harmony parts, we want to limit the movement to get closer, tighter sounds. But with the bass, we really want the bass note of the chord to be the chord-note. It sets up the whole tonality of the chord for us. So, in this case:
See, it’s just C and G, C and G, just like the chords above. Obviously, the intervals are all a lot bigger now, but as you may or may not have noticed, we now have the space to put some other notes in between, like the alto and tenor.
And now, you can literally go through line-by-line following each of those parts: soprano, alto, tenor and bass. And if you look at the note names stacked up along the bottom, you can see those stacked chords, C, E and G make a C-chord. G, B, D and F make a G7-chord. The biggest problem with this, however, is that we have the bassline jumping up above the tenor part, and they’re clashing on that 2nd interval between the F- and G-notes, and we already said how that’s the sort of thing we wanted to avoid if we could help it.
But obviously, it gets a bit hard to read and make sense of the parts when you’ve got that many of them stacked on top of each other like that. Plus, most of the parts are in the middle to top end of their vocal ranges. So, we just drop it all down an octave.
It’s exactly the same as before, we’ve just now got our sopranos and altos together on the higher, treble-clef stave, and our tenors and basses together on the lower, bass-clef stave. But the parts haven’t changed at all. Now, if this looks at all familiar, it’s because this is how 4-part is usually written out, especially in church hymns and choir songbooks.
In most traditional settings, the chords actually won’t be written across the top like that, but they don’t really need to be. Any musician who knows their chords and how they work, and can read music, is able to just identify whichever chord is being played at any given time because of the notes and harmonies. They’ll see that C, G, C and E of that first chord and instantly recognize it as a C-chord. It’s not going to be anything else.
Now, you don’t just need to stop at 4-part harmonies either. You could add a contralto or baritone, or both:
A rule of thumb about that, though: Just like there is a hierarchy of melody notes, there is a hierarchy of doubled harmony notes, following the same pattern, 1, 5, and 3. You generally want to double the 1 before either of the others, if you can, followed by the 5, and last of all, the 3. This is because it’s the 3 that makes your chord major or minor, and because it’s variable, because it changes, it’s inherently less stable. If you have two voices singing that 3-note, they are more likely to sing slightly out of tune and fuzz up your chord, than if they were both singing the 1 or the 5.
As you can see, once you get up to 6-part harmony, it’s rather impossible not to double the 3 sometimes and still maintain good step-motion voice leading in each of the parts. But even when I do double that E-note in those C-chords, I still have 3 C-notes going at the same time, so it shouldn’t overpower the chord.
Now, each of those 6 parts has its own distinct tune, none of the parts are doubled up or down the octave, but you could easily turn those 6 parts into 12 different parts by having the 6 distinct parts sung by men down in the bass, and again an octave higher by women in the treble.
(Yes, some men can sing alto and even mezzo-soprano parts. Yes, some women can sing tenor or even baritone. I’m just talking about the generally applicable, useful rules. There are exceptions to every rule, obviously. But knowing the rule still always helps increase efficiency.)
Now, there’s no way I am trying to cram 12 parts into those two staves, but I will show you an octave harmony:
They’re just doing the same part, an octave apart. One could be a soprano and one could be a tenor. Or you could do like Ed Sheeran and sing the same melody line in two different octaves to avoid over-complicated harmonies.
You see, when you have a very melodic melody line that jumps about a lot, it can be hard to harmonize with traditional harmonies without jumping around an equivalent amount. If you do it right, it can turn out pretty powerfully, but it’s easy to make mistakes with it. It’s not usually the sort of thing you can just do on the fly. You need to actually write it all out like we’re doing and plot out your harmony line(s). Now, at the moment, we don’t have to deal with something overly complicated like that, but maybe we’ll do one in a minute.
First, however, I want to draw your attention to one final element about all the harmonies we’ve done thus far. They’ve all been perfectly stacked up on top of each other, perfectly in sync, just singing different pitches. This is called homophony, or a homophonic passage of music. Polyphony is when harmony lines have different rhythms (and even words) and run around each other like chickens in a coop, instead of stepping forward together like marching ants or soldiers, homophonically.
I’m still basically moving by step, but I’ve thrown a bunch of passing notes in offbeats between more stable chord/arpeggio notes and created a harmony line with a lot more movement. In fact, it has so much going on that it’s almost trying to convince you that it’s the real melody. Luckily, we can handle that by adding another harmony.
Now, I didn’t add a third rhythmically-distinct part, just a second harmony, homophonic with the first one. However, I did create a unique 3-part harmony that’s just got a little bit more going on. What’s more, I have 63 notes and only 11 non-chord notes, meaning that I’m still at 82.5% chord/arpeggio-notes overall.
Okay, how we’ve covered all those basics, I want to give you an opportunity to try it yourself with a different song. If you don’t have manuscript paper or music notation software lying around, you can just draw your own lines on paper.
(FYI, the notes in an F-chord are: F, A and C (F G A B C D E F), and the notes in a C7-chord are C, E, G and Bb (C D E F G A Bb C).
Okay, here you go. This is it. See if you can do it.
It’s a little bit more complicated than Hot Cross Buns, that’s for sure. But still only a few chords.
SO DON’T SCROLL DOWN YET UNTIL YOU’VE GIVEN IT A FAIR TRY!!
Okay, okay, now that you’ve at least tried, let’s look at some options.
By the way, could you tell what song it was?
It was the much-beloved Christmas carol, Silent Night.
This is an easy alto/bass line. With only a couple of exceptions, I’m basically only using the bass notes of the chords to fill in some harmonies. Did you get something similar?
Or did you perhaps get something a little more like this?
As I can’t personally see what you’ve got, I can’t really offer any more specific feedback, but I would recommend going through the article again a few times and trying to really look at everything that’s happening. See if anything went right over your head.
Being able to write good harmony lines can be a really invaluable skill especially in the studio. But again, just like any part of songwriting, all it requires is some dedicated daily effort. Even if you just start with a basic nursery rhyme each day, you’ll be doing pop songs before you know it.
(Nursery Rhymes and other basic traditional folk songs, like Christmas carols, are really great for this because they often imply the chords very clearly with their simple melodies.)
Here’s one last one for you, nice and easy, to get you started.