How To Finish What You Started

Song Structure

Lyrics
If you’re using the verse/bridge/chorus format, in whatever order, there are often common features.
The verse presents the facts and sets the scene. The bridge adds contrast to stop the list of facts from becoming dull. And the chorus comes to a conclusion.
So let’s put this into practice in a dull way, then examine a proper song that is much less so.
The subject matter is the weather. The verse might be talking about how sunny and pleasant it is. The bridge adds contrast by talking about it getting cloudy. Chorus conclusion might be that just because it’s sunny in the morning, doesn’t mean it’s going to stay that way all day. There you have it, the most boring song idea ever.
So let’s look at a proper song.
Northern Lad by Tori Amos. The verse describes hopes and aspirations at the start of a new relationship; “I thought we’d be okay”. The bridge adds doubt in the form of something being not quite right: “But I feel something is wrong”. the chorus comes to the conclusion that if something isn’t working then it’s sometimes wiser to quit: “You’ve got to know when it’s time to turn the page”.
Music
The song is in the key of A major. The verse sequence begins with the I chord, and is mostly major with a minor chord thrown in. This reflects the feel of the opening verse in that it’s essentially optimistic but already there’s a shred of uncertainty there: [A] Had a northern lad [Bm] Well not exactly had.
The other thing to notice is that it does what heaps of chord sequences do in that it starts with the I and works its way to the V which provides the strongest pull back to the I and so is great for turnarounds.
The bridge begins with the vi. This is common, as is beginning bridges with the IV. Tori probably chooses the vi because the bridge is not adding contrast in terms of a brighter change but a darker one, and so the minor chord better reflects this mood.
The chorus in this song is quite unusual. It begins with the vi chord which is the same as the bridge. She manages to make it sound fresh by almost bringing the song to a halt. A lot of chorus’ switch to the V chord to add a contrast and a lift from the I chord of the verse, but Tori most likely chooses the vi because the chorus is talking about something that has failed and so the V chord might not provide the right mood. Also it’s worth noting that she uses the IV to end the chorus. The IV isn’t as powerful as the V when used as a pull back to the I, but still works well.
If you analyse lots of songs, you’ll find some very common practices in terms of chord usage to signal the beginnings and endings of sections, but don’t expect to find hard and fast rules.
Song Dynamics
A very important aspect of song structure is the way songs progress dynamically as they switch between sections, and move through the song as a whole.
The most common overall dynamic feature is that the song starts softly, gradually increases in intensity, builds to a climax, and then quickly winds down for the ending.
This is true of Northern Lad. It mostly involves Tori increasing her vocal volume and pitch before winding down both elements for the ending.
Drums can often play an important role. Drummers vary the ’busyness’ to mark distinctions between the sections. This can be as simple as doubling the amount of hits on the hihat, or just playing a slightly busier pattern. They usually announce the approach of anew section with either a variation to the pattern they had been playing, or with a roll into the next section often concluded with a crash on the cymbal, which falls on the first beat of the next section.
The drummer is softening the blow that might occur if the rhythmic pattern were to just change abruptly.
If you’re not planning on having drums in your song then you’re going to need to create rhythmic variations on your guitar or whatever. You should do this anyway because the bridge and chorus are supposed to add contrast, so you probably shouldn’t just plod along with the same old rhythmic pattern throughout all three sections, but again, no hard and fast rules apply. Sometimes just one variation is enough to refresh the song.

Application

Getting Started
The most important thing when beginning a new song is to be prepared to abandon any ideas you may have about how the song should progress. It’s useful to have a rough idea of what you want to do, but don’t stick rigidly to it.
Approach your project as if you’re discovering an already existing song rather than creating a new one; be a facilitator not a dictator.
Keep your options wide open, and be prepared to be ruthless when an idea or part you like is holding back the project. There are times when you have a really great guitar part, but it’s just too busy, and when you remove it the song takes on greater clarity and spaciousness. It’s a wrench to let that nice part go, but you have to be tough for the sake of the bigger picture. This is true of all the elements of song structure.

Using Motifs

Motifs are hugely important and useful when creating a song. They keep the song ’on track’ and stop it from wandering off and falling apart. If you create a rhythmic motif and apply it to a musical phrase that sounds good, then using the same motif for the next phrase will help to let you know what will and won’t work.

What Is A Motif?

A motif is a pattern that is repeated to create a larger work. In music there are two types of motifs: rhythmic and melodic. Creating rhythmic motifs is simple and can be done without an instrument. Melodic motifs require a little more knowledge and skill, and that thing that can’t be taught; talent.

How To Create A Rythmic Motif

The simplest way is to take an empty bar and fill it with whatever note values you wish, and see how it sounds/feels.
Let’s say the time signature is 4/4. If we choose a whole note, then the bar is filled from end to end with one continuous sound and are motif is done. It’s also incredibly dull, so let’s do something a little more interesting by filling it with a variety of values, e.g, half, quarter, eighth, sixteenth notes, and so on.
You could fill it with 2 half notes; 1 half and two quarters; a quarter, two eights, and two quarters; or whatever. Remember that rests are just as valid as played sounds.
Note that motifs aren’t always the size of a bar, or start at the beginning. They can be as small as just two note values such as a sixteenth and an eighth. It can be useful just to think in terms of short and long notes, or dots and dashes.
The theme from Jaws has a central motif that is a long note followed by a short: daa da (- .). And the example that is always mentioned when talking about motifs is the Beethoven’s 5th: three shorts and a long, da da da daa (. . . -).
It can be useful to just make up a little rhythmic motif with your voice, then work out what’s going on afterwards. Also, rhythmic motifs can be made up of dynamic elements. You could play all sixteenth notes, but by varying the soft and loud, you could create a rhythmic motif.

Motifs In Existing Songs

Analysing existing songs for motifs can have varying results. Sometimes they can be obvious, sometimes they’re well camouflaged, and sometimes they barely seem to exist at all.
Rhythmic motifs are often present in all areas of song structure. From the drums and bass, through the rhythmic elements such as guitar and keyboard, or brass and string sections, to the vocal and guitar solo top line melodies. Different motifs can be assigned to all these elements in order to keep them from treading on each other’s toes and cluttering up the mix. More on that later.
Guitar solos can be particularly tricky. You’re more likely to spot a motif in a Brian May solo than you are in one of Hendrix or Zappa’s. Jimi and Frank do use motifs, but they’re constantly changing them because of their tendency toward seemingly unstructured improvisation. And shredder solos, because of their filling up of all the space, are more likely to use rhythmic motifs that rely more on dynamics.
Vocal melodies are a little less problematic, but still tend to disguise the motifs, either in an effort to make the vocal sound more natural, or because the narrative has created a sentence that doesn’t fit the pattern.
Here’s a song that has a very simple and effective motif underlying the vocal melody.
It is Kathy’s Song by Paul Simon. The motif is simply a two note short long (. -).
When Paul sings the song, he naturalises the phrases in order to make them sound more human and less mechanical, so to properly recognize the underlying motif, we have to reapply it I.e, force the melody to adhere strictly to it.
I hear the driz zle of the rain
. - . - . - . -
Like a mem o ry it falls
- . - . - . -
Soft and warm con tin u ing
- . - . - . -
Tap ping on my roof and walls
- . - . - . -
He uses this . - motif throughout the whole song. Except for the phrase, ’and a song I was writing is left undone’. That’s why it leaps out at you. If you like the song, this is probably the bit you like best. By breaking the rhythmic pattern, it refreshes your interest, but this wouldn’t have been possible were it not for the pattern being established in the first place.
Often it’s necessary to refresh the pattern much more frequently, e.g, every four bars. Sometimes just inverting the motif works well.
As always, don’t expect any hard and fast rules.

Writing Lyrics

When beginning the lyrics, it isn’t such a good idea to immediately try to tell a story, make the phrases rhyme, or even form coherent sentences.
If you start with the sentence, “I am going to the zoo”, you’ve already caused your mind to start thinking of what might happen next, does it rhyme with zoo, and so on. By starting with a coherent sentence, and the intention to tell a rhyming story, you’ve severely limited your options, and the most likely outcome is that you’ll grind to a halt.
Let’s say you have your chord progression. You have created a rhythmic motif that has provided you with a rough melody line. Now you come to the lyrics part, it’s best to sing vague speech sounds that only mimic real speech. Like scat singing but with stuff that makes the phrases sound like real sentences, albeit indecipherable ones.
Record this gibberish, and listen to it several times to see if there’s anything that could be easily converted into something that actually makes some kind of sense. Take it one sentence at a time, still not trying to tell a story.
Narrative songs are quite rare, at least the good ones are, so don’t worry. As long as the music is good, the lyrics don’t have to tell a story. Just try to be interesting. For instance a carpeted floor isn’t as interesting as a floor covered with glass and feathers.

Space

Music has several dimensions that can be utilised to keep your song from becoming cluttered and muddy.
The first is time. If your bar lasts four seconds, and you fill up the first, then you have three seconds of empty space to fill to complete the bar. Using rests to keep it empty is as valid as filling it with sound.
The second dimension is pitch. If you play a very low note on bass guitar, and a very high note on piano simultaneously, even though they occupy the same point in time, they won’t crowd each other because they occupy very different points in audio space.
Stereo recording introduced a third dimension by allowing for panning left and right. So if two instruments occupy the same time and same pitch range, then you can always pan them to different areas of the stereo image.
A common problem when starting a new song is that you feel obliged to fill all that empty space as quickly as possible. So a busy rhythmic pattern is played on guitar. Then you slap on some bass, drums, and a keyboard part.
Up to this point, the different voices of the instruments might have been enough to keep them separate, but this won’t continue indefinitely, and if all your parts are occupying a lot of the same space in terms of time and pitch, then by the time you get to track 6 or so, you’ll have no room left, and you won’t be able to hear anything else you play over the top.
No amount of panning and turning volumes down will help. What will, is going back and simplifying the parts so that they leave lots of room for new additions.
Look at the rhythm guitar part. If it’s occupying a lot of the same area as the keyboard, try shifting it up in pitch, simplifying the chords down into just the most important notes without doubling them up. So just play two note chords that are made up of the 3rd and 7th or whatever. You can do something similar for the keyboard. Have the keyboard use a rhythmic motif that takes up only part of the bar, and the rhythm guitar fill in the gaps. You’ve still kept the musical essence of what they were both doing, but you’ve stopped them from treading on each other’s pitch and time toes, and created a much airier mix.

Listen to Five to One by The Doors for a great example of just how much space you can leave in a song. The lead guitar and keyboard parts occupy similar areas of pitch and time so they are panned hard left and right to keep them separate. Notice how much empty space there is, and how sparse the central motif is.
There will be times when blending different instruments together to create one sound is just what you want. See Leonard Cohen’s Sisters of Mercy for a great use of this technique. By blending different sounds, they create something that is like a weird machine travelling from left to right across the stereo picture.

Capturing Ideas

If you’re gifted enough to be able to transfer ideas directly from your mind to paper, then great, if not, it can be very useful to have a basic recorder that can just instantly record stuff without having lots of setting up to do. You could create a really nice little phrase. But by the time you’ve got your software set up, it’s gone forever, so have something you can just collect little ideas on to work on later.
So that’s about it. Hope I’ve helped. Apologies if this advice leads to cheesiness.