How to write a good song

Some of you will read the title of this post and go looking for some songs I’ve written. You’ll find plenty of tunes by Josh Thompson but he ain’t me.

I’m not a songwriter, despite my aspirations, but I am a writer, and I want you to keep in mind as you read that I’m writing about writing itself.

We’ll look at writing through the lens of songwriting because so many of my friends and readers are so musically inclined, but these principles apply to any story—because a song is, as I’ll show in this blog post, essentially a story.

In the interest of transparency, I’ll say, I’ve written hundreds of poems but I’ve only written one single song, my whole life. Not just one that I’m proud of, no, one total song. And I’m not particularly happy with it, but it’s a song, and it’s mine, it has a melody and chordal structure and song structure—verses, a chorus, and a bridge, and it tells a story. It’s a song.

So yeah, me of all people writing a how-to about songwriting may come across as a bit hypocritical to some of you. But hear me out here.

As I said, I’m a writer. And writers are exactly the same as songwriters, minus a melody. That’s where my poems lack the ability to become songs—I can’t write a melody. Yet.

Part of it is my self-critical nature—that tick that constantly tells me everything’s all been done before, what I’ve written is cliché or tired. You have to work to defeat that nature in order to innovate. I’ve yet to make that inner-voice my slave.

But another part of it is that I have very little interest in music theory, and have learned only enough to get gigs and to have fun and play well with others. You don’t have to be a theory whiz to play well. Don’t let anyone tell you otherwise.

You have to have great ears to play well with others, and you have to know how to speak the language just enough to be conversant. That’s all. Check out this “working musician’s guide” that I wrote. It serves as a cheat sheet for me, for all I need to know about music theory.

So, yes, I have yet to really write a good song. But they say those who can’t, teach. And one thing I’m very good at is filling the role of teacher. Or in the music world—producer. It’s something I pride myself on, having the ability to take existing songs and make suggestions to polish and perfect them.

A good producer can talk you through dynamic changes, install tensions and releases within the song, introduce new and interesting parts or strip away parts that don’t serve the song. That’s where I’ve been most helpful to my friends who are songwriters.

I’ve written verses, I’ve written choruses, and I’ve suggested phrase and word changes. And because I speak the language of a guitarist and an engineer, I’ve served as translator between artist and operator. Between the romantic and the actual. And in so doing, I’ve helped talented people tell beautiful stories.

That’s how I’m qualified to share with you my thoughts on how to write a great song. So let’s get into it. I hope that you are inspired by this, and it drives you to tell great stories.

Point 1: The story is king. Everything serves the story.

Let’s say you’re writing a song about going out west. Are you going to Mexico, are you crossing the Mississippi, or going to L.A.? To just say in your lyrics, “We’re going out west” is weak—it’s not saying much. Tell us where you’re going, why, and how you’ll get there. Answering those three questions alone could give you six verses.

Get specific—at times. Using vague, clichéd, or flowery language is the impulse of all writers. I hear songs that describe heading west and riding a horse, and I immediately think, “Is he going to ride the horse the whole way there?” That’s untenable. Now, there’s an exception to this rule that we’ll dig into in a later point.

And then, in service to the story, have you included thematic musical elements that detract from the story you’re trying to tell? Like, ok, we’re going to Mexico but you’ve included steel drums in the bridge of the song. Steel drums suggest that maybe instead of Mexico, we’re going to the Bahamas.

When telling a vacation story, a lot of guys go for the whole country-reggae Zac Brown Band, “Toes” sound, to convey a general sort of beach-iness. If you’re going to go that route, place a higher lyrical emphasis on going to the beach specifically—not just out west or down south generally. Let’s use “Toes” as an example.

In “Toes,” Zac Brown is clearly telling a story.

The plane touched down, the city’s still on his mind, he’s in the baggage line thinking about his own emotional baggage, but now he’s surrounded by water and not all those Atlanta skyscrapers so the song dramatically changes. There’s a big chord strike and the chorus loudly proclaims, “Adios and vaya con dios!”

He’s saying goodbye to all the stress of cramped city living. Throughout the song, he’s built up the ideas of water and sand and cold beer and the warm sun so much that when we hear a reggae or calypso sort of element, we’ve forgotten that “Adios and vaya con dios” is Spanish, we’re just thinking he’s at ANY beach.

Jamaica, Bahamas, Costa Rica—to an uncultured white boy those thematic elements like the steel drum, the syncopated reggae feel, or a ukulele strumming, are just generic beachy sounding things. They serve the story of a beach vacation.

In our example song about heading out west, we have to find specific elements to drive the story. If we’re going to Mexico, maybe we throw in some flamenco guitar or some mariachi horns—something to glue together lyrical elements.

Zac Brown uses that nylon classical sound with happy major chords to add a Gulfy-beachy sort of feel. If we threw in even just one minor chord strummed on the classical guitar in the bridge of our Mexico song, or adhered to that Spanish phrygian scale sound, that Latin influence could transport us to Mexico in the story instantly.

Have I made that clear? See how even the instrumentation and arrangement serve to tell a story?

Point 2: Be mindful of dynamics.

Not only does the story have to go somewhere, the dynamics have to go somewhere too. A song about the beach ought to ebb and flow like the ocean, with tension and release, rise and fall. Not just changes in volume, but changes in intensity. And structural changes as well—you don’t like songs that are all verse and no chorus.

You’ve heard the phrase, “Don’t bore us, get to the chorus.” It’s true. Few people can pull off all-verse, no chorus songs. Bob Dylan has a few great ones. Neil Young. Joni Mitchell, maybe.

Some songs need to be all gas pedal, no brakes, no slowing down. Some stories require that. One steadily driving song I often think of is “Killing in the Name,” by Rage Against the Machine. That song really has only a couple of places where it lets up, and only slightly.

The words, “Now you do what they told ya,” are repeated 25 times in Rage’s song. What does that and a steadily driving rhythm convey? It conveys slave labor, like prison railroad work, hundreds of hammers being driven to the earth in tandem. You’re only doing what they told you. You’re only doing what they told you.

If you want to convey ‘chill,’ then a fast and steady four beat sound on the drums is working against you. One thing that will let the song breathe tremendously, providing enough air for a story to be told, is to have the drums hang back in the pocket. I know asking a drummer to play less is like pulling teeth, though…

And in our example, in Latin style music, reggae too even, the drums would often play syncopated rhythms and beats on 1 and 3 instead of 2 and 4, with lots of rim shots and rim sounds and not a lot of crash and splash cymbal. Not a lot of force. Chill.

If we’re going to the beach in your song, and the driving rhythmic force in your mix is the piano, you’ve got another problem. Thematically and in terms of intensity, a piano simply doesn’t fit.

You have a phenomenal pianist, but I’m afraid we can’t take the piano down on the beach with us! And if it’s not a beach song, if we’re going to Mexico for some other purpose, your piano likely won’t fit in the car. Your audience will actually become disoriented by the presence of a driving piano.

I know it seems silly but everything serves the song and the song tells a story, so everything in the song is either telling your story or it’s just extra. Unnecessary.

This idea goes back, in part, to point number one. But it’s not just thematic elements that glue a song together, it’s elements by level of their intensity that can dramatically affect a song. Think about it. A concert organ has no place in reggae music, but staccato chops on a Hammond organ can make a reggae song.

It’s the act of holding back a bit, lessening the intensity of the organ, that makes it such a special element in the song. And in the case of the organ, staccato chops are percussive, almost substituting for or supplementing the drums in the rhythmic scaffolding so-to-speak of the song.

Point 3: A great story has only the elements it needs in order to be told, and nothing more.

I can’t remember who told me this, but in writing, they said, “Don’t give people 4. Give them 2+2.” That’s important. Give people a few hints and a few elements and let them connect the dots. But, you say, point number one was all about getting specific! Are you contradicting yourself, Josh? No. And here’s why.

You need to get specific, with little details and thematic elements, in order to tell a story. Without detail, it’s just a list of phrases. Platitudes and clichés. But don’t get so specific that you’re thinking for the audience, or telling them what to think.

Some of our greatest songs say very little, while still telling a beautiful story. Take, for example, George Jones’ classic song, “He Stopped Loving Her Today.” Jones sings, “Soon they’ll carry him away, he stopped loving her today.” That tells us so much with so little.

He stopped loving her because he’s (presumably) dead, but George Jones never tells us the man in the song is dead. Jones writes about all of the things this man did for his lover. So why would he suddenly just stop loving her? We are led to infer that the reason he stopped loving her is because he’s dead.

The most detailed thing that Jones says, that leads us to infer that the man in the song has died, is this: “All dressed up to go away, first time I’d seen him smile in years.” That’s not a lot of detail. But boy, does it tell a story.

Have you heard of the online game of creating six word stories? I don’t recall who wrote this one, but someone wrote, “For sale. Baby clothes, never worn.” Not a lot of detail there, but you immediately understood the inference. We’ve lost a child. That’s taking 4 and handing us 2+2 instead.

Point 4: Never ever compromise. Ever.

This one’s more for the producers than it is the songwriter, I suppose. It’s rare that one person will be songwriter, musician, producer, and engineer all at once, on the same session. You’ve got to learn to work with others in order to deliver the song.

But as a producer, if the engineer or the writer insists on something I can’t abide, I’ll refuse to put my name on it. I believe very strongly in this. Find the dealbreakers—the threshold you’re unwilling to cross with your creative “name,” and never cross them. Never sell out. Never follow. Never put out something you’re less than proud of.

Nothing on earth is worse than having to qualify your project to people. “Go listen to this on iTunes—only, ignore the weird trombone part in the bridge, that wasn’t my idea.” That sucks. Never put yourself in a situation like that.

If you can’t be unreservedly proud to talk about your work then you need to work harder, and if your obstacle is the idea of another person, then you need to be more convincing.

Compromise is not partnership. That’s why I hate, in relationships, when someone tells you you’ve got to compromise. Don’t make concessions just to settle a dispute. If you feel strongly for something, argue for it, win it for yourself, and if it’s misguided then gracefully suffer the consequences.

Partnership is this: Whenever you butt heads with someone, give them an opportunity to argue their point. If they persuade you, then you haven’t compromised. Becoming persuaded means that you no longer feel so strongly about your own point.

If they’re a true partner, they’ll give you equal opportunity to argue your own point, and who knows, you might persuade them.

If neither party is persuaded, neither party is arguing effectively enough. I know how extreme that sounds, but it’s true. If neither party budges, then neither idea has merit. Or, the idea has merit, but its merit has yet to be properly argued. Think about it.

If neither party is persuaded and you both still feel strongly about your own way, then true partnership would have you pivot, and work together on a third and separate idea. That’s partnership.

That’s a good marriage, by the way. If you can argue productively, that’s a good partnership. If you’re sometimes persuaded, and he or she is sometimes persuaded, and when neither are persuaded you’re able to work together on something entirely different, that’s a solid partnership. 

Point 5: More is more to lose. I say this a lot.

Or in the case of songwriting, more is more to go wrong. The more crowded a mix is, the less music is actually being made and the less any voice—be it vocalist, guitar, a drum, whatever—can stand out and be heard. It’s hard to be heard in a crowded room and it’s hard to be heard in a crowded mix.

This is one of my problems with the whole ‘ambient’ guitar trend. Instead of crowding the mix with giant, reverby and overtonesy pad noises, and all these synths and other guitars, why not drop some things and let the music breathe?

It’s in the space, where music is made.

I hope you’ve learned something from my experience, and if you disagree or have something to add, reach out and let me know. My email is always open: joshua.t@icloud.com.