Okay, you’ve got to promise not to laugh. When I was just starting to play guitar, I also was just starting to write songs. I’m not sure why. I’d been playing the piano for years but it never inspired me to write. Something about the guitar just moved me to write (and in all probability, being seventeen and falling in love every other day and, of course, knowing everything there was to know about life already, helped in no small part). I tried to write two songs a month, (which was nothing compared to a guy I later knew in college who wrote one a day) and eventually my limited knowledge of chords and theory quickly caught up with me. Everything I tried to write sounded the same.
Now, I’ve never been a whiz with computers (and remember we’re talking about a time when pockets calculators were just coming out for two to three hundred dollars apiece!) but I’ve always been lucky enough to have friends who were incredibly computer literate.
With the assistance of one such friend I came up with a “song writing” program. Here’s how it worked . I wrote out every chord I could play and randomly assigned it a number by pulling it out of a hat. My friend then ran a “random number” generating program (going only as high as the number of chords I had) which rounded the random numbers into whole numbers. He’d run out a string of them and I’d “decode” the numbers back into chords and play them in the sequence dictated by the computer.
It had some winners, I will tell you that. I got two chord changes that I might never have come up with on my own (especially at the point in my musical life). But nine out of ten times it sounded positively dreadful. And I’m being really polite with this description.
What makes a chord transition “good?” Why are some transitions almost automatic, for lack of a better word? Why is C to G pleasant to the ear while C to Eb minor is worse than fingernails on chalk?
You may find this hard to believe, but a lot of the “science” behind harmony is mere convention. As I’ve stated before, the “theory” in music theory is simply the examination of what has gone on before. Therefore, much of what is a “good” chord transition is the result of centuries of familiarity. If you had been raised on some planet where, oh I don’t know, the augmented fourth was considered good form then it would not seem as jarring an interval as it does to most earthlings.
Add to this the notion that our harmonic conditioning undoubtedly springs from singing. Think about it, people had voices long before they had instruments. When our ancestors first began singing together, what made them feel certain notes went well together and some brought down the wrath of the gods? I can imagine two cave people coaching each other . “No, no Thag. If Og sings the root, you must chant the sixth, okay?”
Be that as it may, we have inherited quite a few guidelines as to what chords work well together. The following is a chart taught to first year theory students. And as always, I must implore you to remember that this is not a be all and end all guide. Chord changes that people may have found harsh in the past might now be the “in thing.” Progressions that we perhaps find trite will possible rule the radio tomorrow (it’s an incredibly easy thing to change “stale” into “style”).
Anyway, here goes. In a major key, the general rule of thumb regarding chord progression would be as follows:
Okay, note yet again that I am ignoring the seventh chord position. Don’t worry. It’s coming up next time.
Now, these charts are all fine and dandy, but I prefer to see things in terms of actual day to day use rather than all the Roman numeral stuff. Let’s look at the keys of C major and G major, shall we?
Some of you will no doubt notice two things about these charts: first that they pretty much confirm the things that A-J Charron has written about in A Simple Song. Secondly, that these progression charts will also help cut out a lot of the guesswork when you’re trying to figure out a song on your own! Consider it an added (albeit late) bonus for getting through these three columns: Happy New Ear, Unearthing The Structure, Solving The Puzzle.
The important thing to remember about coming up with chord progressions is that everything has been done before. You cannot worry about the fact that your song’s chords are not incredibly different from someone else’s. You will be better off spending your energy learning more chords and learning how to use chords to suit the mood of your songs.
Learning more chords is a no-brainer. But again, there are only so many different chords. This is why it’s important to learn different voicings of chords (Multiple Personality Disorder) as well as different tunings for your guitar (Open Tuning Part I). A new voicing can make an old progression sound fresh.
As for learning how to use chords to suit your songs, first let’s take a quick look at how people write songs (hey, by now you should know I can never approach anything in a straight forward manner . chalk it up to being a Cancer.). Of course, we’re talking about the “music” part of the song. Most songwriters tend write the music first, either by coming up with a riff or chord progression or by harmonizing a melody. A-J discusses creating chord patterns in his latest article. Riffs tend to follow the same idea, which makes sense because a riff is simply a pattern of notes (usually) derived from a chord or a scale within a given key. Some “riff” songs consist of the same riff played repeatedly over changing chords. Eminence Front by the Who or Blue Oyster Cult’s (Don’t Fear) The Reaper are examples of this style of writing.
Another method of “riff writing” is to come up with a cool riff and then transpose the notes to fit the chord changes. This is nowhere near as complicated as it sounds. Let’s look at the Beatles‘ Day Tripper as an example (and I’ve copied this version straight from the Guitar Tab site). You can see (or hear) that riff two is exactly the same as riff one except that riff two is played in A while riff one is played in E.
(Oh, yeah, the disclaimer:
This file is the author’s own work and represents his interpretation of the song. It is intended solely for private study, scholarship or research.
If you don’t buy into the riffs being the same, simply write out the notes and then transform them into their Roman numeral equivalents for the respective keys. I’ll do it for riff one and you do riff two if you don’t believe me.
Picking A Color
At some point we all find ourselves whistling a tune that we’re concocting on the spot. Some of us might think, “Hey, this is a cool song,” and set it out but then realize we have no idea of what chords to play with it. Sound familiar to anyone?
To demonstrate how different chord progressions can drastically affect the tone of a melody, I’ve some up with a short and simple melody (and I’ve also provided the guitar TAB for those of you who may not read music)(and yes, you should definitely learn to do so):
When I say “short and simple,” I don’t kid around. Now even though I didn’t put in any key signature (sharps or flats), I’m going to use the first note as my root and play some chords in the key of G major. I’m also going to start out very plain, using just the I, IV and V chords (G, C and D (or D7 in this case in order to emphasize the C note in the melody)). And to keep things easy, I’m changing the chord each measure, using whichever of my three chords contains the note or notes in that particular measure. Here’s what I’ve come up with:
Now if you play the accompaniment while singing or humming the melody you will find that this sounds like a pretty traditional song. Not very hard, was it?
I’d like to use this example to point out something important . namely that it is fairly important to establish a sense of “home,” a feeling of what key the song is in. Playing the G first and coming back to it every other measure certainly fixes it in the listener’s ear, but the thing that really pegs the key of G is the ending, the D7 to G finale. Nothing cements the sense of key as well as a chord shift from V (better still V7) to I. Even a song like Hey Joe, (the Billy Roberts song best known via Jimi Hendrix) that seemingly wanders around aimlessly, takes the time in the introduction to establish itself in the key of E so that when it finally gets back to the E chord, one has a sense of resolution (there will be an in-depth look at this song next time out).
In this second harmonization of my melody, I decide that I will try out the key of C. this is not too hard a choice to make seeing that my melody has neither sharps nor flats. You can see that I’ve used a broader brush, if you will, incorporating many more of the chords available to me in the C major scale. You will also note that the B in the first measure has become a passing tone since it is not part of the C chord that I am using for that measure:
I don’t know about you, but this particular accompaniment feels incomplete to me. Maybe it’s having the G note as the final note, I don’t know. I like it, though. It might work out better as a part of something else. This is how people end up with “pieces” of songs. It’s hard and unsatisfying to explain to someone that a lot of songwriting is “just a feeling” but it is. You know when something works. You definitely know when it doesn’t.
I really go for broke in the third example, first by setting the melody in E minor and then changing the chords with every different note (again picking a chord that contains the melodic note) and tossing in everything but the kitchen sink:
Notice that by establishing the E minor right at the outset and then finishing off with a B7 to E minor really does give us a sense of key. Also notice that even though I use a C to start the second measure and a G to start the fifth measure, just like I did in the first example, the overall tone of this accompaniment is completely different. At certain points I actually worked backwards. As stated, I had decided to use the C to start measure two and I really liked the D in measure three, so how to get from one to the other? Since my melody note between the C and D was E (and that I wanted for the sake of the exercise to use a chord that contains that note), I had a number of choices: C again (no way!), E minor (but I’d already used it and I figured that there might be something a bit more interesting), A minor, which worked and in fact led me to A major and then to A7.
More often than not, and certainly without meaning to, I tend to write songs using a combination of these methods. I may have a chord pattern that I really like for the verses but I’d like to change it around a bit for the chorus. I would then write out the melody and see what chords can be used as feasible substitutes. Often it might mean replacing a major chord with its relative minor. Or using a major seventh. Or sometimes a chord from out of nowhere (kind of like that Bb chord we found in the chorus of John Lennon’s Jealous Guy in Solving The Puzzle).
In reverse, I might have a melody that I really like and then found some chords I think go well with the melody, but I feel a different chord voicing might give the song more of whatever it lacks. Perhaps I might try to come up with a riff that will serve just as well as a strummed chord.
Most songwriters will tell you that even when they have finished a song, it’s not really finished. Somewhere down the road you may learn a trick or hear someone else do something that makes you think, “I could do that in my song.” Paul Simon used to talk of hearing a woman in a coffee shop performing a version of Homeward Bound that he liked so much he used it himself. Hey, it’s his song.
Ultimately what you write is up to you. But the more you know why certain progressions work and certain ones don’t the better chance you stand of liking what you’ve written. By now you know that I’m a firm believer of what I call the “learning more than you need to know” philosophy. If you’re really serious about songwriting, do check out our songwriting page here at Guitar Noise. And as always, feel free to write with questions, either directly to me or on the Guitar Forums. I’m certain we’ll be revisiting the theory involved in writing fairly often in the future.