Hoagy Carmichael’s most famous song, “Georgia On My Mind”, was first recorded by the composer in 1930, and became a hit when it was covered the following year by Frankie Trumbauer. These days, most people associate the song with Ray Charles, whose 1960 recording is often considered the definitive rendition.
Here are the first few bars of “Georgia”, written out in the key of F major. This is the key of Trumbauer’s recording, and the key in which it is most commonly performed today (although Ray Charles sang it in G major, and Carmichael’s original recording is in B flat).
While the song’s melody sticks pretty close to the F major scale throughout, you might have noticed that the third chord seems somewhat out of place. That A7 chord doesn’t seem like it should fit with the F major key signature, but when you listen to the song it doesn’t sound wrong at all. Why is that?
It’s because A7 is a secondary dominant chord in the key of F major. Using secondary dominant chords is a great way to inject a bit of chromaticism into a diatonic composition. Before we look closer at secondary dominants, let’s just define those two words – “diatonic” and “chromaticism”.
The word “diatonic” means that all the notes of a composition (or a section of one) are in one key. If you’ve ever asked “What key is this song in?” you’re assuming the music is diatonic. It’s a reasonable assumption, because almost all of the music that has been made through the ages, including most music being made today, is essentially diatonic.
Take our current example: we’re looking at “Georgia” in the key of F major. F major has one flat in the key signature – B flat – so the notes of the F major scale are: F G A Bb C D E and F.
If you play an F major scale on the piano you’ll be playing mostly white notes – the only black note being that B flat. If you play through the whole melody of “Georgia” you’ll notice that the tune sticks pretty closely to these seven notes. In the whole 32-bar melody there are only two notes that don’t belong to this scale.
But when we say “all the notes of the composition are in one key”, that applies not just to the melody, but also to the harmony – in other words, the chord progression. So how do we know which chords to use?
If we build a three note chord (ie. a “triad”) from each note of the F major scale, we will get the following chords: F major, G minor, A minor, Bb major, C major, D minor and E diminished.
If we compose a song using these 7 chords we can be fairly confident that the chord progression will sound good, because all of the notes in all of the chords are taken from the F major scale. In other words, if we’re careful to stick to those 7 chords the chord progression will be diatonic – and diatonic music sounds good!
The polar opposite to diatonic music is “atonal” music. Atonal music doesn’t have any sense of a tonality or key. In other words, it’s not “in” any key at all. The concept of atonality didn’t exist before the early 20th century. All western music in the hundreds (possibly thousands) of years leading up to that time was essentially diatonic. We generally associate the idea of atonality with 20th century composers like Arnold Schoenberg and Alban Berg. Interestingly, though, Schoenberg disliked the term “atonal”. He preferred the term “pan-tonal” because, to his way of thinking, his music was not “anti-tonal”. Rather, his concept was that the music embraced all tonalities simultaneously. An interesting distinction.
There is a middle ground between completely diatonic music and atonality. That middle ground is what we call “chromaticism”. The word “chromatic” is derived from the ancient Greek word “chrôma”, meaning colour. It’s a good choice of word.
See, the problem with completely diatonic music is that it can sound a bit bland. If all the melody notes and all the chords of your song are derived from the same scale, after a while it can sound a bit “samey”. Even – dare I say it – predictable. One way to spice up your composition is to add notes that DON’T belong to the key that the song is in. In other words, we can step “outside” of the diatonic key – even if only for a moment. This will add some harmonic and/or melodic colour to your composition. (See – there’s that word: “colour”, ie. “chrôma”, ie. “chromatic”.)
What we want to do is to add harmonic and/or melodic interest by adding some chromatic notes to our composition. In other words, adding some notes from outside the diatonic key of the song. But we can’t just go around willy-nilly adding any old sharp, flat or natural. I mean, you could – and it might even sound ok – but you will have arrived at your result by chance, and you won’t understand why it worked or didn’t work. We need a more logical approach.
This is where secondary dominants come in. But before we get to secondary dominants, let’s just have a quick discussion about plain old dominant chords.
The “dominant” note is one of the two most important notes of any scale – the other one being the “tonic”. The tonic is the first note of the scale. It’s the note that the scale is named after. For example, the first note of the F major scale is F – that’s why it’s called the F major scale. When we play a melody in the key of F major, it doesn’t really sound finished until we land on an F. It’s the tonal center of the scale – the “at rest” note. It’s pretty obvious why the tonic is such an important note in the scale.
The second most important note of the scale is the “dominant”. The dominant is the fifth note of the scale. In F major, for example, the dominant note is C. At first glance, it might not be obvious why this is such an important note. You might ask: “Why is the fifth note more important than the second note, or the sixth note?” Well, all of the notes are important in their own way (I don’t want any of them to feel left out), but there is a special relationship between the dominant and the tonic, and it has to do with harmony.
When we play one chord followed by another chord it creates a sense of harmonic resolution which we call a “cadence”. There are different types of cadence, depending on which chords we use. The most common type of cadence is when we play a dominant chord (ie. the chord built on the dominant note) followed by the tonic chord. This is called the “perfect authentic cadence”. Some people refer to this simply as the “authentic cadence” or the “perfect cadence”. I prefer the latter. If you looked through every score of every piece of music ever written and counted up all the cadences (for example, if you were at a loose end for the next 1000 years), you’d find that the vast majority of them are perfect cadences.
Since the dominant chord is the one built on the fifth note of the scale, and the tonic chord is built on the first note of the scale, we can write the perfect cadence like this: V – I. In the key of F major this would be a C major chord followed by an F major chord, or C maj – F maj.
It’s not uncommon to extend any chord beyond the triad by adding more notes. This creates a more complex sound, and we usually achieve this by stacking intervals of a third above the basic triad. The most common chord to extend is the dominant chord. In fact, for centuries the only chord that composers would extend was the dominant chord. If we add a fourth note to the V chord we call it a dominant 7th chord because the new note we’ve added is the interval of a 7th above the chord’s root. In the key of F major, the V7 chord is C7, which is made up of the notes C, E, G and B flat. Therefore, a common perfect cadence in the key of F major is C7 – F maj.
The perfect cadence is a very important sound in western music. The dominant chord creates a certain harmonic tension, and the tonic chord is the resolution of that tension. In fact, it creates such a strong sense of resolution that even if we have a V7 – I cadence in a completely different key to the rest of our song, we will start to hear that new I chord as our home key. This is called “modulation”. We’ll discuss that a bit more later on.
Secondary Dominant Chords
Finally we get to secondary dominant chords, and the explanation of why the A7 chord in the second bar of “Georgia” sounds so good. (I’m sorry it took so long to get here – we kinda took the scenic route.)
We’ve already discussed diatonic chords in a major key, so we know that the chords which sound good in F major are: F maj, G min, A min, Bb maj, C maj, D min and E dim. We’ve also discussed the important relationship between the dominant chord and the tonic chord.
A secondary dominant is when we approach any chord in the key by first playing the dominant of that chord – even when that chord is not the tonic. You see, we normally think of the dominant as being the fifth note or chord of the tonic key. But what if we think of the dominant of a different chord in the key?
For example, in the key of F major, the VI chord is D minor. Let’s imagine for just a moment that we were in the key of D minor. What would be the dominant chord in D minor? It would be A7, which is made up of A, C#, E and G. Now, we’re not in D minor – we’re in F major – and A7 is not a chord that occurs in the key of F major, because it has a C# in it and there isn’t a C# in the F major scale. But since there is a D minor chord in F major, and A7 is the dominant of D minor, we can call A7 a secondary dominant in the key of F major.
Let’s have another look at the first few bars of “Georgia” (above). The song starts with a whole bar of the F major chord, which really establishes F major as our home key. (In fact, it’s an F major 7 – because after all, it’s a jazz song and jazz musicians love complex sounding chords. But that’s not really important here.) Then there’s an E diminished chord (again, it’s actually E half-diminished, which can also be called E minor 7b5. Again, that’s not really relevant – just go with it for now.) All of the notes that make up those first two chords come from the F major scale, so up to this point our chord progression is completely diatonic.
But then, we get that A7 chord that we wondered about way back at the start of this post. What we now know is that the A7 works because it is the dominant of D minor, and therefore it’s also a secondary dominant in the key of F major. So it comes as no surprise that the very next chord after the A7 is D minor – the VI chord in the key of F major. The A7 sets up the D minor because it’s the dominant of that chord, and in the process it gives the touch of chromaticism we were hoping for, adding interest and colour to our mostly diatonic chord progression.
Other Secondary Dominant Chords
That A7 isn’t the only secondary dominant in the key of F major. We found the A7 by working out the dominant of D minor, but there are other chords in F major. What if we look at the II chord in F major – G minor? The dominant of G minor is D7, which is D, F#, A and C. The note that adds chromaticism here is F#. It should come as no surprise that there are no F#s in the key of F major. This secondary dominant is actually very common. It occurs in jazz and other genres as part of the I VI II V turn-around progression. In the key of F major that would be: F major, D7, G minor, C7 leading back to F major.
Other secondary dominants in the key of F major are E7 (leading to A minor), G7 (leading to C major), and F7 (leading to Bb major). All of these chords add a little bit of harmonic interest because they all include at least one note which is not in the key of F major. I particularly like using the F7 leading to Bb major. In this key, F major is the tonic chord, but F7 is a secondary dominant. Cool, huh?
Here are all the secondary dominants in the key of F major. Beside each secondary dominant is the chord it resolves to.
Of course, the other dominant chord you can use in F major is C7, but as that is the actual dominant chord of the key, it isn’t a secondary dominant – which is why it’s been left off the example above.
In theory, you could also have B7 leading to E diminished, but in practice we tend not to see that progression much. The reason is that E diminished is already such an unstable chord that landing on it from it’s secondary dominant doesn’t really feel like any sort of resolution at all. Besides, the dominant chord of E is B7, but E diminished doesn’t have a B in it, it has a B flat, so that progression sounds a bit clunky and odd. I encourage you to play around with it, but it’s really difficult to make this one work effectively. Probably best not to worry about it too much.
Isn’t this just modulation?
I said we would get back to modulation later. That time has come.
As I explained earlier, modulation is when we move from one tonal center (ie. key) to another. It’s another name for a key change. It’s very common for a piece to modulate through different keys. Music that modulates can still be diatonic because in each section of the song the music is in one key, even if that key has changed. In fact, the whole concept of modulation relies on the idea of music being in one key or another – so by definition, music which modulates is diatonic music.
When you were reading the previous section you may have been thinking to yourself “he’s just describing a modulation.” It makes sense that you might think that, because the most common way we modulate to a new key is through the dominant of the new key.
For example, if we were in the key of F major, and we wanted to modulate to the relative minor, which is D minor, the most common way to do so would be to play an A7 chord, which is the dominant of the new key. Then, when we land on the D minor at the start of the next section it really feels like we’ve arrived in the new key.
That description sounds an awful lot like how I explained the A7 as a secondary dominant, doesn’t it? But there is one important difference. When we are modulating we are actually changing to a new key. There is an expectation that we’ll be established in the new key for a while (at least a few bars) and the new key center (in this example, D minor) feels like it has become the tonic. By contrast, when we use a secondary dominant, it is much more transient. In “Georgia” we move through A7 to D minor, but just as quickly, we move away from D minor to the next chord. We don’t stay on the D minor long enough for it to start to feel like the tonic. There’s no sense of “arrival” in a new key. The first eight bar section of “Georgia” definitely sounds like it is all in F major. The A7 to D minor part of that progression feels like a moving harmonic colour, rather than a whole new key center.
Ironically, though, “Georgia” does modulate into D minor later on. The middle section of the song (which jazz musicians of a certain age sometimes call the “channel”, but today is more commonly called the “bridge”) modulates to D minor, and it does it in exactly the manner described above, by first playing an A7. But the difference is that the music stays in D minor for a whole eight bars before finally modulating back to F major. In the middle section it definitely feels like D is the tonic – not F.
In the bridge to “Georgia” (above) you’ll notice the predominance of D minor chords, and a complete lack of F major chords. Notice also the C7 chord at the end of the eight bars, leading us back to F major for the next section.
When should we use secondary dominant chords?
As the explanation above implies, you can use secondary dominants any time you want to spice up a diatonic chord progression and add more colour to your harmony. In this post I’ve mainly discussed secondary dominants in chord progressions, but you can use this idea to add chromaticism to your melodies too.
Stylistically, the use of secondary dominant chords is often associated with Afro-american music – in particular, gospel music. If you listen to old gospel records you will hear that those piano players and organists loved to put lots of secondary dominants in their playing. That influenced many other Afro-american styles, such as jazz, blues and in particular, soul music. A big part of early soul music was taking the gospel style of the Afro-american Baptist church and putting it in a more secular context. So if you want to inject some of that soul/gospel flavour in your music, throw in some secondary dominant chords.
Really, it will work in any musical style, but as with anything, it loses it’s effectiveness when it’s over-used. Just use your ears to work out what is stylistically appropriate. Generally, jazz will use plenty of secondary dominants, and pop music will not use many at all. Other styles will be somewhere in between.
We started this post by looking at “Georgia” in the key of F major. For the sake of simplicity and clarity, I’ve stuck with F major for all the examples in this post. Of course, you can transpose the examples into all twelve keys. Also, just because I’ve stayed with a major key for this article, doesn’t mean that you can’t also have secondary dominants in minor keys. In fact, secondary dominants are probably even more common in minor keys, since chromaticism is generally more common in minor keys.
Like everyone else, I’ll be spending the next few weeks mostly stuck inside, unable to go out, and all my gigs have been cancelled due to COVID19. I’m trying to keep myself busy doing worthwhile things, like writing posts such as this one. If you have any questions about music theory – or just about music in general – send me a message and I might write a blog post about it. At the very least, I promise to reply to your message and attempt to answer your question.
Stay safe and healthy.