Given a melody (say in the key of C, we then know G, F, and Am around C), could any chord progression (always starting with tonic) randomly from I, IV, V and vi still work or work most of time? (e.g. I, vi, IV, V, I, vi, V, I, IV, ..., I)

Why would I ask this question? As I am playing songs by ear, while I am trying to figure out the melody on the right hand, I like to do the chord on the left hand. I will have some idea about the key, major or minor. Thus, I probably know the key and hence I am able to start with tonic. Then I will have to randomly (,or by my instinct,) try the chord. It sounds okay I might say. However, I would like to learn more theory about if it "works" or how well it may work?

  • 2
    Those chords are not really based on the circle of fifths - they're chords which are diatonic to a key - quite different.
    – Tim
    Mar 6, 2023 at 9:48
  • 2
    What has the circle of fifths got to do with the question? Did you accidentally write the wrong title for this one? Mar 6, 2023 at 11:24
  • @piiperiReinstateMonica I think maybe the OP means "G is one fifth above C; F is one fifth below it; and Am is the relative minor." But this doesn't really have so much to do with "the circle"; any relationships between any chords can be plotted on it, without implying any cabalistic causal significance. Mar 6, 2023 at 15:14
  • 1
    I suspect the OP is just starting out learning harmony, and they don't know "diatonic chords" but instead read about "circle of fifth" as a chord progression. Mar 6, 2023 at 15:23
  • Without clarifying what you mean by “work”, this can’t be answered objectively. Even with clarifying it, it still probably doesn’t have an objective answer. Or maybe another way to look at the question is that there is some concept of “working” that would make any chord and melody combination valid. Mar 6, 2023 at 20:18

4 Answers 4


It will fail most of the time, which is why composers exist — to find ways to make otherwise random sounds "work". It's also why music theorists exist — to understand what allows some combinations of sounds to "work" and others "not work".

The first problem, of course, is that what "works" is purely subjective. So, in that sense, a random harmonization of a melody "works" every single time as long as there's someone who thinks so.

But on the other hand, here's a simple example of something that doesn't "work".

  • The melody is "Twinkle, Twinkle Little Star" in C major.
  • The harmony is the IV chord played for the entire song.

This is certainly one possibility that could emerge from a "random" selection of chords, and I think it's reasonable to suppose that most people familiar with the original song would say it sounds "wrong" or "doesn't work."

Another possibility:

  • The melody is "Twinkle, Twinkle Little Star" in C major.
  • The harmony is a random assortment of I, IV, V, and vi chords, all played very quickly to accompany only the first note of the song, then nothing after that.

Again, I think it's fair to say that such a harmonization wouldn't "work".

So, not only does "workable" music require that chords be carefully selected ("carefully" relative to "randomly choosing from infinite combinations"), but they also have to be placed appropriately in relation to the melody.

There are infinite moments at which a chord could be placed and thus infinite possibilities for chord progressions. Compared to infinity, the number of choices that "work" is, I think, vanishingly small.


My answer comes from a slightly different vantage point, but: this absolutely happens, and it's one of the hallmarks of French impressionism (especially Debussy).

We call this pandiatonicism (literally "across the tones" or "through the tones"), and it arises when we use a given pitch collection (like your C major) but we don't necessarily follow the rules of Classical-era harmonic syntax (like predominants moving to dominant moving to tonic, etc.).

As an example, here's an excerpt from Debussy's famous "The Engulfed Cathedral":

enter image description here

We start with a C chord moving to Dm moving to G: a standard tonic, predominant, dominant progression. But then we move V–IV–iii–ii–vi–iii, which is rather backwards from a "textbook" harmonic progression. But in this style, it works.

A similar term here is planing, where we use the same chord shape and move it in parallel motion within the key. Notice that the left hand planes these second-inversion triads while the right hand planes root-position triads.


Surprisingly many chords will "work" in many situations, depending on your definition of what it means to "work". If the audience expects a certain chord progression and will not accept anything else, then I guess nothing else will work than what they expect.

If your chords have been selected from chords usually used in songs in that key, almost anything will work at least to some degree. Maybe it won't tell a certain story that someone expects, but if you have the possibility to force-feed that slightly weird combination to people enough times, it will become familiar and thus expected and thus GOOD, however strange it is. Repetition is the most important thing in making hits.

I tried many different chords for backing Twinkle, Twinkle, Little Star, and I more or less liked all of these variations. Will you like them? Will someone else like them? You decide. Music is mostly about skill, much less about knowledge, so play, listen and experiment to improve your harmony skills.

Twinkle, Twinkle, Little Star

As you see some of the chords aren't even "in the key" like Bb. But it's still nice, it just creates a chromatic alteration for the B scale degree, which the melody doesn't use, so there's no clash.

What doesn't work for me is if there are certain clashes, particularly in the bass range, but as I've got older, I've become much more tolerant for all sorts of note combinations.


The wording of your question could be clearer, but I think I understand it.

Given these diatonic chord: I IV V vi...

If you always start with I...

Can the following chords be IV V vi in any order...

And still "work" ..."most of the time?"

Yes and no. The problem is using vague expressions like "work" and "most of the time."

Just about anything will work most of the time for some reason... but that is not very informative. To me, the question is "do you want to understand functional harmony or do you just want a random chord progression generator?"

The permutations of your set up are:

I IV V vi
I IV vi V
I V IV vi 
I V vi IV
I vi IV V
I vi V IV

All of those will "work" at least "most of the time." But how does that help anyone? Work how? Work when? To continue from there will just be more groping around.

The way to better understand these progressions is through an understanding of functional harmony and the effects of following and not following standard function.

Function harmony is the flow of chord functions: pre-dominant (or subdominant) to dominant to tonic. In Roman numerals that is typically IV V I or ii V I where IV or ii both can fulfill the pre-dominant/subdominant function. Following that standard function is described in various ways like "strong", "moving to home", or "works."

But you don't always have to follow the functional flow. Other progressions can be used. They have different "feels" and can be describe with various musical terms. Let's list a few using just enough chords to illustrate the ideas:

  • IV V | I the standard functional progression, V | I is a "closing" gesture, it's an ending
  • I V is tonally strong, but is an "opening" gesture, it leads to a continuation
  • V IV | I swaps the subdominant and dominant positions, but still has all the primary tonal chords so is tonally clear, could be called less emphatic than the standard functional progression, plagal is the label for the IV I progression at a cadence or important structural point like repeating the progression, heard in rock and 12-bar blues
  • V vi is a deceptive progression the "deception" being rather than V moving to I it goes to vi and deceives our expectation.
  • I vi pairs the tonic and relative minor, the chords are similar as both contain the ^1 tonic and ^3 mediant scale degrees, this progression can be a prolongation of the tonic chord (pairing vi IV is similar but it prolongs the pre-dominant function)

We need to make a general caveat that "functional" and "strong" should not be misunderstood as "good" and "non-functional" and "weak" means bad. "Strong" and "weak" progressions are just the standard descriptions used for progressions in music theory. Each type can be employed for various effects. Weak progression could be used to create a gentle mood for example. Also, "functional harmony" and "tonal harmony" are often used synonymously, and when harmony is non-functional it is sometimes called "modal." (Modal harmony is where the plagal term comes from.) Modal harmony can be just as effective as tonal harmony, but I think that get's beyond the scope of the question, so we should move on.

Now let's list the permutations of I IV V vi again, but with comments and in roughly descending order from most functional...

I vi IV V - standard functional progression, first three chords by roots descend by third, nicknamed "doo-wop" progression.

I IV vi V - same as "doo-wop" but flip the order of the two pre-dominant chords which does not undermine the essential functional flow

I IV V vi - standard functional progress, but ends with deceptive progression

I V vi IV - starts with an opening gesture, continues with a plagal move, nicknamed "Axis of Awesome", with no "ending" gesture it lends itself to repeating

I vi V IV - tonic prolonged, then a plagal move

I V IV vi - opening gesture, then weak ascending thirds, least emphatic, tentative

All of those could be made to "work" but they have different potentials.

I would add in a few more ideas to replace "random" chord use with deliberate handling...

  • harmonic rhythm, how long each chord lasts, the relative importance of a chord in a progression can be altered by duration: one chord per bar, one chord for two bars, two chord per bar, etc.
  • metrical position, placing a chord on beat 1 emphasizes it, other beat positions or syncopation can de-emphasize a chord
  • prolongation or auxiliary effects can be made with neighbor chord one step away by root, ex. V IV | V or vi V | vi, etc.
  • returning to the previous chord will usually repeat one of the devices already covered and add interest to a basic progression
  • to the extent that the non-functional progressions tend toward "weak" progressions, you can try to strengthen or clarify them by just adding a V at the end, or maybe a plagal IV

So, if you apply some of those ideas to one of the "weaker" progressions, you could take (| means a barline)...

| I | vi | V | IV |

...and modify it to...

| I vi | V | IV V | IV | which subordinates vi and emphasizes plagal IV, when reduced it's mostly | I vi | V | IV | IV | which maintains the original chord order, but changes the relative importance of each chord.

...or it could be modified like this...

| I vi | I | vi V | IV V | which when reduced sort of transforms the original to | I | I | vi | V |

The important idea is to take your initial impulse - "what can I do with just four chord" - and replace the "random" groping around with an understanding of how functional harmony, relative strength of progressions, harmonic rhythm, and metrical placement can be used to make a progression "work" to produce a musical effect.

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

Not the answer you're looking for? Browse other questions tagged or ask your own question.