[Music] What comes first, chord progression or melody?

Since this subforum hasn't seen much activity lately, I'll ask a question that's been on my mind for a few days now. Typically, I would always write the chord progression first, starting with a bass line and working my way up, but I've begun to start writing the melodies first, and wrapping a progression around it. And I think I'm liking it a lot more this way. I'm curious as to what methods the other fine composers on this site use, and why.

I usually start with chord progressions for whatever reason, but I think when I write the melody first, that melody typically turns out better than usual.

Whichever I think of first.

And then I change one of the other to fit one of the other.

There is no answer to this question.


edit: Why the fuck do I always type "of" when I mean to type "or?"

Hey, sorry, that came out way more dickish than I ever meant it to.

But I said what I meant to say. I don't think any one facet of the music should follow the other, at least not every time. If you have a chord progression in your head, take it down, then build a melody. If you have a melody floating around up there, tie it down, and build a progression under it.

It's about 50/50 for me, lately. It depends more on what I set out to write. If I'm writing something thematic and melodic, then I'll devote all my attention to a melody, first. Make sure it's catchy, memorable, etc, and then build the rest under it. If I'm writing some, for example, from that set of layered music I did, I usually start with chords. Rather, I'll start with ONE chord, and build something intense over it (using marcato strings to outline the chord, using horns to beef up the bass, adding percussion to fill out the rhythm, etc) and then decide what the next chord will be, and repeat. Once the progression is more or less laid out, I add a melody.

There's other ways, though. Right now, I'm working on chase music. Assassin's Creed style chase music, to be specific. I did a lot of research on how it works ... what the hell, I'll share it. Attached below is a chart. It notates instrument that are heard, and when they're heard. The music is mostly percussion, but there's different types of percussion, and there are some melodic and chordal elements, and some types play every other part, some types only play at certain time, etc. There's more to it, but just look at the chart and it more or less explains it all. I had to take a whole new approach to writing music when I started working on this kind of music, because chase music is actually very stagnant. Sure, it's exciting, but melodically and chordally, it doesn't really go anywhere, so I had to figure out why it still kept the energy up, and the answer to that question is very well explained when you just make a chart in excel like a weirdo.


tl;dr: Never be afraid to change your compositional approach.

P.S. This is the track I used in that chart.

<a href="http://www.youtube.com/watch?v=z4eCp-r0HJU" target="_blank">http://www.youtube.com/watch?v=z4eCp-r0HJU</a>

Whichever I think of first.

And then I change one of the other to fit one of the other.

This. Usually I have a melody first, then find functional or interesting chords (or just the obvious ones that sound nice). Then if I think of a chord that would work well, i.e. better than one I already have, I'll change the melody to go with the chord if it doesn't already.
Unless it's something entirely harmony based, like a chaccone. Or a canon, where I'd start with a melody then expand on it, building the more obvious chords (often with suspensions) later in the canon. I like canons.

Well, I grew to realize that my fascination with unpredictable chord progressions and key changes led me to rely too much on them, and ended up writing really poor and bland melodies to accompany them. My new approach involves writing the catchy, memorable melody first, then using that melody to lead into the interesting chord progressions I love so much, making my music strong on both accounts. That's more or less why I made the thread.