Songwriting: Knowing what you like and learning how to recreate it

I've been thinking about songwriting instruction, and sometimes it feels like it's done all wrong. Or, at least, it starts out all wrong. And maybe it stops short, too. They jump into the elements of good songwriting as though they're objectively derived, or as though there's no subjectivity that goes into it. 

Looking at the songs I love, they mostly violate the rules in some way or another. And I don't mean subverting expectations, which is important when developing hooks, and which can be seen as necessary to good songwriting. I mean, some great songs have plainly bad lyrics. A lot of great songs have only good lyrics and not much in the way of interesting melodies. Some great songs have such strong grooves that they don't really need melody or lyrics. And there are some songs out there that follow the rules to letter, but still feel lacking. So what's up with that?

It seems to me that the first goal in songwriting education should be understanding what you like and why you like it-- what is it that has inspired you to want to write a song in the first place? Bring in 5 songs that you love, then spend time taking them apart. Is this chord change in the bridge what makes your heart sink? Is it that surprising rhythmic flourish just before each verse starts? Is it the internal rhyme scheme that titillates your mind? Or is it just that you like the topic of the song? Maybe you can't resist a sad song about puppy dogs, no matter the presentation. 

Then, once you've determined what you like about music, it makes sense to talk about the tools and techniques that'll get you there. But I feel like most music instruction presupposes the end goal, and that's a mistake. 

After that, I feel like the big question is, would you like your song if you weren't the one who wrote it? This is difficult to ascertain, because a songwriter invests a lot of emotional energy into creating something. You get attached to what you've done, and you feel the emotion that you felt while writing the thing. I think there's a point at which you can say, "hey, I did it-- I made something from nothing; I'm a successful songwriter". There's a lot to be said for that being enough-- you've done your job.

But I think most writers want other people to feel the way they do when they hear their songs. So I think feedback is really important-- not to tell you if the song is good or bad, but to make clear if they're hearing what you're trying to say. They may hate the song because it's depressing, but if you were going for depressing, that's okay. Some people hate happy songs; doesn't mean your happy song doesn't do exactly what was intended. 

So, I'm thinking there are three key parts to songwriting instruction, the first and last often being glossed over: 

1) Assess your taste - What do you love about music? Why does it inspire you to make your own?
2) Tools and techniques - How can you write songs that share the elements that you love?
3) Testing the product - Would you love this song if you hadn't written it? How do you know people are hearing what you're trying to convey?