How did you become a songwriter?

September 6, 2005 8:03 PM

Guitarists (and other musicians), how did you make your move from player to songwriter?

I'm kinda in this stage with my guitar playing where I can play well enough and play covers, but don't really do much else. Can someone recount their experience of shifting from just playing to actively writing and playing their own music? What methods did you use to get there? Instruction? Experimentation? What theory came in especially handy in the beginning? Did you have any issues withing singing and playing at the same time and how did you overcome them? What are the do's and don'ts? Recommend any books, teachers, or methods? etc

I think the best answer would probably be someone saying "go take a class at the old town," but I am interested in other people's experiences and advice.

posted by skallas to media & arts (22 comments total)

I think I started playing because I wanted to write songs.

If I can reconstruct anything about the process, I'd say a lot of time playing and a lot of experimentation. Spending hours of idle time noodling up and down the fretboard, finding runs or chord combinations that sounded cool, and then playing them obsessively over and over perhaps dozens of times until I could hear melody lines starting to emerge. Lots of repetition of simple forms, maybe having one good, five-word chunk of lyric with melody that I'd then try different elaborations on.

Theory that was handy: Chord inversions, because they were more likely to make a basic three- or four-chord progression sound new and different. Playing with modes, as well.
posted by Miko at 8:17 PM on September 6, 2005

I'm not sure there were any "methods" for me - the transition just happened. I went from buying Guitar World and downloading tabs to just not being interested in playing other peoples songs anymore. I can emphasize that a decent understanding of theory helped a lot. Once I had a good grip on being able to play a variety of scales, a variety of chords, and importantly understanding the theory behind how to fit that all together (what chord progressions work, for instance) I was on my way.

So the two most important parts of theory, I found were:

(a) Scales. Particularly, in playing rock music, blues scales. They should be automatical and natural. You should just be able to noodle around and improvise a blues line without thinking about it. But major and minor scales are useful for playing with more pop and traditional melodies. And I'm a fan of the locrian mode, for some easten feeling.

(b) Intervals. Understanding what notes go with what. Knowing the sound and feel of 3rds, 4ths, 5ths and 6ths can be very useful - it helps you know where to go next in a chord progression. It also helps to know what doesn't sound good. This is something you should pick up from playing cover songs - look at the kinds of intervals and chord progressions your favourite band uses, and practice playing around with what you know - different orders, changing major 3rds to minor 3rds.

From there it was just matter of jamming and experimenting. Coming up with interesting chord structures, bringing them together into songs.

I'm not one to help with the singing / lyrics, though, I just write the music.
posted by Jimbob at 8:19 PM on September 6, 2005

Now here's a question I feel qualified to answer.

I think I worked out singing and playing at the same time long before I started writing songs, although rhythmic independence is a difficult skill, and it's something I'm always trying to improve at. It's not easy to play and sing two different, complicated rhythms. This is something that's best to isolate and practice on its own as you would any other musical ability. Especially if you're going to be performing solo, having a strong sense of internal rhythm is very important. Being rhythmically inconsistent is a pretty surefire way to be dismissed as an amateur.

Every bit of theory comes in handy, but I think chord/key/scale theory is probably most vital for starting out. i.e., in a given key, what chords can be used and how are they to be treated, how to modulate between keys, and what notes can be used for melody-writing. Understand how to use all of the diatonic chords in a key as well as their secondary dominants and you'll be in good shape for writing pop songs. But absorb every bit of music theory that you can, it always helps.

I think that for me, understanding the use of non-harmonic tones helped me a great deal in writing melodies. Rather than simply choosing notes out of the air, I often make a decision to use chord tones or non chord tones to achieve a particular effect. For example, singing the third of a chord sounds somewhat "sweeter" and less stable than singing the root, and using appogiaturas often creates a sentimental or "sighing" sensation. It also makes it easier to sing a melody when you're aware of each note and its relationship to the chord you're playing.

Take voice lessons, and practice singing often. Develop good vocal technique, and then add whatever expressiveness or idiosyncracies you'd like to add.

As for how I started writing songs, I don't know, I just did. My songs were really bad when I started, though. I felt like I was just regurgitating other people's songs, writing solely in cliches, pretty much fumbling in the dark. That was before I started seriously studying the things I've mentioned above, though. For me, analyzing the songs of the composers I loved, specifically The Beatles, in agonizing detail was extremely instructive. For that purpose, I can't recommend this site enough. The Beatles did it all in terms of writing pop songs, and Pollack gives you all the details. I've also studied Gershwin, Cole Porter, and many of the other great modern songwriters. Analyzing classical scores helps, too, but not in as direct or practical a way.

But analyze whatever songs you admire. Really figure out what makes them tick. Figure out how to play the melody on your guitar, and analyze the melodic contours and how the melody notes are used against the chords. Analyze th Demystify everything. Don't be intimidated; you can write just as well as any other songwriter can. And keep at it. I don't even remember some of the crap I used to write, but it was probably necessary for me to do it in order to improve. Be critical of your own work, but don't beat yourself up too much. As long as you keep getting a little bit better, a little more fluent, a little more confident, you're on the right track. Feel free to e-mail me with specific questions.
posted by ludwig_van at 8:29 PM on September 6, 2005 [1 favorite]

Er, that bit in the last paragraph was supposed to contain:

Analyze the chord progressions. Figure out which chords and notes occur naturally in the song's home key and which are outside it, and how the outside notes are used. Analyze the rhythmic devices and the arrangement.
posted by ludwig_van at 8:33 PM on September 6, 2005

Here's something simple and practical: I got my hands of some recording software (I use Audacity, it's free). This will allow you to play on several tracks. You can begin to experiment with guitar riffs, repeat them, then sing over them - this is actually the way some bands write music (U2, for example). It's just a way to get started, and when you don't have to think about playing the guitar, you can sing and riff and let the melody take you wherever it wants to take you. It's also a total blast.
posted by crapples at 8:39 PM on September 6, 2005

crapples gives a good suggestion. Multi-tracking software is fun, but for me it's always been more useful for working out arrangements after the bulk of a song (chords, melody, lyrics) has been written. I'll record the main parts, then start adding the vocal harmonies, complimentary guitar lines, etc. It's a great way to conceptualize layered songs when you're by yourself.
posted by ludwig_van at 8:49 PM on September 6, 2005

"Noodling around" is great. It's how you experiment, try new things, be creative, etc. But I find that if I just sit down and "noodle around" that I spend my energy without purpose and often end up without much to show for it at the end of the day.

I have found that the following simple things vastly improve the productivity of a noddle session, if the intent is to actually write something.

1. Use a metronome, click track, or some other rhythm section backing. For me at least, I find that it is way too easy to get distracted by something and keep stopping and starting. With a metronome, you have a rigid beat that you have to adhere to and it makes you somewhat more disciplined. It also forces you to think in the confines of an actual song structure, rather than just kind of "farting around" with some licks that sound cool.
   
   
2. Record your session. It doesn't have to be high-quality or permanent. I use the computer for this. Don't get bogged down with the details because you're not trying to actually produce anything. But hearing what you play is crutial. If you're just experimenting with ideas, trying out licks and runs and so forth, you really don't have a clear picture of what the overall sound of what you're playing is going to be, because you are concentrating on the technical aspects. When you record your practise session and then play it back, you hear it as the listener would. I find that many things that I think sound like crap or are worthless actually have a certain appealing groove when I listen to it again. And things that I thought sounded rad sometimes just don't work when I listen to the recording.
   
   
3. If at all possible, practise with other humans. When you are trying to write a song it is about 300 million times easier to get things moving if you have a drummer and a bass player (or whatever combination of rhythm section you prefer) to fill out the parts. Personally I find it a lot more difficult to try to write a song by playing solo and trying to imagine what the other parts will sound like, compared to actually hearing them. Even a very simple chord progression sounds much better with a good drummer filling in. I find that when I'm practising alone I tend to overcompensate -- because I'm just hearing the guitar and nothing else, I tend to try to come up with elaborate leads and exciting passages (and failing.) But less is often more, and unless you are a Joe Satriani-type you are not going to have success just sitting down and wanking on the guitar alone. Good songs work when all the parts come together, and to write a good song you need to invision all the parts at once. This is hard to do if you are practising alone, compared to jamming with a group.
   

Now, these are just my personal feelings. Everyone is different. YMMV. My context is that of guitar player and four-piece blues/rock/metal type setup. Obviously if you're more into the solo-acoustic type of performance style then a lot of the "practise with others" thing won't apply.
posted by Rhomboid at 9:00 PM on September 6, 2005

I'll preface my comments by first stating that I don't necessarily consider myself a musician but I've been playing guitar for a number of years and I understand what you're asking. I started writing my own songs mostly out of boredom. I got tired of learning other people's stuff but became really energized when I started piecing my own chord progressions together. Pretty soon these coalesced into songs and then songs just started kind of revealing themselves if I worked hard enough to find them. You can probably tell I don't have a lot of theory under my belt but I've written or sketched out dozens of songs so I might have a little knowledge to share. Here's some ideas:

- Write bad songs. Many of them. Songs with totally obvious progressions. Songs that you swear you've heard before. Songs that you would be embarrassed for anyone to hear. In all creative endeavors its those ugly things that eventually fuel something great. 10 bad songs = 1 good song.

- Pretend to be someone when you're writing. If you like Willie Nelson then pretend you're Willie Nelson. Ask yourself "What would Willie do?" when you run up against a wall. If this seems wacky then pretend you're writing a song for him.

- Refine the song to a point you're happy with but make sure it's still very rough around the edges. Then record it. Then listen to it. Then go back and start over. I find that recording the skeleton of a song makes it more malleable for improvements after you've had a chance to listen to it.

- Improvise lyrics where none exist. Sometimes I'll just speak in tongues while working through a song in order to help focus the direction it's moving in. Singing it can bridge you to that next part when you can't quite find the right chord.

- Avoid unnecessary ornamentation. Try writing a song with only three chords. Restrictive exercises are good for setting boundaries. Creativity can sometimes be more inspired by limits.

- Stop learning other people's songs for awhile. Find your own groove.
posted by quadog at 12:42 AM on September 7, 2005 [1 favorite]

I second splitting up singing and playing - learn the parts independantly, practice them until you can do either pretty much unconciously, then put them together. You may find one or two passages you still have to work through, but they'll be a lot less. Pianists break difficult passages up and learn one hand at a time - makes even more sense when you're dealing with two completely different instruments.

Practice your aural skills. Most aspiring songwriters I know get frustrated that they can hear the songs they want to write in their heads, but tend to come up with something more formulaic and less satisfying when they sit down to write - that's becaues we all learn little patterns that tend to dominate our improvisation - they can be handy for getting you out of corners, but these physical patterns can often put the cart before the horse where songwriting's concerned. The better your aural skills are, the more likely you are to be able to get the musical ideas from your head, instead of responding to the physical habit of a particular lick. I know a few people who use an instrument that isn't their main one for the initial stage of working out their ideas, to avoid this. Singing it seems the most obvious way - the physical process is less obvious, and a lot of people find it gives them a more direct connection to their musical ideas.

Persevere with your songs - even if they seem unpromising; it's really important that you learn to refine your ideas, as well as having them in the first place, and the only way to do this is practice. Keep a notebook/harddrive/dictaphone/whatever with your unfinished ideas on - when inspiration doesn't strike, go back to something you left unfinished, and work on it. You'll learn as much from that process as you do from constant improvisation. I've taught quite a few people who've had no problem getting ideas delivered fresh by the ton, but have had no idea how to turn them into finished pieces. Don't be afraid to try your ideas in very different arrangements.

Try a lot of different entry points. You may think you write best a particular way, but it may be, again, that it's the familiarity rather than the quality of the results than leads you to think that. Start with the lyrics, start with a vocal melody, start with a riff, start with a chord sequence, start with a bass line, start with a rhythmic pattern. Richard Rogers used to write new music to existing songs (by other people) then take away the old lyrics, and write new ones to the music. Worked for him - doesn't for a lot of other people. Try setting other people's words to music, and words to other people's music - like singing and playing, don't expect to be able to write great music and good lyrics simultaneously.

Take time away from the music. I often realise stuff when I've been away from a particular piece for a little while. Listen to it critically, as a listener, not the composer or the performer.

Distinguish performance and songwriting in your own critical facilities - just because you (or someone else) can play it really well, doesn't make it a great song. Just because it's hideously out of tune, and played by what sounds like an infinite number of monkeys after an infinite amount of whiskey, blindfolded and with their arms tied together, doesn't mean it's a bad song. Listen to your favourites at a bad karaoke night and you'll see what I mean... Motown used to get a number of different groups in to record each song - they'd then just release the one they decided was best.
posted by monkey closet at 1:43 AM on September 7, 2005

I'm a songwriter and this is how it worked/works for me:

I noodle around on the keyboard, find chord patterns I like and then put a melody with them OR I write my lyrics, set them in front of me on the keyboard and noodle around till I get a "hook" then write the rest of the melody around it.

Also I live by the adage that you have to write 100 bad songs before you write one good one....I keep an "idea file" of phrases, half-written lyrics, and ideas-I don't throw anything out-and when it is time to write I thumb thru that till something jumps out at me. I have songs that I have tried to finish for years and years and then wind up cannibalizing the parts for a new song that actually works.

The main thing is to keep writing. Don't self censor-even if something stinks it still might be good raw material to mine something better from.
posted by konolia at 4:36 AM on September 7, 2005

Stop learning other people's songs for awhile. Find your own groove.

While I'm sure this was well-intentioned, it strikes me as bad advice. Learning other people's songs is one of the best ways to learn songwriting. People who try to deny themselves any influences often seem to me like they're really just shutting themselves out to any new influences and continuing to regurgitate all of their old, possibly unconscious ones. I would recommend, rather, that you stop learning other people's songs by rote; learn them instead by ear, in order to develop your listening. Then write them down and analyze them harmonically, structurally, etc. Only bother to do this with songs that really impress you in some way.
posted by ludwig_van at 8:53 AM on September 7, 2005

Stop learning other people's songs for awhile. Find your own groove.

I say hurrah to this, myself.. you've got a way of talking, walking, a mannerism when you tell a joke, all these personal things.. try and make personal music. You've got a huge amount of subconcious music floating around in your head, if you hear a chord, you'll probably sing notes over it that you've heard a million times. It takes work to get this crap out of your head, and intensly scrutinizing other peoples work will only carve it into your brain in both a creative and analytical manner. Just my feelings, but the 1st thing I look for in music is character. The process depends on what you want to write, and why. I'm a big fan of David Fair "How to play guitar", so that shows my personal angle.
posted by Jack Karaoke at 9:52 AM on September 7, 2005

Harmonic Experience is a great music theory book where you sing over a drone instrument like a guitar to really get a sense of the meaning of all the intervals. Though it won't move you into melody, it really helps you to get what harmonic movement means instead of just winging it.
posted by abcde at 10:37 AM on September 7, 2005

stop thinking about it so much and just do it. write a billion terrible songs.

the most important thing you can do is finish every song you start. no matter how god awful some of them are.

learning what not to do is probably more important to your unique style than imitating what to do

and most importantly, you will get better, through old fashioned practice, at writing complete songs, rather than becoming a master at writing halfies.
posted by Satapher at 1:00 PM on September 7, 2005

and always try to keep your artistic philosophies short and sweet.

music is a phenomenon; dont strip away the mystery.

no one needs to be told what sounds good and what sounds bad.
posted by Satapher at 1:08 PM on September 7, 2005

David Fair's "How to Play Guitar"
posted by Satapher at 2:00 PM on September 7, 2005

Satafer, I disagree about the halfsies. Sometimes ya get stuck and that baby ain't moving. Yet those halfsies are great fields to mine later. One of em gave birth to a song that is going on a cd later this year.
posted by konolia at 2:27 PM on September 7, 2005

Okay, standard YMMv IMHO disclaimers and whatnot.

Don't write songs that you think are bad, just to have written a bad song. Don't take a good verse and attach it to a bad chorus, just to have finished it. Your first songs don't have to be masterpieces, but they at least should be something that you yourself like. Otherwise, you're making crap for the sole purpose of having made something.

Which is not to say that everything you write will be pure gold. Everyone looks back on their first songs and shudders, but hopefully you really liked them when you wrote them, because that's what will keep you going: liking what you do while you are doing it.
posted by 23skidoo at 4:33 PM on September 7, 2005

a finished song is just that; if you don't resurrect you're on the shelf just lying
posted by Satapher at 4:56 PM on September 7, 2005

Oops. Thanks Satapher.
posted by Jack Karaoke at 5:51 PM on September 8, 2005

Nothing personal to Satapher or Jack, but that "How to Play Guitar" article is awful.
posted by ludwig_van at 11:29 PM on September 9, 2005

...that "How to Play Guitar" article is awful

I assumed it was a joke
posted by monkey closet at 8:22 AM on September 12, 2005