I think there are deeper problems with Helen's approach than step nine, the most obvious being that it's most prone to produce formulaic, uninteresting music. For instance:

(1) Accents frequently appear at other places than on the first beat of every measure. In fact, for variety, it's often desirable to place them elsewhere: on syncopated pick-ups, on the second or third beat rather than the first...

(2) The meter of a song is not really dictated by the rhythm of the words; texts in duple meter work wonderfully well as triple meter songs, and vice versa. Often, a song written in one meter is performed by others in a different meter. Nor do you have to stick to a single meter--I've written tunes that flip between 6/4 and 4/4, or between 6/8 and 9/8--these variations were not dictated by the meter of the text, they arose from other considerations.

Even if you're starting at an introductory level, mental fixity is the first thing to be avoided in writing tunes, since tunes often have to compensate in interest for the boring regularity of meter in texts.

The major advantage of Helen's approach is that it chunks down the daunting task of producing a song into smaller, simple subtasks. The major weakness is that it approaches the task from the wrong end: from the details up. One should begin looking at the overall picture and chunk down to the smaller units.

The first task is to select the right kind of text. The best poems to set are the ones which the writer wrote somewhat envisioning them as songs--many of Kipling's poems are good examples: they even include "chorus" verses or interspersed refrain lines. Free verse is a particular challenge due to the lack of regularity in meter, line length, verse pattern or repeated elements.

What attracts you to the text, what parts of it would you like to bring out? If you can't explain this to someone else, you probably won't have a clear idea yourself, and will probably miss the mark.

What is its overall shape? Do you envision it as a simple tune that repeats for each verse, or verse and chorus, or an "A B B C B" pattern, or what? Where does the text build, crest, release? What sort of emotional contrasts exist? For longer texts, how do stanzas group thematically? Do you need the whole text, or are there weak parts to omit?

What's the mood? What musical subgenre does the text suggest or does it need to fit? What other songs in that subgenre would you like this song to resemble? What specifically appeals to you in them? What are the conventions of that type of song? (Conventions are both guides and indications of where you need to break established patterns.)

Look at patterns within the stanzas. What sort of units do the thoughts and phrases comprise? Are there prominent subgroupings to the words? Natural pauses? How do these groupings compare stanza to stanza? Are there words or phrases that beg to be emphasized (for instance, by lengthening or punching)? Texts that have discernable suppatterns tend to be most workable. For strictly regular texts (like broadsides), you may wish to add something to make them more song-like. You might want to repeat a phrase or final line, or extract an epigrammatic couplet to use as a refrain. For longer songs, consider writing two or more tunes to interleave, to maintain interest.

Don't rely on relaxing and "letting your mind float" (as Deckman suggests)--for most people, that's too unmanageable and unproductive: there's not enough direction. Rather, approach writing the song like a newspaper reporter writing a story. He has to get the overall picture, then find out the details, and he does this by asking lots and lots of questions. The answers to the first questions lead you to what other questions you need to ask until you have all the material you need. What parts do you want to echo, what parts do you want to contrast? Where do you need particular musical interest, like a leap, syncopation, unusual progression or modulation? Where do you need to leave space for extra words or syllables in other verses that will have the same tune? How should you tie two bits together? Do certain words suggest certain musical mimicry?

Serendipity and creativity come both in what questions you ask and how you answer them. Your creativity will also charge in full bore once you hit a level of detail small enough for you to handle, and it can be amazing how much flows from that. In the "research" process, you'll discover some musical lines and ideas you'll definitely want to try, and a bunch of other possibilities. Once you get into it, you weigh what seems to work and what doesn't.

When you get to a certain point where the song seems to have taken shape, then you begin editing and re-editing with a more critical ear. Does it really achieve the effect you were hoping for? Where is it too dull or too overdone, where are you not getting the punch or feeling you were after... (Again, more questions--our minds are generally better at problem-solving than at dreaming up out of the sky without direction.) If your inner critic gets too negative--"That's rubbish!"--make it explain why and offer a better alternative. In this way, you make your inner critic a creative collaborator. Editing is the kind of task at which it excels.

To avoid expecting too much from your early attempts, at the outset pose yourself the goal of writing five different tunes to the same text. Regard each attempt as completely disposable, since ultimately you'll only use one. (Whether you actually have to write five, or even more, to hit on one that really pleases you is a different story, but you must be willing to keep trying and not get too attached or critical in the formative stages of a tune.) From each you'll get ideas of what you like and don't like, of what works and what doesn't with this particular text. You'll be more willing to try an experiment that could lead nowhere. If you don't have the willingness and determination at the outset to write at least five tunes, then you'll probably end up with a hack song that no one should have to listen to. Spare us.

Capture musical ideas as they occur to you--don't rely on your memory later. If you can't write musical notation, use a tape recorder. (Yeah, you probably hate listening to your recorded voice, but there are worse things.)

Similarly, if you find certain questions particularly useful, jot them down, keep a list. You can refer to it later when you write other songs.

Now, I don't follow a rigid approach to tune writing, but I find that this is pretty much how I go about things internally. The one thing I don't do enough of is keep searching for other tunes even after I've found one I like. If I did, I'd probably end up with better tunes.