The sweetest raisins won't disguise lumpy oatmeal.
Allow me to explain.
This weekend, I had something like an epiphany. That is, I'd known this concept for a long time; indeed, it's something I think we all know. But now, suddenly, it was clear, sharp and blazing: instead of knowing it, I felt it.
The idea is this: Poems often (usually?) have particular lines or turns of phrase that stand out, that particularly impress themselves upon the reader. Different readers may find different gems in the same poem, but that's not the point.
What hit me is that, when a poem is great, there isn't any bad line. Even if a word or phrase doesn't stand out in a good way, it doesn't stand out in a bad way either. A great poem doesn't have any distractions or specific flaws. Nothing gets in the way, no imperfection distracts the reader.
That may seem obvious (or it may not; for that matter, it may not even seem correct). But it changed how I view the writing process. To me, this idea means that every piece of the poem must be right. It's not enough to have one brilliant line, then not bother too much with the rest.
I think I pay attention to all sections, but there are always parts of the poem on which I focus more attention - maybe because they need it. But often, it's the parts that are already working well that I work on most.
It's not enough for the poem to fit technically; the effect has to be right. The crystal might look perfect on the shelf, but it's only when we hold it up to the light that we know how it splinters and scatters the daylight.
Just a thought - but one that makes me think that nearly every poem I've written could use more work.
By the way, this thought came to me as I was reading 80 Great Poems from Chaucer to Now, by Geoff Page - it's a collection of, well, eighty great poems - not necessarily The Best poems, but all great. He discusses each poem in technical, artistic, and historical contexts. But the poems all impress with their lack of obvious flaws, which is what lit the bulb for me.