In the original draft of A Dark Place, Sorlil gives us 3 stanzas of 5 lines, and a concluding stanza of only 4. Some of the lines in the first half of the poem are quite short. This can feel unsettling if you tend to get obsessed with symmetry when writing, if you're always looking for the most pleasing 'shape' on the page.
I'm not saying that's how Sorlil operates, since I can't possibly know that, but her first draft does have a solid, boxy shape that feels very much on its way to being a final draft. And in her own comment below, she mentions structure as the key element in her revision choices. So why, assuming a quest for 'better' structure, does the poet choose to dissolve her original box shape in favour of looser two-line stanzas?
The first draft perhaps felt a little too close to note form, so she wanted to extend it without having to rewrite. When revising, we usually prefer to work with what's already there rather than write new material, mainly because of natural human laziness but also because revision uses a different set of skills to those we use when creating, and it's not always easy to swop sides, as it were, half way through.
Perhaps she also felt a certain structural gravitas was required to match and contain that seriousness. The five-liner of her first draft may have felt too uneven faced with those four-sided 'slabs' and 'rectangles', yet a four-liner would have presented other problems - such as what to do with the lost fifth line? and would the closed box structure of a four line stanza pull the poem shut instead of opening it up?
Having hit the right note in the two-line stanzas, she extends the line for greater weight, adding 'up' for the sound echo with 'poplars', and for the first few stanzas this new draft feels strong, decisive, uncompromising.
Break-Points & Buffers
Then we hit a snag. The well-known slogan Arbeit macht frei - 'Work shall set you free' - from the original draft has vanished. And this may be Sorlil's first mistake, as it lent an important air of bitterness to the poem, and also provided an important break-point between the inhuman look of the place and the introduction of live human beings - the guard, the poet-narrator herself, the daughter of a survivor, the generic 'you' who flinches at the end.
Structure begins to break down in the absence of that buffering line, which should have separated the two halves of the poem.
The ending feels rushed: 'And now a daughter of a survivor can't stop//talking'. This too-drastic stanza break needs to be rethought. It's not only rupturing the flow of the poem but its arbitrariness actually draws attention to the way the poet has gone about redrafting, rather like a trick that gives away the magician's secrets.
Of course, the endings of poems are notoriously difficult. I've written about them before on Raw Light, most notably here and here, so I won't spend too long on this. In the first draft, the simplicity of those two lines, 'You flinch when I say/I caught the bus from Dachau' works tremendously well within the context of that particular draft. But in this second draft, we're into a new structure where those lines don't fit anymore. So we get that last line in quotation marks - unnecessarily - and placed alone, cut off from the two-line stanza structure as though for additional emphasis. Which it can't carry off.
If this was my poem, I'd be inclined to shift earlier parts of the poem about in order to get back into a position where I could use those original last two lines more or less exactly as they appear in the first draft.
In particular, that unspecified 'you' - a useful poetic device, if somewhat over-used in contemporary poetry - opens the poem up at the close by inviting the reader to identify with it. Yet the second draft obscures 'you' by burying it hurriedly in the middle of a line - a line which is incidentally too long for the established rhythm, making the poem sound breathless and uncertain at that point instead of centred and ready to close.
Shifting the Focus of Revision with each Poem
The closer a poem gets to the real thing, to being fully alive and aware of itself, the less we need an overview of the poem's problems. By that stage, looking at word and line detail becomes the key issue during redrafting. It's particularly vital not to mess too much with structures if they worked just fine in the first draft, or only needed tweaking. Not that I think the shift to couplets was necessarily a mistake here. The poem feels more grave and measured now, less conversational.
But all decisions have a knock-on effect, and in this draft, further adjustments may need to be made in order to compensate for that change. It's also a possibility that the charm of the original draft lay to some extent in that conversational tone - the sight of those inhuman slabs versus the intimate voice of the poet in your head - and a push to regain that might be something for Sorlil to explore in a future draft.
Many thanks to Sorlil for handing over her two drafts to be manhandled in public by such a blunt and insensitive critic. Usual reminder to take everything with a large pinch of poet's salt; another person might say exactly the opposite, and who can be sure which approach is best except Sorlil herself?