I heard Richard Sennett talking on Front Row about his book The Craftsman, and rejoiced (even as my feminism was annoyed by the title, though it's hard to think of an alternative). I confess I haven't read it yet, but he was so cogent in the interview that it's now high on my must-read list. The point he's making is that what we think of as 'craft' - an old-fashioned virtue suitable for gnarled old blacksmiths and batty amateurs with time on their hands - should actually be seen as a thoroughly modern skill-set for modern life at work and at home. Not many people reading this blog would disagree that craft is important, but they might be surprised to discover that one of Sennett's prime examples is software developers. And yet it makes sense. Understanding the materials of your trade, using your tools, exercising your skills, are deeply satisfying activities, and seeking that satisfaction is one of the most basic human motivations, whether you're working on a flat screen or a cave wall.
The beginning of good craft, Sennett says, is problem finding: if you find the right problem, and ask the right questions about its nature, including a willingness to think laterally, then the right solution comes about. My own writing is always kicked off by problem finding in the form of a question, or rather two questions, one about characters ('Why is she standing in this house?') and one about technique ('Can I really pull off a three-way parallel narrative?'). But, more generally, for years I've maintained that good writing is all about getting your process right, not about setting out to produce a particular product. In other words, if you explore and establish how to make your writing happen, the words which end up on the page will be the right words. Of course you then revise - hone, polish, trim, re-shape, polish again - but, like a fractal diagram, each of these little craftsmanly actions is, in itself, a mini version of the problem-finding process. I could argue (though I'm not sure I am arguing) that there's no difference, qualitatively, between getting a sentence right and getting a novel right, even though there clearly is a quantitative difference.
I think that's also what marks the divide for writers between feedback which helps and feedback which doesn't, or even hinders: good feedback helps you towards that problem finding ('What were you trying to do?' 'I didn't believe in her doing that.') whereas feedback that's unhelpful imposes external judgements ('Too long for the genre.' 'Parallel narratives don't work'.) and leaves you to work out what the real problem is. Indeed, as well as technique, one of the craftsmanly skills all writers need is to be able to judge feedback, and decide what to do with it.