Lucy Keer

Will report back. It may be non-representative because this is for a subject that I know rather a lot about, so the main blocker was motivation rather than creation.

I fairly often junk an early version of something and rewrite mostly from scratch, preserving small bits that seem OK. And I’m often aware while writing something that it’s not great and I may have to do that. And sometimes I explicitly think “this is just a throw-away version.”

The overwhelming majority of what I write is written in one sitting, with no more than about 1-3 hours passing between first word and clicking publish (or at least a finished first draft to send to beta readers). It works very well when it's viable at all, but sometimes it isn't.

For the past two years I’ve been supposedly writing a book, but it takes 3+ full days of re-reading it to get the structure back into my head, and two days away means it drops back out so I have to start over, and uninterrupted periods of >3 days have been rare; so slow progress.

I feel like if it takes 3 days to get the structure back in your head then you're writing it in an insufficiently modular way, but I admit I've no actual experience of writing a non-trivial book and this is mostly an opinion derived from writing software.

That is absolutely the case! I’ve written only one finished book before (plus several unfinished ones, ugh) and this one is unique in that way. It’s trying to convey an understanding of the relationships among three things (reasonableness, rationality, meta-rationality) >

> that are densely linked, and the discussion of each aspect of each seems to have to explicitly make the connections to the other two (as well as explaining itself in its own terms). I’ve never come across anything like it, structurally. >

> I do worry that if *I* can’t keep the structure in my head, readers won’t be able to either. Which is part of why the linkages seem to want to be made so explicit. And why the text needs to be extremely clear and simple, which is a job.