My current project is a detective fiction novel set in 1927, where Inspector Gilles Maintenon is on a walking tour of Dartmoor and stumbles on a mystery.
So far I am up to 41,000 words. At about 32,000 words, I kind of stalled out; but the thing is going better now and I have stuff written today, and ideas for tomorrow's writing, which is important and reassuring.
Detective fiction follows a certain formula, because the readers have certain expectations. In a suspense novel, you often know who the bad guys are--the attraction of the tale is the challenges set to the protagonist, and how it is laid out. But in a mystery, the resolution has to be believable, and the readers of the mystery genre expect that if they look back on the book, they will see that all the clues were in fact provided to them. It's a question of seeing the significance. Sometimes you read something, and a little light goes off in your head, and you know a character or fact will be significant later in the story--you just don't know how or why yet.
When I wrote my first novel, a WW I parody, it was in reaction to the way history, especially WW I history, is presented to Canadian audiences or readers by Canadian writers, producers and networks.
In some ways, almost every book or story I have ever written was written to present a thesis, or a premise, or as a reaction, often a gut-level reaction, to something that I either didn't like, or thought was overdone, or too often presented over-simplistically, or in the case of Canadian WW I history, just plain mealy-mouthed Imperialistic bullshit.
My third book was a parody of a space opera, and the basic premise was to put some believable science into the book, although most academics will dispute the possibility of faster-than-light space travel.
The work I am doing now sets out with no thesis, no great social theories, no premise, other than the fact that I think I can actually write a good detective story with a certain tone, a certain feel and some really professional writing.
The thing has to hang together, and it has to make sense, and the characters have to act, sound and feel like people would in a certain situaiton, in southwestern England, in 1927.
So far I have been working on it for about four months, yet my first novel took two and half years to complete a first draft. The next three books took maybe three months to smash out a manuscript, rough as they were at first glance.