A professor in my department used to say that you shouldn't write a linguistics paper like it's a mystery novel: you're not supposed to hold the conclusions until the end. You're supposed to say it first thing, and then walk through the facts and reasoning that support that conclusion. Don't leave the reader in suspense.

I've decided that I beg to differ. Writing a paper is exactly like writing a mystery. Certainly in a sort of extended metaphor of paper-writing, but also in a narrower metaphor of the paper itself. The paper is simply the last chapter of the novel. In a classic mystery, you arrive at the end, and the detective has gathered everyone into a parlor or something, and says boldly and decisively, "You've all been wondering who killed Lord Such-and-so. Well, I can say with certainty that Captain Whats-his-name is the culprit!"

Everybody reacts, either that they never expected it or that they always knew it to be so. And then the detective goes through and explains, step by step, how it all happened. Not necessarily in the order that they discovered the facts, or the order in which they puzzled out what happened, but in the logical order that draws the characters (and the reader) inevitably to the conclusion. Then there's a nice little wrap-up denouement where they muse about something or other: remaining open questions like motivations or whatever happened to this other person who was being shifty, or where some characters will go next.

Put it that way, and that seems a lot like your typical linguistics paper outline to me. You've got your solution stated first, then your supporting facts and reasoning, and finally open questions and next steps. Sure, I've chopped off the first 90% of the mystery novel to get here, but that's how research goes, too. All of the mystery novel's confusion and fact-finding and red herrings and puzzles still happen when you're doing the research, and that's all necessary, and certainly it has to be conducted properly, but most everyone agrees that the part where a mystery novel---and likewise a research project---really needs to stick the landing is in the last chapter, in the paper you'll eventually submit.

The other great thing about this is that I get to be a mystery-solver out of an Agatha Christie novel now, finding the answers out of a muddled mess and revealing them to gasps of surprise and awe.