One of the ways writers make progress in their work is by answering questions: Who murdered the butler? Why did the rabbit stew explode? Where did George get those kneebles, and what is he planning to do with them? But in order to answer a question, one first has to ask it. Sometimes, one does this consciously; other times, unconsciously.
The trouble is, there are a lot of different kinds of questions, which require different kinds of answers and different ways of getting to them. There are information questions, process questions, technical questions, decision questions, and complex questions that combine two or more of the other kinds. Approaching one sort of question as if it were a different sort is a recipe for all kinds of trouble.
Information questions are things like “What kinds of birds live in Thailand?” or “How do I calculate the Lagrange points in a particular planetary system?” or “What did George say about kneebles back in Chapter One, again?” To answer them, the writer has to go find out some information. This can mean rereading Chapter One, educating oneself on celestial mechanics or the ecology of southeast Asia, asking an expert, or even perhaps going off to Thailand to do some birdwatching. Most information questions are relatively straightforward, because the writer knows the information is out there; it’s just a matter of finding it.
A subset of information questions are research questions, which are either much broader or involve something that nobody actually knows yet. “What was life like in Brazil in 1802?” is an example of a broad question; “Can we get human beings to regenerate lost limbs like a starfish?” is something nobody knows yet (so far as I know), and that will take a lot of biological research to answer.
Which brings me to process questions. Process questions fall into two categories: ones that are about the writer’s process (“Do I find it easier to write in the morning, or in the evening? Am I a planner or a pantser?”) and ones that are part of the writer’s process (“Why doesn’t my protagonist want to learn to pilot a spaceship? What do dwarves eat if they live underground without access to the things we think of as normal food plants and animals? What the heck are kneebles, anyway?”) Answering the former requires one to observe one’s work habits and processes and then be ruthlessly honest about what works and what doesn’t, as well as possibly experimenting with different process tweaks to see if something works better than whatever one is doing, even if one doesn’t like it. Answering the latter generally requires the writer to make stuff up.
Technical questions are about the skills, craft, and mechanics of writing, and include everything from “Do I need a comma here?” to “Should this be ‘affect’ or ‘effect’?” to “How can I write a conversation among eight different characters that isn’t confusing?” to “What would change in this story if I wrote it in first-person from Jenny’s viewpoint instead of tight-third from George’s?” A lot of technical questions start with “How do I…” (though that’s not an infallible diagnostic; “How do I get my hero out of this mess?” is really a process question, and not usually a terribly helpful one at that).
And then there are decision questions. These are things where the only answer is whatever the writer decides. Things like “Would it be better to write a YA fantasy novel next, or do a science fiction movie script?”, “Should Jenny be the protagonist, or should George?”, or “Can I really use an infodump here to make this complicated information clear to the reader?” These are also the questions where all the writer’s insecurities come out and party, because a lot of decision questions assume that there is one right answer that everyone agrees on. Many of them either start with “should” or can be easily rephrased to do so, and unlike “Should this be ‘affect’ or ‘effect’?” there isn’t a right answer that one can look up. A bunch of the rest start with “Do I have to…” (and aren’t grammar or punctuation questions where the obvious answer is “yes, that’s how commas work”).
To avoid getting bogged down in “shoulds” and “have-tos,” one has to reframe the question. “Do I want to write a YA fantasy or an SF movie script next?” “Would it be more interesting for me to make Jenny the protagonist, or George?” “How many ways are there to make this complicated information clear to readers, and which one feels like it will fit into my story?”
In addition, one has to recognize that a lot of either-or decision questions are false dichotomies. I could write a fantasy movie script, or a bunch of short stories, or a sequel to something I’ve already done; I could branch out into a totally new genre like police procedurals, or start a collaboration with someone. I could make Jenny and George alternating viewpoints in a braided novel, or part of a strong ensemble cast where it doesn’t really matter which one is technically the protagonist, or ditch both of their viewpoints and make Sam the protagonist.
One thing that confuses the situation is the fact that sometimes the writer needs additional information in order to come up with a reasonable answer to a decision question. These are the complex questions, the ones that need the writer to do two things (do some research or make something up) and then make a decision about it. For example, “Should I take an umbrella with me today?” is a decision question, but to come up with a reasonable answer, one really ought to first answer the research/information question “Is it likely to rain?” by listening to the weather report or looking out the window. Recognizing this is useful, because it allows one to separate the decision part (which has no right answer) from the information part (which does have a correct answer).