Writing and Reflecting in Obsidian — Writing Fiction

This is part 5 of a series about how I write and reflect using Obsidian, an extensible digital note-taking interface with some surprising and unexpectedly useful features. This post reviews the process I use to turn my ideas into prose for both short- and long-form fiction. I’ve been using Obsidian since 2021-02-06, and it has completely revolutionized my writing process. New posts every Wednesday until the series is complete.

Planning and incubating my fiction is all well and good, but how does that translate to actual words on a page? While the majority of my focus is finishing that one novel I’ve been struggling to pass like a kidney stone for past decade, I’ve also been experimenting with short stories recently, especially since Obsidian enabled me to carry more of my ideas to fruition. The process I use is slightly different for both my short and long fiction, but they follow the same ideological practice of breaking down a large, seemingly impossible task, like “Write a story,” into smaller, more manageable chunks, like “Identify the problem or crisis the protagonist faces.” I’ll also highlight the similarities between my short- and long-form fiction where relevant.

The Drafting Process

The process I use for writing short fiction (under 10,000 words) is easier to explain, and builds off of a lot of the stuff I talked about last time. Once I’ve finished my brainstorming, such that everything I need to know about my story is contained within one story worksheet, I have a set of maybe two dozen tasks to build the story out. This will include things like identifying the protagonist’s weakness and need, defining the problem they have to solve, identifying obstacles that prevent them from finding a solution, and working out their moral and psychological self-revelations.

One of those tasks is to build and edit a list of story events. These could be understood as scenes in a short piece or chapters in a longer piece, but I try to confine them to the MICE story quotient(Milieu, Inquiry, Character, Event) as outlined by Mary Robinette Kowal. I make each story event in the list a backlink to a file in Obsidian that doesn’t exist yet so that my story worksheet can serve as a linked outline to the constituent components of the story I’ll be creating later.

Story events, planned for analog drafting, alongside the “Word Sprint” plugin

I also write down this list of story events in my analog notebook so I can write them out—by hand—three to six pages at a time. As you probably know by now, I’m a huge fan of morning pages, and I’m so habituated to writing in three-page stretches that I know exactly what I’m signing up for when I sit down to write. I do all of my principal drafting this way—there’s something about having a physical pen on physical paper that keeps me focused, creative, and productive when I’m writing, and I haven’t had any success writing first drafts any other way.

This also lets me do an initial editing pass right when I transcribe the handwritten pages into their respective story events notes. If I read the pages out loud as I type them up, I can immediately identify turns of phrase that don’t quite work, awkward sentence constructions, and issues with flow that might not be apparent on a purely digital review. (Pro tip: I also embed references to my notebook pages as YAML metadata, such as notebook: XI.58, so I can quickly return to the analog originals when I need to and even see all my analog writing in one place using Dataview.)

Those are real coffee stains from that time I spilled a cup of coffee on this notebook

Once those are all typed up (and remember, they’re linked together via my outline from before,) it’s time to create a “New Manuscript Draft,” which uses QuickAdd and Templater to prepopulate the new note with certain metadata about status, date, and linked project, as Jamie Todd Rubin outlines in this excellent blog post. (Rubin’s site is a hell of a rabbit hole to drop into, and I can’t say enough good things about how much I learned from his “Practically Paperless with Obsidian” series.)


Stephen King recommends stepping away from the first draft for at least six weeks before returning for major revisions, and I’ve found that to be pretty good advice. Try to revise too soon, and you’ll have too much of what you wanted the story to be in your mind to be able to clearly, objectively, and courageously the line of words that actually ended up on the page. As Annie Dillard puts it in The Writing Life:

The line of words is a hammer. You hammer against the walls of your house. You tap the walls, lightly, everywhere. After giving many years’ attention to these things, you know what to listen for. Some of the walls are bearing walls; they have to stay, or everything will fall down. Other walls can go with impunity; you can hear the difference. Unfortunately, it is often a bearing wall that has to go. It cannot be helped. There is only one solution, which appalls you, but there it is. Knock it out. Duck.

Courage utterly opposes the bold hope that this is such fine stuff the work needs it, or the world. Courage, exhausted, stands on bare reality: this writing weakens the work. You must demolish the work and start over. You can save some of the sentences, like bricks. It will be a miracle if you can save some of the paragraphs, no matter how excellent in themselves or hard-won. You can waste a year worrying about it, or you can get it over with now. (Are you a woman, or a mouse?)

Annie Dillard, “The Writing Life”

When I do an editing pass, I’m often reading the prose (aloud) directly in Obsidian, making changes as I go or dropping inline comments with the double percentage signs %% that markdown treats as invisibles. These disappear in “Reading” mode, so I can try to see it as a reader might.

When I’ve cleared all of the comments, I’ll create another “New Manuscript Draft” and manually retype the prose into a second draft. That may sound like a lot of extra work, but I’ve often found that the only way to catch typos is by combing through each line, word-by-word, and reading aloud while transcribing. I don’t have to read aloud loudly, but it’s that extra step of hyper-literalism that lets me catch slips and missteps before my readers do.


Longer pieces are only slightly more complicated versions of the same process. I make heavy use of both the Folder Note and companion Waypoint plugins to automatically generate linked tables of contents for each folder of my writing, as one way of keeping track of drafts. Waypoint automatically updates with the contents of the given folder, so as I create (or delete) story events, it keeps me up to date.

I have not been writing much in this project today, in favor of completing a certain blog post instead

The other plugin I use is Longform which does a fairly decent job at keeping the structure of my longer projects organized. Longform lets me reorder the scenes as needed, while Waypoint goes with the file hierarchy implicit in the folder, and it does its best to keep up when I add new scenes. My favorite feature of Longform is that it lets me set daily writing goals and keeps track of the overall project progress. It’s not quite on par with Scrivener’s ability to automatically calculate daily word count targets based on projected whole project word counts, but seeing the progress bar fill up is motivating enough without all that.

For the most part, the procedure is still the same—I have a Folder Note for the long-form project that contains links to all the individual story events, and I assign myself tasks to write them out, first by hand, and then manually transcribed into their respective notes.


Of course, the canvas plugin changed everything. Now, I can keep track of not only every individual scene, but I can also experiment with different mind map-style configurations of the scenes. This is something Scrivener’s corkboard could never do—it only ever gave me one instance of each scene, and it couldn’t show me nest groupings. Canvas lets me create whatever crazy groups I want, and have multiple instances of the same note allowing me to create almost a choose-your-own-adventure approach to restructuring my novel, which I didn’t even know I was lacking until I had it.

It can render a given canvas as an image, but there was so much prose on my ETM canvas that I had to choose the “Privacy mode” option to convert the text to lines just so my machine could successfully export it. You can see some of my planning and alternate scene sequences below:

Everything’s Totally Manageable


What about when I’m done? Well, this is a sticking point where Obsidian still needs some work. Exporting the finished story works pretty much the same for both short-form and long-form fiction, in that I have a Scrivener project that exists for the sole purpose of compiling a Markdown file into the PDF or DOC format I use to send it out for critique, query, or whatever. I even have compile settings for ebooks or print. That’s the one thing I wish Obsidian were better at, or I wish there was a cleaner process for. I’m not good with PANDOC, I haven’t figured out how to get it to work yet, and so export still requires external tools. But I’m already using Drafts for quick entry most of the time, so there’s no need to pretend to be a purist about it.

Then What?

The next step in my fiction-writing process involves submitting my work for critique. My group meets for an hour every week to discuss the submitting author’s piece, and I have a system of note-taking that allows me to not only collect and collate the feedback I receive, but also link back to it when I go through my editing passes. Check back next week for how I manage critique and submissions using Obsidian.


[…] week, I’ll discuss the missing link between having a story outline and actually getting the prose on the page. It’s simpler than you might think, but that’s not to say it’s easy. Writing is […]