into the depths of google drive
on revising a dark and drowning tide
welcome to the ursprung, the section of my newsletter where I write short essays about how my adult debut, a dark and drowning tide, came to be. if you’re only interested in book updates, you can unsubscribe from all sections except “allison saft” here.
First off, thank you all so much for your very lovely response to the last essay in this series. I’m equal parts glad and sorry to learn my experience resonates with so many of you. May we all prevail over the books that want us dead.
Since my last newsletter focused on the emotions re: drafting and revising A Dark and Drowning Tide, this time, I thought I’d offer you something more concrete. What was I doing during each round of revision, beyond general wailing and gnashing of teeth? To answer this question, I took a dive into my Google Drive, which contains every single draft of A Dark and Drowning Tide. Most of these drafts also happened to be created in Google Drive. I write in Google Docs. :( I know, ok? I’m aware it lags. That’s why I split the document into four parts, one for each act. Anyway.
I expected that opening this folder would dredge up some bad feelings, but it was surprisingly kind of fun. It’s been about a year since I last read the book in full, so time has healed most wounds. It’s fascinating to see what edits were suggested at what point, how I approached them, and what I was working on until the very end. Without further ado, here’s a little (non-spoilery) glimpse at each draft of A Dark and Drowning Tide: the feedback I got, the changes I made, and some stats.
draft zero
There’s not a whole lot to say about this draft, honestly; I did not want to look at it too closely due to some lingering, vague sense of shame. We’re not totally over the pain, after all!! A ‘zero draft’ is a pretty standard part of my process; basically, I’ll write the most bare-bones version of the book, which amounts to dialogue, action, and a very rare inspired passage with real emotion and description. Basically: a very long outline. I don’t enjoy drafting, so this gives me both momentum and some confidence that my external plot hangs together somewhat logically. I don’t like to waste time polishing scenes that are going to get cut early on. The zero draft is just for me; no one else looks at it.
This zero draft clocked in at 35,000 words, and took me a little less than a month to complete. Interestingly, the first chapter does resemble the final chapter, in that the same things happen. I remember struggling to pin down the first scene, so I must have gone back and forth with my agents on some sample pages before I set out to write this draft.
the “real” first draft
Once I finish my zero draft, I read through it, jotting down my ideas on how to expand the skeletal just-for-me draft into something other people can read and give feedback on. Usually, I’ll write myself an editorial letter, with sections for plot, character, worldbuilding, and whatnot. Nothing sums up my mental state better than the “revision notes” document for this first draft of A Dark and Drowning Tide. It contains two items:
Flesh out side characters
Act 3… woof
Very helpful! The only other documents in this folder are a bunch of character sheets, where I was trying to figure out who the side characters were. Besides Lorelei and Sylvia, there are four other prominent characters; the size of the cast was very much out of my comfort zone at the time. Both Down Comes the Night and A Far Wilder Magic are “two people alone in a big house” books, so having six people to juggle overwhelmed me. It doesn’t look I filled out these character sheets in any great depth; I only articulated what each character’s relationship to the theme was.
What strikes me about this draft is that the external plot is already in place; nothing structural has changed from this version to the published version. I suspect this is what made the revision process so miserable! Down Comes the Night was also a nightmare to revise1, BUT the characters were always crystal clear in my mind. Fixing that book was a matter of putting them in different situations until the plot finally clicked into place. With this book, the characters remained very elusive and vague. So much of my time and energy was rewriting the minute details of each scene to try to make them come to life. I would just pick the wrong details every time.
Stats and other facts:
88,000 words
1,400 words of scrap material (basically: passages I didn’t want to delete and would maybe reuse elsewhere. I’m sure there were like 20,000+ words deleted and rewritten in each of these drafts…)
The book was called Vespertine
Part 2 is called “The Lake in the Sky,” which is admittedly kind of silly. In the final draft, part 2 is called “Blood in the Water.” Part 3 is called “The Haunted Isle.” In the final, part 3 is called “The Vanishing Isle.”
Lorelei is a naturalist, not a folklorist, in this draft.
Lorelei was not chosen to be Albrecht’s co-pilot! However, in the opening chapters, she is very confident she will be chosen. When she’s not, she spends some time scheming about how to sabotage Sylvia, which is actually kind of funny. RIP.
Adelheid has a totally different personality. She’s kind of a fretful wet blanket here.
another folder in my first draft folder entitled “second draft” (?)
My revision notes document says “Big Bad Emotions” with no context or explanation. I do not know what this means. Otherwise, it seems I was debating killing off another character. This must have been a final touch-up of the first draft before I sent it off to critique partners.
Stats:
90,000 words
129 words of scrap material
a different folder called “second draft”
This is the real second draft, where I implemented some feedback I’d gotten. In this folder, I have editorial letters a whole bunch of critique partners, as well as from my agents. There is also a revision plan. Just to give y’all a sense of how I approach a massive revision like this:
I compile all the notes I receive into one document and divide the notes into sections (character, plot, worldbuilding, etc.).
I reverse outline the book in an Excel document—basically, I list out each scene. I am not making changes; I’m only recording what exists in the draft I’m about to revise.
I come up with a plan, usually by brainstorming solutions on my own and talking to friends about those ideas. Once I feel good about the vision, I re-outline the book according to how I think it should be. In this new scene-by-scene outline, I’m I essentially describing the book I’m aiming for. This is what my current revision outline template looks like:
After that, I dive in! I go chapter by chapter and try to tackle everything at once. Once I get to the end, I’ll do a final read-through and address any lingering issues.
This round of revisions was probably the most labor-intensive and the most emotionally painful. Here’s a summary of the comments/questions I received:
expand Lorelei’s backstory and motivations
flesh out secondary characters
bring emotion to the forefront, as it’s currently being drowned out by plot
the plot is there but loose and disorganized - clarifying motivations and everyone’s stake in the success of the expedition will help
Swap Sylvia and Lorelei’s roles (Lorelei should be the folklorist and Sylvia should be the naturalist)
There should be a plan to find the Ursprung; without one, the expedition feels very directionless at the beginning
there is too much dialogue :)
The main problem of course, is the side characters! The bane of my existence!!! I got a lot of great advice. One friend suggested giving Lorelei a “solo mission” with each character—and to introduce them one at a time. Another friend suggested that I add the mara-riding scene (IYKYK), which ended up being one of my favorites in the book.
Stats and other facts:
95,000 words
8,700 words of scrap material
The book is now called “Untitled.” I think Margaret Rogerson’s Vespertine was announced or published around the time I was working on this draft.
third (fourth?) draft (the one that went on sub)
After getting my agents’ feedback, I got two completely fresh readers, both of whom have a lot of experience with group casts. They gave me some great notes on fine-tuning the worldbuilding—but also on developing the full cast. They wanted me to focus on the villain in particular. They also suggested I make another character weirder, because that character was fading into the background of every scene the entire cast was in (side note: never write a scene where six people have to be talking. It’s a nightmare). One suggestion—one that really changed my thinking on how to implement these changes—was to have Lorelei approach the cast as folktales she needs to dissect and catalog. I thought it was a great idea, so at this point, I assigned each character a fairytale that Lorelei would associate them with.
Stats and other facts:
94,000 words
60 words of scrap material - just a few images I liked and hoped to preserve. Alas!
This book was still untitled until like, the day before it went on sub. I had a working title of “Deep, Drowning Waters” (lol)
round one of developmental edits
The book sold—hooray! This next revision was undertaken with only notes from my editor, partly because I’d already tapped everyone I could think of to read, partly because I now had deadlines. In this folder, there is a reverse outline; a character map so that I could analyze each character’s relationship to the others; and a timeline of pre-book events.
One of the biggest changes to the book’s external plot happened in this draft. Until now, there was an enemy nation that was threatening to invade Brunnestaad (where the book is set). This is why the king wants the power of the magical spring: to fend off his enemies. In this draft, I decided to make the big threat come from within Brunnestaad’s borders: the king fears one of his confidants will overthrow him if he does not solidify his power. It felt less complicated—and more thematically true.
Beyond that, I continued refining the worldbuilding and developed the Ursprung’s mythology. There are lots of little adapted or invented fairytales that Lorelei tells about the Ursprung scattered throughout the text, which all made it into the final draft. I also added more magical creature encounters and more texture and detail to Lorelei’s backstory. There is also one key scene around the 70% mark that wasn’t quite working. My editor and I went back and forth on this one a few times to work out the logistics and ensure it was as scary/tense as it could be.
Stats and other facts:
105,000 words
250 words of scrap material
My editor asked me to track changes, so:
Insertions: 2,870
Deletions: 1,921
Moves: 44
Comments: 320
round two of developmental edits
I knew this would be the last chance I had to make big changes, and I still had a lingering feeling that something wasn’t 100% right. So, I called in the final boss of my critique partners, who has never met a bookish problem of mine she can’t solve. Between the two of them, I finally got the book where I wanted it to be! Here are the main changes I made during this last major round of revisions:
Clarified the new politics, since it was such a big change in the last draft
Did some fine-tuning of the murder mystery plotline to ensure the murderer’s motives and actions made sense
I think THE note that unlocked the book was my friend pointing out that Lorelei doesn’t really investigate the murder until halfway through. She suggested I have Lorelei come up with some suspects and start poking into things sooner. That gave the plot some drive it was missing.
More tweaks to the group dynamics. Until this point, everyone kind of openly hated King Wilhelm. Now, they all love him complicatedly.
Small adjustments to each scene to incorporate murder mystery investigations and refine the shifting alliances and long-held grudges of this petty friend group
Adding small details to backstory, relationships, and lore
Stats and other facts:
105,000 words
240 words of cut material
finalizing the text
From here, the book went through line edits, copyedits, pass pages, and cold queries—basically, a bunch of different opportunities to get into the weeds of the prose. The final version of A Dark and Drowning Tide is about 110,000 words, which is the shortest of my books. I cannot wait for you to read it!!!!
Thank you so much for coming on this little trip down memory lane with me! If there’s any part of the publishing process (or my writing process) you have questions about, feel free to ask in the comments.
This will likely be the last essay in this series, unless I’m feeling up for a post-publication retrospective (ha… ha). But I can promise at least one more installment: some bonus content! It’ll be a letter. There may or may not be poetry in it.
Lastly, A Dark and Drowning Tide is out NEXT WEEK. Some final housekeeping:
I’m doing a few events and hope to meet some of y’all! Some events are ticketed (including Kepler’s and Capital Books), so be sure to register ahead of time.
If you’d like some art and a signed bookplate, you can enter the preorder campaign until September 17.
You can still order signed copies from Kepler’s, which will come with two exclusive prints in addition to everything below.
I promised myself I would never write another mystery after Down Comes the Night. I clearly did a fantastic job keeping that promise.
So fascinating to read this!! Thank you for sharing your reflections with us
I really loved reading this email. I’m an aspiring author who is also currently trying to revise a draft that just doesn’t seem to be working no matter what I do so it’s nice to know that these kind of struggles happen to already established authors. It’s definitely given me some hope that I can figure this book out. I also really like your reverse outline technique and I think I’m going to try that myself!