Deconstructing "Three Nights" · 2:01am Feb 9th, 2014
I've spent a lot of time talking about writing in the abstract; not so much time talking about particulars. One of the nice things about kicking out a new story—especially one that took me about two months to write—is that it gives me a chance to look back at the process of writing, at places where I made good or bad decisions, at things that worked and at things that didn't.
So how did "Three Nights" do? At the time I'm writing this blog post, it sits at 182 upvotes and 3 downvotes (on just over 800 story views). It spent about a day in the feature box and got all the way up into the #2 position. I think by most standards, that's probably pretty good. For that matter, from a Fimfiction success standpoint, that's probably just about the best I could hope for. After all, this was an out-of-season holiday story starring a character many readers don't particularly like, and it's over 17,000 words long. Okay then, it did fairly well.
But do I count it as a success? That's a harder question, and I'd have to say... yes and no.
I've heard people—probably Orson Scott Card, back when he was an author and not a crazy Archie Bunker parody—say that a writer has to be capable of both loving their work enough to sell it and hating it enough to rip it up and make it better. That's about how I feel with "Three Nights". It took me two months to write, and that was almost entirely because of problems I created for myself early on. So I suppose the best place to start is to ask, what did I set out to write in the first place?
Back at the start of December last year, when I was settling myself back into the fandom, I thought, "Wouldn't it be nice to write a Hearth's Warming story for Christmas about how friendship means none of us are ever really alone at that time of year?" because yeah, Christmas can be a pretty terrible time for some people. Then slowly, the shape of something started to form in my head. Cadance was an obvious character choice because (1) in the semi-canon chapter books, she's established as a foundling and (2) she spends a good chunk of her time in the show being rather isolated and disconnected from other ponies, too. I thought it would be fun to do an interconnecting set of vignettes from Cadance's life, culminating in her actually having a happy Hearth's Warming after other ponies shared the magic of friendship with her.
I really want to defend my process here, but the truth is, my process was terrible. I didn't think hard enough about what I wanted to do, and it cost me a lot in the end. That's the downside. The upside is that I feel like, perhaps more than with anything else I've written for this fandom, I really learned a lot trying to put this story together. And that's where I'd like to go with this post—what went wrong in trying to make this story happen, and what can be learned from it. I'll organize the rest of this post by lessons. Note that there will be plot spoilers below this point.
Stories Need Conflict
This was the number one problem I ran into with this story, in the end. Chapters 1 and 2 were relatively easy to write. They were short, and I was mostly just doing character sketches. That's something I'm comfortable with. Chapter 1 really doesn't have any conflict. Chapter 2 has Cadance trying to be a good foalsitter, which is variously a conflict with Twilight and a conflict with herself, but it's never very strongly outlined.
Chapter 3 took me more than a month to write, and a lot of that was because I didn't know what to do with it. I knew where I wanted to start: Cadance is alone in the Crystal Empire, with Shining and Twilight absent on Hearth's Warming Eve. I knew where I wanted to end: Cadance is reunited with Shining and Twilight and a bucketfull of other ponies, proving that you don't need to be with family for family holidays to be a success.
You may notice that the story doesn't really end on that note anymore.
The problem was, in its original formulation, Chapter 3 was just going to be Cadance moping for however many words it took for the storm to clear and other characters to come rescue her from her own loneliness. In case it's not patently obvious to everyone reading, this is a terrible idea. For one thing, Cadance needs to be active to be a sympathetic character—she can't just be an observer in her own story. For another, it doesn't give her even a shred of a journey to take as a character. If she just sits around and mopes for a while until somebody shows up to save her, she learns nothing and experiences no growth. Clearly, this wasn't going to work.
My first attempt at fixing this problem was to treat the symptom rather than the disease. I latched onto the idea that characters needed to be active, and decided I'd make Cadance do a bunch of stuff to try to make her Hearth's Warming experience better, and then in the end... magic would happen, and her problem would solve itself and she still wouldn't grow any. I wasn't seeing the deeper issue with what I'd decided to do. This first fix was what got written into the first draft of Chapter 3, or at least the first 2/3 of that draft. At that point I sat down for a bit with GhostOfHeraclitus and he very nicely pointed out to me that my ideas for ending the story were crap. So I started to figure out what my issues were, slowly, and checked out a bunch of Bad Horse's recent blog posts. I latched onto a few things he said about story, thinking they were excellent ideas, and finished the last 1/3 of that draft. But I still didn't have very firm ideas of the conflict in Chapter 3, or even in the whole story. I'll come back to this.
Themes Are Hard, Symbols Are Harder
I mentioned that I was trying to essentially write a story with a message here. It's okay to be lonely at Christmas Hearth's Warming because you can't be with your family, since friends are like the family we make for ourselves. Having a message meant trying to work with theme.
I've mentioned many times in blog posts that I think all writing should be directed toward four general targets: character, plot, setting, and theme. The more of these targets you can hit with a scene, or a paragraph, or even a well-chosen word, the better. Theme, though, is not something I'd tried working with much before "Three Nights". I think you can write pretty good pieces that don't deal with all four targets, especially on the smaller scale. "Princess Luna Likes Coffee" is just over a thousand words, pure character exploration, and to this day I love it. "Purple Prose" is basically character and plot (pun not intended) without much concern for setting outside of Twilight's intentionally-overdone self insert romance story (though it does contain one of my favorite 'thing-that-actually-means-another-thing's, the apple Twilight chops up in the second scene). My other three stories play around with character, plot, and setting, but don't really dig down into theme much if at all.
So I wanted to try to work with theme more here. There are two big repeated motifs running through "Three Nights", occasionally because I made conscious choices, occasionally because I think my subconscious picked up on what I was doing and threw the right stimuli my way while I was writing. Those two motifs are family and isolation/connection—which may be painfully obvious and not even bear mentioning. I don't really know how heavy-handed I was.
No, I take that back. I was obviously pretty heavy-handed with family. For transition purposes, though, I'll talk about that second. Let's start with isolation/connection. This is the motif I think I did a much better job with, though some of that may be accidental. Cadance starts every chapter isolated and ends every chapter connected with other ponies. Chapter 1 is too short to really play around with this much except in the narrative arc. Chapter 2 starts to introduce the first of my two attempts at symbolism: Cadance's wings. Cadance as a poor flyer is a bit of a fandom trope (and now semi-confirmed with the show), and I thought that would be a nice way to explore her character and her sense of isolation as a pegasus raised among earth ponies. When she gets uncomfortable in "Three Nights", I almost always make reference to her wings or her feathers. The painting of the pegasi in Twilight's house is explicitly there to plant this symbol in the story and let Cadance think about her own pegasus nature.
The symbol gets paid off to some extent in Chapter 3, with a couple lines in Cadance's conversation with Cinnabar and her flight to the train station. I don't know that it gets paid off well, mind, but I wanted to give some culmination to the use of that symbol. And generally speaking, I think Chapter 3 is where the isolation/connection motif works best, with the juxtaposition of the empty train station and the bustling castle kitchen, the reaction Larimar gets from his foals vs. the reaction he gets from adults (though I know this could be done better), the whole idea of the blizzard, and the fact that all the ponies who create stronger bonds in the story (up to the end, anyway), do so by venturing out into the isolation of the storm. I think I could probably have done better with this, but I don't think I really flubbed it either.
How about family? The first chapter is all about Cadance finding a family. The third chapter has Cadance removed from her family, and a brother-and-sister pair of OCs, and a little family in the caroling scene, and a whole heap of orphans. Oh God, orphans. I can't believe I wrote a Christmas story with so many orphans. It's like a badly written love letter to Charles Dickens and the industrial revolution. Now I know how Skywriter felt with "Beloved".
Note that I'm not talking about how I used this running family motif. For what I was doing in this story, I think using them productively may have been beyond my skill level. I was really more concerned with getting it on the page than figuring out what it was doing. My thought was that all I really needed to do was expose the reader to it a lot and keep it in the reader's mind, rather than try to do any more than that. And (in my serious lack of wisdom on this issue) I still think that's kind of a reasonable approach, though it can be bettered. I'm not a fan of narrative forcing, and I don't think I'd much like thematic forcing either. I'd rather just have the motif there for readers to explore in their own thoughts. But if you're going to go that route, I think you need to make a conscious effort to try to explore it yourself—look at the motif from a number of different directions, and try to give the reader a chance to think about things they might not otherwise think about and form some new opinions. I didn't make any conscious effort to explore the idea of family in Chapter 1 or most of Chapter 3, I just threw ideas at the wall and hoped they'd stick.
Note that I'm also not talking about how I used family in Chapter 2. That brings me to my last lesson.
Kill Your Darlings
And this... This is where I failed. A lot.
The first draft of the story ended in a different way: with Cadance taking Shining aside to tell him about her history of bad Hearth's Warming experiences, and noticing that Octavia (never explicitly named) was gone from the train station. I have a fuller discussion of Octavia and the madness that was my decision to use her as a symbol in the comments to the story proper. Suffice it to say that I latched onto her as a symbol for Cadance's dead parents, and the whole thing made so much sense in my head, and then my pre-readers got ahold of the story—and to a person, everyone was completely baffled by how I ended the thing. Here I had this wonderful, wonderful symbol of Cadance letting go of her misplaced feelings of guilt at her parents' deaths, and nobody got it.
I also received feedback from Skywriter about the whole story after the first time he read it, and while that feedback had all sorts of useful and wonderful things in it, the bit that really jumped out at me was, "Why do you have Chapter 2?" In the first draft, Chapter 2 was wholly divorced from Chapters 1 and 3 both in tone and in narrative arc. It just didn't fit. Skywriter had some great suggestions about changes I could make to the story, and a prominent one involved dropping Chapter 2 and retooling Chapter 3 to make Larimar more significant and have Cadance hash out her parent issues with him, since after all, nopony's going to have more perspective on lost parents than Larimar.
Between the first and second (final) drafts, I had about a week of intense school work that needed tackling, so I spent a lot of time thinking about the central conflict (or lack thereof) in "Three Nights" and what I really wanted Cadance to do. And I realized, awfully late in the process, that I couldn't define a conflict here. I couldn't pin down what I wanted Cadance to learn or accomplish to end the story anymore. I'd had an idea in mind at the beginning—good or not, I'd had one—but it had always been weak and in the process of writing, it had managed to slip away entirely.
In an ideal world, and if I were a better writer, I would have just put the whole thing aside for a couple months to let it sit. But I was very conscious of the fact that this was a Christmas Hearth's Warming story a month and a half late, and that I was increasingly getting distracted from my graduate work by the need to complete it. I really couldn't afford to let it hang around a whole lot longer on my workbench.
I tried figuring out the changes I'd need to make to do what Skywriter suggested, and it didn't work. I'd either have to add tremendous amounts of fluff to move my narrative where it needed to go (and make the whole of Chapter 3 much duller), or I'd need to gut large parts of it and rework both the choice of scenes and the characterization on Cinnabar. Neither of these options appealed to me (though I think Skywriter's suggested course of sticking with the parents, and maybe figuring out how to deliver the Octavia symbol well, might have resulted in the best overall story).
I gave up. I asked GhostOfHeraclitus where he thought things should go. He gave me a different opinion, and I looked at that for a while, and I realized that didn't really wrap up the story in my own mind; he was viewing Cadance differently from me. I spent a lot of time thinking, and I found many new and exciting roads to nowhere.
Eventually, I figured out that the closest I'd come to building a journey for Cadance was through the family motif, only I hadn't really explored it as well as I'd liked. So I went back and retooled Chapter 2, at least, to add the argument at the beginning. The first draft had just provided a brief cameo for Twilight's parents on their way out of the house. This served to recenter the Twilight story a little, and connect her initial isolation to the family atmosphere. It wasn't a great fix, but it helped bring the tone more in line with the other chapters and gave the whole thing a bit more connection to Chapters 1 and 3. Then I went into Chapter 3 and made some modifications to the ending, and to the dialogue in Cadance's bedroom. I also added a short flashback to deal with the letter a couple thousand words earlier in the story (originally, it didn't come up again until that bedroom dialogue), and some more direct use of Octavia as a symbol for Cadance's parents, although I'd worked that bit out of the new ending.
The truth is, I should have been a lot more willing to cut things when I went back into this story, after the problems that cropped up in the initial draft. Octavia really isn't doing much anymore, but I loved her as a symbol too much to cut her. Similarly, Chapter 2 really wasn't serving its purpose with the rest of the story, but I really, really liked it! If I hadn't felt so rushed to get this off my plate and free myself up to deal with school, I don't know what I would have wound up doing. But what I did was look for the best possible set of changes that would let me keep the things I wanted in my story, for the most part.
That's not artistically honest, and I'm sorry.
Like I said before, I'm decently happy with the results, but I feel like "Three Nights" definitely could have been better. The big thing for me, though, is that I feel like I learned a lot about proper planning, about trying to explore themes, and about being willing to sacrifice the things you like for the greater good of a story. Those are lessons I'm going to try very hard to keep in mind as I move forward to working on other stuff. They're also going to be helpful lessons, I think, for another new project—but more on that tomorrow.
One last thing I'd like to mention, I got a great comment from Scramblers & Shadows about the emotional inconsistency in Chapter 3 of "Three Nights". I said I'd try to talk about that some in this blog post, but I never quite saw a good opportunity to hit on it directly. The comment is worth a look, though. I think I agree with the analysis, and I expect you all can see how trenchant it could be here, given the number of things I wound up trying to do with the chapter and the lack of a clear organizing plan to help me approach them. Chapter 3 got plotted out a number of times, and each for somewhat different purposes. There was the "let's make Cadance more active" approach, the theme and symbol approach, the basic "what will make for a decent narrative" approach, and whatever spot revisions I found necessary in my final pass to try to make the whole thing move more decisively toward the ending I wound up with.
tl;dr, plan your stories better before you start writing them.
I really love reading stuff like this, it is fascinating to look at other people's processes and how they work through stories. A lot of my stories end up half written by my subconscious and half by suggestions from Jordanis, who I should probably credit as co-auhor on just about everything at this point, but I've been making a little bit more of an effort to take them apart and think about them lately, and seeing other authors take apart theirs is very helpful.
I gotta agree here. She was a non-entity at best, and a distraction at worst. She wasn't really a symbol anymore, and I could not for the life of me figure out what she was doing there. I disagree with your assessment of chapter 2 though. It basically established her connection with Twilight and Shining, and seemed pretty important in Cadance establishing family ties outside her parents. I think the problem was in chapter 3, to be honest. It kind of got away from you, I think. But it's chapter 3 that would need retooling, not dropping chapter 2.
As for the orphans, that could have gone straight into glurge territory, but I think you handled it fairly well, mainly because the focus was on Larimar and Cinnabar, and not the actually reasonably happy orphans.
I enjoyed the story and glad you finally got it out. I have to say that I think it's good to get out a story, even flawed, and use the lessons on the next one. As opposed to simply trying to write and rewrite the story over and over and never putting it out.
Oh, so that's what the Ocavia thing was!
I was kind of confused about that.
Ah symbolism; and theme for that matter. The darlings of English teachers everywhere. They go round with their "Thou much hath symbolism in thy story. Thou mush haff a theme." It's logical. All their favourite novels probably have symbolism and themes ergo to make a good story, you need to put in symbolism and themes.
Well, 'tis true with the theme thing. Although I always find the word "theme" so horribly vague I try and avoid it. For the purposes of producing literature, rather than analysing it, I prefer making sure I know the answer to the question "why the heck am I even writing this story?"
As for symbolism, for me anyway, its something that happens by accident. Now I have managed to use symbolism, even deliberately, but it's very much a case of it almost happening by accident. I decide what happens without worrying to much. Luck? Not quite. I try and be on the look out for opportunities.
Kind of like the way real people make jokes that fit the natural flow of the conversation vs telling forced, stock "why did the chicken cross the road?" type jokes. They don't force the conversation to go in a particular direction, they just seen an opportunity and take it.
For example, say I'm writing a chapter about a particular day and stuff happens and it finishes up with a conversation in which the mane character loses hope. That conversation, logically, happens at the end of the day.
Now lets say the mane [1] character is Celestia. Ooohh.
The conversation could happen at sunset. Let's do that and put a little extra emphasis on that fact. The conversation happens at the end of the day because that's when it happens. But by paying attention, it can become symbolism too.
[1] I like pony puns too much.
I've been planning a Fallout Equestria spinoff story for the better part of a year and I haven't started actually writing it, at least not really. I think I overdid your advice there, dude.
1814324
That was how both the wing thing and the Octavia thing started. I suppose that's probably the smart thing to do going forward, too—to not try and force the issue but see what may come up organically and stress that in editing. If I don't make a hash of the basic structure to begin with, that'd probably still work fine.
1815239
Brandon Sanderson, my go-to guy for fiction advice, suggests that underplanning and overplanning are the two big traps new writers often fall into. I've got another story I may have overplanned, but I think generally underplanning is the big one for me. The opposite is certainly an issue for a lot of people, though.
1816119
If it helps, most of that time was spent just thinking of the story and talking about it. I've only recently started to make an actual written outline.