Process Imagineering, Part 2: Plans, Plumbing, and Porcelain

Okay. Great. So we hold status meetings for our writing stuff. Same as everyone else who works in an office. Gee, thanks Josh. Truly, this is revolutionary stuff.


We will certainly take the co-authoring world by storm.

My point in talking through our jam sessions (status meetings) was mostly to introduce the idea of continuously clearing snags for the other person. But there are other reasons. Chief among them: to plan work. That is, come up with the stuff you plan to work on. That way, when each of you go off and attend to your own devices, you’ll both have a clear idea of what you’re supposed to be working on.


We probably had, oh, I don’t know… a year? We started outlining mid-2012, but didn’t start principal writing until late 2013. I’ll say there was at least a year of jam sessions where all we talked about was the book’s thesis, the characters, elements of the worldbuilding, and what the hell we would call certain things. Again, if you’re a solo author, all of the decisions are up to you. If you are not, you have to be clear with your intent otherwise you end up going in the wrong direction, or worse, you end up duplicating some amount of work.

Some authors are what they call ‘pantsers’, by which they mean that they ‘fly by the seat of their pants’ when writing a book. I find this wholly alien and bizarre, but that’s just who I am. And while a perfectly reasonable way to crank out a book, it would be completely untenable when two authors are in the mix. If you have two pantsers, you will write two completely different books.

That should be the end goal of the repeated jam sessions: to ensure that you are not writing two different books.

Those first years of jam sessions were spent working our way through the following:

  1. Thesis. Establishing the core questions and concepts that the story should ask/convey. Sort of like a mission statement, but less lame. Find a thesis, or set of theses, that you both agree on, or else you will be writing different books. Do not proceed unless you have a set of theses you both agree on. You will revisit these repeatedly.
  2. Character. Establish the characters that will support your thesis. Get as detailed as you need so that both people feel comfortable writing in their voice. This might involve writing some prototypical scenes to see if a character trait works. Iterate until you have a cast of characters you both feel give you the tools you need. Do not proceed until you have those characters.
  3. Outline. Establish the story outline. Start broad, then get detailed. We had a numbered list of things that we wanted to happen in a book. Things that had to happen. We broke that list down by chapter, and had a list of things that would happen in each chapter. The outline forced us to choose when and where they happened. Both authors should know what each chapter should be about and what thematic and emotional tones should be present. Stuff can always change, but having an outline you can both look at and agree upon is very important. Do not proceed until you have a chapter outline.

There are lots of resources out there on how to approach thesis, character, or outline. And better than I can relate here. More what I am after is getting them down, and agreed upon, so there is no confusion about the core elements to your story.

Once you have both agreed, you’ll have to start in on planning the actual chapter contents (environments, dialog, scenes). It’s not enough to just agree on a thesis and go. You will have to collaborate on the chapter content as well. I found that the best way to convey planned/intended chapter content was to hash out either:

  1. Storyboard. Establish storyboards. It’s relatively easy to write an outline. It’s harder to actually block out locations, dialogue, character movement, and scene progression, keeping all of the variables in your head and making them logically consistent.

    Write out moment by moment storyboards as if you were going to film the thing. That way, both of you can have the same image in your head of what is supposed to happen.

    You can choose to storyboard first, or you can choose to write your dialog first.

  2. Dialog. For me (Josh), cranking out dialog is the best way to lock down what you want from a scene. I try to write dialog with the intent that if you took everything away (narrative, exposition, thought bubbles, etc.) that the dialog would tell the same story. It’s hard to write good dialog. But it’s easy to see bad dialog. Finding things for the characters to discuss, making their conversations believable, and their intentions relatable is core to making someone care about what you write. Everything else is window dressing.

There’s nothing that says you can’t storyboard and write dialog at the same time. The visual elements and character environments tend to inform the dialog, and the dialog informs how characters emote and interact with their environment. Forcing your dialog into a visual storyboard also helps so that you don’t just have talking heads reading their lines back and forth. Characters should never (read: rarely) just be talking.

So, at this point, you are probably regretting getting this far in reading this. I’m telling you what you already know. Like your disappointment in the first paragraph, my pushing of the food around on the writing process dinner plate hasn’t been all that informative. Why have an entire article about planning when you could easily summarize it as “Have a plan, and if not, make one?”

Well. For me, it’s the difference between plumbing and porcelain.


Please. Don’t go. Let me explain. I’ve got something for this.

Mike and I use a number of tools, but at the very core we use Git. Git is a piece of software that is a version control system (VCS). Its lot in life is to track a directory of files, keep past versions, and let us see differences between old and new.

This sounds pretty simple. But among the popular VCS products, Git is pretty complicated and obtuse. It is famous for having hundreds of obscure commands that, while very helpful and powerful in some situations, usually require some degree of Googling to find what you want. Eventually, you get used to the pain. Or you just write more software to make it easier.

There are so many commands that they divide them into two categories: plumbing commands, and porcelain commands. They use the different terms to describe the lower-level, background functions of Git (the plumbing) versus the more human-friendly, human-readable interfaces for Git (the porcelain). The programmers of Git, in their infinite wisdom and wry sense of humor, decided that the best analog for describing the difference between these types of functions was the concept of waterworks in your house. Because toilet humor.

The idea is this: you generally don’t deal with your plumbing every day (unless you need to make modifications or fix something). But you will definitely use your fixtures (your sink, your bath, your probably porcelain toilet) often.

Git plumbing functions are intended to have well-formed interfaces whose format will (hopefully) never change. Output is reported such that another computer program can interact with the interface and get what it needs with little fuss. This makes plumbing functions ideal for consumption by and interaction with other software, given that they are easy to program against and have some guarantee that the interface will not break in future versions.

Git porcelain functions have no such guarantee. Their interfaces are intended for consumption by humans. Fickle, capricious humans, that speak of love and beauty and want computers to know what they want without having to describe what they mean.

So. What does the concept of plumbing and porcelain have to do with planning and the writing process?

Well. The planning process is the plumbing. The writing is the porcelain.

Planning is making sure everything is mapped, and that the right pipes are going to the right places. Your theses need to be present in every chapter, and your characters and their scenes need to be able to convey them. You want a pipe and a drain in every room that needs water, the right valves, and both of you agreeing on where the sink is.

On the other hand, the writing is where all of those pipes meet fixtures, and ideas meet readers. Where your porcelain meets the… hands? Because they’re porcelain knobs… Nevermind. The writing-porcelain is sentence structure, the word choice, the meter, rhythm, and balance. It is the handle you twist to get the hot or cold you want. It is the interface that makes sure all of those things are put together such that a human will read it, understand it, and want to read more. And as you grow as writers, you will eventually say to yourself, “Man. I don’t know why we chose dark gray for the toilet. And why is the tank 6 feet off the ground?” The porcelain is relatively easy to change, and will probably be swapped more than the plumbing it connects to (provided you’ve got good plumbing).

The plumbing and the porcelain of your writing must inform the other. When it comes to good storytelling, neither should exist in a vacuum. But there is a pretty strict order of operations. Plumbing comes first. Because you should absolutely be done with the plumbing before you start putting in any toilets.

You might say, “You started in 2012, but you didn’t start writing until, like, late 2013? What gives? Just write. Figure it all out later.”

It’s because it doesn’t how matter how nice a toilet, nobody should sit on it unless it will flush.