Stop starting from scratch when you write
ianharris.com
ianharris.com
I used to sort of solve this problem by writing it in my head over the course of a day or three and then spitting out some basic language to start things off.
Lately I've come to the realization that almost any level of first draft, no matter how awful or off the mark, is sufficient for me to dive right in and finish writing it quickly. So my go-to tactic is to verbally run down what I need written to someone else, like an assistant or someone who works for me who knows the topic, and then start with whatever they come up with and go from there, effectively eliminating the blank screen problem.
The basic premise of this is insightful, there's really no reason why writing has to start with an empty page. I think that's the more interesting observation than the specifics of how he gets the ball rolling, which could easily vary from person to person.
Dan Harmon (the creator of Community, and Rick and Morty) has phrased it as "prove you suck". From his reddit AMA (https://www.reddit.com/r/IAmA/comments/5azrwi/im_dan_harmon_...) (paragraph breaks added; nothing edited):
> ...if you simply change your goal from "writing well" to "writing badly," you will be a veritable fucking fountain of material, because guess what, man, we don't like to admit it, because we're raised to think lack of confidence is synonymous with paralysis, but, let's just be honest with ourselves and each other: we can only hope to be good writers.
> We can only ever hope and wish that will ever happen, that's a bird in the bush. The one in the hand is: we suck. We are terrified we suck, and that terror is oppressive and pervasive because we can VERY WELL see the possibility that we suck. We are well acquainted with it. We know how we suck like the backs of our shitty, untalented hands. We could write a fucking book on how bad a book would be if we just wrote one instead of sitting at a desk scratching our dumb heads trying to figure out how, by some miracle, the next thing we type is going to be brilliant.
> It isn't going to be brilliant. You stink. Prove it. It will go faster. And then, after you write something incredibly shitty in about six hours, it's no problem making it better in passes, because in addition to being absolutely untalented, you are also a mean, petty CRITIC. You know how you suck and you know how everything sucks and when you see something that sucks, you know exactly how to fix it, because you're an asshole.
> So that is my advice about getting unblocked. Switch from team "I will one day write something good" to team "I have no choice but to write a piece of shit" and then take off your "bad writer" hat and replace it with a "petty critic" hat and go to town on that poor hack's draft and that's your second draft. Fifteen drafts later, or whenever someone paying you starts yelling at you, who knows, maybe the piece of shit will be good enough or maybe everyone in the world will turn out to be so hopelessly stupid that they think bad things are good and in any case, you get to spend so much less time at a keyboard and so much more at a bar where you really belong because medicine because childhood trauma because the Supreme Court didn't make abortion an option until your unwanted ass was in its third trimester.
One of the super fun and satisfying things of being a competent musician is how quickly I can create basic components for my compositions.
That said, musicians tend to have a mental library of ideas and techniques they can employ to flesh out an inspired idea.
I also have a library of partially finished unused themes and ideas that I wrote, which I can bring in to start or expand a piece I'm working on.
So a library of my own partially finished writings would be far more analogous to how I work than a bunch of anecdotes and quotes from other people.
The anecdotes and quotes would be more analogous to other people's music which I've heard and internalized, but would never literally reference while I am composing. The last thing I want is those sorts of cut and paste elements to come anywhere near my work.
For example when you "create" a drum pattern are you really "creating" it or are you simply remembering a drum pattern you've heard or played before that works well with the song or composition you are making. As you probably know brand new drum patterns are rare. Most of the time the specific instrument choices change but the pattern itself is familiar.
I'm not familiar with the parent poster's music, but I am pretty sure that "create basic components" refers specifically to that "copy-and-paste" from a concept into a more tangible form (like a drum loop). So it seems like they have considered this idea explicitly.
I'm not a pro, though I have to play a 3-hour solo gig tonight on piano, and most of it will be improvising ornaments over blues and jazz tunes that I already know.
One of the problems in this gig is that it becomes very easy to fall into rather generic patterns.
That repetition is a real short-coming of a lot of so-called original music that I, and many musicians I have heard and work with run into: a lot of original music is a bad copy of something that already exists.
However, in the process of re-implementing existing stuff (as opposed to consciously reworking a tune, re-performing a piece, or literally resampling audio) I tend to make mistakes, and I find that to be a good source of actually 'new' music.
I can't say for sure if my understanding is correct, but that's how I understand the parent poster's motivation.
Of course, I'm not a "composer" and I often have to create a whole lot of music without a lot of prep, so I am perhapshappier to have a large-ish library of known things in my brain that I can directly quote when I play.
The parent was stating that they don't do the first option, and prefer the second choice. For example, if he/she wanted to make a drum and bass track, they might write their own version of the amen break [1], rather than use a sample from someone else
A hip-hop beat from a certain time period, for example, has certain characteristic quntization and accent patterns that makes it recognizable as from a certain genre.
So my brain hopefully has a huge library of meta-constructs and analysis that I can draw upon to make elements with a controlled amount of familiarity.
Yes, similar in some ways to copy and paste, but really more like understanding meta-templates.
Ideally. :-)
edit
Some people are incredible at recreating things they heard accurately. They can do stuff I can't even pretend to attempt. That isn't my forte. Even if I try to recreate something I heard precisely, there is huge error in the reproduction process and I end up with what I would have done. A Dorian melody becomes harmonic minor. A hi hat pattern becomes a shaker with twice as many hits. Etc.
Well, I dunno. The limitation is that most audiences have little or no context for this kind of work.
On one hand, the audience that is aware of the form usually has little tolerance for messing with the form outside of kitschy appropriations.
Think about how a bluegrass band can play purple rain in their idiom, but if they replaced the upright bass with a synth their audiences would (and I say this from experience) probably not be very interested.
On the other hand, the kind of audience that is okay with that kind of recontextualization usually hasn't invested a lot of time in learning event the generic conventions.
Like, when I didn't know how to play cello, I didn't really appreciate how cheesy and plain most pop music was, until I tried to find some arrangements of pop music for string quartet and realized that a lot of elton john stuff arranged for string quartet is super boring musically...
prior to playing cello I would have thought it was kind of cool to change contexts, now I just want to play Bach as well as I can pull it off.
"Some people are incredible at recreating things they heard accurately."
IME, this is a learned and not inborn, though I think intelligence helps. The combination of having to quickly apply those 'constructs' to specific music in performance help, as does expanding number of constructs....
at some point most folks musicians playing by ear just thinking some version of "okay, that's plastic jesus, melody and all, done really slow, but then the bridge is the stormy monday turn around".
Paul McCartney dreamt the melody for 'Yesterday'. He was sure he'd heard it before, so he went around for weeks humming the tune to people, before deciding it was original. (I just found this story in my swipe file!)
The tools now are so powerful to create new sounds - Figure by Propellerhead, Garageband, etc - that it seems just dishonest to argue sampling is a "necessary" function of art. Maybe it was for the Beastie Boys, but not for some up and comer with even basic access to a phone. Really, it's that low-level and accessible now.
Fundamentally I dislike people claiming that jacking a sample, adding an MPC hip-hop beat over the top, and then rapping / adding vocals is an impressive accomplishment. There's an artistic process in there, sure, but it's built on the skill of others fundamentally. Not a popular opinion, but I'm also able to compose without lifting other people's stuff to start with...because, well, 20+ years of practice and effort is a lot more investment than buying a turntable, records, MPC & a mic.
I think that's missing an important point (which is also a point that TFA doesn't cover). Competent people do what makes them productive. If you start worrying about what specific things "the best" do, you're going to end up with a process that works very well for somebody else, and that very possibly won't work for you. You'll optimize for status, but you won't get near your potential.
The idea in TFA seems to be clearly worth trying if you have the problem it purports to solve. Maybe it works for you, maybe it doesn't. Maybe it doesn't, but it gives you more data on how your brain works, and helps to come up with something that works better for you.
But don't let anyone convince you that because you use this method or that, you're less competent than someone who uses other methods. Results are what matter.
(Edit to add: That doesn't mean you shouldn't pay attention to what competent people are doing. They are competent for a reason. But you shouldn't be discouraged if their techniques don't work for you, or if different techniques do.)
A writer should absolutely have some kind of structure. Writing a novel, here are 4 Scrivener files that encode structure. Writing an essay, here are some structures.
But this author seems to think that the way to cash out the "don't start from scratch" is to collect quotes. I think this is not the best use of your time (you can google quotes on topics easily). Secondly, its not the largest challenge when writing, and thirdly, it sells short his own insight which I think can be taken further.
If any other HN reader has any ideas on how to "not start from scratch" in various kind of writing, I would be interested.
One way is a library of (I called it quotes, you called it stories. I think the right term is) anecdotes, which you gave. I think of that as just an example of a tool set that we can prepare.
But I think there might be other ways. I thought of "structures" off the top of my head, but surely there are even other ways of compiling a library of components so you don't start from scratch.
A library of metaphors?
A library of insights? (I suppose this is why people research and read)
Is a pattern
Argument-> counter argument -> Conclusion?
Or is it
Statement -> Illustration -> Amusing anecdote -> Insight
Or what?
Does anyone know of a comprehensive list of the basic elements in writing?
I don't disagree, but it seems like a minor point.
The piece opens with this part about starting from scratch and the blank page. It goes through these other arts that take what they want and piece them together (graphics, music, programming).
And then the solution is to keep quotes in your back pocket? Ones that's the "type of feeling you're trying to capture"?
It might be useful to take notes (quotes, ideas, data) when you research, but you're still "starting from scratch".
“In quoting others, we cite ourselves.” ― Julio Cortázar, Around the Day in Eighty Worlds
I have it on github, but if you like it, I would recommend rewriting it from scratch because I wrote this in the "Over-Engineering" phase of my career many years ago, over-using design patterns and abstracting away things to the point of making them incomprehensible. Also, I'm only maintaining it for my personal use now, but feel free to clone and make your own if you dare:
https://github.com/ideonexus/memexplex
Another great resource I would love to see made into something I could query for this is the public domain book "Fifteen Thousand Useful Phrases" by Grenville Kleiser.
Do you have any ideas around what tools for editing might look like/do for you?
One useful editing technique is reading your work aloud. But it's slow and annoying to do, so I'd like to see a word processor with visualization of the predicted sound of the words. There are already many good audio visualization techniques, but for full effectiveness this needs another tricky problem solved: speech synthesis with natural prosody. Again, this is an ongoing area of research where machine learning could be very useful.
I'd also like some kind of refactoring support for moving around blocks of text without breaking dependencies between them, eg. accidentally moving a paragraph where something is discussed to earlier in the piece than the paragraph where it's defined. Automatic detection of these kinds of dependencies would be ideal.
Then, your first task when starting a new document can be to delete the irrelevant stuff from the template, if there is any. Next, I like to just start writing bullet points in whatever order/style they come to me in. Once you've done this, organize them in a coherent way that will be reflected in the finished product, cull the points you don't really want, fill in conceptual gaps, and think about whether/how you want to break the document explicitly into sections. Then, finally, start writing, expanding bullet points into prose.
Throughout this process, another important rule applies: if at any point something comes into your head that you actually want to write, even if its a random idea that you know will be be buried deep in the draft, write it immediately. You can work it into the rest of the draft later, once the rest of the draft actually exists.
I revel in the joy of starting from scratch!
This approach works for some people, but creativity and writing are so personal that it's good to have different ways of working.
"As long as you can start, you are all right. The juice will come."
I've found the advice to be very helpful for fiction writing.
[1] https://www.theparisreview.org/interviews/4825/ernest-heming...
Somewhat ironically he explains how he does it.. https://youtu.be/UJgslhoG--M?t=3m21s
As for the blank-paper-scare, my go to tool is the [Most Dangerous Writing App](http://www.themostdangerouswritingapp.com) - for the times you can't get started, it's a good push down the hill. (I'm not affiliated with them in any way.)
> Look at graphic designers, for example. If you’ve ever watched a graphic designer work, most of the backgrounds, icons and badges they use come from stock websites
> An efficient designer will normally buy a funnel diagram template rather than make one from scratch. For $5, why reinvent the wheel?
That being said, the general advice is solid. I'd add two more pieces:
1. Three-sentence drafts are your friend. When you think of something that might make an interesting blog topic down the line, write three sentences about it and let it germinate for a bit. (I've found Trello to be great for this, since you can easily add notes and links.)
2. Blog your email. (This advice is stolen blatantly from patio11.) Especially if you're writing to do content marketing, surfacing customer support + answers to interesting questions is a great way to get a jump start.
The point I'm making is that other creative processes let you start with something to build on. Maybe writing should be that way too.
As a amateur song writer, i try to write snippets of lyrics down as they come. Whenever I'm trying to finish something but can't think of anything suitable, I have my rhyme book with tons of stuff to refer to. In some cases it doesn't even have to be quotes from what I've already written down that I might end up using. It could be word/rhyme/rhythm patterns that i pick up from material that I have available.
So yeah I concur with the title of the article. :)
There are plenty of other more, dare I say mature ways of approaching writing that don't resort to, ahem, swiping other people's shit. Bullet points. Outlines. This example is not a good one in my opinion, and should be considered useful at best for Amateurs intent on developing their own skills over time.
a) The design of the block has been perfected through years other people's work and experience. If I need a spur gear in a design there's almost certainly an off-the-shelf component that's 99% perfect.
b) I'm in a rush and I need something good enough. Slap something down and move on.
In the context of writing it seems like the latter will pop up more often. Did his examples seem trite to anyone else?
If there’s already code that works, a programmer would be insane for attempting to recreate it.
Is it just me, or is "throw it away and rewrite from scratch, get it right this time" so incredibly common that it's a programming trope in itself? Along with "programmers want to write from scratch because it's fun, nobody wants to do maintenance" ?
Indexed notes are kind of like my own private library.
Also a point-form plan can help to keep your thoughts organized. So plan + voice, and only then move to the computer keyboard.
It said you should write or edit but not both at the same time. Writing is supposed to be more on the right side of the brain, editing on the left side so they don't mix well together.
When I cared more about my annual corporate review, for example, I would jot down things I was proud of so I could hype myself up and stand up for myself should I get a sub-par review or raise.
Like many things in life, you learn by doing.
In fairness, it's a neat-sounding technique and I plan to give it a try.