Hold off on the research

I get hung up on research all the time.

I’m writing along when I suddenly need to check up something. It’ll only take a minute, I say to myself. It’s not as if I might be using research as an excuse to avoid thinking through a hard problem. Right?

When next I look away from my laptop, bleary-eyed, hungry, hours have passed. I look back at the screen and realize I’ve been reading some obscure, pointless page on Wikipedia.

What was I looking for again?

Research is a trap

Even if I avoid running down all the rabbit trails the internet has on offer, by the time I turn back to my text editor I’ve lost the momentum. I struggle to get back to that point where, by conjuring more or less the appropriate words in more or less the correct order, typing becomes writing.

If checking some facts during writing can be so distracting, extensive research – poring over articles and books, possibly in an actual library – is one of the most delicious ways to procrastinate a project into oblivion.

The solution is to keep working. Put a placeholder to remind you where the research will fit in and keep writing. Finish the first draft and then do the research.

You thus avoid the cognitive penalty of switching from a writing context to a research context and back again. Moreover, questions often dissolve or change shape as a text develops. At the end of the first draft, you’ll be in a much better position to look back and determine what research is required.


For the longest time I indicated a passage requiring attention with a triple dollar sign ($$$), which is both vague and ugly. It’s better to use subtler markup. If a placeholder stands out too much against the rest of the text, I find it more irritating than helpful.

Wrapping a passage in whatever symbol is not used by your markup language generally works fine. In Markdown I use curly brackets ({}):

An article {that I read last week} claims that...

This may be hard to search for if you have code blocks which include such brackets.

CriticMarkup suggests a syntax for writing comments that is unlikely to interfere with code blocks, though it’s both cumbersome to write and an eyesore:

The article {>> look up in Pocket <<} claims that...

While adding research notes is useful, it’s better to keep the text as free of redundant markup and to write actual copy instead of reminders.

Brief placeholders

For names, short facts, dates or numbers, I use the editorial trick of replacing the unknown words with “TK”:

On July TK three men found themselves on a boat.
The project will cost between $TK and TK.

When you finish the draft and complete your research you can drop in the fact-checked information where appropriate.

“TK” rarely appears in English words, making it easy to locate after you’re done writing. It doesn’t come up that often in code, either.

Make sure that, given the context, what the “TK” stands for is obvious; if explanation is necessary, it’s better to add a note.

Extended passages

Dealing with longer passages is more complicated because the missing information may impact the shape of the text.

The best trick I’ve found is to not use a placeholder at all. Go ahead and write the passage at your current level of knowledge.

Instead of breaking the flow of your text and your writing to create a note about the research, the dodgy passage itself will serve as a placeholder. On subsequent revisions, it will stick out sufficiently due to the drop in detail and sophistication. If for some reason you’re afraid you might miss it, tack a note in markup. I like using mock tags:

{revise} ... {/revise}

Writing an extended passage before you’ve completed your research can sometimes sound thoroughly silly. Don’t worry about it.

I picked this trick up from Brandon Sanderson, who described drafting a novel about a combat surgeon by resorting to such highly-detailed descriptions of field medicine as: “He made things better”. (If memory serves, Sanderson discusses this in his creative writing lectures, but I can’t source it precisely.)

While it’s easier to work with this approach in fiction, the point is not to make stuff up; it’s to write using whatever you already know.

It may seem counter-intuitive, because you have to spend time and effort writing sentences only to later cut or heavily revise them. But if you’re prone to using research as a distraction, this approach will probably save you both time and sanity. Sanity, especially, is hard to come by.

The challenge at the early stage of writing is to figure out what to say, not how to best say it. “First pants,” as Gary Larson famously put it, “then your shoes.”

Finish the draft

When I hit a rough patch in my writing it’s attractive to imagine that there’s a better, easier path available a quick search away. There isn’t. No one has ever solved the particular problem I’m called to solve. Doing research may turn out to be a component of that solution. But in that moment I’m incapable of judging whether a spot of research will turn out to be essential or tangential. Often it’s just an an excuse to goof off.

While switching to research is of dubious value, ploughing through the rough patch never is. Not if I ever intend to finish the draft. It’s in this predicament that research becomes procrastination. Research can provide the ammunition that will turn a fight. But it won’t fight in your stead.

Next time you’re tempted to stop writing and do research, hold off. Put a placeholder if you need to and keep writing using whatever you already know. Later, tomorrow, in a week, you can hit the books. They will all still be there and you will have a complete draft to guide you.

Comments? Tweet or email me.