Work. Finish. Publish.

A couple of days ago, John Scalzi posted a writing advice open thread, asking people to share the best advice they'd gotten on the craft of writing. There's a lot of good stuff in there, much of it fairly specific to fiction writing-- stuff about plotting, the use of synonyms for "said," how to keep track of who's speaking, etc. As someone who's very much an outsider to that side of the writing business, it's interesting to read, but not that directly useful (I do have long stretches of dialogue in the How-to-Teach books, and occasionally needed to worry about the "said" thing there, but that's about it). I'm sure there's a lot of non-fiction writing advice that would seem similarly interesting-but-useless to the fiction crowd.

What's striking, though, is the universality of one particular line of advice, which shows up in lots of pithy forms through that thread. I'm going to repurpose a great quote from Michael Faraday, though, who was once asked what the secret to success as a scientist was, and replied:

"The secret is comprised in three words— Work, Finish, Publish."

The "work" in Faraday's quote is something distinct from the writing that goes into the "publish" part, but you can move the whole thing into the writing portion, sort of like one of those recursive-acronym jokes ("What does the 'B' in 'Benoit B. Mandlebrot' stand for?"). The issues are basically the same whether you're talking about research-leading-to-writing, or just writing.

Work. This is the absolute most essential thing. Whether you're discovering fundamental principles of magnetism or writing a book where you explain science to a talking dog, you need to do the work. Not talk about doing the work, not fiddle with calendar apps and word-processing programs-- do the work. If you're doing science, get into the lab and do some science; if you're writing, park your butt in front of your favorite writing implement, and write.

It's not just that you can't make significant discoveries or produce great books without work, it's also that if you don't work at it, you will never get better. That's true of science, where time in the lab helps you hone new skills and develop expertise that will be useful later, and it's true of writing, where you get better with practice. Even if your early work is crap, you won't be able to produce great work later unless you work.

This is the part of the process where most people fall down. And even people who are successful go through periods where they fail at this step-- God knows, I've had plenty of stretches where I spend more time clicking aimlessly through blogs or playing stupid mobile-phone games than I do working on research or writing. Stuff only really gets done when I consciously decide to do the work.

In a weird way, SteelyKid and The Pip have actually been helpful with this-- the productive activity event horizon they've introduced to Chateau Steelypips has forced me to be a lot more disciplined about my time-- I block out an hour or so every morning after dropping The Pip off at day care when I go to Starbucks and do work on my projects, with no class prep or chair crap allowed, and that has, surprisingly, turned into a really productive block of time for me. I don't recommend having kids just for this purpose-- though there are lots of other good reasons-- but by whatever means, you need to lock up some time as Work time, if you're ever going to get anything done.

Finish. One of the great things about science is that it's never really over. There are still people noodling around with scientific issues that were known in Faraday's time, finding new wrinkles in old problems, and I think that's really inspiring.

It's also a bit of a trap, through, in that there will always be some new thing you can futz around with, and some tiny lingering question you could investigate more. Once you start poking at something in science, there are an endless series of rabbit holes to fall down, all filled with wonderful stuff to explore.

Given that, how can "Finish" be sensible advice? Because there's a point where you have to just declare that you're done. You're never going to wrap everything up completely, and some loose threads will always have to be left dangling for the next person to pick up. Or to form the start of your next project. At some point, though, you have to be finished, or you'll never get anything done.

This is a hard problem, one I've struggled with from both sides. As a post-doc, I was the guy saying "Look, we are finished with this. I'm not re-doing the analysis from scratch with a new paradigm. We can do that for the next paper." As a junior faculty member, I caught myself doing the opposite, saying "I just need to sort out this one more thing..." about a project that was perfectly well publishable as it was at that moment. Recognizing the point when you're Finished is a tricky thing, but essential to success.

The same is true of writing-- there are always a couple of typos, or phrases that are a tiny bit awkward and could be re-written one more time. And there are places you could add asides, or objections that you haven't fully anticipated, or branches of the story you could go off and explore, and a hundred other ways to avoid Finishing. And at some point, you need to Just. Stop. You can never chase down every little lead, fix every little problem. At some point, you're Finished, and need to move on to the next thing.

This is another tricky point to identify. We had a running joke at NIST that the paper-writing process was only over when the latest round of changes was just reversing edits made in the previous round, and that's one indicator. Another good one is when you find yourself saying "Oh, God, I can't stand reading this crap one more time..." At that point, declare it Finished, and move on.

Publish. This is, obviously, critical to academic success, in that you need to have publications if you want to get credit for discovering stuff, and get a job. And it carries over to writing as well-- you're not a success until stuff you wrote appears. In either case, you're never going to be a success that unless you send it off to somebody to publish.

And, as with the Work step, the act of sending it off to others can be essential whether or not it leads to immediate success. Getting editorial feedback is necessary to improve in both science and writing, because that's the only way to find out how your work appears outside of your own head. Just sending an article off to a publication might not bring immediate success, but it will get you feedback that will help you be successful down the road.

This is one of those areas where I've been phenomenally lucky-- as I said over the weekend, I feel a little guilty when reading a lot of writing advice, because I've never needed to hustle the way a lot of folks have. I have a great day job that pays the bills, and the opportunities I've had to publish have mostly come to me, through an utterly bizarre lightning-in-a-bottle process.

But still, this is important advice to heed: your results don't mean anything unless you publish them. And this is another place where people fall down-- they have something all done, but fear either rejection or just hassle, and so good scientific results sit in a (metaphorical) drawer somewhere. There's really no excuse for that in the age of blogs and the arxiv, though-- even if you can't face the slings and arrows of Reviewer 2, you can throw up a blog post or an arxiv preprint with minimal difficulty. And the same sort of thing is true of writing-- if you don't want to deal with pitching stuff to editors (and have some sort of way to keep food on the table without doing that), put it on a blog, or self-publish. It doesn't take much effort, and can sometimes let you do an end run around the hassle-filled parts of publishing (as my publishing career and to some extent Scalzi's can testify...).

I've seen lots of people fail on each of Faraday's steps. I know people who don't have Ph.D.'s because they could never bring themselves to buckle down and do the work, or declare an end to their research and write stuff up. I know people who have been "writing" books for years without ever putting fingers to keyboard, and others who have endlessly expanding manuscripts that they just can't seem to stop noodling around with. And I know people who have done the work, and declared it Finished, but can't be bothered to send it off and deal with referee reports, so good work languishes unread.

So, whether you're a scientist or a writer, Faraday had the core of it: Work. Finish. Publish. They're not always sufficient for great success, but they are unquestionably necessary. Fail to do any of them, and you have no chance of succeeding.

Of course, even better than Faraday's advice might be this comment from Theyis, late in Scalzi's thread:

My advice: Don’t read a column like this just before you write. You will focus more on all the advice than on the writing.

Which is another way of saying "Get back to work." And once you're finished, publish it.

(It's also worth a callback to Harry Connolly's advice that I was talking about back on the weekend, specifically the point where he says "Nothing bad happens if you give up." If you really, truly, can't bring yourself to do one of these steps, that doesn't mean you're a Bad Person, it just means you would rather be doing something else. If you're too busy blogging to finish your thesis research, you probably ought to get out of research and become a writer. If you like the idea of being a writer, but never manage to put fingers to keyboard, you're probably not going to be a writer, and should focus on your day job.)