It has been a little over two months since leaving my job as a scientist and moving across the country, and my day-to-day routine has shifted to a casual blend of [reading | writing | fitness | gardening]. Although I am probably still in the honeymoon phase of unemployment, it has been great. I’ve experienced no remorse, anxiety, or sense of loss from leaving my job. Not only do I not miss it, I don’t even think about it any more.
Since leaving my job, I’ve read about a dozen books - with more than half about the topic of fiction writing (which I hope to try out for the next few months). Altogether, this puts me at maybe 15 books read on “the process of writing” or “the writer’s life”, etc. Some concepts keep coming up in these books, and in looking in the rearview mirror of my professional life, I’ve made a minor personal breakthrough in understanding how my mind works: I don’t necessarily think like a scientist - I think like a writer. Maybe there’s no difference, anyhow.
I’ll explain.
Part I. How I “did” science
The bulk of my life as a scientist revolved around 4 activities: 1) experiments, 2) reading academic papers, 3) writing academic papers, and 4) presentations/teaching.
1. Experiments - For me, actually doing an experiment wasn’t particularly mentally stimulating. If I was lucky, the day’s proceedings involved a mental flow state opportunity (e.g. mine often involved fine tactile skills), but it was usually some physically repetitive drudgery like pipetting solution A into solution B, shaking/centrifuging/precipitating it, then ... Solution C... etc.
The more interesting aspect of experiments was the question: “What is the next one to do?” Early on in a project (typically multi-year), the real answer to this was usually not found in The Scientific Method, but rather, was: “I don’t know, I guess I’ll try this because at least it’ll add another nugget of information that I can add to the pile and maybe that will help.” Often, doing that next experiment provided data that was ambiguous and impossible to interpret statistically. [This doesn’t stop many from drawing conclusions anyway.]
Later on in a project, after a dishearteningly large number of experiments (many of which didn’t bear any fruit at all and were dead ends one way or another), I’d start writing up the research paper (see below). In trying to write a first draft, I always discovered* a whole bunch of things I’d like to be able to state as fact (as part of building an argument), but couldn’t because I didn’t have any hard supportive data. Half the time, these were important control experiments that I should have done years ago at the beginning of the project, were I to have the foresight of what I was really looking for. Sadly, prognostication is a tricky business - in science, doubly so.
*This discovery process didn’t have to be contemporaneous with sitting at my desk with the manuscript right in front of me. More often than not, these “discoveries” occurred while biking to or from work, while going out for a run, or during a morning shower before work. More on this later.
2. Reading papers - This almost never directly helped me figure out something I was interested in. It did, however, deepen the reservoir of tips/tricks/strategies/solutions that seemed to have worked for other people and their topics of interest. (Sometimes these were false positives because not all that publishes is gold.) Reading was an exercise in filling a large toolbox with random tools that you never knew if you were going to need. Often, I would get bored with the material concerning my research focus and read very wide, consuming anything and everything I thought was interesting, taking advantage of the library’s subscription to every journal under the sun. (Nothing but paywalls for me now.
)
3. Writing academic papers - Apparently some people start writing papers knowing what they are going to say. They write very procedurally and linearly, and the final draft contains the still detectable remnants of the first, although some refinement and tweaking has clearly tightened things up a bit. I know these people exist - I have coauthored publications with some of them.
For the life of me, though, I cannot do this. I always
thought I knew what I was going to say, and I’d write it down as the first draft. Often I wouldn’t even get halfway through the first draft before something cropped up. It wasn’t because I hadn’t provided some structure or logical framework for what I wanted to say - I often outlined the main points of my arguments to serve as a scaffold for my draft. What cropped up was the messy business of making sure that what I was saying wasn’t, for lack of a better word,
bullshit. Usually this meant that I had to go do more experiments (see latter-stage Experiments above), which would get me to revisit the thinking behind the earlier experiments (and the assumptions, interpretation of that data, context, etc.).
Inevitably, in revisiting or reanalyzing my data yet again (maybe for the hundredth time), or redoing an experiment yet again (maybe with slightly different conditions), or [insert any number of steps in the process], I would discover something new or a deeper understanding about my problem that allowed me to reframe it in a much clearer, more-to-the-point way. This sounds like it should be a great turn of events, and in the end, I think it always was - from a “good science” standpoint. In practice, it was often very frustrating*, though, because I knew what it meant: another round of reanalyzing data, generating new figures, and writing a new draft that was structurally very different from the previous one. To my great relief, though, it also meant that the paper I had been writing, which I now knew to be flawed in some fundamental way (e.g. framing, logical flow, or Type 3 error) was not going to be put out there with my name on it, and that the paper I was about to write was going to be
less wrong.
Think about me repeating this process for months to years on any given project and you’ll have a decently accurate portrait of how I “did” science. After a decade or so working this process full time, I’ve yielded maybe a few hundred pages of published academic writing, as bullshit-free as I could make it.
*Consider the often conflicting dual career mandates of a “successful” scientist: one must do exceptionally good and thorough work, and also publish frequently. I always sacrificed pace of publication to ensure the quality of my results. I also probably paid the price of this sluggishness on the academic job market, too.
4. Presentations/teaching - Making the slides for a lecture or presentation involved a very similar process to the writing, but with the advantage that I always knew what I was going to say (the content of presentations was always rather formulaic). The rub with crafting a good presentation was 1) to match the level of detail and sophistication to that of the audience, and 2) to make sure time was budgeted correctly*. Because a choice picture (or rather, an schematic animation) was worth a thousand words, I spent way more time fiddling with the grouping and animation tools for Drawings in Powerpoint than I’d like to admit. Here again, I found that just making a presentation straight through was never enough. I’d make it, run through the slides, mentally churning through what I would say and how, and think about how that might play out for the average audience member. In doing this, I’d often find that big chunks of that presentation were too complicated or unnecessary or out of order. Rinse, repeat. Same process as writing a paper.
*About of a third of senior scientists, who have been giving presentations for maybe 30+ years, are clueless about both of these. Their talks are inexcusably dreadful.
Part II. What I’m learning about how fiction writers really write
Much has been written about fiction writing, and a lot has been condensed into punchy little statements that start to sound like cliches. Some of the nuggets that keep popping up:
- Realistic and compelling characters are more important than an interesting plot. The reader needs someone to root for.
- When asked whether they know how a story will end when they sit down to write it, a very large chunk of well-regarded writers say “No.”
- “Writing is rewriting”.
- The point of finishing the first draft is just to have something to work with so that you can start the writing process (which, of course, is rewriting). It is unlikely to be good reading, or more tactfully put:
The first draft of anything is shit. - Hemingway
- There tend to be two kinds of unproductive writers: 1) those that can't generate (e.g. have “writer’s block; have full novels worth of story in their heads, but can’t ever seem to write it down; or simply can’t come up with anything at all); and, 2) those that can't revise (e.g. have several incongruous and typo-ridden tomes which are instantly rejected by any and every publisher that sees them).
The image of writing-process that emerges is that writers 1) have an story idea and maybe a general outline of plot (or a basic plot arc), but 2) by the time they write it, then revise it, then revise it, then revise it, [...], and are 3) finally satisfied with it enough to send it off to the publisher or the presses, the completed manuscript may be entirely different. For example, entire chapters or hundreds of pages may have been added or axed, characters created or eliminated, or the narrative point of view may have been shifted (e.g.
Go Set a Watchman).
Why does this happen? Well, superficially you could say that it is just a function of modifying what was already there in the first draft, but a deeper explanation is that in revising the story, the author discovers new dimensions/details of the characters that they didn’t know when they first started, and those discoveries lead to consequences which are vital to the story and need to be mentioned/modified. In revising a pivotal scene where a character reacts to a traumatic event, for example, the author may realize that the character’s response isn’t quite right - it would be different, given who the character is. While it may be convenient for fitting into the previously outlined plot structure, the character just isn’t being true to themselves. Readers are very keen at sniffing this characters-behave-for-plot-convenience out, and don’t like it. Considering the primacy of character over plot, it comes off an awful lot like
bullshit. So the writer needs to the change the story to let the character be true to themselves. Which may mean a huge reorganization of the story. Rinse, repeat.
One interesting idea that I’ve come across with respect to writer’s block concerns two modes of thought: a creative, associative, generative mode and a critical, logical, deliberative mode (c.f. Kahneman’s System 1 and System 2?). The idea is writer’s block is an overdeveloped critical mind stifling the creative mind before it can get anything down on the page. The remedy, therefore, is to develop the skill of turning off/ignoring the inner critic so that the creative mind is free to meander. An exercise that routinely comes up at this point is to practice writing stream-of-conscious musings immediately after waking up. In the
hypnopompic state, your critical mind isn’t banging on all cylinders and is less intrusive. In hindsight, I was inadvertently benefitting as a scientist with similar critical-mind-silencing methods (see * in 1. Experiments).
Part III. Putting it together
Are either of these processes - how I worked as a scientist and how fiction writers produce stories - really any different?
Both are forms of iterative rumination. Not the rumination that is associated with
negative psychological outcomes, I mean rumination like what a
cow does. Here, that means treating ideas/concepts/problems like mental cud, repeatedly retrieving them from mental storage for a cerebral chewing. After a few rounds (sometimes it takes many), new discoveries or parallel associations to another concept are made, or deeper insights revealed.
This style of thinking is also reminiscent of an evolutionary process, with similar properties: it necessarily requires time but the pace of revision of the ideas can be rapid or sluggish (c.f. punctuated equilibrium). While thinking cannot be squeezed into a schedule with a guaranteed payoff, more time is probably better if it can be spared (of course, at some point, there is a tradeoff there).
I guess I could use the label “The Scientific Method” for this thinking process (“chewing the cud”), but it just feels more organic and broader than the specific practice that The Scientific Method brings to my mind. And in any case, I haven’t heard descriptions of novel writing as “scientific”, although the same problem solving algorithm and insights (perhaps more) seem to be utilized. The physicist Richard Feynman was jokingly reported to adhere to the following algorithm...
The Feynman Algorithm:
1. Write down the problem.
2. Think real hard.
3. Write down the solution.
Interesting discussion of that here.
... which is just a stripped down version of the scientific method and/or “chewing the cud”: step 2 is just acting as a black box for a
do-while loop where the different possibilities are considered from different angles. In science, that might be hypothesis testing. In literature, that might be plot device or character choice testing. [Any so on for many different applications]
Part IV. Application
I think the articulation of this thinking process, and in particular what I’ve learned from viewing it through the lens of a writer, will be helpful to me moving forward.
Where I think writers have made significant headway over the scientists in understanding the process is that writers have the framework of balance-between-the-creative-and-critical, whereas scientists do not. Editing and revision of a piece of writing (i.e. the do-while loop) require a constant seesawing between creation and analysis, and this is explicitly discussed and cultivated in writing circles. Creativity is just not being discussed in the highest science journals at all (as part of my 2. Reading Papers [above], I’ve looked through just about every issue of Science and Nature, among others, for the last decade).
...Which really brings home the point that I should work to develop a balance between creative and critical thinking modes. My creative mode has clearly been undernourished. It really is baffling to consider how many years of my formal education were devoted to developing “critical thinking” (well over 20), while I honestly can’t recall any particular instance (!) of instruction on how to develop practices to enhance creative thinking. Obviously, there is a heavy selection bias in the training of new scientists towards quantifiable analytical skills (i.e. test scores) and conscientiousness rather than creativity (sadly, “working harder” or having fancier equipment is often the solution to scientific problems vs. coming up with creative, novel ideas). But
aren't scientists supposed to be the ones who come up with the brilliant ideas to save the world? They are creatively stunted throughout their education (as, I assume, most every student is). We’re doomed.
Also, this explains why tech innovations are often unoriginal and lame. "It’s like Uber (which is like taxis...), but for helicopters!"
Also, I sense that viewing my thinking through the lens of the writing process will help me troubleshoot and avoid frustration with the challenges of each stage of the process. At conception, I should be mindful to hold back the criticism of nascent ideas and allow them time to be fully expressed. They can be culled later. First drafts don’t need to be perfect. Knowing that the solving of the problem isn’t done all at once, and that new and interesting solutions will likely develop over time if I just keep at it. Going back to the drawing board if something feels like bullshit. Etc.
I don’t think I’m breaking any new ground here intellectually. In all likelihood, I've just reinvented the wheel and there is probably a developed field for all of this... but I’m just excited to have connected some dots in my life and hopeful you might get something out of it as well. Are you a cud chewer too?