I love Game of Thrones. I particularly liked this mini-speech from Petyr Baelish earlier in Season 7:
Don’t fight in the North or the South. Fight every battle everywhere, always, in your mind. Everyone is your enemy, everyone is your friend. Every possible series of events is happening all at once. Live that way and nothing will surprise you. Everything that happens will be something that you’ve seen before.
As I let my mind wander away from work for a bit today, I realized that this is a wonderful quote about science!
Fight every battle everywhere, always, in your mind.
Baelish is a shrewd, obsessive planner. In planning for everything that could possibly happen, he always seems to be prepared, get what he wants, and stay alive. Just like staying alive in Westeros in positions of power, doing good science can be quite difficult because we are set adrift in extremely complex systems. There are so many paths that that can be followed to answer a research question (including the formulation of the question itself!), and all could be the subject of an intellectual battle with a critic. Studying the effect of yearly bonuses for teachers on long-term student outcomes? How do you define long-term? What outcomes will you measure and how? How will you prevent dropout? How do you make differing bonus amounts comparable for teachers who differ in terms of what they teach, where they live, what composition of students they teach from year to year? How would you even define “composition of students”? Also why study student outcomes as opposed to community outcomes? There is no way that a single study could address all of these concerns, or the ones that I couldn’t think of, but these concerns need to be thought about because they need to be answered for us to have any hope of meaningful, actionable conclusions. Just the act of forecasting these hypothetical intellectual battles can motivate the design of better studies.
Everyone is your enemy, everyone is your friend.
Baelish is calculating and knows how effective people can be in various contexts. It can be helpful to think of everyone in the scientific community as your enemy—enemies ready to question every aspect of your work and find every possible hole—but only if it indeed motivates you to do those very things. Only by heavily scrutinizing our own work can we make ourselves the best scientists possible. Acknowledging limitations in private and subsequently making them known to others is key to moving the state of knowledge forward. I do want to de-emphasize any paranoid or hateful connotations of this quote though! Some people in my field take the “everyone is your enemy” part too seriously and critique others in inflammatory ways.
Now the friends part…this probably works out better for science than for Baelish. Scientists form a community, and ideally sharing everything about our work would facilitate a team effort to find even more limitations and address them fruitfully to get leaps and bounds closer to useful answers. But my impression is that things generally don’t happen this way. Groups work somewhat in isolation on different aspects of a problem. Perhaps consortia try to harmonize efforts in some respects, but useful information is still needed from external sources and not able to be integrated easily. Just as it is hard in Westeros to find good allies, it can be difficult in science to find good collaborators. But when it does happen, great deeds are in the works.
Every possible series of events is happening all at once.
In Westeros, livelihoods dance on the whims of nobles and on the breath of armies that can be traded with coin coffers or decimated in an afternoon. This creates a palpable urgency for Baelish to always stay ahead of the game. This immediacy isn’t really felt in science. We don’t gamble with our lives when we submit a paper and wait for the review process to unfold. I think that the scientific community is lured to progress slowly with our recognition system favoring large numbers of publications. Researchers who have large projects are incentivized to break the project up into several publications. I don’t think this is necessarily bad if the scientists have actually completed this larger body of research. The flaw I see is if it incentivizes scientists to publish work that isn’t as complete out of time pressure and fail to follow up with more complete validation because they feel the validation work isn’t “enough” for its own publication. There is some effort to recognize replication attempts, but I wish that there were more urgency and incentive to conduct more complete studies the first time around because it sets the baseline higher for future work.