And yet, waiting in research is perversely productive at times. In measuring my cells’ responses to particular signaling proteins, waiting a longer time before collecting data produces cleaner results. In fitting our data to mathematical models, waiting for the computer to sample larger regions of parameter space yielded a better theoretical understanding of our system. In the year it took to perfect a robotic protocol for preparing hundreds of protein mixtures, I learned how to program and fix nearly every part of our liquid-handling robot, because I had, over the course of many weeks, seen every one of them malfunction. Much of the work I do now benefits from even longer, historical waits in science. A decades-long effort to lower the cost of sequencing DNA means I can understand the evolutionary history of the proteins I study simply by going online. Many of the Bayesian methods we use for analyzing our data are only possible because of a centuries-long wait for computers to catch up with mathematical theory.

Somehow all this history was not enough to teach grad-school me that waiting is not the emptiness before a sudden result, but a mysteriously useful period. I learned this decisively while writing the manuscript containing all the data and insights collected during my PhD. Shaping and distilling many years of work was exhausting. I had created every part of the paper at least five times and kept getting feedback like “I’m just not sure it’s clear,” or “Doesn’t this sentence seem a bit heavy to you?” or worse, “I’ve had a new idea for how to represent the data,” or worst, “Let’s have another look at the abstract.” Entire figures disappeared while new ones emerged, like continents in a sped-up version of plate tectonics. I felt like Atlas trying to hold up the world before it fell apart, or an endurance athlete completing a marathon to nowhere. I joked that the manuscript was “diffusing,” a gradual spreading that paradoxically emerges from the random walk of particles with no net movement. 

But after weeks of styling myself as a martyr to someone else’s perfectionism, I looked back at my early, beloved drafts. I was shocked to see that they were confusing and heavy; I couldn’t imagine explaining the data any other way than the one we had arrived at through the revision process. When I considered all the steps that were necessary to educate myself about the real meaning of the results, to shape the narrative, and to improve the figures, I realized it could only happen with a prolonged wait. I’ve relearned this lesson every time I make something new, whether it’s a manuscript or a hypothesis or a protocol. Anything good takes a long time to come together, and the duration of that wait isn’t measured in something like minutes or days, but something unrelated to time all together, like micro-epiphanies or the number of touches. So much of what I perceive to be true about waiting is false. What feels like emptiness contains imperceptible rotations and shifts that together produce movement. While I feel like I actively engage in this process, large parts of it are outside my control.

The lack of control is hard to accept. My first instinct for filling the emptiness is to reach for relief, like Twitter (people talk about science there!) or organizing my to-do list (meta-work with no stakes!). My second is to work longer hours, based on a bad assumption about how close the finish line is or that waiting is only about time. Neither of these tactics helps much. 

There are some more practical approaches to waiting that can be helpful, like presenting your work to clarify the story and get feedback or breaking the problem into smaller and more informative sub-problems. But lately, I have tried to “nourish” the wait. Leaving the lab to run or read fiction freshens my mind. (Though hobbies don’t exist to support work, nor are they guilty pleasures.) I think the main thing is that it should bring pleasure. I remember a friend saying he traded morning swims for reading his favorite textbook, since learning something new brought more delight than laps did. I also try to find fuel for the small and passive changes that occur while I wait, seeking out random interactions with people or ideas, or welcoming the pivots that come when someone asks, “So what are we really trying to do?” I’ve even found that being open about what I’m struggling with, even with all the social and professional costs of confessing my neediness, has often been the only way out of the forest. I used to think I should only talk to someone about my project once it started working when, in reality, those conversations are most useful during or even before the problems “start.”