Wednesday 25 October 2017

What goes into a paper

I'm currently analysing data on what will probably be the final paper from the project I joined the lab to work on four years ago. This project was to look at the effect of a specific nerve lesion on behavior and function at multiple levels. We've looked at performance (how well the function is achieved), kinematics (how the structures which perform the function move), and neuromuscular physiology (when the muscles which perform the function are active). For this last paper I'm attempting to synthesize these different strands of data to understand something about mechanisms, and if they could potentially be targets for intervention.
On my end, the synthesis has taken months of lining up different data sets, verifying they are properly synchronized, figuring out discrepancies. It's been painstaking. Most days I've had between two or four spreadsheets open, generally half of which are metadata telling me how two different data sources line up. As I was reaching the end of this process, finally getting close to the dataset I needed for my analysis, it struck me that this was data we'd gathered two years ago. I'd been working on it for over a year. Furthermore, the datasets I was working with weren't raw data, they were themselves measurements of muscle activity and performance that former techs and summer students had also spent months to years working on. Cumulatively, the person hours spent on the data set I was assembling was staggering.
There is a paper in review from our lab right now whose entire results section is summarized in one very simple, very elegant graph. Six means with error bars. Those six means with error bars represent two months of caring for over twenty baby animals. Hours of understanding what our measurements meant. Hours of students and techs pouring over videos and chart recordings. Cross checks and visualizations and arguments and sick animals and broken equipment, all distilled down to one figure, six points.
My dissertation papers are the result of four years of work. I know this because I'm the only person that worked on them, and they took me four years. But if that is true, then this paper, that has in some ways also taken me four years, yet builds on the work of close to half a dozen people, is the result of two or three times that many years. And it will be maybe eight pages long, at most?
Academic papers, with their brevity and elisions, their straightforward narratives, are poor monuments to the sheer time consuming, physical work that goes into their production. The only clue in an academic paper of how many hours have been spent on it lies in the length of the author list. Those unsung middle authors are the only monument we allow to the effort that goes into our sleek, polished productions. And maybe, that is not good enough.