There is a strong incentive in science to do something new. Novelty gets funded, gets published and gets talked about. Repetition gets none of those things, and it is repetition that tells you whether anything is actually changing.
In 2013 a colleague handed me a clipboard and a laminated sheet and told me to swim a line. I have swum some version of that line, on some version of that sheet, every year since. It is the single most useful thing I do.
Three properties that make data last
- It is comparable. The method has not changed, even when better methods appeared.
- It is findable. Everything is deposited, versioned and openly licensed within ninety days.
- It is legible. Somebody who has never met us can reproduce a figure from the raw files.
The second point is where most projects fail. Data that lives on a departmental drive is data that will be gone within one staff turnover. We treat archiving as part of the fieldwork, not as an administrative afterthought.
“A ten-year record collected with a mediocre method beats a one-year record collected with a brilliant one. Consistency is a form of resolution.”
The cost of changing your mind
Twice we have been tempted to upgrade the protocol — once for imaging, once for a genetic method that would have replaced two days of counting. Both times we ran the new method alongside the old one for three full seasons before considering the switch. One of them we adopted. The other we did not, and the record is intact either way.
None of this makes for a good expedition film. It is, however, the part of the work I expect to still matter in thirty years.
