Scientific writing: Stop using esoteric operational labels in your manuscript

Stop using operational labels and esoteric short-hands in your manuscript. Keep them in your lab book – that’s where they serve their purpose – but remember they’re meaningless to everyone who’s not you or directly involved in your project.

I most often see this in atmospheric measurement manuscripts with data from several instruments. For example saying:

Sensor X39 recorded lower levels of pollution than sensor X23.

does not get across what saying:

The rural sensor recorded lower levels of pollution than the urban sensor.


See also: long plain text date strings appearing across the top of figures produced directly by computational packages. Your reader doesn’t need to know data were collected:

from 20140301_000000UTC to 20140531_115959UTC

but rather

from March through May 2014

or even simply

during Spring 2014

And on the same note: stop using the plain text labels you used in your computer program as your figure legends. You do know how to do this in your graph drawing program – right?

Instead of:



Original 20th century model
Modified 20th century model

Give your reader a fighting chance!

Steven Pinker tells me this sort of thing results from the “curse of knowledge, in combination with chunking and functional fixity”.