Metaphors in scientific writing: vagueness and confusion, or something better?

One of the foremost virtues – probably the foremost virtue – of good scientific writing is clarity. (I make this argument at length in The Scientist’s Guide to Writing.)  In service of this goal, writers are often advised to be cautious about using metaphors (see Olson, Arroyo-Santos, and Vergara-Silva (2019), A User’s Guide to Metaphors in Ecology and Evolution, for an interesting analysis of metaphors and their strengths and weaknesses). A metaphor names or describes one thing by referring to another, and our literature is studded with them*: the tree of life, the Big Bang, electron shells, biological invasions, chaperonins, tectonic plates – and these are just a few that came to my mind right away.

Critics of metaphors often rest their case on three major grounds. Metaphors are held to be vague, to be misleading, or to resist cultural translation. Do these grounds hold water?**

Butterflies, mustard seeds, and misplaced critical thinking

I’ve just read Graham Moore’s The Last Days of Night, a novel based on the battle between Thomas Edison and George Westinghouse over AC vs DC electrification in the 1880s. This was a fascinating story*, but I’m taking off from it on a tangent today. The epigraph for Chapter 23 is a quote attributed to the architect and futurist Buckminster Fuller: “There is nothing in a caterpillar that tells you it’s going to be a butterfly”. Now, given my cringeworthy memories of what I was like in high school, I should be 100% behind this quote. The problem: as an entomologist, Buckminster Fuller was an excellent architect.

The quote, you see, is nonsense.