Noah Brier | October 12, 2022

The Lumpers & Splitters Edition

On labels, evolution, and the various approaches to building taxonomies

Hey everyone, just a reminder that we’ve moved Wednesday’s edition to be subscriber-only. We have some fun ideas for things to do with WITI, but part of that will be about building up our base of subscribers. As part of being a paid subscriber, you now have access to WITI Subscriber’s Discord, which you should sign up for if you haven’t (you’ll need to verify your sub to get access to the channels). Thanks for reading and subscribing! - Noah & Colin

Noah here. Until recently, the concept of lumpers and splitters—which describes the two most popular approaches to building taxonomies—had somehow eluded me. Lumpers take the tack of trying to categorize by building fewer, larger chunks. Generational categories like Baby Boomers or Millennials and historical periods like the “romantic” are lumpy: they’re big, unwieldy, and imperfect—but they are also simple enough that everyone can understand and work with them. 

The Dewey Decimal system or the North American Industry Classification System (NAICS) codes are clearly the work of splitters: attempts to build exhaustive taxonomies that adequately represent the place of every possible permutation. (If you want evidence, download the 958-page NAICS manual from the Census.) Lumpers think splitters are neurotics, and splitters think lumpers are careless. It’s about as close as you can get to a religious war in the world of taxonomical science. 

Personally, I would describe myself as a reluctant lumper. My natural inclination is towards lumps—likely a function of my time as a strategist and the ways coding and software architecture has rewired my brain—but I also once tried to diagram a marketing strategy in more than three dimensions using a cube of cubes. (It worked exactly as well as it sounds.)

Why is this interesting?

We have Charles Darwin to thank for the distinction between lumpers and splitters. He wrote of “lumpers” and “hair-splitters” in an 1857 letter to his closest friend, the British botanist Joseph Dalton Hooker. Hooker was one of the first people Darwin spoke to about his ideas of natural selection, and, in turn, became the first serious scientist to embrace Darwin’s ideas publicly. 

Further proof there’s an XKCD for everything

Beyond being a friend of Darwin’s, and, therefore, a natural supporter, one of Hooker’s great scientific fights was with the armies of naturalists across the British colonies who were documenting the various flora of their regions—a job Hooker believed should be left to the professionals in England like himself. Hooker believed every new flower and plant didn’t need its own name and branch if things were properly categorized. Natural selection offered an opportunity for Hooker to assert his ideas more fully. From an excellent 2009 Science article, “Lumpers and Splitters: Darwin, Hooker, and the Search for Order”:

If Darwin was right, classification was much more than mere naming; it was uncovering the history of life on Earth—a history that, when combined with the insights of geology, also explained much about the distribution of plants across the globe. Evolution provided the life sciences with laws, explanations for the patterns that had previously been recorded but not explained. That, for Hooker, was the great attraction of evolutionary ideas; they provided a fully philosophical under- pinning for his work, a firm basis from which to tell the splitters he was right and they were wrong. The broadly defined species that Hooker used were, he argued, all the descendents of a common ancestor. 

In the end, lumpers and splitters offers a worthwhile meta-taxonomy: a way to think about the pros and cons of two different classification approaches. But more than that, for me at least, this story of the concept’s birth helps to put classification in a new light. It’s about more than just organizing and labeling—at its best, it’s about describing the evolution of whatever’s being classified. (NRB)

Quote of the Day:

From The Ideal Paper by Inframethodology:

The ideal paper occupies less than hour of the reader’s attention. It consists of about forty paragraphs, each of which takes about one minute to read and supports, elaborates, or defends one of the writer’s beliefs. The ideal paper presents a result that challenges the reader’s understanding of either current theory or current practice on the basis of carefully collected data that represents the world in which we live, framed by the writer’s science, which is the same as the reader’s. The result can be summarized in a single sentence that is both theoretically meaningful and empirically significant. It cannot be understood by (properly) by someone who is unfamiliar with the relevant theory, and it cannot be known to be true wihout the analysis of the empirical data that the paper presents. The ideal paper also proposes a series of implications of the tension between theory and practice that the writer has, ideally, discovered. In the closing paragraphs, the paper returns the reader to the world or the science in which they began with fresh eyes.

Quick Links:

Thanks for reading,

Noah (NRB) & Colin (CJN)

Why is this interesting? is a daily email from Noah Brier & Colin Nagy (and friends!) about interesting things. If you’ve enjoyed this edition, please consider forwarding it to a friend. If you’re reading it for the first time, consider subscribing.

© WITI Industries, LLC.