Reading Sosiak & Barden 2020, Multidimensional trait morphology predicts ecology across ant lineages has me thinking. First, a brief overview of the paper: taking a large sampling of ant species from across all major groups the authors:
a) Classified each species into niches based on their nesting habits (e.g. underground, in wood), foraging habits (e.g. arboreal, subterranean), and functional role (e.g. generalist predator, fungal farmer).
b) Measured many morphological traits such as leg segment lengths and eye placement and entered them into a database.
c) Used machine learning to see which traits, or suites of traits, correlate with different "ecomorphs" (distinct combinations of nesting, foraging, and functional niches).
Now, the findings themselves are of debatable value; although the program can predict with some 77-85% accuracy which ecomorph category an ant belongs in just based on morphological data, in some ways these simply corroborate what had already been observed. However, this is really all just a preamble to frame something else I want to talk about.
| Figure 3 from Sosiak & Barden (2020) |
Binned Findings
"Across our analyses, we find that ecomorphospaces overlap in a core region, where a large proportion of species are clustered. These ants reflect a generalized platonic ant morphology, where they are not strongly morphologically specialized relative to other taxa."
-Sosiak & Barden (2020)
"Collapsed or summarized ecomorph syndromes improve prediction by simultaneously considering multiple ecological niche aspects and accommodating some plasticity."
Why do I find this interesting enough to write on? Because at first glance it is counter-intuitive that a smaller number of categories would be more useful. We would expect that a process which paid attention to every feature and most finely delineated categories would be ideal. After all, isn't this getting closer to the truth? Yet instead their own preliminary results led them to conclude otherwise. Trying too hard to differentiate didn't work because their data itself had imperfect categories, and so it performed better when simplified into fewer groups which assumed wiggle room.
Yet this is, of course, precisely what we do as humans. In a world full of variation we are forced to make arbitrary, but useful, distinctions. It reminds me of Donald Hoffman's "fitness beats truth" theorem, that what natural selection has pushed us for as beings with limited time and processing power is an ability to make generalizations. Useful generalizations. That's the key here. The validity of categorization is linked to outcome, and there is no clear rule for determining how finely it ought to be done except that the two extremes are equally useless: if everything belongs to one category or if everything has its own category we may have perfect predictive power... but it is useless. And in the vast middle area is left us to decide when and were to subsume particulars to a larger classification as we make the universe as simple as possible, but no simpler.
![]() |
| Supremacist Composition: White on White by Kasimir Malevich, 1918 |
Family Resemblance
"We find that multidimensionality greatly aids in the characterization of ecomorphs... Ecomorphs are often defined by integration of traits moreso than individual measurements."
-Sosiak & Barden (2020)
I sometimes have this feeling that we are getting back to where we once were a century ago in terms of appreciating the complexity of the world. That may seem to be a strange thing to say in light of the vast advances in science in this time, but it is a thought that has particularly been on my mind since earlier today as I read a paper involving the history of ecological development. The author notes that nineteenth-century scientists were well-versed in the impact of environment on development, but that later advances in genetics and molecular biology drove such observations from the center stage. That what they offered, and what we wanted, was a clear sort of mechanism, one made only of the simplest parts that would allow us to ignore all the baffling variety and circumstance in natura to reach the real truth underneath. That would satisfy us.
I am hardly the first person to observe or critique the trends of reductionism, but as I thought on it I wondered just how great its breadth and impact was. It is true that science inherited from its physics roots a propensity for breaking things into their components, and it achieved much success with it, yet I feel as though that element of success carried another overtone. A sense of liberating simplicity that cuts through the morass of existence:
“Nature and Nature's laws lay hid in night:
God said, Let Newton be! and all was light."
As I read Art Through the Ages what struck me was how art in the 20th century was distinctly atomized. Rothko's color, Mondrian's composition, and so forth. They were not mature narratives but single thoughts. Another book I was reading (Janson's History of Art for Young People) called Picasso's Bull's Head "a visual pun" and I thought the phrase was incredibly apt. We cannot deny that puns are comedy, yet they are only a fragment of it. A shard of humor. The smallest piece before it ceases to belong to that category. Yet so many of the twentieth-century works fixate on these small pieces so strongly as if desperate to grasp a truth in it, or behind it; that it is only complete simplicity which allows for certainty. There is a quote from Naum Gabo which summarizes this mentality perfectly:
"Such artfully constructed images are the very essence of the reality of the world which we are searching for."
Another, from Kasimir Malevich:
"The essential thing [in pictorial art] is feeling - in itself and completely independent of the context in which it has been evoked."
Context again, or rather its absence. I think there was something at work, an essential drive that ran through the twentieth-century mind to blast away everything and get to the bottom of it all in a frenzy of disgusted disillusionment. In his brilliant essay Iconophobia, Kenneth Clark notes that pure monotheisms are hostile to representational art for they are a form of beguiling idolatry. They aren't the real thing, just something's fake of it, and all they can possibly do is lead one astray. It is the result of feeling that they are in the service of the one core truth of the universe which makes all else distraction. He then goes on to observe that much of modern art is also iconophobic. It detests representation in favor of abstraction, and that it is perhaps for much the same reason: a desire to purify one's vision and so draw closer to the truth.
It causes me to question the place of reductionism in all this. I had always assumed that it, being part and parcel with science, was the driver. That all wings of the human endeavor were attempting to copy science in all its success. Yet that may only be part of the story. Perhaps rather reductionism in science wasn't maintained on the basis of results, but because it appealed to the need we increasingly felt to ground ourselves after having unchained the Earth from the Sun. That truth would be found in a forcible negation of the old view which dissolved all the contextual stories and left us with certainty once again.
To return to ants here at the end, I am reminded of when I read William Morton Wheeler's The Ants. What struck me was his assured holism, that he took it for granted he was observing an organism in its entirety, and that organism has a reality all its own. Like he wasn't looking over his shoulder, afraid that its reality would dissolve into a meaningless assemblage of organic chemistry coded by nucleotides. He knew from hours of long observation that he had a sense for these insects, and that such knowledge, though not quantified or rooted in a lower stratum, was nonetheless real. I do not believe that this is merely because he did not yet have access to those findings but because he was assured of his world and its context.
Yet even as I praise this, his own writings validate the suspicion that led to their demise. His observation was acute, but it was in the service of certain values that have proven untenable. He compares types/stages of human societies to those of ants; that Ponerines are hunter-gatherers, leaf-cutters the agriculturalists, and slave-makers the barbarians. He particularly regards inquilline ants as "degenerate," a phrase he uses not just to denote their reduced morphology but as an indictment; little advanced from Aesop in this manner, there was a cosmic lesson from an animal in how a loss of upright self-sufficiency is ethically reprehensible. I think it is realizing these mistakes, among a thousand others, that conferred on us such a desire to not be fooled again. As a price for stripping away the old beliefs we also stripped away the habits of thought that would see value in context, contingent categories, visual representation, and as the paper puts it, "multidimensionality," the "integration of traits" and "accommodating some plasticity" in our understanding.
I do though have some optimism that we are beginning to return to an acceptance of fuzziness. That after being estranged from categories we have realized they are indispensable, but now we are better aware of their nature. Similarly, that having buried ourselves in a sense that the only thing that mattered is the selfish gene, the simplest unit we can envision, we are crawling out and seeing that it is not a failure when it takes multiple factors in order to contextualize an organism (or a truth). Perhaps even that they can have value in themselves. That, I think, is a good direction to go.




