Spandrels, showmanship, and no Messiahs

One of the first papers we read for my Population and Community Ecology course was SJ Gould & RC Lewontin’s “The spandrels of San Marco and the Panglossian paradigm: a critique of the adaptationist programme“. What an interdisciplinary masterpiece of a title–I remember being a little bit unsure of what I was getting into at that point, but that’s a lot of my memories of the first year of graduate school. First, the use of as-literal-of-a-framework-as-you-get from architecture. Then, a reference to Enlightenment thought giant Voltaire (the incorrigibly optimistic Pangloss). And, of course, a conspiracy theory, the dreaded adaptationist programme.

I’m kidding about that last part–SJG was a noted showman, and several folks have pointed out that, in a lot of ways, it was he who was really behind the sweeping tone of the paper, not RCL. Regardless of who it was, it was likely a conscious decision by at least one of the two famous and well-regarded scientists, and second-guessing the decision is in many ways just as interesting as thinking about their points. But for a couple of reasons, I don’t want to deal directly with the showmanship. First of all, a lot of importance is put on the polish a scientist applies to themself*. The clout that polish holds is all out of proportion to its importance to science or the practice thereof, particularly because (a) polish is entirely orthogonal to science because its efficacy is linked to the culture and subjective life experiences of the audience, and (b) even small factors become extremely hefty as the number of tenure track jobs diminishes and the number of qualified people who want them increases.

There’s not much I can do about general culture nor the small number of jobs, but I can refrain from directly engaging on polish issues (like showmanship and inventing conspiracies). Secondly, neither SJG nor RCL is immediately to hand, and the paper’s both published and famous, so clearly what they did worked for them both at the time and ever since. Part of that is likely their genders, race, and age, but it’s inarguably the case. One other thing to note here is that the sort of polish I’m talking about is distinct from communication, readability, and relatability–if you don’t believe me that it’s worth waiting for another post, just check out the length of this section that deals with ‘what I’m not covering’ in this post. So we’ll skip the polish/showmanship things–and now that I’ve told you why I want to focus on what I’m going to focus on, we can really dig in.

I first became interested in this paper a couple years ago; a labmate of mine countered my adaptationist line of argument by saying “But what about spandrels?” I was thrown and a little annoyed, because I had been enjoying the feel of the wind in my hair as I zoomed along on my train of thought. I couldn’t really remember the general points of the paper, but I had a sense that I disagreed (ah, objectivity!) so I found, downloaded, and read it that weekend. My overall conclusion is that this paper is important as a counterargument to the equally extreme idea that evolution and adaptation are directed processes, able to sense not just the local but the absolute maximum fitness peaks, and head towards them with purpose.

Milder forms of adaptationism are substantially more common and, I agree, are vulnerable to type I errors (false positive, if you get those mixed up–I just googled it). This is important to avoid. But if we were to follow the strategy SJG and RCL advocate, we’d be much more vulnerable to type II errors (false negative). The root of this is that by definition, if evolution functions at all, organisms with any sort of heritable advantage (however rare) will necessarily do at least marginally better on average, over time, than their counterparts. This is because God does in fact play dice with the Universe (sorry, Einstein). That’s why we call it an “advantage”. Our use of this unnerving word to describe this phenomenon means neither that (a) we endorse directed evolution (like the adaptationist racketeers SJG and RCL fear) nor (b) that we think every mutation fixed in a population is advantageous (like the credulous storytellers that SJG and RCL believe many of us to be. I’d like to note here that the language they use to do that is disturbingly, gratuitously racist. Let’s all keep from doing that in our own papers, okay?).

It occurs to me that I could write shorter paragraphs. Would that help?

The rest of the paper hinges on a really odd, almost messianic, invocation of Darwin as the all-seeing expert on Evolution Itself. Invoking Darwin as an expert on the modern understanding of evolution is like claiming that Thomas Edison is the person to ask which of 80 gazillion light bulb choices (CFL? LED? Incandescent? Neon? Halogen?) you should buy for your finicky antique porch light. Both men came up with brilliant (though not entirely unique) ideas, and those ideas were nevertheless so great that we’ve spent a lot of time and resources building on them. Neither came up with a perfect idea at the beginning: neither is a Messiah.

SJG and RCL seem annoyed that other scientists cite Darwin falsely, which is fine. But false citation is a separate issue from misinterpreting the Messiah. And arguing about what the Messiah of evolution may or may not have said or meant doesn’t actually matter now, and nor did it really when the paper was published. Why? Because even if CFL and LED bulbs are new to us right now (think: next-gen sequencing and XYZ-omics), even when they wrote the paper, they had neon and halogen bulbs (think: your favorite classic tool for evolution) back then. Edison would have been thrilled, fascinated, and flummoxed, if you asked him for help on your light fixture choice.

Onwards. SJG and RCL go on to take a couple cheap shots, some of which conceal relevant critiques, and then make a few alternative suggestions. First, the cheap shots:

  • Adaptationists just can’t quit with the adaptationism: if one adaptationist hypothesis is false, they just posit another!
  • When no adaptationist hypothesis works, adaptationists claim that it STILL IS an adaptation, but that they just can’t find it.

I’ve intentionally used the word “adaptationists/ism” a lot up there to illustrate a point: if you substitute your favorite expletive, conjugated (in?)appropriately for those instances, you can see that the critique is pretty general. Yes, scientists in general keep on with the sciencing, and if one science hypothesis is false, they posit another in keeping with their general understanding of their science framework. Furthermore, if they can’t find an explanation, they then put up a conjecture, still in line with their general understanding of the framework, that highlights their lack of understanding. That’s awful. It’s bad science.

Oh my gosh, no. It’s not awful. It’s standard, and not in the sinister way that neofascist politicians like to use the term. Come on, you guys! You don’t like adaptationism, and I grant you that post hoc fallacies are both sloppy and dangerous, but sloppiness is the thing you’re against here, not adaptationism. (T. rex didn’t atrophy its forelimbs just to stimulate its mate, that would be absurd and a sloppy thing to posit. However, once they were lost, they may have retained or developed features that could help with that.). Let’s all take a breath: positing that natural selection occurs is not radical bad science. It’s pretty basic Darwin. SJG and RCL, I think, want us to be conservative in what we attribute to adaptation, but in making their argument, they throw out not just the bathwater, but the baby, soap, towel, loofah, bathtub, toilet, and bathroom door.

Now for the alternative suggestions:

  • Neutral theory is a better explanation: probably there is selection on and for things we don’t observe, and it might be random.

Okay I’m going to say it again. If selection. Is happening. That means. Adaptation is happening. (And that’s not really neutral theory ecologically, not at all. Because selection. Is not. Neutral. But I’ll agree with you that which mutations happen and when they happen is totally random, so in that sense, yes, Motoo Kimura’s neutral model is definitely happening.). Back to adaptation and selection, though. This is crucial. Those two go together like … leading strand and lagging strand. Or any other things you can think of that are inextricably linked. Peanut butter and jam. Powwows and drums. Lacciri and kosan. Water and life. Picking and your favorite.

  • Plasticity is also a better explanation: sufficient elasticity removes the requirement for adaptation

As someone who studies transgenerational plasticity, I like this one. A lot. And I like that it invokes standing genetic variation, and I like that it relies on Kimura’s neutral model. And I like all of this because I’m interested in environmental stability (or lack thereof) and what that does to organisms. But look, we can’t just ignore all that fantastic work done by the Grants on Darwin’s finches–that excellent natural experiment shows that, while there is a lot of standing variation, and there’s certainly some plasticity, adaptation and speciation happen via genetic mechanisms. So, yes plasticity! But not only plasticity.

In conclusion, natural selection is a thing, and SJG and RCL really didn’t like the word “adaptation”. It worried them when they wrote the paper, worried them enough to make the the leap from observing heritable change happening over time in response to environmental pressures to warning everyone off the heady drug that is storytelling. They worried that people would forget the utility of seeking out the similarities among organisms (the Bauplane) to explain phenotypes. I think they worried that the larger structure of science was going to be lost in a tangled knot of flashy, poorly supported narratives. But the way they wrote the paper makes it hard to find the good points, and good intent, underneath the blistering scorn. Straw man arguments make it harder to find the real point, and when you heap up enough of them in one place (perhaps in hopes of an illuminating bonfire) you risk losing the needle of truth in your haystack.

A final thought: Ironically, given the use of the term ‘spandrels’ by biologists after this paper, architecture isn’t much use for thinking about biology. Architecture is a study of how, most elegantly, to arrive at a structure with predetermined characteristics. Biology is a messy process that ends up, sheerly through an amalgamation of chance and luck, with marvelously complex things that were in no way predetermined. Architecture is built by a designer. Biology happens. In architecture, spandrels may indeed be pointless. But in a biological structure, anything may be an artifact of a previous adaptation, the result of a tweaked Bauplane, or, yes, something we just don’t understand yet.

* A goal I have, and one that I hope I’ll meet throughout the blog, is to use the “they/them/their” pronouns to refer to nonspecific individuals and groups. For those of you for whom a singular “they” is a little bit painful because of your internal grammar-o-meter, I feel you (English major that I am). However, it’s MUCH more important to me to normalize inclusion and disrupt sexist binary gender norms than it is to follow the grammatical norms that are the product of an antiquated colonial patriachy, let alone one steeped in hawkish misogyny, homophobia, racism, transphobia, ableism, and orientalism (among others).