The twilight of fantastical science

A post arguing that we shouldn’t give bonus points to off-beat and bold theories just because they are off-beat and bold.

The allure of deep, hidden connections

Probably we all had moments when we were seduced by the idea that there are unknown, hidden, long-ranging connections between seemingly distant parts of reality.¹ The examples range from silly superstition to thoroughly validated science. From astrology, The Da Vinci Code, and modern-day conspiracy theories of all kinds, to established science, like . . hmm, difficult! Maybe, I’ll come back to that. In between these extremes, there is a grey zone that is the topic of this blog post. A zone that is both a source of creativity and a problem for scientific credibility and ethics.

There is something profoundly human about the allure of deep, hidden connections. Maybe because they hint at a larger, more colorful reality than taught in school. Maybe because this brings a sense of unity—we face an unknown, vast universe together on “a mote of dust suspended in a sunbeam.” Nonetheless, our soft spot for grandiose, speculative, and causally intricate stories makes science vulnerable to exploitation and going awry.

Scale separation = why reductionism works so well

Science, at large, progresses in a divide-and-conquer fashion—by defining objects of study, describing their components, how they interact, and bring about function. This would clarify everything except how the components work, so one can zoom in on them and iterate the procedure. 

In complexity science, we like pointing out that the focus on zooming in to components, reductionism, does not complete the picture. We also need holism—to explain how the interaction of components builds up functioning systems.³ But it is remarkable how successful reductionism is. Just how well explaining organs by their tissues, tissues by their cells, cells by their molecules, etc.,⁴ tells us something fundamental about our world—it is fundamentally organized into a hierarchy of scales, with very few exceptions.⁵ We don’t require chemistry to explain political economy, etc. Indeed, this hierarchy of scales is probably a reason why the academic division of topics—physics, chemistry, biology, psychology, etc.—is so stable.

Across too many levels

A special case of fascinating and fantastical theories is those that cross several levels of the hierarchy of scales: Quantum theories of mind are an obvious example. Theories about how Gödel’s incompleteness theorem puts constraints on biology.⁶ A hyped recent paper arguing that the “logic of biology” can be traced to voltage-gated ion channels also comes to mind. The Gaia hypothesis, too. Probably some everything-is-cellular-automata-style complexity science as well.

These mentioned theories, as such, are not bad as such (being hard to falsify, a.k.a. “not even wrong“). They might inspire people to think big, and we certainly need that sometimes too. However, I worry that by evaluating theories not only by their predictive or explanatory power but also by how creatively, mind-bendingly intriguing they are, we undermine the (sound, scientific) grounds for denouncing politically loaded racist, misogynist, homophobic science. Note that biological determinism (and its extensions), underpinning much of the scientific racism of the early- to mid-last century, is another example of a theory with long causal chains, linking genes to the fate of nations.

Note that there is nothing wrong with mapping out complicated networks of causal effects. Or connecting them into models for proof-of-concept or scenario testing. To explain the Covid-19 pandemic,⁷ we can invoke scales from the biochemistry of the original zoonosis to the politics of mask use. Hmm, we could maybe even throw in Gödel’s incompleteness theorem in a discussion of the computational basis of simulations. But not all this at once, in one theory from Gödel to masks, bypassing complications along the way. Any other explanation for a complex scenario than one following the hierarchy step by step would be the kind of fantastical theory this post is about. Not necessarily wrong, but calling for an exceptionally careful validation.

Closing thoughts

Maybe the problem I describe is not such a big issue. At least, from a long-term perspective, things have become better. But scientists are still willing to lower the bar for theories that tickle the human fascination for a shadow world luring behind our reality. We don’t need that.

You might say that the scientific process will take care of everything problematic. Look at the replication crisis in behavioral science, for example. But that is an exceedingly slow response, in particular, with theories that claim to be scientific based on the premise that they could be validated, pending future development of experimental techniques.

I love toying around with semi-serious ideas and silly computational models. The mind works best when it is free. I don’t want to discourage such chaotic creativity, and this post shouldn’t either. There is an enormous space for ingenuity even without dubious theories.

But what if reality actually has spectacular, long, and hidden causal chains? Of course, it’s the duty of science to find them, but such claims need to be presented with due support—extraordinary claims need extraordinary evidence.

Finally, back to the question of whether a theory consisting of long-ranging connections between seemingly distant parts of reality ever went on to become mainstream science. I think we need to go back very far in time to find an example, the germ theory of disease, maybe? This is, of course, a good reason for caution.

Notes

1. As a high-school student, I liked Gary Zukav’s The Dancing Wu Li Masters. (It was one of several books following Fritjof Capra’s The Tao of Physics in trying to connect modern physics and various aspects of Eastern thought. A common feature of all these books is that the description of physics and Eastern philosophy is much better than the connections between these topics. In fact, the linkings are little more than juxtapositions of various quotations and stories.) As a university student, I was fascinated by self-organized criticality. Now I enjoy reading about the Langlands program (a research direction in math seeking non-trivial similarities beyond the transitional divisions) with a similar fascination. ↩︎
2. It might be my least favorite Maddin movie. Do prioritize My Winnipeg, Cowards Bend the Knee, or Archangel. ↩︎
3. Note that complexity science is not against reductionism; just saying that it is not enough. Indeed, complexity science assumes the validity of mainstream, reductionist science. ↩︎
4. Note that reductionism does not mean striving to predict climate by the Schrödinger equation. We can explain and predict A by B, B by C, C by D, but not A by D, and that’s fine. ↩︎
5. Is this profound or not? That’s a question without a scientific answer. ↩︎
6. I can’t find the link, but I’m thinking of Seel and Ladik’s “The tragicomedy of modern theoretical biology”—a book chapter from the 1980s. But there are also other papers on the same theme. ↩︎
7. Thanks to Luis M Rocha for pointing out this example and calling for clarification. ↩︎