draft

In defense of positivity (36 Views of Mount CritRat)

I continue with the theme of looking at what types of science the critical rationalists think are boring or wrong. The things people believe aren’t worth thinking about are – I believe – often well worth thinking about.

This time, I wonder what it means that they are so repulsed by Kuhn’s “normal science” – or, as I prefer to call it, “augmenting science.”

About this series

The critical rationalists give the impression of talking about science, writ large:

However, it’s quickly apparent that they treat experimenting and theorizing very differently:

They care enough to regulate the actions allowed to theorists, but let experimenters run wild, so long as they send simple signals to theorists. (“Prediction X has been confirmed or refuted.") I covered their attitude to experiment in the last post. This one, I’m going to slice science up in an orthogonal way.

Background: In 1962, Thomas Kuhn published The Structure of Scientific Revolutions. It famously distinguished between “normal science” and “revolutionary science.” Science spends most of its time doing normal science, solving what Kuhn calls “puzzles.” During this process, oddities and anomalies accumulate until they can no longer be ignored. Then there’s a period of revolution that produces chaos that consolidates into the ideas, tools, practices, institutions, and habits that underpin another period of normal science.

In 1965, there was an academic symposium that devoted a day to Kuhn, later published as Criticism and the Growth of Knowledge. I characterize that day as “a bunch of critical rationalists gang up on Kuhn and he defends himself in a rather bemused way.“ I should note that I agree with many of the critical rationalists' criticisms of Kuhn. His distinction between normal and revolutionary science is too clearcut. He discounts the fractal nature of science: that, during a period of relative calm, smaller-scale revolutions are happening. His notion that rival theories are incommensurable (that you can’t effectively translate between them, and indeed can’t understand the old theory once you have “gestalt switch” to a new one) seems to me to be way overdoing it. They really dislike his idea of normal science. Here’s Popper:

[The bad kind of students, namely engineers] wanted to know only those things, those facts, which they might apply with a good conscience, and without heartsearching. I admit that this kind of attitude exists; and it exists not only among engineers, but among people trained as scientists.

Okay, that’s an opinion. It seems to me a typically-overwrought Romantic sensibility – it’s weird to have “heartsearching” as a job requirement. Popper could fairly say the following:

What the vast majority of scientists do is not what I think of as the essential progressive core of “science.”
The subset of scientists I care about follow a particular methodology. Let’s talk about its logical basis…
Someone else will have to provide a methodology for rest of the scientists. Just as I don’t care about providing a methodology for plumbers, I don’t care what “normal scientists” do.

Instead, he says:

I can only say that I see a very great danger in [the idea of normal science]. […] [It is] a danger to science and, indeed, to our civilization. And this shows why I regard Kuhn’s emphasis on the existence of this kind of science as so important.

Just what are these normal scientists up to that’s so bad?

The success of the ‘normal’ scientist consists, entirely, in showing that the ruling theory can be properly and satisfactorily applied in order to reach a solution of the puzzle in question. […]

This seems pretty innocuous to produce a reaction on the scale of “I tremble for my country when I reflect that God is just.” It’s easy to just note that Popper especially, and the critical rationalists in general, are moralistic methodologists of the sort that sees no possibility of a morally neutral action. If you don’t do it precisely right, you’re doing it wrong. And they are certainly temperamentally drama llamas, so you’re not just doing it wrong, you’re doing it catastrophically wrong.

Having noted that, we could just move on. However, I think digging into just what the critical rationalists are reacting against sheds some light on why their methodology is only narrowly applicable. It’s because they have an overly narrow view of what science is for. Maybe, just maybe, we’re allowed to have more than one goal. While normal science might not be a good fit for Popper’s goal, it’s a nice fit for another one.

Normal theory

Normal science can be theoretical, experimental, or a mixture of the two. Let’s start with theory.

A weird thing about critical rationalist historical examples is that apparently nothing of theoretical consequence happened between the publication of Newton’s theory of gravitation (1687) and Einstein’s general relativity (1916). There were predictions (the orbits of Halley’s comet and an unknown planet that would account for oddities in the orbit of Uranus) and observations (Halley’s comet’s return on schedule, the discovery of Neptune, the anomalous precession of the perihelion of Mercury, but all of those are about the 1687 theory, not about new theories.

⊕

The pendulum makes for an interesting case study of theory development. Newton’s laws define “entities” like mass, force, the acceleration due to gravity, and so on. The pendulum adds new entities used to constrain the motion of a mass: the idea of a pivot point, and the distance from the pivot to the mass in question (the bob). Using Newton’s laws, you can [derive]( a differential equation for the bob’s motion. It’s one that’s hard to solve, but if you restrict the starting position of the bob so that its angle from the vertical is less than six degrees, an approximation allows a closed-form solution. In that solution, the starting angle drops out of the equation, which explains Galileo’s observation that the time for a swing of the pendulum is independent of its starting angle. That’s why pendulums are useful for clocks and measuring the force of gravity in different locations. I imply that the pendulum’s period’s independence from starting amplitude was explained using Newton’s laws. Huygen’s explanation of the pendulum was actually published 14 years before Newton’s Principia. However, he was a correspondent of Newton’s, and used Newton’s calculus in his derivation of a less approximation-ridden formula for the pendulum, so maybe…? I appeal to Lakatos’s kind of dodgy idea of a “rational reconstruction” of scientific history. Pendulum history rationally could have happened my way.

I think that’s neat. I think the various scientists who elaborated on the theory of the pendulum (greater angles, the compound pendulum, etc.) were doing good work. Yet this doesn’t fit into a critical rationalist framework:

No one was trying to refute Newton; rather they assumed his theory was right.
Even considering pendulum theory alone, refutation didn’t play a role. There was no bold predictions about the period of a pendulum; rather, they were explaining something known to Galileo.
As far as I know, there were no bold predictions made, then tested.

Such examples do not appear in the critical rationalist literature that I’ve read.

Normal experiments

There’s a paper in Nature called “Intestinal interoceptive dysfunction drives age-associated cognitive decline.” I learned of it from a podcast episode, “Could boosting gut–brain communication prevent memory loss? A tale of microbes, memory, and our internal senses.” The whole story is intricate and fascinating, but the short version goes like this:

It is known that old mice, like old humans, have worse memory than youngsters.
Someone noticed that when young mice were housed with old ones, the young ones developed worse memory. It was as if they were infected with bad memory.
It could be “social contagion,” the young mice learned to be forgetful by observing the young mice. I suspect the experimenters didn’t believe that plausible, but they did an experiment to rule it out.
Another hypothesis was due to the background knowledge that mice eat the feces of other mice. Maybe young mice are literally being infected by microbes in the feces of old mice.
That was tested. Mice with no observational contact with old mice but fed their gut bacteria (the gut [microbiome]) developed memory loss.
That’s consistent with [a current trend](https://edyong.me/i-contain-multitudes() in biology and medicine: the gut microbiome, once thought to be all about the gut, actually has effects in lots of bodily systems.
Okay, fine, but how does this actually work? What’s the connection between the gut and the brain?
Experiments ensued. It turns out the Parabacteroides goldsteinii bacterium (found much more often in old mice) causes a particular kind of inflammation in the big-deal vagus nerve, which means the brain gets weakened signals, which hurts neuronal activation in the hippocampus, a brain region important for memory.

I think that’s neat. I will shift my diet toward the feces of youngsters.

My point is that this has nothing to do with critical rationalism. They are trying to apply a theoretical/conceptual tool to see if helps explain something that happens in the world.

These experimenters aren’t playing that game. They don’t place any great significance on observations/experiments.

Of course

Once again, this expansion of human knowledge has little to do with critical rationalism. There’s a general idea (Lakatos would call it a “research programme”) that weird bodily effects might be caused by gut bacteria. So when scientists are faced with bodily oddities, they reach for that idea to guide their investigations. I’d say that, rather than worrying about confirmations and refutations of a theory, they’re using the theory to answer “how does this work?” questions. They’re not trying to confirm some theory of the microbiome – rather, they’re trying to use the theory. If they can produce results that add evidence in favor of the microbiome theory – “hey, it gives a good explanation for this” – that’s fine. If not, they likely wouldn’t jump to the idea that the relevance of the microbiome: but that’s not what they’re trying to do. Nor are they trying to refute it. If their experiments don’t pan out, they’re unlikely to

Or, rather, we are allowed to care about it, but we shouldn’t call it science but rather “applied science.”

The trick here is that the critical rationalists claim to be providing rules for “science.” By saying that, they capitalize on the respect that post-WWII societies have for scientists (especially physicists, the favorite science of the rationalists), while not exactly admitting they’re providing rules for a small minority of what people – including scientists – consider science.

They handwave away the difficulties of experimentation. Or, perhaps more precisely, they don’t think there’s a need for their methodology to say much about experimentation once it’s established that the purpose of experimentation is to send simple signals to theorists.