The Machine Stops: Science and Its Limits
lareviewofbooks.org
lareviewofbooks.org
Strevens’s Machine, the engine of modern science, runs on a single principle, a kind of speech code that he dubs “the iron rule of explanation.”
No. That's wrong. As Fred Hoyle once wrote, "Science is prediction, not explanation". This is key. Successful prediction means you got something right. That's objectively testable. Explanation is not testable in that way.
If you can't predict, you can't make an engineering technology. Electronics exists because physics is predictable.
The frustration is that the hard sciences have been so successful that most repeatable phenomena have been figured out. Now what? Physics is stuck trying to figure out cosmology, where you can't experiment, layers below the subatomic level such as superstrings, where you can't experiment, and how gravity interacts with other forces, where you can't experiment. Real progress is being made at very low energy levels down near absolute zero, though, where experiments are possible.
The "soft sciences", psychology, sociology, and economics, seem to be unable to make reliable predictions. Even getting repeatable experiments is tough. Are they even sciences? Maybe psychology should be shipped off to the medical school, economics to the business school, and sociology to the school of public administration.
String theory?
Perhaps science is those explanations confirmed by accurate prediction.
It is only when we found scientific explanations, like those of Galileo, we were able to predict and understand correlations.
Makes me wonder if the Mayans knew about the stars and the planets, even though they could very well predict the sky above them.
At least in my field (mechanical and materials engineering), understanding means explaining in basis to well established fundamental laws, a prediction is not enough.
>> ...scientists [] resolve their differences of opinion by conducting empirical tests rather than by shouting or fighting or philosophizing or moralizing or marrying or calling on a higher power....
His usage of "empirical tests" lines up with what you (and others) call "prediction".
Summary: I don't think you're in disagreement with the book.
As a comparative, this review [1] notes "Strevens argues that modern science owes its success to the relinquishing of deep philosophical understanding in favor of the shallow power to predict empirical observations. [...]"
[1] https://blogs.sciencemag.org/books/2020/12/28/the-knowledge-...
When that situation arises, it's the fault of the writer. The classic remark on this is by Lewis Carroll: “When I use a word,’ Humpty Dumpty said in rather a scornful tone, ‘it means just what I choose it to mean — neither more nor less.’. ’The question is,’ said Alice, ‘whether you can make words mean so many different things.’. ’The question is,’ said Humpty Dumpty, ‘which is to be master — that’s all.”.
Of course, this isn't typically why science gets funded - we want the engineering applications that are enabled by our ability to calculate - but a version of science that's all prediction, no explanation seems very unappealing (not to mention sterile for further investigation).
What the scientists are in it for is one thing. But what confidence non-scientist members of the public should have in claims made by scientists is another. The latter is what prediction is for: the better the predictive track record of the scientific claims, the higher the confidence they deserve.
As you indicated, the "human scale" things that we can experiment with to make predictions have been "plucked."
Now we have to make increasingly more complex tools (LHC etc...) for physical things we can't experience directly and our ethics prevents us from experimenting with those "soft" things.
The lacunae in this model are far more subtle, and it is difficult to gauge when or how this might break down.
If (for eg) I wanted similar (analogous to how well we understand mechanics) understanding of the health implications of going vegan, or dropping fertility (ref: post on HN front page currently) where would one even begin to research that question? Even assuming one can unilaterally bring to bear all of humanity’s collective resources on this problem, the scientific method is not powerful enough to generate timeless generalizable truths for this problem. It’s very possible that the ideal diet might easily change when we’re talking about humans living in a Martian colony, or a century into the future with different climate, or living in a different continent different from their native ethnicity.
Questions of that nature/complexity cannot strictly be answered “scientifically”. We continue with traditions (as a baseline) and occasionally experiment with tweaks and gather empirical evidence when we can, to iteratively improve our dietary practices (as our food technology improves).
https://www.cs.ucdavis.edu/~koehl/Teaching/ECS188/PDF_files/...