John Brockman has collected his “angels”: all of the many scientists, philosophers, psychologists, techno-geeks, and mathematicians that he either is an agent for or whom he simply knows, and posed to them a provocative question: “What scientific idea is ready for retirement?” The results, in the form of 1-4 page mini-essays, are compiled in a new book, This Idea Must Die: Scientific Theories that are Blocking Progress. You can buy it for only $11.81 on Amazon.
Although I’m not a fan of “idea anthologies” in general, this one is good, and well worth reading. For one thing, you’ll be surprised at the ideas that people say must be deep-sixed, including “Theories of everything” (Geoffrey West), “Entropy” (Bruce Parker), “Falsifiablity” (Sean Caroll, and I disagree with him), “Humans are by nature social animals” (Adam Waytz), “Mind versus matter” (Frank Wilczek), “Culture” (Pascal Boyer), and “The illusion of scientific progress” (Paul Saffo, whose essay I again disagree with). You can see the entire list of contributors, which number about 150) at the Amazon page, simply by clicking on the bookcover link here.
As you can see, there are many provocative essays, and the authors are a veritable cross-section of scientific thinkers, young and old (full disclosure: I have an essay on why we should get rid of the idea of free will). You will be surprised at many essays, puzzled by others, and disagree strongly with still others. But that’s the point of a book like this: it shakes you out of your complacency and makes you think. I particularly enjoyed the short essays on statistics (among the things we should eliminate are our obsessions with the mean and the standard deviation), and those on physics, arguing about whether time really began with the Big Bang, whether string theory is of any value, and whether the wave function really collapses.
In my own field, where I can really evaluate the quality of the contributions, the essays are mixed. The contributions by Dawkins (who advocates the retiring of essentialism) and Seirian Summer (get rid of the notion that life evolves via a shared genetic toolkit) were good, but there were several I saw as misguided in one way or another, including Roger Highfield’s claim that we shouldn’t say that evolution is “true”, Kevin Kelly’s claim that mutations aren’t really “random” (I’ve criticized their earlier versions of these two essays in a New Republic article), Nina Jablonski’s idea that race is just a social construct and the whole notion should be scrapped, and Martin Nowak’s criticism of inclusive fitness. But I found most of the essays intriguing and many counterintuitive. For a mini-education in contrarian thinking in science, this book is essential.
52 Comments
The one-number summaries of sets were particularly useful in the past when they were the best analytical tools we had, and there’re still lots of instances where they remain useful.
But just a quick glance at an histogram will tell you far more than any single-number summary ever possibly could.
b&
sub
I think that we should get rid of histograms, except as a means to set up the calculus of continuous distributions. Order plots contain far more information, and multiple plots can be usefully combined on a single chart.
Mean and standard deviation are fine as long as you are aware of their limitations, especially the potentially large effect of outliers.
In the last couple of years I’ve given a lecture on “the trouble with means” to one of our engineering courses, mainly emphasising that you shouldn’t base decisions just on summary measures, you should always look at the full distribution.
Any statistician will always advise you to graph the data first (histogram, boxplot, scatterplot, matrix plot etc etc etc).
“Falsifiablity” (Sean Caroll, and I disagree with him)“Falsifiablity” (Sean Caroll, and I disagree with him)
Good, because unfortunately, Carroll doesn’t understand the first thing about falsifiability. Most of the things he says that proper science should involve are actually perfectly compatible with falsifiability. If he had actually done a bare minimum of research and read at least the first 70 pages of The Logic of Scientific Discovery, he would write a little more intelligently about the topic.
Seriously: Get the book! Read it! (If you’re really desparate, try googling for a PDF version. ;>)
I see Carroll’s point on falsifiability. You can read it here:
http://www.preposterousuniverse.com/blog/2014/01/14/what-scientific-ideas-are-ready-for-retirement/
He says:
“My answer was “Falsifiability.” More of a philosophical idea than a scientific one, but an idea that is bandied about by lazy scientists far more than it is invoked by careful philosophers. Thinking sensibly about the demarcation problem between science and non-science, especially these days, requires a bit more nuance than that.”
You can read most of the essays at The Edge:
https://edge.org/annual-question/what-scientific-idea-is-ready-for-retirement
You must really dislike Ken Kelly’s piece, Jerry; you mentioned it twice!
/@
Sign right here!
b&
I’ve fixed that; it was an error that came up on revision.
The entire set of responses are also available on Edge’s website here: http://edge.org/responses/what-scientific-idea-is-ready-for-retirement
I imagine this will be a controversial suggestion, but I think religion as a protected class is an idea that should die, unless we extend the protection to political ideologies as well, and I believe some states do. Unlike race, gender, and almost every other protected class religion is a choice, not an inherent part of who you are.
Mike, although I’m with you, the question was about *scientific* theories that need to die….
150 entries, and not a single one about geology. I feel marginalized.
That just tells us that all the ideas in geology are rock solid.
/@
Which is good.
As long as they are not stumbling blocks.
As long as new ones can erupt anytime…
I get all sedimental, thinking about it, but the geology questions require a more layered interpretation.
No schist!
/@
be gneiss now!
Aaaaannnnnnd they’re off!
I really don’t want to know about geology. All that metamorphic and sedimentary nonsense just goes over my head.
Igneous is bliss.
So much for science being a religion.
Oh dear. One more addition to my must-read list which expands much faster than my ability to get through it.
I have no idea what you mean. 😁
/@
We should get rid of the idea of “free will?” Why? How could it possibly make any difference one way or the other?
“With earth’s first clay, they did the last man knead,
And there of the last harvest sowed the seed,
And the first morning of creation wrote,
What the last dawn of reckoning shall read.”
The incomparable Edward Fitzgerald in his “translation” of the Rubaiyat. It still counts as the best critique of Islam ever in my book.
There are some nasty ways of thinking that come along with believing in free will, which Jerry has posted about before. I believe the idea of a “Just World” and the belief in retribution were a few mentioned.
If there’s no such thing as free will, those things are inevitable. If there’s no free will, it’s a matter of complete indifference whether you believe in it or not, as far as consequences are concerned. It only matters if there actually is free will. It seems to me the most rational choice is to believe in it. That way you’re covered either way.
Without (libertarian) free will our choices still matter. The big difference is that we cannot assign moral responsibility to our actions, so justifying punishment becomes a lot harder.
My point: it can have practical consequences.
See f.i.
Covered against what? The parallel to Pascal’s wager is faulty.
The summary sounds more like a list of open ideas that promote progress than harmful ideas that could block progress.
My list would include:
– The unsubstantiated idea that philosophy and/or mathematics says something on empiricism. Of course those two areas can suggest research, as can religion and astrology and so on, but shouldn’t be used for purposes of peer review.
– The idea of Big Bang as the beginning of the universe. It is tied to “Theories of everything”, as a beginning in a singularity could pin parameter values.
But it is historically and scientifically confusing, having bait-and-switched a hypothesis of a singularity to a later phase of reheating. Max Tegmark, who hails from Stockholm I believe, was revisiting around New Year and held a seminar. I saw part of the video, and he calls the eras
“Cold Little Whoosh” vs “Hot Big Bang”
since inflation is much colder, happens in a smaller volume and so is initially ‘slower’ than when everything reheats.
It is the “whoosh” that makes our universe…
– Falsification as a naive model describing actual hypothesis testing of measurement theory.
That it is philosophy and not empiricism is more a stain than a block, but then there is competition between theories that pass initial tests and the F word doesn’t help understanding that.
Interesting. How do we measure and predict “harmful”? It should be more than opinion, surely?
+1 on your first suggestion.
(I’m sure you’re right about the next ones, too, I just have no grounds for saying anything about them.)
Jablonski has it exactly backwards: we need to kill the idea that there is no such thing as a biological race, so we can concentrate on distinguishing between socially constructed races and biological races.
I found disturbing the idea of the return of Lamarckism via epigenetics. As if DNA methylation processes were not possibly be subject to natural selection. This collection is mixture of good and mediocre articles.
The essay:
Natural Selection is the Only Engine of Evolution by Athena Vouloumanos
Second that. Although I think ‘epigenetecists’ do sterling work, the notion of Lamarckian inheritance is spurious, if not ridiculous.
Note, I suspect most epigenetecists -despite the hype in popular press- would not adhere to the notion that these mechanisms are really mechanisms of adaptation.
I agree. It would be really interesting to read an article about long term effects (>10 generations) of these methylations to discuss their supposedly adaptive nature.
If epigenetic marks turn out to be heritable, there is no reason to not call such inheritance Lamarckian, though obviously different from what Lamarck envisioned.
Does anyone disagree that such marks are subject to natural selection?
But are they *traits* or *characters*?
I wouldn’t say that a Theory of Everything should die because there are none out there to kill. People still try to come up with a viable Theory of Everything but so far none have succeeded – indeed it may be an impossible task. Even utterly bizarre concoctions such as String Theory have failed (despite what its proponents may claim) and unless it comes up with some verifiable prediction which can’t be explained by the other current models (or a newly invented but simpler model) it shall remain a zombie theory.
In general I don’t believe there is any reason to ‘kill’ any ideas in science; ideas are abandoned when better ideas come along. The best we can do is point out why the other idea isn’t as good.
But while we’re at it, I think one idea that should be relegated to the history books is this notion that we can somehow predict everything if we only knew enough of the initial state of a system. If only we knew about every little butterfly flapping its wings we could predict the weather. If only we knew what every quantum was up to at the point of the Big Bang we could predict [insert favorite despot]. If only we had more data we could predict climate better… The reason such a notion must be discarded is that it is not necessary, not practical, and indeed provably not possible. After all, how can you possibly ever represent the full state of a system. Furthermore, given our fundamental limitations to observations as summarized by such natural laws as Heisenberg’s Uncertainty Principle, we can’t actually ever demonstrate that we can predict single events based on better knowledge of some fundamental state of nature. So, the old idea: the currently widespread notion of absolute determinism; the better idea: absolute determinism is not necessary and cannot even be proven, so don’t waste time pursuing it.
Very cool – this is going right on my reading list!
I have enjoyed some of his compilations in the past.
If you equate entropy with disorder, as Bruce Parker does in proposing its demise, then you do not understand it.
So how !*should*! we understand entropy?
/@
Short answer: through the mathematics.
Longer answer: there are several interpretations appropriate to the system under discussion,but no unique one. “Unavailable energy” is very useful. “Disorder” is occasionally valid, as in the third law of thermodynamics, but note my very careful use of “equate” in commenting on Parker’s essay. The big problem with “disorder” arises when people such as creationists think that it encompasses all that there is to be said, and use it in arguments against evolution.
Correction: I had meant to write “a measure of unavailable energy”.
“Disorder” in a colloquial sense generally means things like a pile of dirty dishes in the sink and yesterday’s socks and underwear lying on the floor — or, perhaps, an unruly protester at a city council meeting, even if the protester is upset at the city’s intention to bring back “Whites ONLY!” drinking fountains.
Physicists who use the term mean something rather different…something, in fact, that’s a dead ringer for entropy….
b&
Glad you asked, Ant!
That struck me as well. There are so many ways in which entropy can be interpreted (and I cannot claim to understand all of them). I haven’t read his piece yet but it seems to me that he is proposing that some interpretations of entropy need to die. But entropy (in the abstract sense as a measure of the number of microstates in a macrostate) is not going to go anywhere … ever.
Somehow when entropy’s under discussion I keep thinking of the way money comes in so many different denominations, so that it can easily leak away in quanta of various magnitudes and in different directions. But I’m just a biologist, what would I know about thermodynamics?
Since Freudianism is already on its deathbed, I vote for the entire discipline of nutrition (until somebody finally gets it right). Every bit of nutritional advice given during my 60 years has turned out to be wrong. People around the world thrive on an amazingly diverse variety of diets. Quit telling me what to eat!
Woody Allen might have been more accurate than he realised when he had the doctors in _Sleeper_ recommending the “deep fat” diet : http://www.zoeharcombe.com/2015/03/saturated-fat-chd-in-europe/
/@
Agree!
When I saw the title of the Edge email about this in my inbox I automatically knew what idea Jerry was going to write about. 🙂