Edge.og 2014: What scientific ideas need to be retired/ part 3 - Input Junkie
Edge.og 2014: What scientific ideas need to be retired/ part 3|
Every year, Edge
publishes answers to a hopefully interesting question. 2014 was about scientific ideas that are ready for retirement...
I've pulled out the bits that I think are interesting, and the result is long enough that I'm posting it in several chunks-- this is the last one. I was running out of attention now and then, so if you think I might have missed more good stuff, I suggest reading the Edge link from the bottom up.
My text is what I agree with. If you want a description of the idea that someone thinks ought to be retired, go to the link.Evidence-based medicine shouldn't be standard practice because the "evidence" frequently isn't very good. Mice are so different from people that it's simply useless to use them to study cancer treatment or sepsis treatment. More detail about how mice are different from people.
This could be an example of how one foundational mistake can have a huge effect. Anyone have information about whether countries other than the US depend on mouse studies?Calculus should not be a prerequisite for studying computer programming.Experiments trump multiple regression analysis (perhaps especially in social science) because multiple regression analysis doesn't do a good job of teasing out factors which are correlated with each other.
Sounds reasonable, but I might be missing something.Essentialism doesn't capture the complexity of the real world.
I agree, and I'll add that essentialism makes it very tempting to ignore information you've got if you've already concluded that you know the essence of something you're looking at.Malthusianism is wrong
. This seems to be true currently-- and did anyone predict a considerable drop in oil prices?
"This nasty idea—that you had to be cruel to be kind to prevent population growing too fast for food supply—directly influenced heartless policy in colonial Ireland, British India, imperial Germany, eugenic California, Nazi Europe, Lyndon Johnson's aid to India and Deng Xiaoping's China. It was encountering a Malthusian tract, The Limits to Growth, that led Song Jian to recommend a one-child policy to Deng." Anyone know whether all of this is historically accurate?
However, the author claims that Mathusianism was never true, and I have no idea whether that's the case.
Thoughts?One genome per person oversimplifies matters.
There's a lot more mosaicism and mutation than you might think.
Now that I think about it, there's a lot of evidence that women pick up cell lines from their sons-- it seems to be easy to check for Y chromosomes. Shouldn't it be just as likely that women pick up cell lines from their daughter?
Also... Timothy Leary thought that women were more mentally flexible than men because of re-imprinting during childbirth. Assuming he was actually on to something, maybe it was actually genetic changes.Different languages might have effects on cognition, but it's not enough to add up to different worldviews.
Dammit, Sapir-Whorf is so convenient for science fiction! However, even pretty weak Sapir-Whorf might be overblown.Looking for simple explanations is a disaster in social sciences.
Seems likely to me.Maybe the Hard Problem isn't consciousness.
Maybe it's coming up with thought experiments that make sense. How much is a philosophical zombie (something that behaves exactly like a person but isn't conscious-- it can do philosophy, but isn't limited to it) like a perpetual motion machine?Mutations aren't fully random.
Sounds good to me.
This entry was posted at http://nancylebov.dreamwidth.org/1078450.html
. Comments are welcome here or there.
comments so far on that entry.
|Date:||January 20th, 2016 04:41 pm (UTC)|| |
The article critiquing essentialism is very well-argued, but it has a couple errors. It says, "American official forms require everyone to tick one race/ethnicity box or another: no room for intermediates." That's not true: "mixed race" is a category, or even more than one category, and many people use it. True, you're only ticking one box, but the box indicates more than one race.
Also, there is nothing built into the rules of the Electoral College that requires all the votes in one state to go for one candidate. That's a political decision made by states to maximize their influence, and a couple smaller states have bucked this. It would make more sense to criticize the essentialism inherent in the popular vote when someone who gets 51% of the vote gets all the winnings, or even, say, 43% of the vote if there's a third candidate with any popularity whatever.
There isn't even a requirement that electors vote as they promised to. It's rare, but electors have bolted and voted for a different candidate.
|Date:||January 20th, 2016 09:24 pm (UTC)|| |
Some states have laws to that effect, but their enforceability is doubtful.
|Date:||January 20th, 2016 08:44 pm (UTC)|| |
And the article attempting a rebuttal of Saphir-Whorf is simply asinine.
Would you care to be a bit more specific?
|Date:||January 21st, 2016 01:34 pm (UTC)|| |
I could, but if I had the time and energy to devote to unraveling such trivial nonsense, I'd have done it already. I was just putting down a marker. Suffice to say that Sapir-Whorf is about framing reality, and the writer mocks it because it doesn't shape reality, which is a different thing.
|Date:||January 20th, 2016 11:25 pm (UTC)|| |
"But any evolutionist knows there must have existed individuals who were exactly intermediate."
Not a sentence you would hope to see in someone arguing for science.
I'm with you on all of these, with one qualified partial objection. I think Malthus was right in that the rate of human population growth will, if uncontrolled, exceed its food supply eventually, but not before exploiting all the 'resources' of the earth and sea. Sure, we could turn Earth into one big factory farm by killing off everything we don't find useful, which would allow us to pack billions more breeders into concrete high-rise tenements. Or we could limit our numbers to levels that leave some space and resources for the rest of the inhabitants of this pretty little planet.
Malthus didn't take genetic diversity into account at all. Yes, we want to reduce our population to about a tenth of what it is now, but we also want to keep as many genes in the pool as possible. The simplest, fairest way to accomplish this without coercing anyone would be to provide birth control free to all, and pay a fat cash bonus for vasectomy or getting tubes tied. From a genetic standpoint, it would be most desirable for our species if most women had two children by two different fathers, one of them with genes as different as possible from her own, but just try getting that idea to fly: hoo boy!
Malthus also didn't consider the effect of access to birth control and education for women.
At this point, I think anyone who makes strong predictions about population is a fool. I also believe that genetic engineering is on the horizon, and that's going to make accurate prediction a lot harder.
Have you looked into Superforecasting
? In spite of the flashy title, it's a serious effort to improve forecasting.
Your two children per woman with genetically different fathers is at least good enough for science fiction. I want to see the society where the second father has some personal connection to his child so that a lot of children have two fathers.
That's certainly true, and even if he had, I very much doubt he would have favored leaving the choices entirely up to individual women. But it's our bodies that do the child-bearing, individually, one baby (usually) at a time, at serious risk to our lives and health: no man has the right to choose for us. Anyhow, with access to birth control, education and financial independence, very few women want more than two children, and many don't want any.
Do you remember Heinlein's The Moon Is A Harsh Mistress, where the protagonist's family is a 'line marriage' - about six husbands, six wives, whose children may be fathered by any of the husbands; all the children are the children of all of them. That's a way to maximize genetic diversity in a small, closed society while also providing maximum family stability and personal liberty. (And of course Heinlein is at pains to stress how, in Luna Colony, mating and reproduction is 100% Lady's Choice, for inescapably logical reasons.)
The calculus requirement very likely comes from computer programming's roots in electrical engineering, where calculus is very useful.
|Date:||February 14th, 2016 06:14 pm (UTC)|| |
Evidence based medicine is a bit of a joke in intensive care (I'm an ICU nurse). Some new treatment comes in every few years (activated protein B was the last one) and then the research is demonstrated to have been manipulated or just lucky and then whatever it is goes out of fashion. Much of intensive care has no research base at all, because you can't do random controlled trials on critically ill people who can't consent.