Triple Fail: Big Science Blunders
Three strikes in one day: scientific theories and assumptions in unrelated fields have blundered big time. What are the lessons?
Here are three news items that all appeared on June 3rd, 2020, showing big fails in unrelated sciences. They all show that things they thought they knew are wrong. One shouldn’t expect scientists to bat a thousand, but these blunders look pretty embarrassing, in spectacular ways.
Astronomy and Cosmology Fail
Hubble makes surprising find in the early universe (NASA Hubble Space Telescope). Instant stars? It cannot be. According to Big Bang theorists, stars needed time to grow. The first atoms had to form, then clump together, then form stars, then galaxies. This is known as cosmic evolution, and stellar evolution, and galactic evolution. It was so Darwinian; simple to complex, gradually over billions of years. The first stars needed to be composed of almost pure hydrogen – the so-called Population III stars, theoretical but necessary for the theory to be plausible. The “Pop III” stars* were supposed to blow up in supernovas, and seed the galaxies with the first heavy elements. NASA’s announcement moves the “bang” from the big bang to the theory itself. It’s blowing up! Listen to the big fail:
New results from the NASA/ESA Hubble Space Telescope suggest the formation of the first stars and galaxies in the early universe took place sooner than previously thought. A European team of astronomers have found no evidence of the first generation of stars, known as Population III stars, as far back as when the universe was just 500 million years old.
The exploration of the very first galaxies remains a significant challenge in modern astronomy. We do not know when or how the first stars and galaxies in the universe formed.
This really messes things up. The Big Bang was only supposed to produce hydrogen, with small amounts of helium and lithium. Elements up to iron could be cooked inside stellar cores over millions of years, but all the higher elements above iron had to be formed in supernovas. Now what are theorists going to do? They were wrong – there is no evidence for hydrogen stars!
“These results have profound astrophysical consequences as they show that galaxies must have formed much earlier than we thought,” said Bhatawdekar….
These results also suggest that the earliest formation of stars and galaxies occurred much earlier than can be probed with the Hubble Space Telescope.
In any other occupation, workers who fail this spectacularly would be fired and ashamed. Scientists have a bad habit of covering their failures with the announcement that they are “excited” to be wrong. “This leaves an exciting area of further research for the upcoming NASA/ESA/CSA James Webb Space Telescope — to study the universe’s earliest galaxies.” If the Webb telescope doesn’t find them, will they finally have to admit that it looks like creation, not evolution?
*Population III stars were named according to historical traditions. “Population I” are metal-rich stars (where ‘metal’ is astronomer lingo for elements heavier than helium). “Population II” stars are metal-poor. Theoretical Population III stars are supposed to be pure hydrogen, with no heavy elements.
Studies of Brain Activity Aren’t as Useful as Scientists Thought (Duke University). For years, neuroscientists have merrily observed their patients perform tasks in MRI machines, figuring out how the brain works. The MRI machine records the data obediently, yielding nice images of blood flow, a proxy for brain activity in a well-known, highly-trusted process known as “functional MRI” (fMRI). Well, that proxy orthodoxy has just collapsed. Look what the press release from Duke U says:
Hundreds of published studies over the last decade have claimed it’s possible to predict an individual’s patterns of thoughts and feelings by scanning their brain in an MRI machine as they perform some mental tasks.
But a new analysis by some of the researchers who have done the most work in this area finds that those measurements are highly suspect when it comes to drawing conclusions about any individual person’s brain.
Even readings taken on the same person a few weeks or months apart can be so different as to be useless for drawing conclusions. Look how bad this is. The correlations don’t even rise to “fair” like a C grade; they’re more like a D or F, according to Ahmad Hariri, a professor of psychology and neuroscience at Duke University who led the reanalysis.
Functional MRI measures blood flow as a proxy for brain activity. It shows where blood is being sent in the brain, presumably because neurons in that area are more active during a mental task.
The problem is that the level of activity for any given person probably won’t be the same twice, and a measure that changes every time it is collected cannot be applied to predict anyone’s future mental health or behavior.
Hariri and his colleagues reexamined 56 published papers based on fMRI data to gauge their reliability across 90 experiments. Hariri said the researchers recognized that “the correlation between one scan and a second is not even fair, it’s poor.”
The best that can be done is to test a large number of people and compute averages, the article says, but how can that be? Bad data based on wrong presumptions averages out to bad averages, it would seem. Hariri is flabbergasted by this revelation.
“This is more relevant to my work than just about anyone else’s!” Hariri said, his voice rising. “This is my fault. I’m going to throw myself under the bus. This whole sub-branch of fMRI could go extinct if we can’t address this critical limitation.”
Hariri has been using fMRI data as part of a long-term study of 1,300 undergraduate Duke students. By combining brain scans, genetic testing and psychological assessments, Hariri is searching for biomarkers of individual differences in the way people process thoughts and emotions, such as why one person comes away from a traumatic event with PTSD or depression and another does not.
“We can’t continue with the same old ‘hot spot’ research,” Hariri said. “We could scan the same 1,300 undergrads again and we wouldn’t see the same patterns for each of them.”
In short, fMRI is completely unreliable for what neuroscientists and psychologists thought it was showing. All those papers go out the window. Should Hariri be fired? Pulling himself out from under the bus, he tries to think of ways to carry on somehow. His colleague Russell Poldrack, a psychologist at Stanford who is also guilty of publishing phony fMRI “scientific findings” tries to help:
“There’s three things you can do,” Poldrack said. “You can just up and quit, you can stick your head in the sand (and act as if nothing has changed), or you can dig in and try to solve the problems.”
It’s an admirable spirit. But neither Poldrack or Hariri seem to know what exactly to do. Maybe they could follow the adage, “When in trouble, when in doubt, run in circles, scream and shout.” That should cause some interesting images of blood flow in the MRI scanner.
Humans and Neanderthals: less different than polar and brown bears (University of Oxford). This big failure is not really new; evidence has been growing for years that Neanderthals were fully human. It’s more like a coup de grâce, the final nail in the coffin showing that paleoanthropologists were wrong, wrong, wrong about human evolution, and misled generations of students about Neanderthal Man – and Denisovans, too. The subtitle reads,
Ancient humans, Neanderthals and Denisovans were genetically closer than polar bears and brown bears, and so, like the bears, were able to easily produce healthy, fertile hybrids according to a study, led by the University of Oxford’s School of Archaeology.
Polar bears and brown bears easily hybridize, and could be considered variations of the same species. In his third book Darwin Devolves, Michael Behe considers them brown bears with “broken genes” – a case of devolution, not evolution. For the polar bears, the devolution worked out in the Arctic environment, allowing them to subsist on a high-fat diet that would normally be unhealthy for brown bears.
So if modern humans, Neanderthals and Denisovans are all genetically closer than the bears are, does that not demolish the evolutionary tale that they were separate species? For decades, impressionable students, misled by artwork, have been led to view Neanderthals as dumb brutes not as high on the “fitness” scale as us moderns. They were slotted into their own species, Homo neanderthalensis.
Was there any apology at Oxford for this failure? Any remorse for bad science and false teaching? No; Darwinians just move the boundaries of species a little bit, so that they can keep telling new evolutionary tales.
Professor Greger Larson, Director of the Palaeogenomics & Bio-Archaeology Research Network (PalaeoBARN) at Oxford and senior author of the study says, ‘Our desire to categorise the world into discrete boxes has led us to think of species as completely separate units. Biology does not care about these rigid definitions, and lots of species, even those that are far apart evolutionarily, swap genes all the time. Our predictive metric allows for a quick and easy determination of how likely it is for any two species to produce fertile hybrid offspring. This comparative measure suggests that humans and Neanderthals and Denisovans were able to produce live fertile young with ease.’
Now wait a minute. What does this do to Darwin’s “Origin of Species” if species cannot be defined? And what happens to natural selection if “lots of species, even those that are far apart evolutionarily, swap genes all the time”? Larson can’t have it both ways. Either species are real, or they are not. He cannot talk about “species… that are far apart evolutionarily” when he has redefined what a species is. Additionally, what happens to the meaning of “hybrid” if different organisms can “produce live fertile young with ease”?
These changes to definitions and concepts are way at odds with what evolutionists believed in the 20th century, when genetic mutations were naturally selected and passed on through the germ line. Is it even proper to talk about “Darwinian evolution” any more, in a world of rampant hybridization, epigenetics and non-random variations?
For good measure, here’s a fourth scientific failure that is more ominous, because it touches on political bias. When President Trump advocated hydrochloroquine (HCQ) as a possible treatment for coronavirus, he was savaged by the news media, in spite of the drug’s decades-long reputation for safety as a treatment against malaria and lupus. Scientific papers and news stories, out to prove Trump a fool, were uniform in their condemnation of his statements (even though he was just saying it was worth a try, since there was no other drug available at the time). Almost all science news and journal articles said his advocacy of HCQ was irresponsible and could harm people because of potential side effects. And when Trump started taking the drug himself (on advice of his physician) as a prophylactic, scientists, newspapers and politicians were outraged.
Many of Trump’s critics relied on the “Lancet study” that warned of dangerous side effects from HCQ. Because of the Lancet study, the World Health Organization (WHO) halted clinical trials of HCQ immediately. Trump, in turn, was portrayed as anti-science, and was criticized widely for setting a bad example that could harm people, or even kill them, if they did what he did. Trump insisted, however, that people should only try it under advice of their physician. As a prescription drug, it cannot be taken over the counter anyway.
Well, that Lancet study, it turns out, has big flaws. It has come under fire by scientists who looked at it and found it full of procedural errors and questionable evidence. Dozens of scientists signed an Open Letter disputing the methods and conclusions of the Lancet study, reported The Scientist on May 30. Not only that, the leader of the study, and the institution he founded, named Surgisphere, have had a shady past. Medical Xpress said June 3 that WHO is now resuming clinical trials with HCQ.
“Scientific publication must above all be rigorous and honest. In an emergency, these values are needed more than ever,” he said.
He added, however, that decisions to halt clinical trials on the basis of an observational study were “completely unjustified”.
Whether HCQ proves to be effective against COVID-19 remains to be seen after the clinical trials are completed. But for now, have any of the news media apologized to Trump for relying on a flawed study in their criticisms? Those who know the wrath of his Democrat opponents already know the answer; they just moved on to other subjects in their relentless attacks. Trump got through his two weeks of HCQ medication with no apparent ill effects. Although medical effectiveness cannot rely on anecdotal evidence, there are some who had COVID-19 who were helped dramatically by HCQ. Michigan state representative Karen Whitsett claims she was pulled back from the brink of death by HCQ (video). She thanked President Trump profusely for issuing his order that people had a “right to try” it if their doctors thought best. It should not be called a “cure” but rather a medicine that can help in certain situations. Not enough is known at this point.
Update 6/04/20: The Scientist reported today that Lancet and the New England Journal of Medicine (NEJM) retracted the flawed study. Bad science fooled the world: “Surgisphere’s studies have contributed to shaping the global pandemic response,” Catherine Offord reported. Science Magazine was similarly frustrated. It quoted one medical expert saying, “Here we are in the middle of a pandemic with hundreds of thousands of deaths, and the two most prestigious medical journals have failed us.”
One has to wonder why so many politicians and reporters jumped on flawed study and relied on it: was it possibly because of hatred for President Trump, and eagerness to question his judgment?
Incidentally, a preprint on bioRxiv reported that, since HCQ comes in stereoisomeric forms, one form gives 60% better results than the mixture of both. This could explain why some early results looked partially promising. If administered in the more effective form, the effectiveness might be much higher; “we recommend that future clinical studies should employ S-HCQ as a potentially superior drug substance for the treatment of COVID-19 for improved therapeutic index,” they concluded. This, too, is an early report that requires clinical trials.
Important lesson about science: scientific theories are tentative, not final. They are always subject to being overthrown by new discoveries. Another lesson: assumptions distort science. The more experiments are removed from repeatable observations like those performed by Michael Faraday, which could be reproduced by others, the less reliable they become. Faraday didn’t use proxies in his experiments on chemicals, electricity and magnetism; he tested the real deal. Molecules-to-man evolution is about as far removed from reality as one can imagine.
Third important lesson: science is mediated by fallible humans. Worldview beliefs, expectations, assumptions and emotions cannot be eliminated from science. They can be minimized through force of carefulness and character, but that takes integrity. Integrity does not evolve.