The Science Conundrum
Updated: Jun 9, 2022
Rupert Sheldrake has been called the most controversial scientist on Earth. His best-selling book, The Science Delusion, tackles what he calls the dogmas within conventional science which can blind us to deeper discoveries about the way the world works.
This is an interesting presentation on the dogmas that plague science. While I quibble with some of his comments, in the larger sense he ain't wrong. The Scientific Method is, IMHO, among, if not the single greatest achievement of mankind. That said, science is practiced by humans, and we all have our failings. Often those failings manifest themselves in a faux science that is little distinguished from the superstitious dogma it seeks to replace.
The next time someone invokes the holy name of SCIENCE to justify some illogical action, point them at this presentation. Note that the TED folks were so upset that they censored him from their site.
In my upcoming book Undercover Alien, I address the subject of human failings and the limits of science in this scene:
"Scientists!" she interrupts him with a snort. "Most 'scientists' are bottle-washers and button sorters. Scientists come in three flavors: Politicians, bottle-washers and button-sorters, and competent scientists. The latter is exceedingly rare."
Her unexpected passion startles him. His mouth begins forming words, but voice rising, she charges on before he can engage.
"Science, regardless of the question, is the one true path to knowledge for those able to walk it, but it is a rigorous and demanding path. The scientific method is humanity's most outstanding achievement, although it is only valid when practiced without bias or agenda by those competent with the tools and methodology. Few are able.
"Many so-called scientists are merely disguised politicians! They invoke holy 'science' to hammer those with an alternative viewpoint into submission. They invoke the sacred name of 'science' to silence opposition, end discussion, and browbeat opponents into accepting unsupported claims on faith and without evidence."
He stares at her as though she has grown two heads. She doesn't notice.
On a roll now, her voice takes on a tinge of passion. "The politicians who disguise themselves as scientists are easy to recognize once you understand them. Anyone who dismisses data without examination, insists the 'science is settled' and attacks those who question pet conclusions; no matter what advanced degrees they may hold, no matter how impressive their title, such a person is not a scientist. A Ph.D. does not a scientist make.
"Science is, or should be, a collaborative, self-correcting process from observation to testing predictions based on a hypothesis, to reproduction by the scientific community." She continues. "When this happens, you eventually arrive at a theory you can begin to trust. Sadly, it does not occur as often as we imagine. Mistakes, human error, and political machinations often confound the results.
"Take, for example, the University of Hawaii' Math glitch' some graduate students stumbled across."
His eyebrows tilt in question.
She dismisses his puzzlement with a wave of her hand. "It was in the news. A graduate student attempted to verify his study's calculated data prior to publication. His results did not match the results of a similar study by another professor. When they dug into the math, they discovered 'professor bottle-sorter' had used an ordinary spreadsheet. Spreadsheets are suited, barely, for business arithmetic but never for scientific work. Remember, computers cannot do math with their built-in instructions. Even math co-processors or accelerators are merely streamlined and optimized variations on conventional instructions. At best, computers only do limited arithmetic, addition, subtraction, etc. Complicated programming is necessary to turn those elemental capabilities into real math, and there are specialized tools and math libraries for doing so.
"Common business tools such as spreadsheets perform their calculations using double-precision floating-point as defined by the IEEE. This fact is well-understood."
He nods as he realizes where she is going; he interjects, "Double-precision floating-point math can hold the equivalent of about fifteen decimal digits. This digit limit leads to meaningless rounding errors in a business calculation but deadly errors when doing scientific math. Any competent scientist understands this; button-sorters, not so much."
"Right," she says. "The student used a more sophisticated set of tools based on quadruple-precision floating-point, able to work with twice as many digits before encountering rounding errors. Unfortunately, rounding errors had ruined the professor's previous calculations." She makes 'bunny ears' with her fingers as she continues, "This 'glitch' invalidated the conclusions of over a hundred research studies and an equal number of scientific papers. Of course, the popular press blamed it on the computer, but the scientist was the glitch. There was nothing wrong with the machine; it was the professor who failed."
He continues, "Once you weed out the blatant politicians-in-scientist-clothing, those who remain, who appear sincere, still often fail due to confirmation bias and other human failings."
"Science is incredibly hard," she says. "Science is a process of jumping from one wrong hypothesis to another. Each new hypothesis requires more effort to disprove than the last, but inevitably, each falls, replaced by a more resilient one. Eventually, we develop a hypothesis that survives all our falsification efforts. We eventually call this most challenging hypothesis a theory and begin using it to make predictions. So, the discovery of the math glitch in Hawaii disproved many solid-seeming theories."
He nods and says, "Hypotheses and theories are fragile things. Anyone who claims a pet theory or hypothesis is proven and not subject to debate is so nutty their farts smell like Jif and is certainly no scientist."
She grimaces and interjects, "Thanks. One more comfort food that I will never again find comforting!"
Whether he ignores her jab or is merely too focused for the interruption, nonetheless, he continues, "Trust the science, not the scientist. Judge a scientist by the questions asked, not the answers given. Science may light the way, but engineers are more honest. Engineers find ways to improve things, but engineering is never disproven, unlike science. The engine either works, or it does not, the bridge stands, or falls, the airplane flies, or it does not. So you can trust an engineer."
She rolls her eyes. "I know a few bridges which have fallen, but I take your point. Science is always subject to disproof, whereas engineering is the definition of empirical proof. Engineering may be refined and improved, but not disproven."
He nods. "Engineering proves science, and math explains engineering. The three are inseparable. Until an engineer builds a bridge, an engine, an airplane, or a computer program that reliably operates as predicted by theory and conforms to the math, conclusions based on theories alone are but unicorn farts — pretty ideas and idle fantasies unsupported in the real world. Math and science give us the tools to understand why engineering works. Math is useful in describing the empirical world but fails at truly proving anything, and science without derived engineering is mere speculation. With a collective body of engineering, a theory may become a law, as in the laws of thermodynamics or the law of gravity. Still, until a theory successfully predicts the real world, it is suspect, provisional, and untrustworthy."
She nods and adds, "But such flawed theories are the currency used by the political class for partisan ends, usually some form of power-grab, but whatever the purpose, such theories are favored tools of all politicians everywhere."