Amid fears this first diary would be greeted with a mass yawn, I have hesitated to polish and publish. Recent remarks by my physical therapist plus a pushed BBC Science video prodded me close to the edge, and then I heard this gem one time too many:
“Oh, that’s just a theory.”—John Q. Public, usually dripping with sarcasm
We have all heard and read such nonsense, and many variations of it. To emphasize the galactic stupidity of this statement, let us remember that gravitation (which is a local special case of general relativity) is also just a theory. A number of important and wealthy Russians have empirically tested via defenestration this “so-called-theory of gravitation” to great publicity, and with notably the exact same result every time, despite it being just a theory.
In truth, the pinnacle of science is a theory.
The root of the word science is the Latin scientia, which means knowledge or learning. In practice, a mature theory is the limit (boundary) of human knowledge in its field. We can and do occasionally extend/generalize/discard extant theory by finding exceptions/errors that its successor explains, but even so just end up with a new theory bounding our new limit of knowledge.
In this universe it does not and can not get any better than that, ever, for humans. No one is particularly fond of this state of affairs, but there it is. Blame Karl Popper. Or God. I certainly do.
“Words mean things.”—Rush Limbaugh, disingenuous as always
Rush is right: theory means a very specific thing. Any old coot’s crazy idea, what the general public colloquially and incorrectly refers to as a “theory,” is actually a conjecture or a hypothesis. A conjecture or hypothesis is very different from a theory.
A scientific theory is a hypothesis with all three of the following:
1) Substantial and persuasive evidence in favor, in the form of repeatable experimental data.
2) Explanatory power.
3) Predictive power.
Without all three it is not science. Without all three it is not a theory. But I repeat myself.
[Digression: It is very important to understand that any old crazy-coot hypothesis that manages to achieve all three becomes theory. Poof! No matter how ugly or awful or unpopular, it becomes theory. Ugh. Even if it is something so totally ridiculous that no one could ever possibly take it seriously—like, say, quantum mechanics. At its debut, no one was happy about quantum mechanics. Many talented scientists tried to find palatable, classical work-arounds or disproofs of that quantum nonsense, devoting years or decades to the quest. They all failed. Partly as a result of that, the Standard Model of quantum mechanics is now the most intensely tested and quantitatively verified scientific theory. The sheer volume of corroborating evidence accumulated for the Standard Model is shocking. As I recall, that mountain of data is orders-of-magnitude larger than all DNA sequencing data ever collected, worldwide.]
So, what do these three defining characteristics of science specifically mean?
Item 1) is pretty obvious: without corroborating evidence very significantly in excess of contradictory evidence, the hypothesis could simply be false. Note that in most fields there is normally some contradictory evidence, for a number of excellent reasons such as uncorrected measurement or methodology bias or error, large inherent measurement variance, confounding factors, a higher-order phenomenon afoot, etc.
Item 2) is merely anthropocentrism codified: if it doesn’t help us explain and understand something, who cares? Why bother? Humans are human-oriented, even in their science.
Item 3) is just as important as Item 1). A theory must make predictions about the real world that can be tested by experiments, and then it must also pass those experimental tests.
However, there is much more depth implicit in these requirements.
For example, concerning item 1), insufficient contradictory evidence can be highly suspicious in many fields. If the results seem too good to be true, gosh, maybe they are. Harvard Business School just found this out the very hard way with the world’s leading “researcher” in honesty. Honest! What sweet irony.
[Digression: In further irony, particle physics is finding out that the opposite very hard way is also problematic. The distressing strength of the Standard Model is such that, in hundreds (thousands?) of terabytes of data, there is exactly the amount of each possible outcome predicted. After looking really hard for decades, including the Higgs boson discovery/confirmation, there certainly appears to be no contradictory evidence and indeed no controversy left in this field. A jaundiced eye might conclude that the field is a dead end for future discoveries.]
I anticipate item 2) to generate some debate because neural networks are true black boxes. Are the results of a neural-network analysis in any way explanatory? Can it really qualify as science? My hunches are “no” and “maybe.” The rapidly expanding use of neural networks in peer-reviewed work should provoke this discussion soon.
Popper adroitly observed that definition item 3) is actually a constraint on the set of allowed scientific hypotheses. A well-posed scientific hypothesis must make empirically testable predictions. In other words, a valid scientific hypothesis must be falsifiable. If a hypothesis is not falsifiable, then it is ill-posed; it is not science.
[Digression: This ‘Is it falsifiable?’ test is the source of the curious cloud surrounding String “Theory.” After 40 years of prodigious, clever effort yielding some beautiful mathematics and intriguing unifications, the field has yet to produce a single falsifiable hypothesis. String “Theory” satisfies only the least important of the three criteria for a scientific theory: explanatory power. This week a BBC Science video auto-queued up in my feed wherein Dr. Brian Greene constantly refers to the field as a theory (defensible nomenclature for a mathematician, but wildly inappropriate for a scientist) while extolling its unifying and explanatory powers only. There are multiple examples on YouTube of significant public friction between celebrity-physicists Neil DeGrasse Tyson and Brian Greene on this very topic. Dr. Tyson is correct. While Dr. Greene may be my superior as a theoretical physicist, he utterly fails Popper’s exam on being a scientist. Persistently. In public. On video.
Just as it has for 40 years, the String “Theory” investigative field still contains fascinating ideas with potential.
Just as it has for 40 years, the String “Theory” investigative field still does not rise to the level of science. It is not a theory. It is Mathemagic, perhaps, but definitely not science yet. Maybe next decade.]
“Science is an argument, the final arbiter of which is the result of experiment.”—Richard Feynman, being wise
Hegel’s Dialectic* is usually stated: thesis, antithesis ==> synthesis; or A, not-A ==> B.
As Feynman quips, the much-vaunted Scientific Method, in action, works much like that.
Some kook or crank or scientist-guy, possibly deluded (or drunk), stands up at a meeting and claims “Phenomenon A works this way!” That’s a hypothesis: any old coot’s crazy idea. This is the thesis in Hegel’s Dialectic.
Some other guy, also possibly deluded (or drunk), stands up and replies “No! You fool! You aren’t looking at the data correctly at all. Phenomenon A works this other way!” That’s the antithesis part of Hegel’s Dialectic. Those two, and usually a bunch of others, argue further about how to tell the difference, and then go home to do experiments. While this script may seem corny or contrived, I have been there as a minor player. This really happens, exactly like this! Ego and rivalry are often disappointingly influential drivers of science. It is a human endeavor, after all.
Most of the time those experiments determine clearly who is correct, and the loser limps home after the next meeting and looks for something else to investigate. Occasionally, if you are very lucky, those experiments reveal that things are not clear at all, neither is necessarily correct, and everybody in the field reboots. That yields what is called an active field of research, which is where many successful dissertations happen.
In either case a synthesis occurs, which is the conclusion of Hegel’s Dialectic.
This is how scientia takes another small step forward, even if some of the original individual players may have been deluded or even dishonest. That is the virtue of Feynman’s Argument.
*a misnomer, since Hegel never stated a dialectic in this form, but ‘Chalybäus’s Dialectic’ never caught on.
“Just the facts, ma’am.”—Detective Joe Friday, always deadpan
A huge and society-wide misunderstanding concerns the word “fact.”
The popular notion of “fact” means something that is True in some absolute (possibly divinely-imbued) sense. With that definition, it is clearly impossible for two “facts” to be in conflict. Yet it happens, often! There is almost always some contradictory evidence. The world is just that complicated. In fact, as discussed above, conspicuous paucity of contradictory evidence is often very suspicious.
Competent scientists speak of hypothesis, accepted theory, predictions, data, probabilities and confidence, error-bars, evidence, conflicting evidence, and defensible conclusions. I have never heard a competent scientist use the phrase “scientific fact” despite its popularity among ill-informed legislators and so-called pundits.
Consider the situation in a court of law. Even there the “facts of the case” is just the evidence presented. It is normal for some pieces of evidence to be contradictory, for a number of excellent reasons the biggest of which is our reliance upon human witness testimony, which is well-documented to be a terrible evidentiary source.
In its common usage, the word “fact” is ill-defined. “Scientific fact” is an oxymoron.
“A Ph.D. is not an inoculation against stupidity.”—Robert L. Park, cracking wise
Discernment is a serious issue in all fields of human endeavor. As the example of Dr. Greene shows, there is pervasive confusion over who really is a scientist.
It can be hard to tell.
The conclusive answer as to ‘who is a scientist?’ boils down to whether Karl Popper’s ideas inform one’s thoughts and decisions. But given the difficulty of gazing into the hearts of men—it can be hard to tell.
Education in science topics, even a lot of it, can impart the lingo but does not make the recipient a scientist. Engineers and physicians are excellent examples of this—there are certainly some engineers or physicians who are scientists, but they are a tiny minority. A few scientific results are important tools in their toolkits, but science is not the job of normal engineers and physicians. They have neither the training nor experience. A number of ambitious physicians who discovered they had PR skills during the covid-19 crisis are themselves very confused about this point.
Confusion over who is a scientist grows further because numerous so-called academic scientists are prima facie not so; examples abound among the Merchants of Doubt and those occupying laughable politically-appointed posts such as Alabama State Climatologist. Absent the involvement of one of these well-known quacks, assessing the import of some piece of science news is a tricky exercise in caveat emptor.
The easiest clue in assessing veracity (or lack thereof) is that true scientific results of import are never announced at press conferences. Never. That is not how science works. On the other hand, that is precisely how fund-raising works. The example of Pons & Fleischmann and “cold fusion” springs to mind. The recent “room-temperature superconductor” press conference in South Korea seems eerily similar (no one has reproduced the results despite many trying). The publicity surrounding a Big Data release (such as a new set of James Webb Space Telescope images) is not an exception to this rule because the Big Data itself is very rarely the true scientific results—rather it is the fodder for the true scientific results.
Another easy tipoff is vanity publishing. If someone has to pay his/her own money to publish, hmmmm, there might be a darn good reason for that. The oft-heard plaint “The world isn’t ready for my ideas!” is never true. Not if your ideas are any good.
As discussed above, any mention of the term “scientific fact” should automatically trigger questions about the speaker’s background and motives.
The best clue is content: real science has lots of mathematics or numbers or both. Science is quantitative. Even the life sciences have become incredibly data-driven. If it doesn’t have at least a few equations, or several tables of data and some derived statistics, or some impressive graphical data displays, well then it probably isn’t 21st-century science. The vast majority of interweb pseudo-science propaganda fails this test instantly. Paragraph after paragraph of emotive text is never science, even if packaged in compelling, well-produced, fast-paced video. Testimonials are never science. Anecdotes are never science. Absorbing large amounts of narrative is not “doing your own research.”
My final test is cui bono (who benefits?). If you see an interesting result published, ask yourself: who paid the paychecks that led to this result, and why? For example, if it is published by a major university or recognizable research institution such as the Howard Hughes Medical Institute, then it might be quite trustworthy.
On the other hand, if it is from some institute you have never heard of or, even worse, some of the ones you have heard of (such as the Tobacco Institute or the Petroleum Institute or Hillsdale College or Harvard Business School), then it probably deserves grave suspicion of multiple lies in every word.
Thank-you for your time and attention. Hope this has been informative, entertaining, and thought-provoking.
Acknowledgements: much of this is merely dilute Karl Popper according-to-DrBoomer. Thanks to Witgren’s Bootcamps for example and inspiration.