“Right is right even if no one is doing it; wrong is wrong even if everyone is doing it.” -Augustine of Hippo
Science isn't the easiest endeavor you can undertake. Sure, the rewards are tremendous: you can wind up understanding any phenomenon in the Universe as well (or better) than any human has ever understood it before. But on your way there, you're going to have to do some of the most difficult work you've ever done. It isn't just mathematical and scientific work, either, but internal work on your own psyche. You'll need to learn how to be wrong.
No one comes into a scientific field already knowing all the answers; that's why we do the science in the first place. When you're just learning it, you put an incomplete number of puzzle pieces together, and your incomplete picture is usually incorrect. Or at least, less correct than the best picture. This means it's up to you to challenge your assumptions, revise your internal beliefs, and draw superior conclusions. The reward, if you can make it, is not just a better understanding, but the lesson of how to be wrong, and how to be better in the future.
It's a lesson that goes far beyond science, and can be applied to all areas of our lives. Imagine what the world could be like if we all did it.
The world of physics may seem bleak now to some but I think we maybe really close to TOE.
I'm afraid not. They thought they were very close to a TOE in the eighties and early nineties. Since then, Super Strings, M - Theory, etc...None of them have produced anything except excuses and more math, not new physics. This doesn't mean there isn't new physics, it just means they are going to have to back up the car (start discarding assumptions) until they are back on the road again. Once that happens, then you will see something new.
When you stack too many assumptions, if anything is even slightly off with your earlier assumptions, everything on top sort of...falls down.
Check it out!!:
"How To Be Wrong" is very simple. Don't assume you "know it all" already. Imagine being an unbiased scientist.
Then start a new blog.... not ***assuming*** a "start" or "beginning."... or that your opinions are established scientific facts.
Imagine an eternally oscillating universe. Forget your instruments and math expertise for that moment of un- programed contemplation
That's all I can give you for advice. I'm fairly sure that you will ignore it as always. "Too close to home" is the defense mechanism (just a minimum of "psychology".) I'll be gone soon. Not that interested in sci-fi.
There are many "real science" sites and discussions out there on the net. Best wishes for your sci-fi projects and clique of fantasy followers and pursuit of fame and fortune.
This may be my last post. Or not. Open the box to see if I'm dead or alive... or both until observed!
It's been "interesting."
Michael, why do you take a blog so personal? Ethans blog here is not meant to be be a "Final Solution" answer to everything.
Ethan's strength is his story telling ability within the science community. He just tells a story. he is not trying to prove he is the smartest person who knows everything.
I think you may be taking this blog a we bit personal..
Relax, appreciate the "story"...
I agree completely to everything you just said there.
Of course we should always remember lessons of our past to guide our future.
But I would also caution to always remember past experiences cannot be absolute proof for what can or cannot happen in the future.
@Ragtag Media #4,
I agree completely to everything you just said there.
Also I think we became like a family here, just like you had said before. I also think we are always having good time here. Why leave?
Can you imagine a world where humanity valued learning and revising your opinions as much as strongly-held convictions that were unshakable, no matter what the evidence indicated?
I'm in the process of teaching my kid lots of games. Like many small children, he doesn't like to lose. But the more games he plays, the thicker the skin he gets. And the more he does it, the more he thinks about the overall activity rather than the outcome of any specific game. "Daddy won, I'm upset" becomes "Daddy won 6 of the last 10 and I won 4...pretty good" hopefully will become in the future "I have no idea who's won more games this week. Play on!"
I think that's a lot like science. People who do a little of it, or who have one single idea they focus on, tend to worry about whether it's going to 'win.' Professional scientists, OTOH, tend more towards the attitude of "hey, 2 of my 50 papers have stood the test of time. Cool!" Or even "what, that paper of mine is still kicking around? I lost track. Who knew?" The activity becomes the focus, rather than the success or failure of any individual effort's outcome.
I think it's lovely to think about scientists considering as you speculate, but I would add one more thing. An experiment to test the speculation. Until you test it, all you really do have is an unfinished idea that can't carry it's own weight, much less some other theory. If it can't be tested, don't stack anything else on top of it, perhaps the scientist should try moving the research in a direction that can be tested. Right now we have a preponderance of speculation, almost sixty years of built up theory that may come crashing down because so many scientists kept treating the citing of each others papers as evidence of anything.
Okay CFT, where is the temporal cut-off? If I develop an hypothesis that can't be tested for another week, is it science? How about a year? How about 10 years? How about 100 years?
And let's say nobody can figure out how to test my idea right now. Looks completely untestable to human science. Is it science? But then next week some smart guy comes up with a method to test it nobody had thought of before. In that week, did my idea magically transform from non-science to science?
I agree that testability is a good criteria for science. But I don't think there's any bright line distinction. I think it's within reasonable science behavior to push the boundaries; to consider ideas that are currently untestable or that look to take decades to test by current technology. Because we can be wrong about 'untesable' or 'takes 50 years' just like we can be wrong about an hypothesis. Now no, this doesn't mean every single crank idea should be funded because it might pay off. We still need to prioritize what gets funding using reasonable criteria (and testability is one of them). but I don't think it's the only one or always absolutely necessary. A good national science portfolio should include some high risk, high payoff research. Not a lot, but not none either.
1. Thanks Ethan, for STEALING MY IDEA and writing it up better than I could.
2. Interesting that none of the comments here address being wrong. (Including this one.) Sure, it's epistemology, but....?
3. Paging Dr. Dunning & possibly Dr. Kruger. Dr. Dunning, white courtesy telephone please.
GregH: " 2. Interesting that none of the comments here address being wrong."
My comment above:
How To Be Wrong” is very simple. Don’t assume you “know it all” already.
Look a little closer.
Okay CFT, where is the temporal cut-off? If I develop an hypothesis that can’t be tested for another week, is it science? How about a year? How about 10 years? How about 100 years?
Perhaps a better word to use would be "falsifiability". If your hypothesis can be falsified, even if we can't test it now (but may be able to later), then I believe we can consider it scientific.
By definition, if its not testable it is not a hypothesis.
Doesn't mean its not science though. Non-hypothesis driven science is certainly viable.
Science- "the intellectual and practical activity encompassing the systematic study of the structure and behavior of the physical and natural world through observation and experiment."
For example, Ethans recent posts on multi-universes being maybe inevitable is science, even though apparently there is no way for us to know if they are real.
String theory research is science, but there is no evidence it is 'right'.
The list is endless.
Let us, for the sake of argument, assume that Science is the intellectual and practical activity encompassing the systematic study of the structure and behavior of the physical and natural world through observation and experiment.
What is the purpose of observation and experiment, if not to test a model/hypothesis/theory? I will lump these together with the understanding that a hypothesis is a weaker conjecture than a theory, and that a model is one manifestation. Ref here Sabine's recent critique of Inflation as producing more models than insight.
As I understand the process, if an experiment produces observations (data or numbers) that are consistent with predictions drawn from some theory, then the experiment can be said to corroborate, or even confirm the theory. Conversely, if the data is inconsistent with what one could expect (predictions) from a theory, then the theory is, if not actually disconfirmed, qualified.
As an example, I recall reading in a link from this blog that the discovery of CMB was the measurement that confirmed the Big Bang theory and disconfirmed the Steady-State theory. If that example is challenged due to the "experiment" not being assembled by humans, perhaps the human measurements of "the fluctuation in the CMB, of the large-scale structure of the universe galaxies clump, cluster and form" have tested and confirmed Cosmic Inflation. If these too are inadequate due to their naturalness, surely the LIGO (and now VIRGO) experimental apparatuses tested the prediction of Gravity Waves by theory of General Relativity?
If Einstein’s theory of General Relativity, Lemaître's "Cosmic Egg" (Big Bang theory) and the universe’s preceding Inflation as proposed by Guth (and fixed/enhanced by others) are Science, and they are Science because they could be tested (Karl Popper et al.), then models/hypotheses/theories that cannot (at least in principle) be tested are suspect as not being Science.
You're mixing words up here that are well defined and need to be careful. A hypothesis isn't a conjecture. Nor is it a theory. I use wiki quotes in the below so I don't mess up.
A hypothesis must, by definition, be testable - otherwise it isn't a hypothesis. Its is a proposition of fact based on observations/information. Its testability is the requirement of the scientific method as proposed by Popper. A theory is "a coherent group of tested general propositions, commonly regarded as correct, that can be used as principles of explanation and prediction for a class of phenomena:" When you have the theory, you can form hypothesis (or better still a null hypothesis) and test them, and as more of them are proven true the theory becomes stronger. A conjecture is "an opinion or conclusion formed on the basis of incomplete information." These terms have strict mathematical meanings, and there are proposed separate theories and conjectures and they are named as such.
Science itself is "the intellectual and practical activity encompassing the systematic study of the structure and behavior of the physical and natural world through observation and experiment." It encompasses all of the above.
A theory as a whole may not be testable, but parts of what it implies may be giving it credence. When there is enough evidence and no contradictions it eventually transitions to fact. If any observations contradict the theory, and the observation is sound and reproducible, then the theory must be abandoned.
As would call building equipment as part of science, and you don't need a hypothesis for that.
You said "What is the purpose of observation and experiment, if not to test a model/hypothesis/theory? " The purpose of observation and experiment may be to test a hypothesis (not a theory). The observations can be used to develop theories and conjectures. From those you then develop hypothesis and there we are full circle - the scientific method. (A model is not a theory or a hypothesis).
I'm more comfortable with the O.E.D.'s definition of theory (https://en.oxforddictionaries.com/definition/theory), where "hypothesis" is given as a synonym. If you use the O.E.D.'s definition of hypothesis, you'll see "theory" given as a synonym. Although you do agree with the O.E.D.'s definition of conjecture, you may be surprised to learn that both "theory" and "hypothesis" are given there as synonyms.
Let us, for the sake of argument, accept as true that a conjecture is an opinion or conclusion formed on the basis of incomplete information.
As no science I'm aware of claims to have complete information (TOE), then all science truths are provisional and tentative, some more than others. If you’re comfortable with that, then we’re in agreement about what a conjecture might be.
"… If any observations contradict the theory, and the observation is sound and reproducible, then the theory must be abandoned." While that may be true in theory (pun intended), I doubt many physicists have abandoned GR because of the non-locality of QM.
If it is true that "A theory is 'a coherent group of tested general propositions, commonly regarded as correct, that can be used as principles of explanation and prediction for a class of phenomena:' ", then it is false to claim "The purpose of observation and experiment may be to test a hypothesis (not a theory)". I side with Popper on this issue.
We may also have differing POVs about the word "model". I see models as being derived from a theory or hypothesis. The myriad models derived from the Cosmic Inflation hypothesis are what Sabine Hossenfelder took such grave exception to in her guest post here 29-SEP-17. While Ethan remains a stout defender of the hypothesis, even he ended his synopsis with "… the phenomenon of 'infinite model-building,' where theorists churn out model after model after model, predicting every imaginable outcome, and therefore, predicting nothing at all."
Hmmmm. Thanks. Food for thought! Later.
@ Steve Blackband #16
Well said. I agree.
"Theoretical Physics" has abandoned, on a large scale, the evidence required (eventually) to make it science. It's now modern metaphysics ("what if" conjecture) wearing the certified robes of science with doctorate degrees. (Just and "opinion piece".)
Michael Mooney #11:
I was wrong.
Theres a difference between words as just words, which are flexible and changing, and words we use in science that we assign (or at least try to assign) stricter definitions to.
Bucket is a synonym of pail, but it is not exactly the same. When there is a difference, Pail is actually a hypernym of bucket. Yea, I had to look that up too. After all its why we use math and not words to predict in science.
So as this guy says:-
"A theory is a set of statements that is developed through a process of continued abstractions. A theory is aimed at a generalized statement aimed at explaining a phenomenon.
A model, on the other hand, is a purposeful representation of reality.
As you can see, both share common elements in their definitions. What differs one from the other (in my opinion) is that one is aimed at generalized statements(theory) while the other is aimed as a helpful tool to understand specific phenomena(modeling).
Another way to link the two and point out differences is, a model is often used to describe an application of a theory for a particular case. Sometimes it involves a given set of initial and boundary conditions.."
Thus theory is a hypernym of model. Similarly Theory and hypothesis may be synonyms, but in science that does not mean they are the same thing. Clearly they are not (and so to with conjecture). In fact the best dictionary definitions say synonyms are the same or NEARLY the same.
All semantics I know, but important for the structure and process of science. As Mr Popper taught us.
Models have to be used carefully and 'sparingly' which I think is more along the lines of what Ethan meant when he wrote “… the phenomenon of ‘infinite model-building,’ where theorists churn out model after model after model, predicting every imaginable outcome, and therefore, predicting nothing at all.” I apologize for putting words into his posts.
I knew a guy who modeled a complex biological process using around one hundred differential equations - carbon metabolism actually. The problem was the model only had a few inputs, and a few outputs. We would tell him the inputs and he would use the model to predict the outputs. We would then do the experiment and get different results than his outputs. He would then wiggle some of the equations and voila, his model (post)predicted the results. The process would repeat. Totally useless. Clearly underparameterised in this case but less extreme cases have the same problem.
A theory is much more robust and must predict the observations, if not it must be abandoned. Models however get tweeked. Or probably more descriptively, twerked, endlessly. Oh no. Mental Chlorox please!
Whoops I stated that wrong way round. Sorry.
Bucket is the hypernym of pail.