home | index | units | counting | geometry | algebra | trigonometry | calculus | functions
analysis | sets & logic | number theory | recreational | misc | nomenclature & history | physics

Final Answers
© 2000-2020   Gérard P. Michon, Ph.D.

The Scientific Method


 Coat of arms of 
 Sir Francis Bacon  Coat of arms of
 Galileo Galilei
The strongest arguments prove nothing so long as
the conclusions are not verified by experience.

Roger Bacon   (1214-1292) 
 
Experimentum summus judex.
Albert Einstein   (1879-1955) 
 
Science is built up with facts, as a house is
with stones.  But a collection of facts is no more
a science than a heap of stones is a house.

Henri Poincaré   (1903) 
 
At the heart of science is an essential balance between two seemingly contradictory attitudes;  an openness to new ideas  (no matter how bizarre or counterintuitive they may be)  and the most ruthless skeptical scrutiny of all ideas, old and new.  This is how deep truths are winnowed from deep nonsense.
  Carl Sagan  (1934-1996) 
 
... You should always ask  "why not?"  but realize
that usually there's a goddamn reason why not.

  Murray Gell-Mann  (1929-2019) 
 Michon
 
 border
 border
 border

Related articles on this site:

Related Links (Outside this Site)

Cargo Cult Science  by   Richard P. Feynman   |   Pseudoscience (video)
How to Become a Good Theoretical Physicist  by   Gerard 't Hooft.
On the Allison Magneto-Optic Method  by   Michael Epstein & al.  (1999).
New Science Theory  by   Vincent Wilmot.
String theory:  The ultimate dead end? by   Robin McKie  (Oct. 8, 2006)
 
Alhazen (985-1040)   =   Ibn al Haytham: First Scientist  by   Bradley Steffens.
The First Scientist: A Life of Roger Bacon (1214-1292)  by  Brian Clegg.
 
Wikipedia :  Scientific Method   |   Pseudoscience   |   Cargo-cult science
Evolution of the Scientific Method  &  The Next Fifty Years of Science  (Video)  by  Kevin Kelly 
The Scientific Method and Ben Stein's 'Expelled' Movie

A Hiccup in the Scientific Method (13:29)  by  Jade Tan-Holmes  (2019-05-10).
Is Most Published Research Wrong?   Derek Muller  (Veritasium, 2016-08-11).
Biggest Idea #24:  Science (2:10:41)  by  Sean M. Carroll  (2020-09-01).

border
border

The Scientific Method

A scientific body of knowledge is called a  theory.  What makes a theory scientific is that it makes definite predictions which can be  disproved  experimentally.  The Scientific Method is a process which allows theoretical knowledge to evolve under experimental constraints.

A physical theory is not an explanation.  It is a system of
mathematical propositions, deduced from a small number of
principles, which aim to represent as simply, as completely,
and as exactly as possible a set of experimental laws.

Pierre Dûhem   (1861-1916)


whitehorse456   (Yahoo! 2007-08-08)   On the laws of Nature
Will the law of gravity always remain confined to a theory ?

Nothing can be stated clearly unless proper meanings are assigned to crucial words like  theory  and  law  in a scientific context.  In particular, the word "theory" is not an insult  (as in the  silly  saying "it's just a theory").  A  theory  is simply a consistent body of scientific knowledge not yet disproved by experiment  (at a given level of precision).

In experimental sciences, a theory can never be  proved  for certain; it can either be  disproved  by experiment  or be consistent with it.  This is precisely was makes a theory  scientific.  A statement that can't possibly be disproved by experiment may be highly  respectable  but it does no belong in an experimental science...  It might be mathematics, philosophy or religion, but it's not  physics.

Now that we have the basic vocabulary straight, we may discuss the example of gravitation.  Gravity is a physical phenomenon which is obvious all around us.  As such, it's begging for a  scientific theory  to describe it accurately and consistently.  A  law  is a rule within a theory  (like Newtonian mechanics).  Newton's own inverse square law of  Universal gravitation  describes gravity extremely well in terms of Newtonian mechanical forces.  Loosely stated:

Two things always attract each other in direct proportion of their
masses and inversely as the square of the distance between them.

We do know that the Newtonian theory does not provide the ultimate law for gravity.  General Relativity (GR) provides more accurate experimental predictions in extreme conditions  (e.g., a residual discrepancy in the motion of the perihelion of Mercury is not explained by Newtonian theory but is accounted for by GR).

Does this mean Newton's theory is  wrong ?  Of course not.  Until we have a "Theory of Everything"  (if such a thing exists)  any  physical theory only has a limited range of validity, where its predictions are accurate at a stated level of precision  (stating the claimed precision is  very  important in Science; an experimental prediction is  meaningless  without it).  The Newtonian theory predicts the motion of planets in the Solar System to many decimal places, over eons...  That's all we ask of it and that's what makes it so valuable.

Even General Relativity is certainly  not  the ultimate theory of gravitation.  We know that much because GR is a  classical theory, as opposed to a  quantum  theory  (i.e., a theory obeying quantum logic).  So, GR is not mathematically compatible with quantum phenomena which are so obvious at a small scale.

Science is just a succession of better and better approximations.  This is what makes it nice and exciting.  If we insisted at all times on "the whole truth and nothing but the truth" no meaningful scientific statement could ever be made.

As a consistent body of knowledge, each  theory  allows you to make such statements freely, knowing simply that the validity of your discourse is only restricted by the general conditions of applicability of that theory.  Without such a framework, scientific discourse would be crippled into utter uselessness.

Video :  Newton's laws can not be proven  by Walter Lewin  (8.01 MIT lecture, Fall 1999)


(2007-07-14)   Controlled Experiments   (Francis Bacon, 1590)
It's  easier  to isolate the causes of a phenomenon when we create them.

Experiments are now so intertwined with Science that we may forget that good Science can be performed without experiments.  There are no experiments in  Astronomy.  Astronomers have to be content with observing phenomena over which they have no control.

The same can be true, albeit less systematically, of other sciences as well.  There was a time when particle physicists relied on random cosmic rays of unknown origin for something worth observing.

Similarly, experiments in evolutionary biology can only be performed on simple lifeforms that reproduce so rapidly that a large number of generations can be observed in a reasonable amount of time.  The power of natural selection established under such controlled conditions can be extrapolated to more complex lifeforms with longer lifetimes just like physical laws established in the laboratory are extrapolated to astronomical objects.

Paradoxically, the aim of Science is to provide ways to  dispense with  countless experiments.  For example, it is now pointless to play with objects like gears, wheels or magnets in the hope of producing  perpertual motion  when we have established to a fabulous degree of precision that all the simple components of such systems obey conservation laws that preclude it  (including relativistic mechanics, electromagnetism, thermodynamics and quantum theory).  Scientific advances come from discovering new aspects of reality  (e.g., radioactivity in 1896)  new paradigms  (e.g., quantum logic)  or clever ways to harness nature  (modern technology).

From a pedagogical standpoint, it's useful to repeat simple-minded experiments to stress the fact that Science is ultimately based on observation, not dogma.  Of course, no one will ever be able to actually repeat  all  the experiments on which basic scientific knowledge is based, but we should at least be satisfied that we  could  reproduce any key experiment we like.


Occasionally, a pedagogical demonstration turns into a flop which the teacher may or may not explain correctly...  Watch carefully the videos quoted in the footnote below, where Julius Sumner Miller (1909-1987)  makes the following statements pertaining to the time it takes for things to roll down an inclined plane, starting from rest:

  • All hoops roll alike  (hollow cylinders).
  • All disks roll alike  (homogeneous cylinders).  They beat all hoops.
  • All spheres roll alike  (homogeneous spheres).  They beat all disks.

He then proceeds to demonstrate this on a wooden inclined plane.  In some cases, he  fails  miserably but allows the showmanship to prevail over his scientific training, using apologetic sentences like:

  • "It will be a little difficult to check this." 
  • "If our provisions for Nature were of an adequate sort, I could show you the following phenomena..."
  • "We have a little trouble with this board, but that's another piece of evidence that, when you deal with Nature, you must meet Her requirements...  A little difficult to show it  (the hazards of experiments)  but I'm going to tell you: All spheres beat all disks."

Can you discover the fallacy that  Julius Sumner Miller  chose to ignore?

Video :  Julius Sumner Miller - Physics - Rolling Things   Part 1  |  Part 2


(2007-07-15)   History of the Scientific Method
The key ideas behind modern science.

Mathematics serves as the language of all natural sciences, but it's not itself a natural science.

 Come back later, we're
 still working on this one...
  • Mathematical Theorems.  (Thales, 600 BC)
  • Library science and information management.  (200 BC)
  • Scientific methodology.  Alhazen / Ibn al-Haytham (965-c.1040).
  • Collaborative Encyclopedia. (1000)
  • Observation rather than received wisdom.  Roger Bacon (1214-1294).
  • Skepticism  (Francisco Sanches, 1581).
  • Controlled Experiment  (Francis Bacon, 1590).
  • Experimental Laboratories. (1600)
  • Observational Tools.  Microscope, telescope (1609)
  • Society of Experts.  (1650)
  • Scientific Repeatability.  (Bolton, 1665)
  • Peer review.  (1675)
  • Hypothesis / Prediction.  (Newton, 1687).
  • Falsifiable Testability.  (Karl Popper, 1920).
  • Randomized Design.  (Fisher, 1926).
  • Control Placebo (1937) and double-blind clinical protocol (1950).
  • Computer Simulations.  (1946).
  • Investigating the Scientific Method.  (1962)

 Come back later, we're
 still working on this one...

(2007-07-16)   Make-Believe Science:  Pseudo-science
Nonscientific approaches which look respectable  (but aren't).

Allow me to pay tribute to my younger self.

The French  Concours Général  was instituted in 1744.

More than a quarter of a millenium ago, the year 1744 marked the 101st anniversary of the birth of Isaac NewtonJames Stirling was 52, Maupertuis was 46, Linné, Euler and Buffon were 37, Franklin was 35, Jean-le-Rond d'Alembert was 27, Maria Agnesi was 26.

That year, an academic competition was created in pre-revolutionary France which is still in existence:  Every year, selected graduating high-school students compete in what's called the Concours général  (short for  Concours général des lycées et collèges).

The institution temporarily disappeared during the French revolution and, again from 1904 to 1922, because some top  lycées  reportedley placed too much emphasis on it.  It was threatened again in the early 1980s, as the socialist government of France didn't consider it  politically correct  to stress excellence.  (In 1983, the French minister of education didn't see fit to attend the award ceremonies of the  International Mathematical Olympiads,  which were held in Paris.)

In recent years, the  Concours Général  has regained much of its former prestige, and then some...

Concours Général  of 1973  (philosophy for science majors):

At the age of 17, I had the honor to represent my school in mathematics and philosophy  (I was coached after hours by my philosophy teacher,  Bernard Lefebvre).  Against some expectations, I didn't do very well in mathematics but I nailed it in philosophy and made the front page of the local paper  (although I didn't win the top prize;  two other people, somewhere in France, were ranked ahead of me).  The question we were asked was:

In what sense can we distinguish between true and false sciences ?

Thus, even at that young age, I must have known already...  I must have known that there ought to be more to Science than scientific appearances !

As Science progresses, the time it takes for anyone to obtain a working knowledge of the state-of-the-art in any field keeps increasing but, arguably, motivated teenagers can still get a fairly good grasp of what distinguishes good science and  pseudoscience.  Yet, adults and trained scientists may go astray...

It's probably uncharitable to expose people who are trying to make a scientific contribution without the benefit of a proper scientific education.  Unfortunately, it seems next to impossible to provide some wannabe-scientists with the elements that would make them wise critics of their own  theories.  Even at my own modest level of notoriety, I am regularly confronted with that nagging problem.  A few gurus and luminaries chose a humorous approach to the issue:

On the other hand, some trained scientists who make outrageous claims may fool everybody except their better peers.  They can even fool themselves  If they don't fool themselves, they are simply liars or crooks and things are easier to sort out!


(2011-10-22)   Faster-than-light neutrinos?  Not ready for  prime time.
Scientific investigation goes on  in spite of  distorted media reports.

Extraordinary claims require extraordinary proof.
Marcello Truzzi (1935-2003)

In a CERN press release dated  2011-09-23  (which has been corrected and updated several times since)  it was originally announced that the OPERA team at Grand-Sasso was reporting the detection of neutrinos traveling faster than light, as part of a experiment conducting jointly with CERN.

This careless release  (whose original content went against the very core of modern physics)  started a media circus to which virtually all scientists with a modicum of media visibility had to reply while the error was being corrected.  Eventually, at least two OPERA leaders lost their jobs.

Faster-than-light neutrinos?  in "Cosmic Variance"  by  Sean M. Carroll  (Caltech)
Science Now  |  Scientific American  |  The Guardian  |  Voice of America  |  New Scientist  |  Chicago Tribune
Videos :   Sixty Symbols   |   Tony Darnell   |   Marcus du Sautoy

Epilog :

border
border
visits since April 2, 2010
 (c) Copyright 2000-2020, Gerard P. Michon, Ph.D.