Where do all these formulas really come from?

  • Thread starter Gersty
  • Start date
  • Tags
    Formulas
In summary, physicists do not rely solely on empiricism or curve-fitting when developing theories. They start with an idea and then use mathematics and logic to predict and explain experimental results. The goal is to find a mathematical equation that can be used to accurately describe and predict physical phenomena. This equation must be able to be calculated from an underlying theory and fit with experimental data in order to be considered a valid theory.
  • #1
Gersty
47
1
So how did people like Einstein, and Newton really do it on a day to day basis? How do physicists really do physics? Is it a bunch of scientifically minded mathematicians sitting around doing lots of math until the results they get match experimental ones, or is it more a matter of doing things like plotting data on graphs, coming up with best fit lines and then finding the formula for the line like we all do in high school math class?
 
Physics news on Phys.org
  • #2
I suppose a strong interest and aptitude in mathematics is most important, but I also see as vital a strong desire to find and define mathematical relationships.
Perhaps those people see a repeatable phenomenon, find out that that it has not previously been described mathematically, and then go for it... days, months even many years. Sometimes with collaboration.
 
  • #3
Gersty said:
So how did people like Einstein, and Newton really do it on a day to day basis? How do physicists really do physics? Is it a bunch of scientifically minded mathematicians sitting around doing lots of math until the results they get match experimental ones, or is it more a matter of doing things like plotting data on graphs, coming up with best fit lines and then finding the formula for the line like we all do in high school math class?

Well, no. Curve-fitting would be pure empiricism. While that's sometimes useful, you can only find really obvious relationships with it. Otherwise you're just stuck with a re-statement of what you already knew: "The curve looks like this". Empirical formulas can sometimes help inspire theories (e.g. Rydberg's formula inspired the Bohr model of the atom), but they're not scientific theories themselves. They're just a sort of summation of observed information.

Physical theories start with a theory - an idea about how things work. Basically "What if...". This idea can then be put into mathematical form, and using maths/logic you work out the consequences of this assumption and get some predictions. If they match the experimental results, or at least tell us something about the experimental results, then it's a useful theory. If not, then it gets thrown in the wastepaper basket. (It's easy to get the idea from textbooks that Science follows a nice straight path. In reality it's a very twisted path, full of dead-ends and failed attempts. It's just that those get forgotten.)
 
  • #4
Here is one specific example. In 1696, Isaac Newton was challenged to solve the following problem. Consider a bead sliding without friction on a curved wire under the influence of gravity from some starting point x,y to a lower point x', y'. Find the shape (path) of the wire such that the transit time of the bead is a minimum.

Newton thought about it, and (supposedly) solved it in a day. (How long would it take you?) The problem is now known as the Brachistochrone problem. See

http://mathworld.wolfram.com/BrachistochroneProblem.html

Because the proper math had not yet (in 1696) been developed, it is also rumored that Newton also invented the method called the Calculus of Variations (in the same day?).

So empirical curve fitting was not used; the solution came from using advanced calculus.

Bob S
 
  • #5
A theory is still just a theory until it gets compared to data points. At some point, you need curve fitting or some other kind of comparison with experimental data. Whether it's empirical or not depends on how the fitting function is defined. If it is defined only by the fit, it's empirical. But if it can be calculated from an underlying theory, and the resulting curve fits the data points, then we're getting serious. Additional fitting curves from the same theory will serve to validate it. This is physics.

The game is often to construct an equation that, when when solved with the proper boundary conditions, gives functions that can be fitted with experimental data.

Of the top of my head, this applies to at least to Newton's laws, Maxwell's equation's, and Shrodinger's equation. Einstein's GR would also be useless if hadn't at some point been compared to experimental data (astronomical observation in this case).

But IMO, finding the right equations is simply a matter of educated guess (or well-thought out) work by people who are familiar with the mathematical tools.
 
  • #6
"Physical theories start with a theory - an idea about how things work. Basically "What if...". This idea can then be put into mathematical form, and using maths/logic you work out the consequences of this assumption and get some predictions."

This is the part I'd like to examine. Turning an idea into a mathematical equation. At first, it seems pretty straight forward when you're dealing with simple and easily observable phenomena: "If I push twice as hard, the object will move twice as fast." But the universe rarely seems so simply, and yet it seems that many physics formulas boil down to simple stuff like inverse square laws and straight direct proportionality. I mean most of Newtonian kinematics is simple multiplication and division, half of this times the square of that, and so on...
 
  • #7
Dr Lots-o'watts said:
A theory is still just a theory until it gets compared to data points. At some point, you need curve fitting or some other kind of comparison with experimental data. Whether it's empirical or not depends on how the fitting function is defined. If it is defined only by the fit, it's empirical. But if it can be calculated from an underlying theory, and the resulting curve fits the data points, then we're getting serious. Additional fitting curves from the same theory will serve to validate it. This is physics.

The game is often to construct an equation that, when when solved with the proper boundary conditions, gives functions that can be fitted with experimental data.
That's not what "curve fitting" is. Curve fitting is a mathematical method for building an equation to fit data. If you already have the equation and are trying to check it, you simply plug-and-chug and compare the predicted result to the experimental one.
 
Last edited:
  • #8
Gersty said:
"Physical theories start with a theory - an idea about how things work. Basically "What if...". This idea can then be put into mathematical form, and using maths/logic you work out the consequences of this assumption and get some predictions."
Agreed

This is the part I'd like to examine. Turning an idea into a mathematical equation. At first, it seems pretty straight forward when you're dealing with simple and easily observable phenomena: "If I push twice as hard, the object will move twice as fast." But the universe rarely seems so simply, and yet it seems that many physics formulas boil down to simple stuff like inverse square laws and straight direct proportionality. I mean most of Newtonian kinematics is simple multiplication and division, half of this times the square of that, and so on...
When I first came across Newton's solution to the Brachisochrone problem (post #4), I could not believe that math could be so powerful. It certainly involves a lot more than simple multiplication and division. Newton had to "invent" calculus of variations in order to solve it. And he did it in (rumored to be one day in) 1696.

Bob S
 
  • #9
russ_watters said:
That's not what "curve fitting" is. Curve fitting is a mathematical method for building an equation to fit data. If you already have the equation and are trying to check it, you simply plug-and-chug and compare the predicted result to the experimental one.

If you have a theoretical function that fits the first time, then great. If not, curve fitting may help make corrections to the model. Either way, the goal is to find the function that fits the data, and to be able to derive that function from general postulates. How to do that is up to you.
 
  • #10
Dr Lots-o'watts said:
If you have a theoretical function that fits the first time, then great. If not, curve fitting may help make corrections to the model.
"Model" = "function"
Either way, the goal is to find the function that fits the data, and to be able to derive that function from general postulates.
You're getting the order of the steps in the scientific method mixed-up. If you have a theory to test, you have already derived a mathematical model/function to which you are looking to match data.
 
  • #11
I suppose it could be said that a repeatable phenomenon is required for postulate formulation.
 
  • #12
Richard Feynman said:
First you guess. Don't laugh, this is the most important step.
Then you compute the consequences. Compare the consequences to
experience. If it disagrees with experience, the guess is wrong.
In that simple statement is the key to science. It doesn't matter
how beautiful your guess is or how smart you are or what your name is.
If it disagrees with experience, it's wrong. That's all there is to it.
Thats how they do it ;)
 

1. Where do all these formulas really come from?

Formulas are derived from scientific principles and theories that have been tested and proven through experiments and observations. They are used to describe and predict natural phenomena in a concise and mathematical way.

2. How are new formulas discovered?

New formulas are discovered through scientific research and experimentation. Scientists use data and observations to identify patterns and relationships, which can then be expressed in the form of a formula. Collaborative efforts and advancements in technology also contribute to the discovery of new formulas.

3. Are all formulas universal?

No, not all formulas are universal. Some formulas are specific to a certain field or topic, while others may only apply under certain conditions. It is important to understand the context and limitations of a formula before applying it to a problem.

4. Can formulas change over time?

Yes, formulas can change over time as new evidence and research emerge. Scientists are constantly refining and improving formulas to better describe and explain natural phenomena. However, fundamental principles and laws in science remain unchanged.

5. Do scientists always create new formulas?

No, scientists do not always create new formulas. Sometimes, existing formulas are sufficient to explain and predict a phenomenon. However, when new discoveries are made or when existing formulas are found to be inadequate, scientists may create new formulas to better represent the natural world.

Similar threads

Replies
5
Views
2K
Replies
2
Views
8K
  • STEM Educators and Teaching
Replies
5
Views
544
  • STEM Academic Advising
2
Replies
60
Views
3K
  • Introductory Physics Homework Help
Replies
20
Views
2K
Replies
2
Views
8K
Replies
204
Views
2K
Replies
7
Views
1K
  • Other Physics Topics
Replies
9
Views
3K
  • STEM Academic Advising
2
Replies
43
Views
4K
Back
Top