Engineering: Success Through Accuracy, Not Popularity

[K. Doc Holiday]

The scientific process.

An engineer once told me a joke:

"A scientist is considered a success when his peers agree with him. An engineer is considered a success when he actually gets it right."

After a physicist's comment today I am starting to think it wasn't a joke. The physicist said if enough people can be convinced of the validity of someone's theory it is accepted and propagated as true, regardless of whether or not it is true.

I thought science was supposed to be objective. Doesn't this method sound more like a popularity contest? Is science a democracy? Get enough people to vote for your theory and that somehow makes your theory right?

Gotta say, I am in shock.

Maybe I should spend my free time studying engineering?

 

[mfb]

Physics is not about "true", it is about having a good model. In order to get accepted, a theory has to pass all experimental tests. If a theory passes all experimental tests, it is a good model. And people will continue to do experiments trying to contradict the theory.
There is nothing democratic about it.

How does the engineer know that he got it right? The results pass all experimental tests. Oh wait...

 

[dkotschessaa]

Physics is not about "true", it is about having a good model. In order to get accepted, a theory has to pass all experimental tests. If a theory passes all experimental tests, it is a good model. And people will continue to do experiments trying to contradict the theory.
There is nothing democratic about it.

How does the engineer know that he got it right? The results pass all experimental tests. Oh wait...

Ideally this is how it is supposed to work. In reality there are untestable theories, plenty of politicized science, and junk papers get published with alarming regularity.

The scientific process is not to blame, but the people involved are.

-Dave K

 

[mfb]

There are approaches that do not yet produce testable predictions - they are still in development. No one says "string theory is right".

There is junk everywhere, but in physics the rate is very low.

 

[Baluncore]

Physics is not about "true", it is about having a good model.

Which reminds me of the reductionist's statement; "Science is truth. Do not be mislead by fact."

 

[SleepDeprived]

I was having dinner with a friend of mine a long time ago and we were discussing religion, and then he said well science is like a religion too. I told him, although not directly disagreeing, that in science people are willing to change if a better theory comes along.

I read the book "The Schrodinger's cat" a long time ago, and in the book, it talks about a true occurrence that happened once in a quantum physics conference. The scientists were playing a game in which one can only ask "yes" or "no" question. So everyone was supposed to agree on an object and the person has to guess what's the object is based on the "yes" or "no" answer. Once time, they decided to change the rule and that they will not agree on an object before hand, but whoever answer the "yes" or "no" question has to be consistent with the previous answer. After 25 questions later, the contestant was able to guess the "correct" object. I guess this story illustrates the nature of our scientific progress as viewed from a perspective that we can never understand the truth nature of our universe but can only "guess" based on how it affects our experience.

Anyway, back to the OP original post, I don't think we can discount any aspects of science whether it is physics, mathematics, or engineering ... they all play a part in the modern world. I would like to use an example.
In electrical engineering, all the recent progress could not be realized without the help of simulation software. For example, to design effective circuits, circuit engineers need to simulate using either SPICE or or full 3D analysis. SPICE or 3D are based either directly indirectly from Maxwell equations, but Maxwell equations alone is not enough because there are a lot of other types of math needed to carry out the simulations. And obviously Maxwell, math themselves are not enough because you would need to understand a lot of software to put these all together. But obviously that is not enough because you need very fast CPU and hardware to be able to realize these complex simulations. So you can see they all sort of tied together. One interesting thing is people who created these software are not necessary circuit design engineers (as a matter of fact they probably don't have the mind set to do circuit designs) but their end products are indispensable in circuit design.
And obviously circuit design engineers use these software to create even faster hardware to enable even more sophisticate software ...

 

[Baluncore]

Is science a democracy? Get enough people to vote for your theory and that somehow makes your theory right?

No. A theory is never “right” it is only ever “accepted” as part of the current paradigm by your peer group. There is nothing to say that any accepted paradigm will ever be absolutely “right”.

A science student's answers in an exam may be “right” or “wrong”. But the exam questions are testing the students understanding of the current paradigm, not the absolute truth of anything.

Physicists are never satisfied. They either want better students or they spend their time trying to find better models of observed reality.

There is junk everywhere, but in physics the rate is very low.

“Junk science” is any ill informed or paralogical attempt at making sense of the real world.

“Private theories” is the term given to the recurrent tendrils sent out by scientists in search of “better” paradigms. The community of scientists try to keep them private because public discussion would confuse the science and the students, who must learn to crawl the current paradigm, before they start to walk or teach it.

Judging by the “many truths” of post-modernist analysis, and from the history of science, the paradigm of the day will later be seen as no more than a happy fairy tale. The paradigm is accepted until it has served it's purpose, that purpose is to provide a wobbly stepping stone to another, apparently more secure paradigm.
And so science advances.

 

[dkotschessaa]

And so science advances.

"...one funeral at a time."