Theoretical vs. Experimental Physics

[Proton Soup]

Yeah, I just think experiments in general are boring (including in chemistry, biology, and physics). They are very tedious and time consuming when I have to do them in school. And you know, the reason for doing experiments in school is to solidify what you've been taught on a specific subject (so I've heard), but actually, experiments never solidify it; they just waste my time. I'm doing an experiment when I could be looking at the textbook some more and thinking about the concept.

that's a bit like thinking you can become a better driver simply by reading more and thinking about it. there is skill involved. and like driving, lots of little nuances that simply don't show up in the textbook. you may not think you are learning anything, but you are, even if the point is not apparent to you now.

 

[isabelle]

Although, as some others have said, being a lead experimentalist is easily as exciting as being a lead theorist. In fact, the reason that I choose physics instead of math, and in particular to specialize in condensed matter theory, is that as exciting as it is to prove a new theorem it is even more exciting (for me) when the new theorem implies the existence of something in the real world which you then go and find.

The most beautiful moments in the interaction between theory and experiment in physics have no parallel for me in mathematics, here are two examples:

(1) The prediction from Maxwell's equations of E&M waves that travel at the speed of light.

(2) The QED first-order correction to the magnetic moment of the electron.

In (1) we can measure the strength of electrical and magnetic forces between known charges and currents and relate these to the speed of light, measured independently!

In (2) we use what at first appears to be an absurdly elaborate scheme, involving dozens of integrals that appear to diverge to infinity until you fix them with a bizarre prescription, and at the end of the day you end up with the most precise agreement between any measurement and any prediction in all of history.

There are many others, found in the early days of the statistical theory of gases, or the development of quantum mechanics (which is obviously the best tested theory of all time), critical exponents for phase transitions, etc. As a joke, the allure of modern theoretical physics is to make predictions that are so crazy that no reasonable person would believe them, and that people only take seriously after they are shown by experiments to be true!

 

[humanino]

development of quantum mechanics (which is obviously the best tested theory of all time)

Can you back that up with the best measurement you know for quantum mechanics, and the best measurement you know for general relativity ?

 

[humanino]

Well, maybe experiments will be more enjoyable once I get into college, but for now, experiments just take up time in the day.

Make sure you get into a decent college first.

 

[Englishman]

As for experiments in school, I agree with you 100%. As a TA I have to waste my time babysitting college kids in lab classes that they hate and derive no benefit from. When they ask me a question, they don't have any curiosity, they just want to know how to follow the procedure so they can finish and get on with their lives. College physics labs (at my major state university) are nothing more than busy work to make students feel like they are getting something for their (outrageously large) tuition fees.

Then, at the graduate level, my classmates who are pursuing their PhDs in experimental physics spend years in a basement laboratory assembling electronic circuits and calibrating optical benches (lens, mirrors, lasers, etc). One day they hope to become Principal Investigators with their own laboratories, but that day is far in the future and cannot happen for everyone. On the otherhand, at least they get paid to assemble circuits while I have to teach to support my theoretical studies.

By the way, I really like teaching, but only when the students care about learning, which doesn't happen in physics labs...three hours of pointless torture several times a week, and then I spend hours grading the lab reports that the students could care less about, it's a pathetic system.

Wow, even in college. That's exactly how I feel; it's just an absolute waste of time.

What is a TA ?

 

[Englishman]

Make sure you get into a decent college first.

Well, I'm planning to go to Stanford, and that's definitely an above average college.

 

[humanino]

One thing I don't understand, is a lab should be given by a theorist.

Englishman, did you ever pay a visit to a real research lab ? Those guys build things that have never been built before, and discover stuff that has never been seen before. Down in the experimental lab, you can meet real artists, engineer who it so happens also display in national museum on week end, or have patents they sell to their fellow world-known professional athletes. I know personally an engineer who designed windsurfing boards. We receive visits several times a week. I don't remember any visitor not having their eyes wide open.

 

[Englishman]

that's a bit like thinking you can become a better driver simply by reading more and thinking about it. there is skill involved. and like driving, lots of little nuances that simply don't show up in the textbook. you may not think you are learning anything, but you are, even if the point is not apparent to you now.

It is not like thinking I can become a better drive by simply reading more and thinking about it. This is completely different. I believe that you understand that there are two ways of doing physics: theoretically and experimentally. As a physicist, one can solely be doing theory (being what is called a theoretical physicist), and accomplish great things in physics (Einstein didn't do one physical experiment). And besides, I did not say that I am not learning anything; I said that doing experiments does not solidify what I have learned.

 

[Englishman]

One thing I don't understand, is a lab should be given by a theorist.

Englishman, did you ever pay a visit to a real research lab ? Those guys build things that have never been built before, and discover stuff that has never been seen before. Down in the experimental lab, you can meet real artists, engineer who it so happens also display in national museum on week end, or have patents they sell to their fellow world-known professional athletes. I know personally an engineer who designed windsurfing boards. We receive visits several times a week. I don't remember any visitor not having their eyes wide open.

I have never been to a real research lab.

 

[Proton Soup]

It is not like thinking I can become a better drive by simply reading more and thinking about it. This is completely different. I believe that you understand that there are two ways of doing physics: theoretically and experimentally. As a physicist, one can solely be doing theory (being what is called a theoretical physicist), and accomplish great things in physics (Einstein didn't do one physical experiment). And besides, I did not say that I am not learning anything; I said that doing experiments does not solidify what I have learned.

maybe the point is not simply to solidify what you have learned.

anyway, i think we have come to the point: that you simply don't like experiments. which is fine, you may also hate taking your humanities courses. but there also, you will be learning something whether you see the value in it or not.

 

[whybother]

maybe the point is not simply to solidify what you have learned.

anyway, i think we have come to the point: that you simply don't like experiments. which is fine, you may also hate taking your humanities courses. but there also, you will be learning something whether you see the value in it or not.

Even as someone who doesn't really enjoy experimentation, in high school, and even as a underclassmen in college frankly, you just haven't seen enough of anything yet to know if you really like or dislike experimental physics. English shouldn't judge yet, because honestly, there is so much out there.

 

[humanino]

I have never been to a real research lab.

Well you see, you might change your mind about

I just think experiments in general are boring (including in chemistry, biology, and physics). They are very tedious and time consuming when I have to do them in school.

If you had been lucky enough to have a passionate professor whose eyes shine when he explains to you in late hours an experiment he especially designed to trick you into thinking it would be trivial, you might have had a different conception of what experiments feel like.

 

[isabelle]

Can you back that up with the best measurement you know for quantum mechanics, and the best measurement you know for general relativity ?

For me QM is the best tested theory because it has been applied in a huge variety of experiments that go far beyond the experiments that were envisioned by the creators e.g. fractional quantum hall effect, GHZ experiments, etc.

I think GR is theoretically strong, since SR is very well tested and GR is basically the unique diffeomorphism invariant extension thereof. As for tests of anything lik the full theory (and not just the first post-Newtonian term in the potential which goes as 1/r^2 whose coefficient was correctly guessed empirically before Einstein to solve the problem with the precesion of mercury), we still have not detected classical gravitational waves :(

To answer the question directly, I would say that the anomalous magnetic moment of the electron is the most precise measurement of QM (sure it involves QED, but QM is an essential part of relating this to what is measured so if QM was wrong to 1 part in a billion then this would ruin the measurement of the magnetic moment). As for GR, afaik the best strong-field precision tests depend on observations of binary pulsars and are not more accurate then 2 or 3 sig figs.

One thing I don't understand, is a lab should be given by a theorist.

As I explained, the experimental students get paid to do research i.e. assemble circuits and tune lenses (their words, not mine!), and I get paid to teach undergraduates; getting grant money for doing theory as a grad student is difficult (certainly not impossible but there is more money for experimental 'research' i.e. grunt work in the beginning).

What is a TA ?

A teaching assistant, which means I roughly have the same role for college students as your high school teacher in your classroom (professors at big universities do not do most of the teaching, it is done by graduate students like me).

Proton Soup, I disagree with your 'you always learn something, whether you realize it now or not.' After all, there is such a thing as a total waste of time. For me, a total waste of time occurs when people are doing something they don't want to do and no one in the world is benefiting. I have had many hours to sit in lab and reflect on the extent to which school labs match this description. So far I have not found any students who like being in lab, or who will do anything beyond what is necessary to 'get the grade.'

 

[Englishman]

maybe the point is not simply to solidify what you have learned.

anyway, i think we have come to the point: that you simply don't like experiments. which is fine, you may also hate taking your humanities courses. but there also, you will be learning something whether you see the value in it or not.

You don't understand. You keep saying that I believe I'm not learning anything. No, that it not it. I said that it does not solidify what I have learned. And I never said that there was zero value in experiments. Please don't twist my words.

 

[George Jones]

For me QM is the best tested theory because it has been applied in a huge variety of experiments that go far beyond the experiments that were envisioned by the creators e.g. fractional quantum hall effect, GHZ experiments, etc.

I agree, but, according to Roger Penrose (who might not be unbiased), when it comes to best measurements, GR wins.

As for GR, afaik the best strong-field precision tests depend on observations of binary pulsars and are not more accurate then 2 or 3 sig figs.

In his Road to Reality, Penrose writes about the original system discovered by Taylor and Hulse:

"a confirmation of Einstein's general relativity to about one part in 10^14, an accuracy unprecedented in the scientific comparison between observation of a particular system and theory."

 

[humanino]

I agree, but, according to Roger Penrose (who might not be unbiased), when it comes to best measurements, GR wins.

I read the same, was surprised at that time, and hoped that maybe a gravitational expert might comment

 

[Proton Soup]

You don't understand. You keep saying that I believe I'm not learning anything. No, that it not it. I said that it does not solidify what I have learned. And I never said that there was zero value in experiments. Please don't twist my words.

well, there's always mathematics

 

[robphy]

http://arxiv.org/abs/astro-ph/0407149
"Relativistic Binary Pulsar B1913+16: Thirty Years of Observations and Analysis"
J.M. Weisberg, J.H. Taylor
and
http://www.astro.cornell.edu/academics/courses/astro201/psr1913.htm

might be helpful