Transterrestrial Musings

Comments

Someone quick, pull Jane's foot out of her mouth on the first quote. And stick around, she just may have done it again. ;-)

45 microseconds Jane, 45 massive microseconds of relative clock error per day. Try locating a mobile without this having been accounted for in the design. Why, you might even think terrestrial triangulation was a better solution. You might even by a fuddy duddy old fashioned Geometrix Wireless Location System instead of the fancy GPS Nav system in your car.

Oops, but that may not work either. Oh, Oh, the Time Reference for synchronization was picked up from a GPS signal!

Posted by Toast_n_Tea at May 17, 2007 04:15 PM

Mmmm. Yummy, tasty toes.

Okay, can I hide behind "practically nothing?" And I hedged a little in my response in the other thread.

And toast, you forgot the 7 additional microseconds of special relativity correction as well.

I don't deny for a moment that General Relativity is essential in solving some sophisticated orbital mechanics problems, but I don't think it HAS to be taught at the undergraduate level. I think it's fine to defer to graduate school. Although it's really up to the department at your particular university.

Regarding decisions about whether you need to incorporate General Relativity into your problem, I think we can rely on people who are specialists in the are to understand the problem thoroughly. Although if NASA is giving flight trajectory responsibility to undergraduate new hires, then maybe that explains a lot.

Posted by Jane Bernstein at May 17, 2007 04:37 PM

if NASA is giving flight trajectory responsibility to undergraduate new hires, then maybe that explains a lot.

Jane, not (necessarily) to excuse NASA hiring, but if I'd been hiring at JPL, I would have expected graduates, even undergraduates, with physics degrees, to understand at least the implications of General Relativity, and its historical context, even if not being expert at applying it, and not being experts at the time. My impression is that even this is not even being taught (even if you and the other posters objecting aren't physics majors).

You remain completely wrong. It's OK, it happens in the best of families. Even I have admitted it, occasionally...

Posted by Rand Simberg at May 17, 2007 04:45 PM

This is really funny (in a negative way) because the Gravity B probe has confirmed part of the frame dragging, with the other part to be verified by December. If this checks out (which I think it well), it will conclusively "prove" GR as much as a scientific theory can be proved.

I would think that anyone interested in space development would want to be intimately knowledgable of GR, because any potential method of FTL (or wormholes) must be consistant with GR.

Also, it is worth noting that Einstein got his inspiration for deriving GR from Mach's principle.

Posted by Kurt9 at May 17, 2007 05:26 PM

I think the reason GR is not taught to physicists is that the math is generally used only for GR. In the absence of astronomers, you can often learn more GR from the math department than from the physics department. I don't remember any physics courses where the implications of anything were taught in a qualitative sense. It was math first. No math, no physics.

I have a little project on the back burner to teach some of the math behind GR in the context of problems that undergraduate physics majors normally study. You can do some cool things. If the students get the math, I can con an astronomer into teaching the GR. Unfortunately, projects like this don't pay the bills, especially for nontenured faculty like me.

My understanding is that many cell phone operations now use timing signals provided by the GPS system. Oddly enough, the equations used to correct the raw satellite signals for the effects of GR are often trade secrets. No telling what will happen if the secret of GR gets out.

Steve Langford
Department of Physics and Astronomy
Washington State University

Posted by Steve Langford at May 17, 2007 09:11 PM

Steve, where is WSU located? Spokane?

Second, FTL applications aren't particularly useful since we currently don't have an inkling whether it can be done at all. Doesn't seem a good idea to me to say "Learn GR kid so you can use all that FTL stuff we haven't discovered yet and maybe never will."

Posted by Karl Hallowell at May 18, 2007 02:17 AM

General Relativity would also be important for accounting for timing errors in very accurate terrestrial clocks. One can detect the skew due to gravitational time dilation on atomic clocks, and the current most accurate clocks (like NIST-F1) are accurate enough to see the effect for vertical displacements of about 15 feet, IIRC.

Posted by Paul Dietz at May 18, 2007 10:03 AM

Rand,
Obviously general relativity is very important to you. Certainly there are a couple of applications where it's critical. However, I still don't think you've proven that it's worth dedicating a couple of courses in an undergraduate curriculum to it. It's the sort of thing you need to touch on in a survey course so that students are aware of when general relativity is a consideration. For example, when you took aerospace engineering, I'm sure that you covered the Navier-Stokes equations, then had them simplified with the assumption of an irrotational flow. You needed to know where the assumption broke down, not several more courses on turbulent flow theory - not at an undergraduate level. I'm more concerned about students getting a good grounding in mathematics and thermodynamics than with them picking up in-depth knowledge of the more esoteric aspects of our professional specializations. I understand your stake in general relativity; I might make a similar case for the value of electrochemistry. I think that the specialized knowledge is best acquired through advanced education beyond the basic undergraduate curricula.

Posted by George Skinner at May 18, 2007 11:02 AM

I still don't think you've proven that it's worth dedicating a couple of courses in an undergraduate curriculum to it.

1) I'm not sure that it would require two full courses, or even one, to give a good enough understanding of it to at least evaluate whether or not is was needed for the task at hand and

2) You miss the point that it's not even required for a physics graduate degree.

Posted by Rand Simberg at May 18, 2007 11:11 AM

We covered the linearized forms of general relativity in my high school physics classes - that's probably enough to tell you when you're going to get into trouble without getting further into relativistic effects. You've also had a couple of physicists comment that it would probably take a couple of courses to do a proper treatment of general relativity. Continuum mechanics is also pretty useful for fluid mechanics and strength of materials, but it's a brutal course to take if you're not going to be using tensors on a regular working basis.

Requiring people to take general relativity at a graduate physics level makes even less sense. Graduate studies require specialization to achieve the in-depth knowledge required, not a broadened field.

This debate is getting tedious. I don't have any vested interest in the debate over general relativity education, but just wanted to point out that there are some fairly good explanations for why it might not be included in the general curriculum. I'm sure you'll continue to advocate that it should be; I'm just trying to suggest reasons for why it isn't.

Posted by George Skinner at May 18, 2007 01:48 PM

Arrrgh!!! I think everyione is missing the point. IMO, a physics degree should give one a broad understanding of how the universe works. How can you get that without GR?

And, as others pointed out, to achieve progress/discoveries you need to understand what you are modifying/incorporating, don't you?

Posted by mockmook at May 19, 2007 07:47 AM

From the prospective of someone outside the field, most of the negative comments seem to indicate that physics education is currently meant to develop engineers not physicists. What do you care whether general relativity or the Standard Model have uses in this or that technology - if they are fundemental parts of physics they should be learned.

Engineers are concerned with applications right now. Physics, if it cares about applications at all should be concerned with those two or three decades away, applications that may very well involve these areas.

One obvious question for all of you faculty experts: If these subjects are not being taught, what subjects are covered? What else is so important that it leaves no time for these things?

Posted by Doug Collins at May 19, 2007 08:00 AM

I was a graduate student in physics at UC Davis and neither the Standard Model or General Relativity were required courses. We did have courses in the Standard Model but only those students heading in that area were required to take any of them, and the only reason we got a GR course was because a recent faculty hire was a Quantum Gravity researcher--it was a specialty course only and not even offered every year. I used to rave that it was absurd that any PhD in physics not know the basics of GR and the Standard Model--it doesn't matter if you're going to "use" it. I hear that same lame excuse about learning algebra from 13 year olds. Any PhD in physics who actually argues that it's not necessary--and should not be required--to have some basic knowledge of GR and the Standard Model is an emabarrassment to physics

Posted by PC Martin at May 19, 2007 08:30 AM

In my view (and I'd be interested to hear the views of some physics professors) the purpose of physics graduate school is to prepare the students for doing professional physics research. In this respect, there is simply too much to learn to get you up to date on ALL the major physics sub-fields. Such as

-Condensed Matter
-Particle Physics
-Atomic Physics/Quantum Optics
-Non-linear Dynamics
-General Relativity

This list is of course by no means exhaustive and there is simply no way for a program to prepare you to actually do research in all these fields. You eventually have to choose to specialize, and time spent devote to one of these categories necessarily takes away from the others.

I would agree that a good physics PhD should at least have some done some rudimentary studies in all these fields, but the purpose of this is mostly so you can effectively field questions you are asked by the general public, but this really isn't going to amount to a level much above what can be explained in a simple conversation.

I have not read Mr Tipler's book, but i find his assertion:

I show that, across all disciplines, a collapse of belief in Christianity over the past several decades among university faculty has been accompanied by a collapse in the belief that there is fundamental truth which should be imparted to students.

to be somewhat ridiculous on its face. I mean what exactly is he concluding here?

If he's saying that you need Christianity to understand physical law, well then I'd probably have to point him toward the innumerable numbers of non-Christians who have been doing physics since time immemorial.

If on the other hand he's saying the decline of Christianity among university staff and a lack of sufficient breadth of physics knowledge both share a root cause... well, all I can say is that would be a rather difficult thing to prove; and it certainly does not seem to me that the remarkable pace of physics research has in any way slowed in the last few decades.

Adam Steele
Dept of Physics
Georgia Tech

Posted by Adam Steele at May 19, 2007 08:39 AM

I have a physics BS from MIT, MS and PhD from the University of Minnesota. Did all of this during the Sixties.

I suppose I've been introduced to Special and General relativity, not to mention the Standard Model. Didn't like a one of 'em. You hand me a rubber ruler and a stretching clock, and I start feeling like I'm being painted into Dali's "The Persistence of Time". This did not stop me from doing well in nuclear physics.

All minds have limits. "Don't try to teach a pig to sing. It wastes your time and annoys the pig." In my case, I know relativity and the Standard Model well enough to avoid them.

It is a large sign for me: "Here Be Pigs". Somehow, I have had a satisfying life nonetheless.

Posted by Ellen Kuhfeld at May 19, 2007 09:51 AM

There are two separate issues here: should GR and standard model be required for undergraduates, and should it be required for a PhD?

The answer to the first is clearly "no". There is just too much more basic material to be covered at the undergraduate level, to fit in GR and advanced particle physics. My undergrad physics degree at Stanford was one of the most demanding degree programs at the university, and it didn't include this material. There just wasn't room to fit it in. More fundamental topics like mechanics, electromagnetism, thermodynamics, quantum mechanics were (and should be) already full-year required course sequences.

Perhaps what the laymen are missing here is that both GR and particle physics require a lot more mathematical and physics background than they realize, a huge amount more than shows up in, say, a Scientific American-level presentation of either.

At the Ph.D. level, most universities take the view that Ph.D. students are grownups who can decide what they really need to know to have a research career, with advice from their graduate advisors. Just because it's not a formal requirement of the degree doesn't mean they aren't studying it.

Posted by Andrew Myers at May 19, 2007 10:00 AM

A Ph.D. in physics already takes on average about 7-8 years beyond the bachelors. Most students spend most of that time in poverty. Ever lived in Boston on $10K/year? Eating meat once a week while your college classmates are getting stock options gets old fast. Go ahead, add a few more years of coursework and watch the number of Americans in physics--already much less than 50% of grad students--drop even further.

If students are interested, they'll audit the class or just pick up a book. That's what I did. Maybe Tipler's ticked because nobody wants to take his class.

There are very few/no physicists today who are at a research level of competence in all fields. Does Stephen Hawking do condensed matter physics? Does Phil Anderson do GR? It's called "division of labor".

As for general culture, most schools think two years of coursework is the limit. Getting a Ph.D. is not about taking classes, it's about doing research, and you have to draw the line somewhere. So what are you going to take out to make room for GR? The first year is general courses and quals: QM, E&M, stat mech, math methods (complex analysis, PDEs, group theory). The second year is specialized coursework and carrying out initial research to prepare your dissertation proposal. This is where students entering fields requiring GR take it, and those entering high energy begin to study quantum field theory, eventually getting into the standard model. For those going into condensed matter, you study solid state physics, lots of advanced stat mech leading to renormalization group theory, and maybe some complex fluids.

And why the sudden concern? Has GR EVER been a requirement in most grad programs? Is there a sudden shortage of relativists? No and no.

T. M. Kelley, Ph. D., Condensed Matter Physics

Posted by at May 19, 2007 12:56 PM

The discussion here raises some interesting questions.
General relativity is related to Einstein's effort to find a unified field theory that could describe all physical behavior; and the standard model is a model which attempts to describe all physical behavior other than the force of gravity, which is the main subject of general relativity. Both of these subjects are part of an attempt by physicists to find one single theory that explains the entire universe and all its behavior.
The quest for such a theory has been pursued by physicists who ardently believe that their goal exists, that there is such a theory and indeed that they have some hope of discovering it.
This belief is a form of monotheism, monotheism in a Spinozan sense. There is one law that governs everything in the universe.

Of course we all believe that the world is governed by physical laws which we hope are comprehensible to us. But the pursuit of one unified theory of everything goes beyond that.
Many of the physicists who pursue this goal consider themselves atheists, but their goal is without meaning without a belief in its existence.
It is probably not an accident that the development of modern science occurred in a society imbued with monotheism.

There is a problem though with pursuing the search for such a theory, and that is: the closer one comes to it, the further one gets from mundane reality. The more we know the less relevant to anything is what we do not know.

In theoretical physics, there are very few phenomena that we can experiment with that depend in any way on what we do not know about gravity or elementary particles.

The consequence of this is that a physics student who wishes to do research related to these things runs the risk of meeting the fate of Icarus: by flying to close to the sun, (actually too far from earthly concerns) he may have his wings melt and he may perish (actually fail to accomplish anything appreciated in this world and ruin his career)

In fact students of physics are attracted to hot areas, areas in which lots of new things are happening. They want to contribute to society and further their own careers through success in research. Areas such as general relativity and improvements on the standard model, do not now seem to be active areas about which either goal seems practical. And this more than anything else is responsible for lack of interest in learning about or teaching these things.

Posted by Daniel at May 19, 2007 01:51 PM

Well, I speak from the perspective of someone who has just finished the (rather exhaustive) MIT undergraduate physics education.

I have some comments:
In a physics undergraduate education is expected that you will learn:
Classical Physics
Electricity and Magnetism
Statistical Physics and Thermodynamics
Quantum Physics
Special Relativity
Mathematics (Calculus, Diff. Eq's. Linear Algebra)
Laboratory Physics

This is quite a lot of material. And ALL of it is completely essential towards every sub-field of physics you can go into.

It would be really just cruel to add a dollop of GR onto the top of this.

GR is extremely hard. At MIT it's considered a second or third year graduate course. And it isn't like it's easy material to dumb down and retain much of the higher-order features. Not only is it hard but it's also mathematically intensive. You could take it no earlier than a junior. Considering the core of the major is taught sophomore and junior year, that basically puts you in for senior year. When you might want to be taking a course relevant to your subfield.

GR, though a beautiful theory, isn't relevant to the majority of sub-fields of Physics. If you want to do astrophysics, cosmology or theoretical physics, than of course you ought to take it. But if you're in condensed matter work, you are far better off taking a Solid State Physics course. Or in my case, I am in atomic physics so I took a course on Atomic and Optical Physics

Posted by David at May 19, 2007 01:55 PM

as well as electrical engineering courses (good to know if you're in an experimental field). And if you're in high energy physics and theory, Quantum Field Theory seems important.

The same objections apply to the Standard Model.

Likely, I will take GR when I go to Caltech for my graduate studies. But it will just be for the hell of it, because I'm curious, not because it's necessary or ought to be necessary for my degree. It would be a bit like forcing biologist to learn special relativity.

And what exactly does Frank Tipler do as a physicist? As far as I can tell (from his website) he writes articles about Physics and society. That's a perfectly reasonable thing to do, but it hardly gives you the standing or perspective to state what physics undergraduates or graduates need to know. His implication that people do not study modern physics if they don't study the standard model is simply facetious: plenty of work has been done in other sub-fields in the last fifty years (Quantum information processing in my case).

Alright,

David, B.S. Physics

To T.M. Kelley
7-8 years? 10k. I assume you're exaggerating for effect. At top tier schools its 5-6 years, 20-25k usually.

Posted by David at May 19, 2007 01:56 PM

Statistics for time-to-degree via NSF

http://www.nsf.gov/statistics/infbrief/nsf06312/nsf06312.pdf

cf Table 3, p. 5, 2003 academic year time to degree for all physics and astronomy doctorates was 7.6 years total, 7.0 years registered. If you dig around in NSF's various reports, it varies some by subfield and by experimental vs theoretical.

Do you have data for "top tier schools" (however defined) time to degree?

I haven't found similar data for stipends, but a quick google suggests it probably has risen faster than inflation since I started in the mid 90's. Let's say it's 25K--that's a lot more manageable than 10-12K, but I bet it still sucks, and it still gets old fast.

TMK

Posted by at May 19, 2007 02:33 PM

PS David--have fun at Caltech, I postdoc'd there, it's a blast. Caltech is (I think) fairly unique in that all grad students are expected to finish in five years, and any longer than that requires some explainin'. Too bad more schools don't have that attitude.

Posted by TMK at May 19, 2007 02:43 PM

According to MIT Physics meantime to graduation is 5.7 years

According to Caltech Physics:
"FIFTH YEAR: The Institute expects graduate students to complete their program of study by the end of the fifth year. Students beyond the fifth year must petition the Dean of Graduate Studies for excessive registration and provide a timetable for completion of the thesis requirement. All that's left is to write a fabulous thesis!"

It's hard to give a list of schools that are "top tier" because it depends on the field, but generally its equivalent to well-funded, and therefore the same as those schools that give stipends of 18K+. (You could probably also calibrate it with paper citations, etc.)

Though it seems possible that smaller stipends correlates with less funding which tends to make it harder to get research work done, meaning longer graduation time.

Posted by David at May 19, 2007 03:09 PM

It's always easy to suggest others should do more work, but there are tradeoffs. I was waiting through Tipler's article for what courses he would cut back on, unless he wanted to lengthen the PhD process even further. Sure enough, he cut electromagnetism to a semester. You cannot possibly teach graduate-level E&M (or quantum for that matter) in a single semester and hit all the important points! And those areas of physics are simply more broadly useful to the great majority of research areas.

I agree with the comprehensive exam idea (at UCLA it covered quantum, E&M, and stat mech), but you have to let people get on to research, which as others have noted, can take quite a while. I'd love to see a comparison of average PhD times across all academic disciplines, I'm guessing physics is already among the highest.

Still, the most absurd thing Tipler says concerns his 'lack of belief in fundamental truth' nonsense. I've yet to meet a physicist (or any other hard scientist, for that matter) who does not believe in the underlying reality of the universe and its laws. That's what attracts most of us to the field in the first place, that curiosity. That's a humanities issue, not a physics issue. Sure, there must be a couple nuts, someone has to write these quack quantum mysticism books, but these people are few and far between. And to link this bogus 'trend' to a weakening of faith in Christianity is so nonsensical that I wouldn't take this guy's word for it if he observed it was raining outside.

Posted by Dave at May 19, 2007 05:44 PM

In the 1980s at SUNY-Stony Brook, it became possible to earn a Bachelor's degree in Mathematics without ever seeing an ε-δ proof.

Posted by Eric Jablow at May 19, 2007 06:56 PM

This discussion reminds me of one a classmate initiated during the first day of Gross Surgical Anatomy my first year of med school in 1994 at UCLA.

Him: "I'm going to be an ophthalmologist. Why do I need to know how many veins drain the liver?"
Professor: "Three reasons. First, since you're not an ophtalmologist yet, you have no idea what effect hepatic disease has on eyesight. Second, understanding the human body is fundamental to everything we do in medicine. And third, the discipline of learning all of this stuff is essential to teach you how to think like a doctor. If this is a challenge you're not up to, and you one to specialize in one part of the body, consider being a dentist. Or perhaps podiatry."

The relevance to this discussion is that you could make a case that mastering General Relativity is necessary not for its own sake but for learning how to think like a physicist. Maybe that's the point Rand's trying to make. It sounds as if the physicists on the thread disagree somewhat. But it's enlightening to read the range of opinions all the same.

Posted by Jane Bernstein at May 20, 2007 05:32 PM

I'll second the earlier comments about the math required for GR. It'd be a lot more work to introduce enough tensors to do the GR, just for doing the GR at that level of school.

I've got a Ph.D. in astrophysics, and the only place I remember having tensors in undergrad school was briefly, in sophomore classical mechanics. In grad school, we had them in some E&M problems, and in quantum field theory (ooh, lots of 'em there!). And if I hadn't taken cosmology, that's all I would have had of them.

Cosmology was chock-full of GR, of course. And having had it, I really appreciate the experience. Since I do quasars and therefore black holes, I need to use a little of it from time to time. But beyond knowing the qualitative effects, I can't imagine undergraduate physics majors *needing* to know GR to really understand physics.

But yes, the qualitative description of GR should be taught in a sophomore modern physics course. I got a solid background in special relativity then, and it would be ideal to follow it up with a descriptive summary of GR. But quantitatively? I'm not convinced.

Posted by Tim Hamilton at May 23, 2007 05:48 AM

Post a comment