PDA

View Full Version : Thought experiments, "what-if"s, imagination, and science



Nereid
2010-Jun-13, 07:15 PM
The trigger for this thread is some lengthy discussion in this Q&A (now Astronomy) thread (http://www.bautforum.com/showthread.php/104771-What-would-happen-to-space-time-if-something-huge-suddenly-accelerated-disappeared), and this (now locked) About BAUT one (http://www.bautforum.com/showthread.php/104852-Is-BAUT-s-Q-amp-A-section-becoming-intimidating).

Many, perhaps most, of us think of poor Schödinger's Cat (http://en.wikipedia.org/wiki/Schr%C3%B6dinger's_cat) when we hear "thought experiment". And what an excellent example of a thought experiment it is!

In physics, a thought experiment is a real experiment, only you can't actually do it because of limitations of technology, or money, or similar. In the case of the Schödinger's Cat thought experiment, there are certainly technological constraints (and, perhaps, ethical ones too!), but it is an experiment which could, in principle, be carried out. Note that you'll sometime find thought experiment written as it was in German, when the idea was first published - Gedankenexperiment.

Another example*, taken from a recent Q&A thread: Would it be possible to destroy a black hole if you threw enough anti-matter into it? (http://www.bautforum.com/showthread.php/104455-Destroying-a-Black-Hole-with-Anti-Matter) We do not have a black hole to hand, and we probably couldn't make, or get, enough anti-matter to do this experiment properly even if we did.

An example of thought experiments turning into real experiments: the EPR paradox (http://en.wikipedia.org/wiki/EPR_paradox). In this (or these, there is more than one of them) thought experiment - proposed by Einstein, Podolsky, and Rosen - some essential characteristics, or features, of quantum mechanics (QM) are tested. At the time, doing real experiments to test QM in this way was technologically impossible; several decades later it became possible, and many "EPR" tests have been done (the result? QM 1, EPR 0).

An absolutely critical feature of these, and all, physics thought experiments is that the "laws of physics", as we know them today, are assumed to remain the same; a thought experiment does not arbitrarily assume away any well-established theory. In fact, one of the key points of thought experiments is to test one such theory! The EPR paradox thought experiments were designed explicitly and specifically as tests of QM, which, at the time, had amassed a great deal of experimental support, but which had implications that were deeply troubling to really, really smart physicists like Einstein.

Obviously, all thought experiments are "what-if" questions!

However, not all "what-if" questions are thought experiments.

For example, "what if Hitler had won WWII, would Sweden still be neutral?" is a "what-if" question, but it is not a thought experiment. Not only does it have nothing to do with physics, but it assumes that the past is mutable^.

What about "what if the Sun suddenly disappeared, would it radiate gravity waves**?"? As it stands, one cannot tell if this is, or could be, a thought experiment; to tell, you need some details of how the Sun might suddenly disappear.

It could, for example, simply stop radiating visible light - it might be being eaten by a black hole in its core, and the eating suddenly reaches the photosphere.

Or it could, perhaps, turn into a lot of photons, by having a solar-mass of antiparticles stream into it, radially, and isotropically.

In each of these examples - and many more - we have a thought experiment (nearly; some further details would be required, but that's a quibble).

But what about "well, the Sun's mass simply vanishes; one moment 2 x 1030 kg of matter, the next, zero; how would space-time be affected?"? Such a "what-if" is certainly imaginable! And it might form the basis of a really good scifi story (a lot of scifi rests on deliberately breaking a known law of physics, and exploring some of the logical consequences of that). However, it is not, and cannot be, a thought experiment.

Why? Because it assumes a well-established "law of physics" is violated (conservation of mass-energy), in a particularly spectacular fashion; remember that thought experiments keep the laws of physics fixed, unbroken.

Further, space-time is a technical term which has meaning (in physics) only within General Relativity (or similar theories, or extensions to them); in GR, mass-energy is conserved, so if you assume it's not, you no longer have GR, and so 'space-time' becomes meaningless.

Now this might be hard to grasp, especially if you aren't really familiar with how science works, or what the well-established physics theories really say (or mean), so perhaps an example from biology might help.

What if all the dogs in the world suddenly, instantly, turned into oak trees; would they still have litters of ~six?

We can certainly imagine that! After all, our imaginations are much more powerful than that simple example! We can - well some of us can - imagine "what if colourless purple dreams didn't sleep furiously, would they still have six impossible thoughts before breakfast?"

But is that what-if sensible, in biology? Is it even meaningful?

I don't think so, and "what if the Sun's mass-energy instantly becomes zero, would it radiate gravitons?" is equally meaningless, in physics.

One more example, from finance (courtesy of Pengwuino (http://www.physicsforums.com/showpost.php?p=2757633&postcount=9)): "what if money had no set value (as in, a $1 bill is a $1 bill and can't decide to be a $10 bill at will), how much would my monthly payments on my 30-year mortgage be?"

* note that this is not expressed in a form typical of serious physics thought experiments
^ under the many-worlds interpretation of QM, this question is not meaningless; however it is not testable, even in principle, even within this interpretation.
** or gravitational wave radiation, or gravitons, or ...

aastrotech
2010-Jun-14, 06:34 AM
Another example would be death. Oft mentioned is the imaginative question of whether a person could die and come back and tell us what the other side looks like. This imagines a definition of death that can be come back from. That's not a scientific definition of death. If you "come back" it's not death. It's near death. No matter how close you come to death or how long you are "dead" if you come back you were never dead. Sound familiar? "No matter how close you come to the speed of light..."

Since no one, by definition, has ever come back to tell what the othe side looks like we can't even know if there is an other side. The imaginative question is scientificly meaningless.

A problem in this area is that when "imagination questions" are posed that go against scientific understanding then the imaginer is saying "come with me on this imaginary ride with me as the driver". The appropriate answer to these questions is either "Ok lets go on this ride and see where your imagination goes" or "No thanks".

JohnD
2010-Jun-14, 10:12 AM
Nereid,
The one about money isn't meaningless, because money doesn't have a fixed value.
Anyone from a country that has suffered inflation will tell you so, and tell you what the effect on your mortgage payment will be.
One reason for property booms is that they occur in inflationary times. Your repayment of the loan to buy the property is made in currency that is progressively devalued, and you can sell the property for the inflated price to finance more purchases.

Were you trying to define a 'thought experiment'?
You used the 'disappearing sun' example as a counter example, an event that is impossible, whatever happens in the future so does not illuminate the situation (!). But Einstein's original insights on relativity came from imagining what it would be like to 'ride on a beam of light', another event that can never happen. So I'd suggest that a valid thought experiment is one that provides an insight that advances theory.

Except that Schroedinger's Cat was explanatory, to show what he meant, not to give him any further insight. So maybe the definition must be expanded to include insights in the the minds of others as well as the experimenter.

John

Nereid
2010-Jun-14, 04:23 PM
Nereid,
The one about money isn't meaningless, because money doesn't have a fixed value.
Anyone from a country that has suffered inflation will tell you so, and tell you what the effect on your mortgage payment will be.
One reason for property booms is that they occur in inflationary times. Your repayment of the loan to buy the property is made in currency that is progressively devalued, and you can sell the property for the inflated price to finance more purchases.

Good point.

I took Pengwuino's example ("no set value") to mean that much more than "doesn't have a fixed value" ... he (she?) meant it had any arbitrary value anyone, at any time, chose to give it ("a $1 bill is a $1 bill and can't decide to be a $10 bill at will").

So inflation is not at all what was meant.


Were you trying to define a 'thought experiment'?
You used the 'disappearing sun' example as a counter example, an event that is impossible, whatever happens in the future so does not illuminate the situation (!).
Another good point.

Since we cannot say what successful theories (of physics) will come along, at some arbitrary time in the future, we cannot invoke any such to test them, using a thought experiment.


But Einstein's original insights on relativity came from imagining what it would be like to 'ride on a beam of light', another event that can never happen.
Actually, I think this is a very good example of applying our present-day understanding to a past event (an anachronism?).

At the time Einstein did his thinking, the prevailing theory of light - electromagnetic radiation, per Maxwell's equations, etc - did not necessarily preclude riding on a beam of light (in principle!). Indeed, the experiments to establish, unambiguously and comprehensively, the modern view (essentially SR) were not done until well after Einstein, at age 16, did his thinking*.

But I may have my historical facts wrong; if so, please set me straight.


So I'd suggest that a valid thought experiment is one that provides an insight that advances theory.
That's certainly a widely-held view of what a thought experiment is.

However, as I tried to show, it is not what is usually meant, in terms of the way science (well, physics at least) is done. For one thing, it's too broad.



Except that Schroedinger's Cat was explanatory, to show what he meant, not to give him any further insight.
That may be so (I don't know what Schrödinger thought of this); however, thought experiments - in the sense of my OP - are not the property of just one person (whether the author or not). Rather, they are defined by what the community of scientists (physicists in this case) thinks of them. And I think Schrödinger's Cat was, and still is, regarded by physicists as a thought experiment.


So maybe the definition must be expanded to include insights in the the minds of others as well as the experimenter.

John
Good point.

If I did not make that clear in the OP, perhaps I should go edit it; what do you think?

* The famous Michelson-Morley experiment was done before, but it did not, at the time, by itself, cause the then accepted theory of light to be abandoned

HenrikOlsen
2010-Jun-14, 04:49 PM
That may be so (I don't know what Schrödinger thought of this); however, thought experiments - in the sense of my OP - are not the property of just one person (whether the author or not). Rather, they are defined by what the community of scientists (physicists in this case) thinks of them. And I think Schrödinger's Cat was, and still is, regarded by physicists as a thought experiment.
Schrödinger's Cat was a though experiment pointing out a missing piece in the quantum thinking of the time, since it (the thinking) lacked any idea of how to get from a quantum superposed state to an actual macroscopic effect, its apparent paradox triggered the development of such ideas as decoherence and collapsing wave functions.
It shows one very important use of thought experiments, that of pointing out places where current hypotheses are lacking because they can't make predictions for the outcome.

JohnD
2010-Jun-14, 10:02 PM
Careful Hendrik, you make a good point, I think, but are approaching the stage referred to in the recent thread on Q&A about how discussions of tiny points put off people.

No, don't edit your original post, Nereid, that would confuse everyone, as references to your OP may then refer to phantom points.
But I would recommend editing severely any post that occupies more than one screenful.
We are not writing essays here. If the idea cannot be expressed on a single sheet of paper/screen then the idea needs to rethought and clarified. Churchill used this discipline.
The exposition and detail can come later (and always does, especially on a board like this)

John

Nereid
2010-Jun-14, 10:45 PM
Careful Hendrik, you make a good point, I think, but are approaching the stage referred to in the recent thread on Q&A about how discussions of tiny points put off people.
I think the point being made is wrt Q&A threads alone; one of the actions that the Q&A guidelines state is for "extended discussions" to be moved to a more appropriate section, where, by implication, they can go on for as long as the participants like (subject to the usual caveats, of course).


No, don't edit your original post, Nereid, that would confuse everyone, as references to your OP may then refer to phantom points.
Thanks; good suggestion.


But I would recommend editing severely any post that occupies more than one screenful.
Hehe ... that could be quite a challenge!

Long gone are the days when 'a screenful' was easily defined, and didn't vary much; today, with mobile phones, netbooks, large screen monitors, etc, 'a screenful' can vary by as much as a factor of 10.


We are not writing essays here. If the idea cannot be expressed on a single sheet of paper/screen then the idea needs to rethought and clarified.
As a general principle, this is good advice.

However, when it comes to complicated and subtle points of physics, I think it can lead to disaster.


Churchill used this discipline.
Fortunately, Churchill was not required to explain general relativity, or quantum mechanics! :p



The exposition and detail can come later (and always does, especially on a board like this)

John
Personally, I've had bad experiences trying to get across difficult stuff in the manner you describe.

And here, in the S&T section - as also in the others under Space & Science, except Q&A - we can, and should (IMHO) take as much space as we need.

Ken G
2010-Jun-15, 03:15 AM
The way I would put it is, the difference between a thought experiment and a real experiment is that a thought experiment poses a question to some theory, whereas a real experiment poses a question to nature herself. People get all confused when they think a thought experiment is a "what if" question posed to nature-- that mixing of the two actual possibilities results in nought but a confused muddle, and is often the largest problem we find in the Q&A section. A thought experiment is exactly a "kicking the tires" of a theory, and so makes no sense unless the theory in question is specified explicitly (which is why it showed such complete lack of insight for the mods to fault Nereid for asking that technologically impossible "what if" questions be framed in terms of some theory, rather than just posed in a vacuum). Any question like "what would happen if" that is intended to apply to nature, and not some theory, can only be answered by reference to similar experiments, or cannot be answered at all if no similar experiments currently exist.

Jens
2010-Jun-15, 06:34 AM
Long gone are the days when 'a screenful' was easily defined, and didn't vary much; today, with mobile phones, netbooks, large screen monitors, etc, 'a screenful' can vary by as much as a factor of 10.


I don't know if you've ever seen the original version of Willie Wonka, but the contract (http://www.kenmark-inc.com/downloads/wonkacontract.pdf)shows that even before computers, a "screenful" or "pageful" could be loosely defined!

JohnD
2010-Jun-15, 05:53 PM
I think you know what I mean, Nereid.
For instance, and because it is a convenient example, your OP contained 850 words.
That's just over two pages of 12 point text, though your frequent new paras bulk it up.
Despite your prolixity, I still didn't know what you were saying, though I assume you were seeking to define 'thought experiment'.

And later UK politicians pursued the same policy as Churchill. "One-page" Waldegrave, the Science Minister in 1993, issued a challenge to explan the Higgs Boson on one page, with bottles of vintage champagne as prizes. One of the five winners was Simon Hands, and his winning page, Ripples at the Heart of Physics, can be read at http://pyweb.swan.ac.uk/~hands/higgs.html. I leave you to find the others, to admire what can be done on one page and to emulate that succinct style.

John

Nereid
2010-Jun-15, 07:09 PM
I think you know what I mean, Nereid.
For instance, and because it is a convenient example, your OP contained 850 words.
That's in the ballpark of what I generally aim for; namely ~500 to 1,000 words.


That's just over two pages of 12 point text, though your frequent new paras bulk it up.
That's somewhat new to me; I used to write in paras about the same length as Ken G's, but I have found, from considerable experience, that on the internet, in discussion fora, an almost bullet point presentation style tends to get the point across better than the alternatives, in many cases.


Despite your prolixity, I still didn't know what you were saying, though I assume you were seeking to define 'thought experiment'.
Clearly, then, my post fails in its intent.

Can I, as they say, get back to you on that?


And later UK politicians pursued the same policy as Churchill. "One-page" Waldegrave, the Science Minister in 1993, issued a challenge to explan the Higgs Boson on one page, with bottles of vintage champagne as prizes. One of the five winners was Simon Hands, and his winning page, Ripples at the Heart of Physics, can be read at http://pyweb.swan.ac.uk/~hands/higgs.html. I leave you to find the others, to admire what can be done on one page and to emulate that succinct style.

John
I wonder how long it took Simon to write that, counting all the 'thinking' time, drafts, etc!

And I wonder how many excellent Wikipedia entries - the ones longer than a couple of screensful (screenfuls?) - would be better if reduced to ~500 words?

01101001
2010-Jun-15, 07:35 PM
Nereid's OP.
Midway between tweet and blog.
Perfect. Just perfect.

But that's just about the form. Please don't get me started on the content.

mugaliens
2010-Jun-15, 08:20 PM
Thought experiements serve a very useful purpose in helping to explore questions while eliminating obvious dead-ends, such as "an elephant" as an answer to "how large an animal can we lift with a 50-lb monofilament test line?"

EigenState
2010-Jun-15, 08:43 PM
Greetings,


The way I would put it is, the difference between a thought experiment and a real experiment is that a thought experiment poses a question to some theory, whereas a real experiment poses a question to nature herself. People get all confused when they think a thought experiment is a "what if" question posed to nature-- that mixing of the two actual possibilities results in nought but a confused muddle, and is often the largest problem we find in the Q&A section. A thought experiment is exactly a "kicking the tires" of a theory, and so makes no sense unless the theory in question is specified explicitly (which is why it showed such complete lack of insight for the mods to fault Nereid for asking that technologically impossible "what if" questions be framed in terms of some theory, rather than just posed in a vacuum). Any question like "what would happen if" that is intended to apply to nature, and not some theory, can only be answered by reference to similar experiments, or cannot be answered at all if no similar experiments currently exist.
Not to quibble, but while real experiments may pose a question to nature itself, good experiments are formulated within the context of some theoretical framework, if only implicitly. Only purely descriptive observations lack the need for some foundation of theory.

Beyond that, I must agree that "what if" questions must be formulated within some theoretical framework. It appears to me that Nereid is struggling to ascertain some pragmatic mechanism to address the inherent difficulties that most laymen have in formulating such questions. If so, that quest is an age old one with which every teacher is more than familiar. Communicating via the internet does not facilitate the process.

If I may make a suggestion, prepare and add a Sticky to the Q&A Board that discusses how scientific questions are correctly posed, the need for a contextual framework, and support it with carefully chosen illustrative examples of well posed versus poorly posed questions. Far from a cure all, but it might prove helpful.

Best regards,
EigenState

Ken G
2010-Jun-16, 12:14 AM
Not to quibble, but while real experiments may pose a question to nature itself, good experiments are formulated within the context of some theoretical framework, if only implicitly.The way I would put that same point is, real experiments are questions for nature, but the more useful questions to ask nature are guided by theories. The theory helps us interpret the answers we get, but the question is strictly for nature. This is to be contrasted with "thought experiments", which are posed strictly to theories, and we never have any idea if they have anything to do with nature. They simply never tell us anything about nature (it would be quite impossible for a thought experiment to tell us something directly about nature), they only tell us things about our theories of nature, and so only have anything to say about nature if we assume our theories are correct-- unless we are referring to real experiments, and then we are not strictly talking about thought experiments.


Only purely descriptive observations lack the need for some foundation of theory.
I think we agree-- the way I would put that is that all observations are purely descriptive at some level, but we feel "safer" applying certain theories in a confident way, such that other theories, that we are less confident about, are the ones we are "testing." I agree that in the absence of any theory we are confident about, observations are not terribly informative, but there's usually a pretty clear distinction about the theories we are confident about (that a meter reads what it is designed to read, that we are seeing the meter reading correctly, etc.), and those we are testing (like the equations of general relativity or something).


Beyond that, I must agree that "what if" questions must be formulated within some theoretical framework.That is the crux of the matter, for sure.


If I may make a suggestion, prepare and add a Sticky to the Q&A Board that discusses how scientific questions are correctly posed, the need for a contextual framework, and support it with carefully chosen illustrative examples of well posed versus poorly posed questions. Far from a cure all, but it might prove helpful.That's a good suggestion, it's up to the mods. There was some threads elsewhere about that, I don't know if they'll see the suggestion here.

EigenState
2010-Jun-16, 12:55 AM
Greetings,

We are good on the philosophy I think.



That's a good suggestion, it's up to the mods. There was some threads elsewhere about that, I don't know if they'll see the suggestion here.
I feel certain that one of you π × 103 fellows can make the idea known to management. I believe that it would prove to be a good learning tool as well as an established resource that responders can point to when first encountering a poorly posed question.

Best regards,
EigenState

EigenState
2010-Jun-16, 06:04 PM
Greetings,

Some potentially useful resources should the decision be made to prepare the suggested Sticky:

An Introduction to Science: Scientific Thinking and the Scientific Method (http://www.freeinquiry.com/intro-to-sci.html)

Scientific Thinking and the Scientific Method (http://teacher.nsrl.rochester.edu/phy_labs/AppendixE/AppendixE.html)

Best regards,
EigenState

Nereid
2010-Jun-17, 09:33 PM
Despite your prolixity, I still didn't know what you were saying, though I assume you were seeking to define 'thought experiment'.

As promised, getting back to you on that.

All thought experiments are "what if?" questions.

However not all "what if?" questions are thought experiments.

Unfortunately, we have seen several examples of the related logical fallacy - of thinking that "IF A is a B" leads logically to "THEN all B's are A's" - in recent BAUT threads. Specifically, when limited to some physics (or cosmology) questions, one recent thread contained several examples of a "what if?" question being assumed to be a thought experiment, because, implicitly, all "what if?" cosmology (and physics) questions are thought experiments. Or, more precisely, all such "what if?" questions can always be re-expressed as thought experiments, without further qualification.

A central characteristic of a (physics) thought experiment is an assumption (sometimes implicit) that all the established "laws of physics", relevant to the thought experiment, apply.

A corollary is that if even one such law does not apply, then the "what if?" question is not a thought experiment.

That's it.

Questions? Thoughts?

Ken G
2010-Jun-18, 01:07 AM
I'd say that's right. And so sometimes the answer to "I'd like modern science's best answer to this 'what if' question" is "modern science has no answer to that 'what if' question, because science cannot do the experiment, nor includes any accepted theory in which that could even happen in principle." Of course, we all know what happens if you say that-- you get accused of lacking imagination.

01101001
2010-Jun-18, 01:15 AM
I imagine Carl Sagan said (http://en.wikiquote.org/wiki/Carl_Sagan):


The truth may be puzzling. It may take some work to grapple with. It may be counterintuitive. It may contradict deeply held prejudices. It may not be consonant with what we desperately want to be true. But our preferences do not determine what's true. We have a method, and that method helps us to reach not absolute truth, only asymptotic approaches to the truth — never there, just closer and closer, always finding vast new oceans of undiscovered possibilities. Cleverly designed experiments are the key.

EigenState
2010-Jun-18, 01:26 AM
I'd say that's right. And so sometimes the answer to "I'd like modern science's best answer to this 'what if' question" is "modern science has no answer to that 'what if' question, because science cannot do the experiment, nor includes any accepted theory in which that could even happen in principle." Of course, we all know what happens if you say that-- you get accused of lacking imagination.

Agreed, for what that is worth. I still fail to understand the lack of presence of any obvious treatment of the Scientific Method in the Q&A Board. Some basic understanding of that, and appreciation for it would be a great asset in my opinion.

Best regards,
EigenState

Nereid
2010-Jun-18, 11:42 AM
Good point, sufficiently good that I posted it to the Feedback section's Q&A policy changes thread (http://www.bautforum.com/showthread.php?p=1748088#post1748088).

Ken G
2010-Jun-18, 01:15 PM
Also, we should clarify that we are not objecting to questions like "what would happen to the Earth if the Sun were replaced by a star with twice the mass", because although it would not be possible to replace the Sun without already having altered the Earth's orbit in a way not being considered, the question is really asking what Newton's laws say would happen in a situation with a particular initial condition that is different from the situation now in a particular way, and the answer does not care about the history before that. So it's a proper thought experiment, framed in terms of a particular theory (the "initial conditions" and "action at a distance" of Newton's laws). The problem is when the question is expressly about what happens in the transitional period as the Sun is getting replaced, like how long would it take for the signal to reach Earth, or what kind of gravitational radiation would be produced, which are fundamentally about how an impossible process itself is happening. That's the kind of question that boils down to "what do we think would really happen if something we don't think could happen happened?" The nonscientific nature of that question is clear.

01101001
2010-Jun-25, 04:34 AM
That's the kind of question that boils down to "what do we think would really happen if something we don't think could happen happened?" The nonscientific nature of that question is clear.

For the benefit of denizens of the future in which my sig may become altered, I reproduce here its current commentary on the matter, direct from the fingertips of our beloved leader Phil Plait:

If you can chuck reality into the dustbin, then all manners of silliness seem equally plausible. --The Bad Astronomer (http://blogs.discovermagazine.com/badastronomy/2010/03/22/stepping-off-the-narrow-path-of-reality/)

He's my hero.

BigDon
2010-Jun-25, 05:48 PM
Hey Nereid, I've been paying attention to how this has been bugging you for the last couple of weeks, at least.

Obviously you and I have different roles in this forum. I ask a lot of questions, some downright fanciful, and you and the other major physicists here answer them. What exactly would you like from me?

A couple of months back I asked a question in QA concerning the difference between the Sun as it actually is and how different its output would be if you managed to strip off the mantle and expose the core. That seemed straight forward enough and I got good answers and a huge new appreciation for the Sun's mantle.

I don't see the difference between that and having the Sun disappear altogether, as far as thought experiments go. (Well okay, you can't make a star disappear, and there are mechanisms by which stars can be stripped of their mantles.)

And now I find myself experiencing a chilling effect, dividing future questions into "regular" and "Nereid" class questions.

Ken G
2010-Jun-26, 05:26 AM
A couple of months back I asked a question in QA concerning the difference between the Sun as it actually is and how different its output would be if you managed to strip off the mantle and expose the core. That seemed straight forward enough and I got good answers and a huge new appreciation for the Sun's mantle.

I don't see the difference between that and having the Sun disappear altogether, as far as thought experiments go.That is a good question, but it also has a good answer-- and I believe this answer cuts to the heart of the whole issue. The point is that every question that cannot be put to an experiment must be put to a theory, and the question has to be possible in regard to that theory. That's it, that's the whole point. Thought experiments are kicking the tires of theories for which the thought experiment is possible. That is why your question about the Sun is a perfectly acceptable thought experiment, because it is asking about the transfer of radiation in the Sun, and the transfer of radiation in the Sun allows us to strip its mantle (nitpick: it's called the "envelope" for a star). One could also take that question a step further and ask what would happen to the pressure balance in the Sun if you strip its envelope, and get an answer consistent with both radiative transfer and force balances, which is also fine because those theories still allow the envelope to be stripped.

So the problem only appears when the question is fundamentally inconsistent with the answer sought. We could ask what happens in Newton's theory of gravity when we strip the envelope of the Sun, and the answer would be, the pull on the Earth would instantly decrease. That's also a fine thought experiment for kicking the tires of Newtonian gravity, because Newtonian gravity is an action-at-distance theory that depends only on the mass of the Sun, and the mass can be changed without creating any problems for that theory. But if we want an answer that includes some kind of light travel time for the Earth to find out what happened to the Sun, then we are asking a question that is fundamentally about general relativistic spacetime. If we want an answer to a question put to general relativity, then we have to frame the scenario in a way that is consistent with the equations of general relativity. It makes no sense to stipulate that the Earth cannot receive an instantaneous signal from the Sun yet the Sun can be instantaneously disappeared. It's a bit like claiming that 2+2=4 in a situation where 1+1 does not equal 2-- for a scenario to be put to a particular theory as part of thought experiment designed to understand better that theory, then the entire scenario must be consistent with that theory, or the answer will depend on things that were not self-consistently stipulated in the question.

Note that doesn't mean the questioner should be criticized for a "bad question"-- instead, it means that part of the educational process stimulated by that question should include consideration of how the question could be made relevant to the theory for which it was intended to probe. If our goal is to learn about a theory, it might be even more informative to think about what kinds of questions that theory can answer, then to try and use the theory to answer questions for which no self-consistent answer exists.

Nereid
2010-Jun-26, 12:40 PM
Hey Nereid, I've been paying attention to how this has been bugging you for the last couple of weeks, at least.

Obviously you and I have different roles in this forum. I ask a lot of questions, some downright fanciful, and you and the other major physicists here answer them. What exactly would you like from me?

A couple of months back I asked a question in QA concerning the difference between the Sun as it actually is and how different its output would be if you managed to strip off the mantle and expose the core. That seemed straight forward enough and I got good answers and a huge new appreciation for the Sun's mantle.

I don't see the difference between that and having the Sun disappear altogether, as far as thought experiments go. (Well okay, you can't make a star disappear, and there are mechanisms by which stars can be stripped of their mantles.)

And now I find myself experiencing a chilling effect, dividing future questions into "regular" and "Nereid" class questions.
Ask all the astronomy and space questions you like, at any time, and in any way. :)

However, assuming you have understood the point I am trying to make here (well, one of them) - namely that 'thought experiment' has a clear, and special, meaning - I would hope that your questions might become somewhat more clearly stated. After all, the whole point of the Q&A section is to give readers a better understanding of some aspect of astronomy or space ... if you don't end up with a better understanding, then the BAUTians answering your questions have failed.

The deeper point - about the nature and limits of science - are not easy get across, not least because (as has been pointed out, many times) many of those who ask questions have a rather, um, confused view of what science is and what it can (and can't) do.

01101001
2010-Jun-26, 02:16 PM
The deeper point - about the nature and limits of science - are not easy get across, not least because (as has been pointed out, many times) many of those who ask questions have a rather, um, confused view of what science is and what it can (and can't) do.

As well as a not small enough number who answer questions.


What exactly would you like from me?

Don't censor yourself from asking questions. Ask. But be open to an answer that boils down to "Unask the question (http://en.wikipedia.org/wiki/Mu_(negative))," and hope it's accompanied with reasoning why.