ChatterBank1 min ago
What Scientific Idea Is Ready For Retirement?
Question posed at Edge magazine website.Among them some fairly controversial recommendations, I think, either because to discard them seems counter-intuitive, or because others are somewhat cherished, like Moores Law.
Have a browse at your leisure, see what you think :)
http://
Have a browse at your leisure, see what you think :)
http://
Answers
Best Answer
No best answer has yet been selected by LazyGun. Once a best answer has been selected, it will be shown here.
For more on marking an answer as the "Best Answer", please visit our FAQ.
Quite a lot of ideas there indeed. Naturally I'll focus on the Physics-related ones; biology, Chemistry, Economics and Computer Science I know less about than most A-level students do.
* * * * * * * * * * * * * * * *
"Naturalness" (Maria Spiropulu, Peter Woit)
It is beginning to look like the LHC will turn up a great deal of precision confirmation of the Standard model, most spectacularly in the form of showing the existence of a Higgs-like particle, and also in improved estimates for the 17-odd free parameters. Equally, though, it's looking like nothing new is turning up. The one most people banged on about is Supersymmetry, that is an enormously powerful and (mathematically at least) convincing idea for resolving some of the outstanding issues in particle physics, but it's not the only idea out there that is being tested, and squeezed out of relevance, according to current data.
That said, these theories are tough beasts to kill, if indeed they do end up dead. There are two reasons for this. Firstly, there are experimental problems with the Standard Model as it exists, so it necessarily follows that something new will have to emerge at some point. Secondly, these theories have so many parameters that it can be difficult to rule out a theory totally -- too many options to choose from, and in practice that means that really only the simplest "new physics" models are being ruled out, ones in which, say, the many parameters are reduced to a small handful. For now, at least, I think this means that Supersymmetry is still alive, but might have to be retired as an idea sooner rather than later*.
Anyway, returning to naturalness which was the title of this section. I think Naturalness in its current form is (on the verge of being) ready for retirement. The naturalness condition roughly speaking works in the following way (with apologies for the technical language):
-- take a physical quantity that is small, e.g. the mass of an up quark is about 2 MeV (which is small in the sense that masses are usually closer to the scale of GeV = 1000MeV).
-- Is there some associated symmetry that is "almost true" but broken because of this small number?
-- If yes, then the small value is "natural", because you expected it to be zero.
-- If not, then the small value is "un-natural", because it could just as easily be a very high value.
In this sense, though, what you end up having to do is find a new way of trying to make the un-naturally small naturally small, by introducing a new symmetry. And yet, as Guido Altarelli (CERN, University of Rome) put it, in the absence of this new physics that would appear to end up suggesting that "Nature is un-natural". Which is evidently a pretty stupid thing to say.
Nevertheless, the history of Physics in recent years is the history of ever-greater amounts of symmetry. Are we at the end of that particular part of the journey to understanding our world? I don't know, but I expect we'll have a good chance of finding out in the next decade, once the next run of the LHC is completed and we have all the data to analyse.
(*Thankfully my own career won't be crushed in three years if and when Supersymmetry etc. is ruled out at the 13-14 TeV run of the LHC, as my work is likely to focus on a decay that is already known to exist, has been observed, but isn't understood very well yet.)
* * * * * * * * * * * * * * * *
"Naturalness" (Maria Spiropulu, Peter Woit)
It is beginning to look like the LHC will turn up a great deal of precision confirmation of the Standard model, most spectacularly in the form of showing the existence of a Higgs-like particle, and also in improved estimates for the 17-odd free parameters. Equally, though, it's looking like nothing new is turning up. The one most people banged on about is Supersymmetry, that is an enormously powerful and (mathematically at least) convincing idea for resolving some of the outstanding issues in particle physics, but it's not the only idea out there that is being tested, and squeezed out of relevance, according to current data.
That said, these theories are tough beasts to kill, if indeed they do end up dead. There are two reasons for this. Firstly, there are experimental problems with the Standard Model as it exists, so it necessarily follows that something new will have to emerge at some point. Secondly, these theories have so many parameters that it can be difficult to rule out a theory totally -- too many options to choose from, and in practice that means that really only the simplest "new physics" models are being ruled out, ones in which, say, the many parameters are reduced to a small handful. For now, at least, I think this means that Supersymmetry is still alive, but might have to be retired as an idea sooner rather than later*.
Anyway, returning to naturalness which was the title of this section. I think Naturalness in its current form is (on the verge of being) ready for retirement. The naturalness condition roughly speaking works in the following way (with apologies for the technical language):
-- take a physical quantity that is small, e.g. the mass of an up quark is about 2 MeV (which is small in the sense that masses are usually closer to the scale of GeV = 1000MeV).
-- Is there some associated symmetry that is "almost true" but broken because of this small number?
-- If yes, then the small value is "natural", because you expected it to be zero.
-- If not, then the small value is "un-natural", because it could just as easily be a very high value.
In this sense, though, what you end up having to do is find a new way of trying to make the un-naturally small naturally small, by introducing a new symmetry. And yet, as Guido Altarelli (CERN, University of Rome) put it, in the absence of this new physics that would appear to end up suggesting that "Nature is un-natural". Which is evidently a pretty stupid thing to say.
Nevertheless, the history of Physics in recent years is the history of ever-greater amounts of symmetry. Are we at the end of that particular part of the journey to understanding our world? I don't know, but I expect we'll have a good chance of finding out in the next decade, once the next run of the LHC is completed and we have all the data to analyse.
(*Thankfully my own career won't be crushed in three years if and when Supersymmetry etc. is ruled out at the 13-14 TeV run of the LHC, as my work is likely to focus on a decay that is already known to exist, has been observed, but isn't understood very well yet.)
String Theory (Retirers: Eric R. Weinstein, Frank Tipler, Marcelo Gleiser, Peter Woit, Paul Steinhardt, Lawrence M. Krauss; Keepers: Andrei Linde, Gordon Kane, Sean Carroll.)
I wondered if this was going to come up. Here's my take on it: String Theory is the closest thing to physics that a pure mathematician can do. Since we have no problem with people doing pure mathematics, for its own sake, we should have no problem with people studying and trying to understand String Theory. It's likely that there will be a very long wait before experiments can truly put the theory to the test, but on the other hand people have studied all sorts of weird and wonderful things that had apparently no role in nature until it was found or invented years later. One example would be the maths of General Relativity, laid down some 50-odd years before Einstein put it to use in Physics. Perhaps the time scale for String Theory to become true physics will be measured in centuries, but if its time does come, then those who mocked it now will end up being criticised in the future for lacking imagination, or something along those lines.
In itself there is a useful lesson here, I think, about how Science works. Ultimately, you can only know what the right idea is now, based on what theory and experiment tell you now. Is it wrong to back the best fit to the available data, even if that best fit ends up being shown to be wrong because it fails a future experimental test? I'm not sure that it is, or at least, people should allow that sticking to the line, "Based on the available evidence, Theory A is the best description of problem X rather than Theory B" is a sensible and reasonable position, provided it's qualified.
Anyway, String Theory looks like it risks being an idea that was born far too soon, well before its time, and is unlikely to be part of the experimental physicists' world for many years or even decades to come. But I don't think that it's worth retiring, because while the theory has little physical relevance, it's proven to be extraordinarily useful in a purely mathematical sense (most notably in the so-called AdS/CFT correspondence: see http://
So no, String Theory is absolutely not ready for retirement -- although, perhaps, it should be set aside as the fundamental theory of nature for the time being, and efforts used instead to see how it can be used in other branches of physics.
I wondered if this was going to come up. Here's my take on it: String Theory is the closest thing to physics that a pure mathematician can do. Since we have no problem with people doing pure mathematics, for its own sake, we should have no problem with people studying and trying to understand String Theory. It's likely that there will be a very long wait before experiments can truly put the theory to the test, but on the other hand people have studied all sorts of weird and wonderful things that had apparently no role in nature until it was found or invented years later. One example would be the maths of General Relativity, laid down some 50-odd years before Einstein put it to use in Physics. Perhaps the time scale for String Theory to become true physics will be measured in centuries, but if its time does come, then those who mocked it now will end up being criticised in the future for lacking imagination, or something along those lines.
In itself there is a useful lesson here, I think, about how Science works. Ultimately, you can only know what the right idea is now, based on what theory and experiment tell you now. Is it wrong to back the best fit to the available data, even if that best fit ends up being shown to be wrong because it fails a future experimental test? I'm not sure that it is, or at least, people should allow that sticking to the line, "Based on the available evidence, Theory A is the best description of problem X rather than Theory B" is a sensible and reasonable position, provided it's qualified.
Anyway, String Theory looks like it risks being an idea that was born far too soon, well before its time, and is unlikely to be part of the experimental physicists' world for many years or even decades to come. But I don't think that it's worth retiring, because while the theory has little physical relevance, it's proven to be extraordinarily useful in a purely mathematical sense (most notably in the so-called AdS/CFT correspondence: see http://
So no, String Theory is absolutely not ready for retirement -- although, perhaps, it should be set aside as the fundamental theory of nature for the time being, and efforts used instead to see how it can be used in other branches of physics.
Ideas in Quantum Mechanics: "no reality" and "collapse of the wavefunction" (see Anton Zeilinger, David Deutsch, Kai Krause, Freeman Dyson)
Firstly a shout-out to Freeman Dyson as one of the giants of 20th-Century Physics, so nice to see him still around and contributing to the discussion, although at 90 he's presumably not got long to go.
Anyway, I think all the contributors above have good points. Quantum Mechanics is a mathematical description of reality, and I think several people have taken the mathematics too literally. My own interpretation runs along the lines of "Quantum Mechanics is probability in nature". Things are not in two places at once, but they could be in either place and you don't know which. The cat wasn't dead and alive simultaneously; you just don't know which one it is, standing outside the box (remember that the cat has to know!). And so on. Also quantum mechanics isn't a complete theory anyway, since it doesn't include Special Relativity. You need to turn to Quantum Field Theory for the "more complete" (though still not final) theories of how things work; and certain paradoxes disappear in that theory or become less of an issue.
At any rate, some of the interpretations of Quantum Mechanics that are kicking around are borderline dubious and ought to be kicked into the long grass. One that wasn't mentioned, but ought to be, is the "many-worlds" interpretation. This one states that all quantum choices result in a new universe, or set of universes, being created, ad nauseam, so that every possible outcome occurs in at least one of these Universes. But the thing is that, again, Qunatum Mechanics was about probability -- and we don't need six Universes to explain the outcome of a dice roll. "Oh, it came up a four here, but in five other Universes the duplicate yous saw a 1, 2, 3, 5 or 6 respectively." It made no sense there; it makes no more sense in the Quantum World.
The Uncertainty Principle is another one. It's not an experimental limit but a theoretical one, and the difference is massive. At any rate, it's badly named -- a better name, but it's too late now, would be the "Non-commutativity principle" -- and it states that certain pairs of quantities cannot be simultaneously determined, and that the order in which you measure them matters.
* * * * * * * * * * *
Those are the three that stick out. There's also a lot of topics that could be called "New Ideas on the Philosophy of Science", and arguably are far more important, since the ones I have dealt with above are specialised subjects.
Anyway, the 20th Century marked a huge leap forward in Scientific understanding, and hopefully the 21st Century will too. But possibly -- indeed, almost certainly -- in directions that are unexpected.
Firstly a shout-out to Freeman Dyson as one of the giants of 20th-Century Physics, so nice to see him still around and contributing to the discussion, although at 90 he's presumably not got long to go.
Anyway, I think all the contributors above have good points. Quantum Mechanics is a mathematical description of reality, and I think several people have taken the mathematics too literally. My own interpretation runs along the lines of "Quantum Mechanics is probability in nature". Things are not in two places at once, but they could be in either place and you don't know which. The cat wasn't dead and alive simultaneously; you just don't know which one it is, standing outside the box (remember that the cat has to know!). And so on. Also quantum mechanics isn't a complete theory anyway, since it doesn't include Special Relativity. You need to turn to Quantum Field Theory for the "more complete" (though still not final) theories of how things work; and certain paradoxes disappear in that theory or become less of an issue.
At any rate, some of the interpretations of Quantum Mechanics that are kicking around are borderline dubious and ought to be kicked into the long grass. One that wasn't mentioned, but ought to be, is the "many-worlds" interpretation. This one states that all quantum choices result in a new universe, or set of universes, being created, ad nauseam, so that every possible outcome occurs in at least one of these Universes. But the thing is that, again, Qunatum Mechanics was about probability -- and we don't need six Universes to explain the outcome of a dice roll. "Oh, it came up a four here, but in five other Universes the duplicate yous saw a 1, 2, 3, 5 or 6 respectively." It made no sense there; it makes no more sense in the Quantum World.
The Uncertainty Principle is another one. It's not an experimental limit but a theoretical one, and the difference is massive. At any rate, it's badly named -- a better name, but it's too late now, would be the "Non-commutativity principle" -- and it states that certain pairs of quantities cannot be simultaneously determined, and that the order in which you measure them matters.
* * * * * * * * * * *
Those are the three that stick out. There's also a lot of topics that could be called "New Ideas on the Philosophy of Science", and arguably are far more important, since the ones I have dealt with above are specialised subjects.
Anyway, the 20th Century marked a huge leap forward in Scientific understanding, and hopefully the 21st Century will too. But possibly -- indeed, almost certainly -- in directions that are unexpected.
I have been sampling a few of the proposed theories due for debunking with a degree of scepticism.
You have some proponents of this debunking who have little to no expertise in the relevant field, for instance ( Paul Davies, astrophysicist, wanting to relegate the somatic theory of cancer to the dustbin of scientific history). You have others wishing to refute the notion of scientific advancement,claiming instead that each advance in our knowledge leaves us more ignorant; Or Ornish, who has an agenda of his own, namely the promotion of the Ornish method of weight loss, who wants us to reject the idea of a large randomised controlled trial and being the gold standard of evidence in medicine/nutrition.
So, many of these ideas that are to be dumped come from individual enthusiasts, often with an alternative agenda of their own.
Some of them do make for interesting reading though; Cannot comment on those that Jim has singled out specifically since it tends to make my head hurt :) - but I thought I understood that the latest discoveries about elemental particles, such as the Higgs Boson, tended to reinforce the standard model rather than weaken it.
You have some proponents of this debunking who have little to no expertise in the relevant field, for instance ( Paul Davies, astrophysicist, wanting to relegate the somatic theory of cancer to the dustbin of scientific history). You have others wishing to refute the notion of scientific advancement,claiming instead that each advance in our knowledge leaves us more ignorant; Or Ornish, who has an agenda of his own, namely the promotion of the Ornish method of weight loss, who wants us to reject the idea of a large randomised controlled trial and being the gold standard of evidence in medicine/nutrition.
So, many of these ideas that are to be dumped come from individual enthusiasts, often with an alternative agenda of their own.
Some of them do make for interesting reading though; Cannot comment on those that Jim has singled out specifically since it tends to make my head hurt :) - but I thought I understood that the latest discoveries about elemental particles, such as the Higgs Boson, tended to reinforce the standard model rather than weaken it.
Generally speaking, yes, the Standard Model has been passing the LHC test with flying colours. Even more flying in fact than most people were expecting. The most non-trivial known mistake that I am aware of is that it cannot cope in its current form with the fact that neutrinos (appear to*) have mass. The SM insists that they must be massless, and it's not the easiest thing in the world to fix this discrepancy. This is the strongest hint of new physics, anyway, beyond the Standard Model, but there are a few others. In some sense at least, as I understand it, the assumption of new physics at some point is "built in" to the Standard Model.
Naively, whenever you calculate anything in the Standard Model, you keep getting infinity. This problem arises because the naive calculations assumed that the Model extends to all energy scales, so is fixed by removing this assumption. But then, of course, that implies that there is some energy scale at which the Model is no longer valid, which in turn means that something new, we know not what, starts to kick in at that scale. This scale is assumed to be certainly not more than the so-called "Planck Energy", which may be the point at which gravity starts to become Quantum. Theories such as Supersymmetry and others tend to assume that the scale was somewhat less, and nearer to the energy scale being probed at the LHC. Apparently, this isn't true -- but still, there will be some point at which the Standard Model stops working properly and we need something new.
*Technical note on neutrino masses: what's actually been observed is the so-called phenomenon of "neutrino oscillations". Neutrinos come in three types (electron, muon and tau), and these can be seen to turn into each other over long distances. This can be explained in a standard Quantum-Mechanical way by the neutrinos having masses. A short note in a lecture I attended, given by Carl Bender, suggested that a new type of theory can explain these oscillations without needing massive neutrinos. I have no idea if this is true or not.
Naively, whenever you calculate anything in the Standard Model, you keep getting infinity. This problem arises because the naive calculations assumed that the Model extends to all energy scales, so is fixed by removing this assumption. But then, of course, that implies that there is some energy scale at which the Model is no longer valid, which in turn means that something new, we know not what, starts to kick in at that scale. This scale is assumed to be certainly not more than the so-called "Planck Energy", which may be the point at which gravity starts to become Quantum. Theories such as Supersymmetry and others tend to assume that the scale was somewhat less, and nearer to the energy scale being probed at the LHC. Apparently, this isn't true -- but still, there will be some point at which the Standard Model stops working properly and we need something new.
*Technical note on neutrino masses: what's actually been observed is the so-called phenomenon of "neutrino oscillations". Neutrinos come in three types (electron, muon and tau), and these can be seen to turn into each other over long distances. This can be explained in a standard Quantum-Mechanical way by the neutrinos having masses. A short note in a lecture I attended, given by Carl Bender, suggested that a new type of theory can explain these oscillations without needing massive neutrinos. I have no idea if this is true or not.
In the Great Gatsby, in the final chapter the father hands the inquirer a bit of paper that Gatsby has written things to do
and instead of hanging around burlesque shows he has substituted the activity
" study needed inventions "
if we knew which ideas were duff
er
wouldnt we already have junked them ?
and instead of hanging around burlesque shows he has substituted the activity
" study needed inventions "
if we knew which ideas were duff
er
wouldnt we already have junked them ?
//if we knew which ideas were duff wouldnt we already have junked them ? //
Nope bad ideas carry on an awfully long time
For example look at the continuing resistance to the MMR vaccine
But the medal *has* to go to catching a cold by being too cold!
With its roots in Aristotle's 4 humours and the middle ages how many people still think that you 'catch a cold' by going out without a jumper.
This 350 years since we looked down microscopes and discovered bactreia and later virus'!
Nope bad ideas carry on an awfully long time
For example look at the continuing resistance to the MMR vaccine
But the medal *has* to go to catching a cold by being too cold!
With its roots in Aristotle's 4 humours and the middle ages how many people still think that you 'catch a cold' by going out without a jumper.
This 350 years since we looked down microscopes and discovered bactreia and later virus'!
The postulated existence of an infinite number of parallel universes.
I am not an expert in anything relevant, but whenever I hear this proposed I imediately think of Occam's Razor.
Actually I'm not sure at anyone really believes in this, or what it is that is meant by it, but the notion keeps rearing its head. It is at once unprovable and infinitely complex. Why not just postulate a supreme deity while one's about it?
I am not an expert in anything relevant, but whenever I hear this proposed I imediately think of Occam's Razor.
Actually I'm not sure at anyone really believes in this, or what it is that is meant by it, but the notion keeps rearing its head. It is at once unprovable and infinitely complex. Why not just postulate a supreme deity while one's about it?
There are a number of different theories which invoke parallel Universes. While each has different origins, the basic idea behind them is broadly the same, in that in my opinion they emerge out of a fear of "small probabilities". If, for example, there is a huge number of different options for the shape of the Universe, and no way in which to choose between them, or expect one to be more likely than the other, then the way to explain why we see this one is because all the others did actually occur elsewhere, and we just happen to be in this particular one.
As it happens the ideas all tend to have a certain amount of mathematical support, in that they can naturally emerge from the maths. Also, if you have all sort of Universes bouncing around in some meta Universe supporting them, it's possible that two universes will collide and this is likely to leave some sort of signal that can be detected. So it may not quite be total garbage after all.
I am nevertheless very sceptical because the original argument might be based on the same faulty thinking behind, say, the "747 argument" that tries to support a designer because of how apparently unlikely it is that complex proteins can form. But this argument fails because the structures of complex proteins are energetically favourable rather than entirely random. It might be the case that some higher theory constrains the structure of the Universe more tightly than it appears, the large numbers disappear, and therefore there is no need for a multiverse after all.
I've already discussed the separate "many-worlds interpretation" o Quantum mechanics, that is rather closer to the parallel universes idea. This too is based on a fear of probability, I think, and is therefore equally un-necessary. It also is usually untestable in most constructions, although apparently some theorists have suggested ways in which the separate worlds might interact afterwards, which could again lead to a signal.
In my opinion, as long as an idea is untestable and unfalsifiable it's unscientific (though even falsifiability is up as an "idea ready for retirement" in the above link -- I can't agree with Sean Carroll, who seems to be discarding the idea of falsifiability deliberately in order to allow the multiverse idea to be considered Science). In this sense then, the multiverse/ parallel Universe ideas belong to philosophy, or metaphysics, rather than Science. At least for the time being.
Still, it's almost incredible that we could even conceive of such ideas and have some way of justifying them. Perhaps in time the ideas will turn out to be far more relevant than it appears. In turn this probably suggests that the entire question should be answered in the simple way:
"The scientific ideas ready for retirement are those, and only those, which the data rule out definitively."
All other ideas could turn out to be dead ends, but one other feature of the History of Science is that some ideas turn out to be declared dead only to come back with a vengeance, or are ignored until suddenly some experiment comes along that spectacularly confirms it. As long as Science is not at an end, this possibility of sudden resurrection of forgotten ideas remains possible, although in some cases highly unlikely.
As it happens the ideas all tend to have a certain amount of mathematical support, in that they can naturally emerge from the maths. Also, if you have all sort of Universes bouncing around in some meta Universe supporting them, it's possible that two universes will collide and this is likely to leave some sort of signal that can be detected. So it may not quite be total garbage after all.
I am nevertheless very sceptical because the original argument might be based on the same faulty thinking behind, say, the "747 argument" that tries to support a designer because of how apparently unlikely it is that complex proteins can form. But this argument fails because the structures of complex proteins are energetically favourable rather than entirely random. It might be the case that some higher theory constrains the structure of the Universe more tightly than it appears, the large numbers disappear, and therefore there is no need for a multiverse after all.
I've already discussed the separate "many-worlds interpretation" o Quantum mechanics, that is rather closer to the parallel universes idea. This too is based on a fear of probability, I think, and is therefore equally un-necessary. It also is usually untestable in most constructions, although apparently some theorists have suggested ways in which the separate worlds might interact afterwards, which could again lead to a signal.
In my opinion, as long as an idea is untestable and unfalsifiable it's unscientific (though even falsifiability is up as an "idea ready for retirement" in the above link -- I can't agree with Sean Carroll, who seems to be discarding the idea of falsifiability deliberately in order to allow the multiverse idea to be considered Science). In this sense then, the multiverse/ parallel Universe ideas belong to philosophy, or metaphysics, rather than Science. At least for the time being.
Still, it's almost incredible that we could even conceive of such ideas and have some way of justifying them. Perhaps in time the ideas will turn out to be far more relevant than it appears. In turn this probably suggests that the entire question should be answered in the simple way:
"The scientific ideas ready for retirement are those, and only those, which the data rule out definitively."
All other ideas could turn out to be dead ends, but one other feature of the History of Science is that some ideas turn out to be declared dead only to come back with a vengeance, or are ignored until suddenly some experiment comes along that spectacularly confirms it. As long as Science is not at an end, this possibility of sudden resurrection of forgotten ideas remains possible, although in some cases highly unlikely.
Latest Posts
Quizzes & Puzzles1 min ago
Quizzes & Puzzles3 mins ago
Quizzes & Puzzles8 mins ago
Law14 mins ago
Related Questions
Sorry, we can't find any related questions. Try using the search bar at the top of the page to search for some keywords, or choose a topic and submit your own question.