"How would I have known that looking at the picture revealed the error instantly if I didn't read the article?"
Good point. I was speaking generally though and not picking a fight. I raised the SDSS because looking at the pictures is exactly what they're getting people to do - they're just not doing it themselves. It would be an ideal; we're not really built to see patterns in numbers although we *can*. But we are built to see patterns in images, and to hear them in sounds and that's a very quick way of spotting mistakes or errors. (Of course in this case it involves cross-checking a catalogue with the image, but we can easily imagine quite a few ways of setting that up over the internet without forcing someone to actually wade through the catalogue.)
And yes, I think there's a difference between "if one reads the article" and "if you read the article" - both in meaning and in how it makes you sound. "One" is an extremely useful word but unfortunately - on this side of the Atlantic at least - I associate it with days long past and over-wealthy upper middle-classes who long so much to be upper class. Wrong and unfair, of course, but yes.
i doubt you're being entirely serious but i'm going to vent spleen anyway:) despite what people think every large astronomical survey that i've ever heard of makes its data fully public -- sometimes with a year's delay to allow project members to work on the data (which is reasonable; these are the universities that paid to run the damned thing in the first place so they should be able to get first crack at it) but it still all gets released. if you dislike what nasa have done with the cmb data, for example, you can feel perfectly free to go and download seven year's worth of raw time-ordered data and trawl through the entire fucking process. a couple of chinese guys did that a year or two back and found some minor differences from nasa (and a couple of major ones). since it seemed to hinge on where the satellite was pointing and whether that gets identified with the centre or edge of a pixel it's most likely that nasa, with access to the engineers who built and programmed the satellite, are right... but maybe not. and in any event it's useful for someone to do it. i'd just rather it's not me.
the same comments all go for something like kepler, too.....
kepler isn't really a project to sit there and find specific examples. i don't even know why they released this one except i guess they thought it of great interest (and wanted publicity, no doubt - which might not even have been the project's idea, the same as the entire cold fusion debacle came from a press release the researchers didn't want to go out but were forced into by the university who were grubbing for publicity). it's a project to collect vast numbers of planets so that we can say that we expect such-and-so-many earth-like planets around.
it's similar in a way to something like the sloan digital sky survey (which i know more about) which basically takes photos of enormous numbers of galaxies in an attempt to find out about the universe. it's not actually looking for details of a galaxy at a redshift of 0.9 though you can certainly use sloan's catalogue to do so. sloan itself doesn't give a shit; it uses a somewhat inaccurate way of estimating the redshift (because collecting the data to do it accurately would take a lot of time and expense... and then analysing it would take much longer again; photometric is the only sane way to go when you get to the number of galaxies sloan looks at) and piles galaxy upon galaxy to beat down the statistical error. kepler does the same, except that instead of finding the mass content of the universe and the baryon acoustic oscillations, it's looking for earth-like planets.
mod +1 insightful. kepler is fucking *huge* and oddly enough they don't have the money to employ people to go through and double-check everything that appears in it against a few different catalogues and then visually against an image. if they did, people would rightly be making a scene at the enormous waste of tax-payer's money.
Hey, sci-fi still can be fun sometimes, even though I feel going back to the Moon is pointless given its cost when there's so little money around and so many more important things to do with that (breaking our dependence on Middle-Eastern oil, finally defeating cancer, defeating AIDS would be my three main targets), going to Mars is seriously pointless, going to asteroids is more or less pointless, space escalators will never be built, humanity will never colonise another planet without prohibitively expensive ties to Earth being maintained, and we're doomed to die either when we pollute our planet enough that it kills us, when we exhaust every last exploitable resource or, ultimately, when the sun becomes a red giant.
I like the idea of massive solar panels in space, but getting that energy to Earth would be enormous fun. Either you tie them to Earth with insulated cable, which would be horribly expensive, wasteful, prone to terrorist attack or accident (or planes flying into it) *and* require the panels to be under more-or-less continuous thrust to keep them in geosynchronous orbit; or you try and beam it down. Forgive me for not wanting to stand beneath a beam of energy enough to provide power even for the UK, let alone the USA. Or we split it into loads of little beams, necessitating hundreds and thousands of receivers around the country... which birds shit on. A bit of a waste of all that power if it goes to heating up pigeon shit. Even worse, the first rainy day and all that energy gets blasted into the clouds, scattered, diffused, and bathes the entire country in low-level radiation. Sure, those problems could be overcome, but I'm sure there are thousands that I haven't thought of.
I *can* see the point on some manufacturing on the Moon though. If it was possible to get things back from the Moon cheaply (it's not) then it would be a brilliant environment for it. We manufacture on Earth and we shit in our own backyard. On the Moon, who cares what crap we pile onto it? We can dose the whole thing with uranium and not give a shit, although we might have to put the goods that come down from it into quarantine for a few centuries if we were silly enough to do that.
Though saying that I'm sure that some Diana-loving maniacs would complain about us polluting a barren, pointless ball of dust and rock.
Anyway. I agree with a lot of what you say... except that I'm employed to study cosmology, which is in all fairness amongst the most narcissistic of human sciences. It's simply an attempt to understand why the entire universe looks the way it does. Sure, there are knock-on effects that are useful - cosmology is a big part of the drive for evermore accurate telescopes which have helped drive everything from hyper-sensitive electrical motors to the CCDs sitting in your digital camera... but seriously, all that would have come anyway. Whether from spy satellites, where a lot of it was probably developed first anyway, or otherwise, even just from more sane, rational astronomy (which can actually have practical benefits) it would have come. Nah, cosmology is pointless. But it's of interest and governments around the world, from the USA to Iran, judge it of interest and use enough to fund its research. And who am I to argue if someone pays me to basically pursue my hobby full-time?
Also, if you read the article (or knew about Kepler) you'd also know that they deal with hundreds of thousands of these cases -- Kepler isn't about targeting individual stars and finding Earth-like planets, it's about getting weight of numbers on your side to beat down the statistical error. I don't know why they picked this exact case to look at when they admitted themselves that it was a questionable one (maybe because it had turned out so nicely inside the habitable zone with a sane mass?) but that's not the aim of the project anyway.
My point is that no-one at Kepler is employed to sit and stare at a few hundred thousand images saying "Yes, that star's brighter than that one" or "Yes, that looks like an isolated system" or "No, that looks like CCD noise, bin it". *Should* there be someone employed to do this? Maybe so, but you'd have to admit it would be one seriously boring job -- and you'd need a good few sets of eyes given the bad human judgement in it. There just isn't enough money going around to hire someone to stare at images full-time like that, let alone to hire three or four people to do it... In similar circumstances the Sloan Deep Sky Survey set up Galaxy Zoo over the internet to get people around the world to classify hundreds of thousands of images of galaxies for them. Get thousands of people assessing the same galaxy and the hope is that they clump around a decent classification. (I may not necessarily agree with that unless you can guarantee all of them have gone through the test cases carefully enough, but I know a few people involved in setting it up and they did think of that kind of thing; there are a good number of known test cases thrown out to assess how accurate people's judgement is and use that as a weighting factor.)
NB: I'm not saying *you* didn't read the article, I've no idea. But saying "If one reads the article" makes you sound seriously stuck-up so I avoid doing that...
Einstein got his PhD based on theories that he later overturned, in as fundamental a manner as is possible in physics. The same goes for Dirac, Schroedinger, de Broglie, Compton, Heisenberg and Bohr. Should their PhDs all have been revoked because they were found to be based on an inaccurate theory?
I don't think anyone sane claims that scientists are without bias. They might claim "science" (in some unreal Platonic ideal sense) is unbiased, but scientists never can be because no-one ever is. The *data* is unbiased; you just need to be aware of what it actually means. So for example if you're measuring the temperature of an iron rod under increasing electric load and fail to mention that one end of it is in broad sunlight, your data may be unbiased but you're going to interpret it totally wrongly (and perhaps deliberately). And scientists... no, of course they're not. It's too political and grant availability is too based on fashions and the whims of partisan selection committees to be unbiased. Most scientists I know tend to be as unbiased as they can within those constraints, but anyone with their head turned on knows they're there. Anyone with their head turned on and a high-school level of history or sociology is also aware that *no-one* is unbiased and any time anyone interprets anything they add their own bias to it, inevitably.
Ultimately the ideal is that you collect data and understand it as best as possible, remove the effects of all known systematics, build a model and make a prediction which can then be tested. Frequently someone comes along and points out another systematic, which then changes your model, predictions and tests. This can repeat for decades (or, in the case of something like evolutionary theory which is still, err, evolving, centuries). It's simply the nature of science.
That isn't what seems to have happened in this case. It *looks* like slack groundwork, right at the start of the calculation. That's not a good image to have. To their credit, they checked, they acknowledged their fault, they repeated their calculation and they published the result knowing full well how silly it would make them look. (Also, in all fairness, knowing full well that someone else would do the same and make them look both silly *and* like they were hiding something.) One thing to take away from that is that the calculation itself appears to be totally sound -- you can bet that someone checked through the entire thing, from the data pipeline to the final calculation. So a typical case of garbage in, garbage out...
And any way of looking at it it makes the Kepler team look daft.
Yes, except their motivation for saying that exotic matter exists is the widely-accepted notion that dark energy exists -- and much as you may dislike it (and I'm no fan myself, as a quick browse through my comments on slashdot would reveal), there is some very strong evidence that *something* of the sort has to exist. So let's take that evidence, which is both observational and theoretical, and assume that there's some basis behind it. That's the very nature of science: make repeated observations, use them to build a predictive theory, test those predictions and then use it to make further predictions. We've done the repeated observations part (from various cosmological observations), made predictions (specifically the location of the baryon acoustic oscillations in the large-scale structure) and seen them validated. All these guys are doing is saying "OK, so dark energy appeals to be real. If we assume it's not a cosmological constant -- and no-one likes the idea of making it a cosmological constant if there's a dynamical alternative -- then we've got a type of matter that violates the weak energy condition. There has been work studying matter that violates the null energy condition, and it fits cosmological observations, so let's apply it to the interior of stars and see what effects you get."
Yes, of course it's speculative. Mainstream cosmology doesn't employ this type of matter because it's very contentious. Even so, it *has* been used in cosmology and hasn't violated any observations, only encountered theoretical prejudice. So it's a valid thing to do to apply it to different situations. They just picked a situation that's of intrinsic interest -- wormholes. And that's only natural.
My point isn't really to argue, it's to point out that it's not as wild and crazy as it seems -- it is, somewhat, since they're deliberately choosing ghost fields, but that's not something new. They've been studied for a very long time. So they've taken something that's been applied to physics and applied it to a new field to see if there are any predictions they can make. Sounds fair enough to me.
Well, to be fair, this "phantom matter" is something that looks suspiciously like a dark energy. It's just a fluid violating an energy condition. If you read the original paper, they even use dark energy as the first motivating example -- although the matter they choose to employ is something that wouldn't normally be used as a dark energy because it's a "ghost" field and can cause some major problems with stability.
... "would solve" is a very strong statement. "Might solve" would be closer, and you'd have to find a way of getting stars to form before the formation of the CMB, or for the rebalancing of the CMB between its formation and the present epoch to not fuck up the anisotropies, which are now very well observed and fit predictions from inflation to a very good degree.
"If you accept A (which not only has nothing to support it at all, but actually has strong theoretical reasons to assume to be false)"
Given you appear to be talking about "phantom matter", care to back that up? Sure, it violates a couple of energy conditions but those energy conditions themselves are pretty arbitrary in the first place. "Phantom matter" is also known as "dark energy" and like it or not, there's a hell of a lot of support for something like dark energy. You have to be very clear about what you're objecting to if there's a lot of evidence -- observational evidence -- for something.
That said, do I believe any of this is actual physical reality? No, I don't.
Alternatively, get hold of a copy of Snow Leopard and put that on your Mac Pro, it'll still be supported for a good few years. And while I agree that I don't like the look of some of Apple's "features" in Lion, on the whole it looks like a very minor change over Snow Leopard and I suspect it'll be possible to ignore most of the new stuff. If not, well, I can always swap to Linux or FreeBSD.
I've never actually liked focus-follows-mouse since I tend to fiddle with the mouse a lot while I'm reading something onscreen, or knocking it out of my way otherwise. It frustrates me, so its lack in OSX isn't an issue for me. Obviously it's something other people like, but if the lack of focus-follows-mouse is actually a deal-breaker then I think you're probably inventing reasons not to go for OSX instead of just admitting that you don't want to swap to OSX, which would be just as reasonable. (It's only a fucking OS. Seriously, why is this such an important issue?)
As for "How am I supposed to get any work done on this thing?", well, depends what work you're doing. I've not found anything that I can do on Linux that I *can't* do on OSX. It's still a UNIX-like system (yes yes I know it's certified UNIX and Linux isn't; but who actually cares? All we care about is that it's UNIX-like), it's got a GNU-based userspace, the versions of things like gcc might be a couple of steps behind the bleeding edge but they're that in something like Debian, too, and I can still get hold of software I want and compile it up. It's just another UNIX. Drop to command-line and I can basically do what I want.
Maybe you're doing something really niche that seriously *does* need Linux, in which case great - don't go for OSX, why would you? That's like someone who uses Windows-only software swapping to AmigaOS.
Feynman was very open about how he enjoyed teaching because it gives you something to do in the time between ideas. It's kind of a way of living with the pain of breaking new scientific ground for a living:) The rest of us -- those not on permanent contracts but stuck in postdoctoral hell -- simply can't afford to do it. Step too far from the mainstream and you get no publications and, more importantly, no citations. That means you get no job. And unfortunately if Einstein were living now, relativity wouldn't be accepted unless he cut a deal with a university letting him put them as his affiliation. No university name = no publication. The system's even set up that you have to declare your affiliation. Certainly you can talk with a local department and might get permission to affiliate yourself with them, but then they're not going to be happy if you start publishing crackpot theories under their banner, so you'd better damn well make sure you've come up with the new relativity in your spare time from working as a mechanic because otherwise you won't get published and you sure as hell won't get back into academia.
(Seriously, it's very hard to leave and come back. I know one or two people who've done it. Otherwise you're left behind way too quickly.)
As spiralx said, so far as I know no-one claims that Einstein was right in that. Entanglement appears to be a fact of life. Arguing over what it *means* continues, of course, but I don't know much about that.
Hawking radiation hasn't been completely disproven -- I'd like any physicist claiming that to resubmit their papers to the arxiv so we can all see their claims. There's a kind of equivalent argument that helps demonstrate radiation from horizons (though you'll have to take parts of it on trust). It's actually relatively quick to see that if you accelerate an observer (with constant acceleration), he sees radiation coming from the normal (unaccelerated) vacuum. It's called the Unruh effect (http://en.wikipedia.org/wiki/Unruh_effect). The interesting parts come because if you draw a space-time diagram of constantly-accelerated observers, you see "horizons" forming -- space is split into four patches and a constantly-accelerated observer can never cross between them. Well, unless he slams on the brakes and changes direction but then he's not a constantly-accelerated observer anymore. So we've got horizons of some form. (This is called Rindler geometry, by the by.) The other interesting part is the equivalence principle, locally equating gravity with acceleration. Via the equivalence principle, the Unruh effect predicts that radiation is emitted in the vicinity of horizons.
That radiation is also Hawking -- that is, Planckian -- though demonstrating that involves a very different calculation to Hawkings' original.
I'd be very suspicious of anyone proving to disprove Hawking radiation. It's always possible, I guess, that it doesn't occur in quantum gravity (and hence not in reality) because it's a semi-classical calculation involving quantum fields on classical spacetimes, but I'd be surprised. I don't know that anyone's developed any QM theory of gravity enough to test anything, though. So far as I know, they haven't (loop quantum would be the immediate suspect, and I've not heard that someone's been able to calculate the radiation or lack of it in LQG) but I'm not in the field so I can miss things.
Gravity waves I believe is a matter of time... or that they really will be so weak as to be practically unobservable. Personally I think GR is too successful for there not to be gravitational radiation flying around, we just may never see it (though I think we will).
Gravitons are a different matter. They're a quantum field-theoretical construct ascribed perhaps more status than I'd like. You can derive (classical, linearised) GR by considering interactions of massless spin 2 particles, so you call them gravitons. Quantising them has turned out to be an enormous wild goose chase. I think it's not the way to go... btu I'm not a quantum field theorist, and one of the initial drivers of string theory was that one of the fundamental modes of the string turned out to be massless and spin 2. That wasn't in the original game-plan, it dropped out of it. That excited a lot of people and I can see why. But going with them means you more or less drop the geometrical interpretation of gravity (at least as a "fundamental" thing). In the mid 70s Weinberg wrote (in "Gravitation and Cosmology") something like 'no-one takes the geometrical interpretation of gravity seriously anymore'. I've a lot of respect for Weinberg (well, duh) but I disagree with him here - plenty of people do, just maybe not all the other particle and field theorists he had coffee with each morning...
But what marks the CMB out as special in your mind? The only thing it's got in your argument is that it's not rotating (which again involves basically just sticking to its monopole; some of the CMB anisotropies *are* from vorticity). But the same goes for the galaxy distribution on very large scales -- neither all that hydrogen nor the CDM we assume to sit there appear to be rotating either. I could pick either of those as my frame (and I normally pick CDM because it's convenient). "Rotation" is an intrinsic quality, it can be very well defined within the bounds of GR. You can make entire spacetimes that rotate, sure (the Goedel spacetime is probably the best known not least because it contains closed timelike loops, but more mundanely a Kerr spacetime is a rotating black hole), or you can have fluids that rotate in a spacetime (inevitably dragging some of the spacetime with them), or both.
There's no one "preferred frame" for rotation, I can pick any number of frames. I can attach them to particular fluids (such as the CMB, baryons, CDM, white-haired dogs) and then when I separate my fluid quantities out with respect to it, a particular rotation term emerges. It's all fine, I assure you.
Of course, if you don't want to trust the underpinnings of relativity it's a bit different, but the same happens with any metric-based theory of gravity so you'd have to be looking at something pretty exotic (at least for modern physics).
Don't absolutely rely on me for that - as I say I'm not a particle physicist. I'd more or less agree, except that you can do science degrees in pure maths, which is not only not falsifiable but it in principle has nothing to do with reality. You can play annoying word games and point out that any time you try and apply it to reality you're an applied mathematician. The same goes for things like string theory. But unfortunately high-energy physics will always be like this; there are limits that we can feasibly probe, but it doesn't mean the universe doesn't operate on those scales because clearly it does (or did). Unfortunate but true:( We may never really know how things behave. And I'm often swayed by Feynman's point that there may *be* no theory of everything and we shouldn't expect to find one because ultimately physics is an ensemble of algorithms. Set up a situation (say an electron and a photon approaching one another) and see what you expect to come out (scattering). Plug in the numbers and get a prediction. That's ultimately physics. Why should one set of algorithms play nicely with another? It's a bleak view but sometimes I feel there's validity in it, especially when I look at the long, fascinating and unhappy history of quantum gravity.....
"But you have a strong hint, don't you, from the fact that the numbers seem to work out. You have a number of independent ways of getting at the dark matter in clusters of galaxies -- (M/L), the cluster baryon fraction, weak lensing, etc. -- which are fairly consistent with each other."
Sure:) It's a problem for people like me who more or less oppose Lambda CDM -- anything that replaces it has to work as well as it. And it does work - extremely well. I'm very uneasy about the physical content of it, but up to first-order perturbations it does work well. I guess my main objection to it is that at the current epoch (say redshifts lower than 1) the universe is increasingly badly-described by Lambda CDM.
And the philosophical underpinnings are pretty shaky. The phenomenology is extremely good - not perfect, but extremely good - but the underpinnings are wobbly.
"nobody ignores the large scale structure of the universe in arguing for homogeneity/isotropy. They argue instead that we converge to it as larger volumes are considered."
Yeah, sure -- but write down *how*. That's where you come very unstuck. The way I normally set up the model (because it allows for arguing an alternative) is that first you take the CMB, which is totally isotropic around the Earth (well, to one part in 1000 if you count the dipole, one in 100,000 if you subtract it though doing that is itself an assumption that's purely due to our velocity w.r.t. the CMB) and then assume the Copernican principle. That tells you the universe is homogeneous and isotropic "on average". You can impose that directly, which puts the universe in de Sitter (no), Minkowski (no) or anti de Sitter (no), all of which violate observations dramatically. So instead you foliate spacetime with maximally-symmetric spatial slices. That gives you the RW metric. Then assuming GR forces you to put in perfect fluids and recover the Friedman equations.
All well and good, but that's ignoring, for example, all the anisotropies on the CMB. Obviously at last scattering the CMB was extremely isotropic and the universe very close to homogeneous and isotropic. That's no guarantee it's anything like that now; the only constraints come from very large inhomogeneities fucking up the large-scale CMB (which is, by the by, fucked up but to be fair recovering the low quadrupole from inhomogeneities isn't that straightforward). Remember that the large-scale CMB is created by two things: primordial perturbations, those which were outside of the horizon when the CMB was formed; and late-time effects. The ISW is the best known of them and it's a massive problem for LCDM (well, that's a lie, the theory is only about 1.5 sigma away from the observation, but it *looks* very serious and the low quadrupole does worry a lot of people). They're on large scales because they're effects very near to us, and when things are near to us they look big. Sorry to sound patronising if I do - totally unintended. It's just it's that simple. Obviously it has to be something on the sky that subtends quite a lot of it - the ISW counts because it's the effect of "dark energy" so it's pervasive, and a Gpc scale void counts because, well, it's a fucking gigaparsec big, it'll subtend pretty much everything...:)
Anyway. The standard approach (to setting up the background equatons - I'm not knocking how people find baryon fractions and so on in clusters) also ignores the structures in the universe today. We've got an enormous problem with averaging in relativity -- you can't do it. You can average scalars, of course, but the average of a tensor field breaks covariance because it's coordinate dependent. That immediately should raise suspicions about the Friedman equations. Let's say for example that we want to do an average on the metric. That averaged metric may or may not turn out to be FLRW. We actually have *no clue*. Quite possibly it will so let's for the sake of argument say that on large scales, the universe is actually described by FLRW metric(s). But
I wrote a long reply to this and lost it by being an idiot:( Anyway, I said firstly to dig out a copy of Joao Magueijo's book which may or may not be called "Faster than the Speed of Light" which postulates a changing speed of light and lets you do away with inflation. I think a lot of people would be happy if we could do away with inflation completely, but anything that replaces it has to repeat its successes, which are many.
As for the CMB, it doesn't violate relativity, it's *predicted* by relativity. All it is is a thermal bath of photons left over from the big bang. If you put in a big bang cosmology, that's what you get out. It doesn't pick a physically preferred frame any more than saying "This is MY car and that's a preferred frame!" It is a preferred frame - it's the one you're observing from, but it's not physically preferred. GR is all about being able to look at things from any frame of reference -- that's how it wipes the force of gravity; it notes that it's as fictional as centrifugal force (which it is) and changes reference frames to study it properly. The CMB isn't an absolute inertial reference frame, it just happens to exist. It's not imposed on or by relativity, but if you use a Robertson-Walker solution and put any photons in it at all it's forced on you.
Also it might interest you to know that in the standard cosmological model (phenomenological as it is) we don't even lock ourselves to the CMB. We actually lock ourselves to CDM most of the time, or lock ourselves to some frame that makes gravity look as Newtonian as possible.
As for simultaneity and so on... well, GR is wrong. Unfortunate, but true. How wrong and in what way we don't know, but yes, there seem to be issues.
You can get Schroedinger's equation in a totally deterministic (if rather contrived) manner by taking a totally classical system and adding a modification to the potential. If you then take the momentum potential (or action, however you want to phrase it) and the density you can bundle them together as basically the phase and the amplitude of a wavevector that... solves the Schroedinger equation. Totally deterministic and totally indistinguishable from standard QM. It's also very ugly and is riddled with conceptual difficulties, not least how you go about second-quantising to get something like a field theory, but even so, it's a totally deterministic formulation of standard QM.
Also, as basically a gravity theorist, it pains me to say it but GR is less reliable than the Standard Model. We can test it down to maybe 0.1mm and up to Pluto's orbit and beyond that, nah. It's because gravity's so damned weak. The other forces are much stronger and much better probed, and electromagnetism is the best known of all. QED is still one of our best theories. The problem with taking that too far is that then people have a bad habit of assuming gravitons must form the basis of a successful quantum theory of gravity and throw out geometry altogether -- knowing all the time of course (because they know this shit much better than I) that GR is non renormalisable and you're doomed to failure going at it in anything like a typical way. That way leads to string theories and supergravities and 30 years of intensive maths with absolutely zero solid predictions. A nice alternative is loop quantum gravity, which takes a much more conservative approach, takes geometry seriously, finds a way of quantising spacetime without gravitons and 7 extra dimensions, and has absolutely zero solid predictions for about fifteen years less intensive study. Other alternatives are brilliant, and all of them have absolutely zero predictions, to my knowledge.
As for the black hole question, it's not answerable within "big bang cosmology" because GR would have broken down before you get to the point that all the matter is within its own event horizon. Before GR is at all valid (even assuming we could use it that early) you can't extrapolate. So there's no way of saying that a black hole would inevitably be there because there might not *be* black holes in quantum gravity and on those scales probably aren't. (QG is expected to wash out singularities in some manner, and we have no idea what it's gonna do to event horizons on such small scales but if you've, say, granulised spacetime in some manner you've immediately got quantised areas and volumes as in loop quantum gravity, and that puts limits on horizons.) But it's all speculation... as is saying that there'd be a black hole back then because that extrapolates GR well beyond its regime of validity.
I wouldn't say "it didn't, we live inside still". I've heard people suggest that (I've even seen a Creationist cosmology that used external coordinates to chart the interior of a Schwarzschild hole, put us bang in the centre on top of a singularity and then used gravitational time dilation to prove that the universe is only 6000 years old. No shit, he did it. Of course, it relies on using the wrong time coordinate and the wrong coordinate patch in general, and then assuming that the singularity on the event horizon is real -- it's not -- and then that we can live in the centre of those coordinates, but it was still a stroke of genius.) But I wouldn't. The geometry is weird if you believe the extended Scwarzschild solutions and it's even weirder if you think we live in a charged or rotating hole. I don't believe any of it. I don't really have any firm beliefs one way or another, in all honesty, but definitely not that we live inside some big black hole.
I think I've made my opinions about dark matter and dark energy fairly clear somewhere in these comments, three times I think:) They are band-aids, and very ugly ones at that. MOND is an interesting one. No-one, including Milg
SUSY *is* a hack-job in my opinion. But what I meant is that in, say, MSSM, you've got well more than a hundred free parameters. Some of them are well-constrained; others aren't. Change a few of these and you can change the phenomenology. But as I've said a few times today, I'm not a particle physicist so I'm also happy to be told I'm wrong -- I talk with particle physicists sometimes, and the rest I get through the kind of academic osmosis that causes confusion worldwide...:)
1: Pure phenomenology. No-one constructing inflationary models that I know of actually seriously believes that it's fundamental physics (at least, not after the second or third year of their PhD). What they *do* believe, frequently, is that the phenomenology can help guide a more fundamental theory. Personally I don't always agree with that; I think it can shroud a fundamental theory (in a similar vein to how cosmology is built on phenomenology that basically shrouds a very serious and neglected underlying issue).
Unless you're using the Higgs itself to drive inflation -- Guth's first model did this but it ran into problems with a graceful exit; it's recently been reawakened and re-examinded, though -- you're going to have a massive problem identifying an inflaton. We've not observed *any scalar fields whatsoever*. Even the Higgs remains elusive, though that might change in the near future. (Don't hold your breath.) So you immediately have a problem that what you're doing is specious. You can then either ground your inflaton in a well-reasoned model of high-energy physics or, and this is the standard approach, just invent a scalar field, call it the inflaton, and give it an arbitrary potential. So long as you make the potential flat enough that scalar field is an inflaton.
Basically it's phenomenology. But the people who do it are convinced it gives *suggestions* about what lies underneath, and in some ways they've got a point. Inflation works extremely well and it's standard to assume there was an inflationary epoch. You solve the horizon problem, the flatness problem and (if you believe in various GUTs) the monopole problem. (Basically -- why does the CMB look identical in opposite directions when the universe is too young for them to have ever interacted; why is the universe so fucking SMOOTH; and why do we not see any of these magnetic monopoles that GUTs produce in abundance?) Even more importantly, though, the quantum fluctuations of a scalar field coupled to gravity in the early universe produce tiny seeds that are basically exactly right. You can make models that get them exactly wrong but actually you have to work a bit; basic inflation made a prediction of those seeds, and when WMAP came along and looked at the CMB in unprecedented detail, it was dramatically confirmed. Basically those ripples had to be almost exactly Gaussian random, and "scale-invariant" meaning that the extremely large wavelengths were massively more powerful than the shorter wavelengths. That maps through to the formation of the CMB, when electrons condensed into protons to form hydrogen and light rays could suddenly free-stream carrying with them a photograph of the early universe. And it maps through even further, to the large-scale structure of galaxy clusters where we can look at those very same wavelengths. Much of a shift from those early imprints and that distribution is changed actually quite dramatically.
2: Dark matter is a big issue (well, duh). Basically "dark matter" is a catch-all term for whatever is causing rotation curves to deviate from the Newtonian prediction. I get irritated when people immediately assume it's a new exotic species of particle. I've put a couple of rants on this thread aimed at this kind of thing. My feeling is that dark matter in galaxies (and galaxy clusters) is made up of five or six different effects, *all* of which act as "dark matter", ie to flatten rotation curves: exotic particles perhaps, if supersymmetry is true; massive neutrinos since we now know that they are massive even if we don't know the mass, and neutrinos are so abundant that with *any* mass they form at least a dark matter even if it can't be the full dark matter (attributing the entire dark matter to massive neutrinos badly washes out structure on galaxy cluster scales); relativistic corrections coming from our naive assumptions that galaxies inhabit Minkowski (ie normal flat) space, since they don't, and that may -- *may* -- be able to account for up to roughly a tenth or more of spiral galaxies' dark matter; i
Slashdot Mirror
"How would I have known that looking at the picture revealed the error instantly if I didn't read the article?"
Good point. I was speaking generally though and not picking a fight. I raised the SDSS because looking at the pictures is exactly what they're getting people to do - they're just not doing it themselves. It would be an ideal; we're not really built to see patterns in numbers although we *can*. But we are built to see patterns in images, and to hear them in sounds and that's a very quick way of spotting mistakes or errors. (Of course in this case it involves cross-checking a catalogue with the image, but we can easily imagine quite a few ways of setting that up over the internet without forcing someone to actually wade through the catalogue.)
And yes, I think there's a difference between "if one reads the article" and "if you read the article" - both in meaning and in how it makes you sound. "One" is an extremely useful word but unfortunately - on this side of the Atlantic at least - I associate it with days long past and over-wealthy upper middle-classes who long so much to be upper class. Wrong and unfair, of course, but yes.
knock yourself out, someone already posted the currently-public data
http://science.slashdot.org/comments.pl?sid=2028214&cid=35422006
i doubt you're being entirely serious but i'm going to vent spleen anyway :) despite what people think every large astronomical survey that i've ever heard of makes its data fully public -- sometimes with a year's delay to allow project members to work on the data (which is reasonable; these are the universities that paid to run the damned thing in the first place so they should be able to get first crack at it) but it still all gets released. if you dislike what nasa have done with the cmb data, for example, you can feel perfectly free to go and download seven year's worth of raw time-ordered data and trawl through the entire fucking process. a couple of chinese guys did that a year or two back and found some minor differences from nasa (and a couple of major ones). since it seemed to hinge on where the satellite was pointing and whether that gets identified with the centre or edge of a pixel it's most likely that nasa, with access to the engineers who built and programmed the satellite, are right... but maybe not. and in any event it's useful for someone to do it. i'd just rather it's not me.
the same comments all go for something like kepler, too.....
kepler isn't really a project to sit there and find specific examples. i don't even know why they released this one except i guess they thought it of great interest (and wanted publicity, no doubt - which might not even have been the project's idea, the same as the entire cold fusion debacle came from a press release the researchers didn't want to go out but were forced into by the university who were grubbing for publicity). it's a project to collect vast numbers of planets so that we can say that we expect such-and-so-many earth-like planets around.
it's similar in a way to something like the sloan digital sky survey (which i know more about) which basically takes photos of enormous numbers of galaxies in an attempt to find out about the universe. it's not actually looking for details of a galaxy at a redshift of 0.9 though you can certainly use sloan's catalogue to do so. sloan itself doesn't give a shit; it uses a somewhat inaccurate way of estimating the redshift (because collecting the data to do it accurately would take a lot of time and expense... and then analysing it would take much longer again; photometric is the only sane way to go when you get to the number of galaxies sloan looks at) and piles galaxy upon galaxy to beat down the statistical error. kepler does the same, except that instead of finding the mass content of the universe and the baryon acoustic oscillations, it's looking for earth-like planets.
hahaha
mod +1 insightful. kepler is fucking *huge* and oddly enough they don't have the money to employ people to go through and double-check everything that appears in it against a few different catalogues and then visually against an image. if they did, people would rightly be making a scene at the enormous waste of tax-payer's money.
Hey, sci-fi still can be fun sometimes, even though I feel going back to the Moon is pointless given its cost when there's so little money around and so many more important things to do with that (breaking our dependence on Middle-Eastern oil, finally defeating cancer, defeating AIDS would be my three main targets), going to Mars is seriously pointless, going to asteroids is more or less pointless, space escalators will never be built, humanity will never colonise another planet without prohibitively expensive ties to Earth being maintained, and we're doomed to die either when we pollute our planet enough that it kills us, when we exhaust every last exploitable resource or, ultimately, when the sun becomes a red giant.
I like the idea of massive solar panels in space, but getting that energy to Earth would be enormous fun. Either you tie them to Earth with insulated cable, which would be horribly expensive, wasteful, prone to terrorist attack or accident (or planes flying into it) *and* require the panels to be under more-or-less continuous thrust to keep them in geosynchronous orbit; or you try and beam it down. Forgive me for not wanting to stand beneath a beam of energy enough to provide power even for the UK, let alone the USA. Or we split it into loads of little beams, necessitating hundreds and thousands of receivers around the country... which birds shit on. A bit of a waste of all that power if it goes to heating up pigeon shit. Even worse, the first rainy day and all that energy gets blasted into the clouds, scattered, diffused, and bathes the entire country in low-level radiation. Sure, those problems could be overcome, but I'm sure there are thousands that I haven't thought of.
I *can* see the point on some manufacturing on the Moon though. If it was possible to get things back from the Moon cheaply (it's not) then it would be a brilliant environment for it. We manufacture on Earth and we shit in our own backyard. On the Moon, who cares what crap we pile onto it? We can dose the whole thing with uranium and not give a shit, although we might have to put the goods that come down from it into quarantine for a few centuries if we were silly enough to do that.
Though saying that I'm sure that some Diana-loving maniacs would complain about us polluting a barren, pointless ball of dust and rock.
Anyway. I agree with a lot of what you say... except that I'm employed to study cosmology, which is in all fairness amongst the most narcissistic of human sciences. It's simply an attempt to understand why the entire universe looks the way it does. Sure, there are knock-on effects that are useful - cosmology is a big part of the drive for evermore accurate telescopes which have helped drive everything from hyper-sensitive electrical motors to the CCDs sitting in your digital camera... but seriously, all that would have come anyway. Whether from spy satellites, where a lot of it was probably developed first anyway, or otherwise, even just from more sane, rational astronomy (which can actually have practical benefits) it would have come. Nah, cosmology is pointless. But it's of interest and governments around the world, from the USA to Iran, judge it of interest and use enough to fund its research. And who am I to argue if someone pays me to basically pursue my hobby full-time?
Also, if you read the article (or knew about Kepler) you'd also know that they deal with hundreds of thousands of these cases -- Kepler isn't about targeting individual stars and finding Earth-like planets, it's about getting weight of numbers on your side to beat down the statistical error. I don't know why they picked this exact case to look at when they admitted themselves that it was a questionable one (maybe because it had turned out so nicely inside the habitable zone with a sane mass?) but that's not the aim of the project anyway.
My point is that no-one at Kepler is employed to sit and stare at a few hundred thousand images saying "Yes, that star's brighter than that one" or "Yes, that looks like an isolated system" or "No, that looks like CCD noise, bin it". *Should* there be someone employed to do this? Maybe so, but you'd have to admit it would be one seriously boring job -- and you'd need a good few sets of eyes given the bad human judgement in it. There just isn't enough money going around to hire someone to stare at images full-time like that, let alone to hire three or four people to do it... In similar circumstances the Sloan Deep Sky Survey set up Galaxy Zoo over the internet to get people around the world to classify hundreds of thousands of images of galaxies for them. Get thousands of people assessing the same galaxy and the hope is that they clump around a decent classification. (I may not necessarily agree with that unless you can guarantee all of them have gone through the test cases carefully enough, but I know a few people involved in setting it up and they did think of that kind of thing; there are a good number of known test cases thrown out to assess how accurate people's judgement is and use that as a weighting factor.)
NB: I'm not saying *you* didn't read the article, I've no idea. But saying "If one reads the article" makes you sound seriously stuck-up so I avoid doing that...
Einstein got his PhD based on theories that he later overturned, in as fundamental a manner as is possible in physics. The same goes for Dirac, Schroedinger, de Broglie, Compton, Heisenberg and Bohr. Should their PhDs all have been revoked because they were found to be based on an inaccurate theory?
I don't think anyone sane claims that scientists are without bias. They might claim "science" (in some unreal Platonic ideal sense) is unbiased, but scientists never can be because no-one ever is. The *data* is unbiased; you just need to be aware of what it actually means. So for example if you're measuring the temperature of an iron rod under increasing electric load and fail to mention that one end of it is in broad sunlight, your data may be unbiased but you're going to interpret it totally wrongly (and perhaps deliberately). And scientists... no, of course they're not. It's too political and grant availability is too based on fashions and the whims of partisan selection committees to be unbiased. Most scientists I know tend to be as unbiased as they can within those constraints, but anyone with their head turned on knows they're there. Anyone with their head turned on and a high-school level of history or sociology is also aware that *no-one* is unbiased and any time anyone interprets anything they add their own bias to it, inevitably.
Ultimately the ideal is that you collect data and understand it as best as possible, remove the effects of all known systematics, build a model and make a prediction which can then be tested. Frequently someone comes along and points out another systematic, which then changes your model, predictions and tests. This can repeat for decades (or, in the case of something like evolutionary theory which is still, err, evolving, centuries). It's simply the nature of science.
That isn't what seems to have happened in this case. It *looks* like slack groundwork, right at the start of the calculation. That's not a good image to have. To their credit, they checked, they acknowledged their fault, they repeated their calculation and they published the result knowing full well how silly it would make them look. (Also, in all fairness, knowing full well that someone else would do the same and make them look both silly *and* like they were hiding something.) One thing to take away from that is that the calculation itself appears to be totally sound -- you can bet that someone checked through the entire thing, from the data pipeline to the final calculation. So a typical case of garbage in, garbage out...
And any way of looking at it it makes the Kepler team look daft.
Yes, except their motivation for saying that exotic matter exists is the widely-accepted notion that dark energy exists -- and much as you may dislike it (and I'm no fan myself, as a quick browse through my comments on slashdot would reveal), there is some very strong evidence that *something* of the sort has to exist. So let's take that evidence, which is both observational and theoretical, and assume that there's some basis behind it. That's the very nature of science: make repeated observations, use them to build a predictive theory, test those predictions and then use it to make further predictions. We've done the repeated observations part (from various cosmological observations), made predictions (specifically the location of the baryon acoustic oscillations in the large-scale structure) and seen them validated. All these guys are doing is saying "OK, so dark energy appeals to be real. If we assume it's not a cosmological constant -- and no-one likes the idea of making it a cosmological constant if there's a dynamical alternative -- then we've got a type of matter that violates the weak energy condition. There has been work studying matter that violates the null energy condition, and it fits cosmological observations, so let's apply it to the interior of stars and see what effects you get."
Yes, of course it's speculative. Mainstream cosmology doesn't employ this type of matter because it's very contentious. Even so, it *has* been used in cosmology and hasn't violated any observations, only encountered theoretical prejudice. So it's a valid thing to do to apply it to different situations. They just picked a situation that's of intrinsic interest -- wormholes. And that's only natural.
My point isn't really to argue, it's to point out that it's not as wild and crazy as it seems -- it is, somewhat, since they're deliberately choosing ghost fields, but that's not something new. They've been studied for a very long time. So they've taken something that's been applied to physics and applied it to a new field to see if there are any predictions they can make. Sounds fair enough to me.
Well, to be fair, this "phantom matter" is something that looks suspiciously like a dark energy. It's just a fluid violating an energy condition. If you read the original paper, they even use dark energy as the first motivating example -- although the matter they choose to employ is something that wouldn't normally be used as a dark energy because it's a "ghost" field and can cause some major problems with stability.
... "would solve" is a very strong statement. "Might solve" would be closer, and you'd have to find a way of getting stars to form before the formation of the CMB, or for the rebalancing of the CMB between its formation and the present epoch to not fuck up the anisotropies, which are now very well observed and fit predictions from inflation to a very good degree.
"If you accept A (which not only has nothing to support it at all, but actually has strong theoretical reasons to assume to be false)"
Given you appear to be talking about "phantom matter", care to back that up? Sure, it violates a couple of energy conditions but those energy conditions themselves are pretty arbitrary in the first place. "Phantom matter" is also known as "dark energy" and like it or not, there's a hell of a lot of support for something like dark energy. You have to be very clear about what you're objecting to if there's a lot of evidence -- observational evidence -- for something.
That said, do I believe any of this is actual physical reality? No, I don't.
Alternatively, get hold of a copy of Snow Leopard and put that on your Mac Pro, it'll still be supported for a good few years. And while I agree that I don't like the look of some of Apple's "features" in Lion, on the whole it looks like a very minor change over Snow Leopard and I suspect it'll be possible to ignore most of the new stuff. If not, well, I can always swap to Linux or FreeBSD.
I've never actually liked focus-follows-mouse since I tend to fiddle with the mouse a lot while I'm reading something onscreen, or knocking it out of my way otherwise. It frustrates me, so its lack in OSX isn't an issue for me. Obviously it's something other people like, but if the lack of focus-follows-mouse is actually a deal-breaker then I think you're probably inventing reasons not to go for OSX instead of just admitting that you don't want to swap to OSX, which would be just as reasonable. (It's only a fucking OS. Seriously, why is this such an important issue?)
As for "How am I supposed to get any work done on this thing?", well, depends what work you're doing. I've not found anything that I can do on Linux that I *can't* do on OSX. It's still a UNIX-like system (yes yes I know it's certified UNIX and Linux isn't; but who actually cares? All we care about is that it's UNIX-like), it's got a GNU-based userspace, the versions of things like gcc might be a couple of steps behind the bleeding edge but they're that in something like Debian, too, and I can still get hold of software I want and compile it up. It's just another UNIX. Drop to command-line and I can basically do what I want.
Maybe you're doing something really niche that seriously *does* need Linux, in which case great - don't go for OSX, why would you? That's like someone who uses Windows-only software swapping to AmigaOS.
Gentoo and Arch are old-school Linux....?
Group-theoretical epicycles. Marvelous...
Feynman was very open about how he enjoyed teaching because it gives you something to do in the time between ideas. It's kind of a way of living with the pain of breaking new scientific ground for a living :) The rest of us -- those not on permanent contracts but stuck in postdoctoral hell -- simply can't afford to do it. Step too far from the mainstream and you get no publications and, more importantly, no citations. That means you get no job. And unfortunately if Einstein were living now, relativity wouldn't be accepted unless he cut a deal with a university letting him put them as his affiliation. No university name = no publication. The system's even set up that you have to declare your affiliation. Certainly you can talk with a local department and might get permission to affiliate yourself with them, but then they're not going to be happy if you start publishing crackpot theories under their banner, so you'd better damn well make sure you've come up with the new relativity in your spare time from working as a mechanic because otherwise you won't get published and you sure as hell won't get back into academia.
(Seriously, it's very hard to leave and come back. I know one or two people who've done it. Otherwise you're left behind way too quickly.)
As spiralx said, so far as I know no-one claims that Einstein was right in that. Entanglement appears to be a fact of life. Arguing over what it *means* continues, of course, but I don't know much about that.
Hawking radiation hasn't been completely disproven -- I'd like any physicist claiming that to resubmit their papers to the arxiv so we can all see their claims. There's a kind of equivalent argument that helps demonstrate radiation from horizons (though you'll have to take parts of it on trust). It's actually relatively quick to see that if you accelerate an observer (with constant acceleration), he sees radiation coming from the normal (unaccelerated) vacuum. It's called the Unruh effect (http://en.wikipedia.org/wiki/Unruh_effect). The interesting parts come because if you draw a space-time diagram of constantly-accelerated observers, you see "horizons" forming -- space is split into four patches and a constantly-accelerated observer can never cross between them. Well, unless he slams on the brakes and changes direction but then he's not a constantly-accelerated observer anymore. So we've got horizons of some form. (This is called Rindler geometry, by the by.) The other interesting part is the equivalence principle, locally equating gravity with acceleration. Via the equivalence principle, the Unruh effect predicts that radiation is emitted in the vicinity of horizons.
That radiation is also Hawking -- that is, Planckian -- though demonstrating that involves a very different calculation to Hawkings' original.
I'd be very suspicious of anyone proving to disprove Hawking radiation. It's always possible, I guess, that it doesn't occur in quantum gravity (and hence not in reality) because it's a semi-classical calculation involving quantum fields on classical spacetimes, but I'd be surprised. I don't know that anyone's developed any QM theory of gravity enough to test anything, though. So far as I know, they haven't (loop quantum would be the immediate suspect, and I've not heard that someone's been able to calculate the radiation or lack of it in LQG) but I'm not in the field so I can miss things.
Gravity waves I believe is a matter of time... or that they really will be so weak as to be practically unobservable. Personally I think GR is too successful for there not to be gravitational radiation flying around, we just may never see it (though I think we will).
Gravitons are a different matter. They're a quantum field-theoretical construct ascribed perhaps more status than I'd like. You can derive (classical, linearised) GR by considering interactions of massless spin 2 particles, so you call them gravitons. Quantising them has turned out to be an enormous wild goose chase. I think it's not the way to go... btu I'm not a quantum field theorist, and one of the initial drivers of string theory was that one of the fundamental modes of the string turned out to be massless and spin 2. That wasn't in the original game-plan, it dropped out of it. That excited a lot of people and I can see why. But going with them means you more or less drop the geometrical interpretation of gravity (at least as a "fundamental" thing). In the mid 70s Weinberg wrote (in "Gravitation and Cosmology") something like 'no-one takes the geometrical interpretation of gravity seriously anymore'. I've a lot of respect for Weinberg (well, duh) but I disagree with him here - plenty of people do, just maybe not all the other particle and field theorists he had coffee with each morning...
But what marks the CMB out as special in your mind? The only thing it's got in your argument is that it's not rotating (which again involves basically just sticking to its monopole; some of the CMB anisotropies *are* from vorticity). But the same goes for the galaxy distribution on very large scales -- neither all that hydrogen nor the CDM we assume to sit there appear to be rotating either. I could pick either of those as my frame (and I normally pick CDM because it's convenient). "Rotation" is an intrinsic quality, it can be very well defined within the bounds of GR. You can make entire spacetimes that rotate, sure (the Goedel spacetime is probably the best known not least because it contains closed timelike loops, but more mundanely a Kerr spacetime is a rotating black hole), or you can have fluids that rotate in a spacetime (inevitably dragging some of the spacetime with them), or both.
There's no one "preferred frame" for rotation, I can pick any number of frames. I can attach them to particular fluids (such as the CMB, baryons, CDM, white-haired dogs) and then when I separate my fluid quantities out with respect to it, a particular rotation term emerges. It's all fine, I assure you.
Of course, if you don't want to trust the underpinnings of relativity it's a bit different, but the same happens with any metric-based theory of gravity so you'd have to be looking at something pretty exotic (at least for modern physics).
Don't absolutely rely on me for that - as I say I'm not a particle physicist. I'd more or less agree, except that you can do science degrees in pure maths, which is not only not falsifiable but it in principle has nothing to do with reality. You can play annoying word games and point out that any time you try and apply it to reality you're an applied mathematician. The same goes for things like string theory. But unfortunately high-energy physics will always be like this; there are limits that we can feasibly probe, but it doesn't mean the universe doesn't operate on those scales because clearly it does (or did). Unfortunate but true :( We may never really know how things behave. And I'm often swayed by Feynman's point that there may *be* no theory of everything and we shouldn't expect to find one because ultimately physics is an ensemble of algorithms. Set up a situation (say an electron and a photon approaching one another) and see what you expect to come out (scattering). Plug in the numbers and get a prediction. That's ultimately physics. Why should one set of algorithms play nicely with another? It's a bleak view but sometimes I feel there's validity in it, especially when I look at the long, fascinating and unhappy history of quantum gravity.....
"But you have a strong hint, don't you, from the fact that the numbers seem to work out. You have a number of independent ways of getting at the dark matter in clusters of galaxies -- (M/L), the cluster baryon fraction, weak lensing, etc. -- which are fairly consistent with each other."
Sure :) It's a problem for people like me who more or less oppose Lambda CDM -- anything that replaces it has to work as well as it. And it does work - extremely well. I'm very uneasy about the physical content of it, but up to first-order perturbations it does work well. I guess my main objection to it is that at the current epoch (say redshifts lower than 1) the universe is increasingly badly-described by Lambda CDM.
And the philosophical underpinnings are pretty shaky. The phenomenology is extremely good - not perfect, but extremely good - but the underpinnings are wobbly.
"nobody ignores the large scale structure of the universe in arguing for homogeneity/isotropy. They argue instead that we converge to it as larger volumes are considered."
Yeah, sure -- but write down *how*. That's where you come very unstuck. The way I normally set up the model (because it allows for arguing an alternative) is that first you take the CMB, which is totally isotropic around the Earth (well, to one part in 1000 if you count the dipole, one in 100,000 if you subtract it though doing that is itself an assumption that's purely due to our velocity w.r.t. the CMB) and then assume the Copernican principle. That tells you the universe is homogeneous and isotropic "on average". You can impose that directly, which puts the universe in de Sitter (no), Minkowski (no) or anti de Sitter (no), all of which violate observations dramatically. So instead you foliate spacetime with maximally-symmetric spatial slices. That gives you the RW metric. Then assuming GR forces you to put in perfect fluids and recover the Friedman equations.
All well and good, but that's ignoring, for example, all the anisotropies on the CMB. Obviously at last scattering the CMB was extremely isotropic and the universe very close to homogeneous and isotropic. That's no guarantee it's anything like that now; the only constraints come from very large inhomogeneities fucking up the large-scale CMB (which is, by the by, fucked up but to be fair recovering the low quadrupole from inhomogeneities isn't that straightforward). Remember that the large-scale CMB is created by two things: primordial perturbations, those which were outside of the horizon when the CMB was formed; and late-time effects. The ISW is the best known of them and it's a massive problem for LCDM (well, that's a lie, the theory is only about 1.5 sigma away from the observation, but it *looks* very serious and the low quadrupole does worry a lot of people). They're on large scales because they're effects very near to us, and when things are near to us they look big. Sorry to sound patronising if I do - totally unintended. It's just it's that simple. Obviously it has to be something on the sky that subtends quite a lot of it - the ISW counts because it's the effect of "dark energy" so it's pervasive, and a Gpc scale void counts because, well, it's a fucking gigaparsec big, it'll subtend pretty much everything... :)
Anyway. The standard approach (to setting up the background equatons - I'm not knocking how people find baryon fractions and so on in clusters) also ignores the structures in the universe today. We've got an enormous problem with averaging in relativity -- you can't do it. You can average scalars, of course, but the average of a tensor field breaks covariance because it's coordinate dependent. That immediately should raise suspicions about the Friedman equations. Let's say for example that we want to do an average on the metric. That averaged metric may or may not turn out to be FLRW. We actually have *no clue*. Quite possibly it will so let's for the sake of argument say that on large scales, the universe is actually described by FLRW metric(s). But
I wrote a long reply to this and lost it by being an idiot :( Anyway, I said firstly to dig out a copy of Joao Magueijo's book which may or may not be called "Faster than the Speed of Light" which postulates a changing speed of light and lets you do away with inflation. I think a lot of people would be happy if we could do away with inflation completely, but anything that replaces it has to repeat its successes, which are many.
As for the CMB, it doesn't violate relativity, it's *predicted* by relativity. All it is is a thermal bath of photons left over from the big bang. If you put in a big bang cosmology, that's what you get out. It doesn't pick a physically preferred frame any more than saying "This is MY car and that's a preferred frame!" It is a preferred frame - it's the one you're observing from, but it's not physically preferred. GR is all about being able to look at things from any frame of reference -- that's how it wipes the force of gravity; it notes that it's as fictional as centrifugal force (which it is) and changes reference frames to study it properly. The CMB isn't an absolute inertial reference frame, it just happens to exist. It's not imposed on or by relativity, but if you use a Robertson-Walker solution and put any photons in it at all it's forced on you.
Also it might interest you to know that in the standard cosmological model (phenomenological as it is) we don't even lock ourselves to the CMB. We actually lock ourselves to CDM most of the time, or lock ourselves to some frame that makes gravity look as Newtonian as possible.
As for simultaneity and so on... well, GR is wrong. Unfortunate, but true. How wrong and in what way we don't know, but yes, there seem to be issues.
You can get Schroedinger's equation in a totally deterministic (if rather contrived) manner by taking a totally classical system and adding a modification to the potential. If you then take the momentum potential (or action, however you want to phrase it) and the density you can bundle them together as basically the phase and the amplitude of a wavevector that... solves the Schroedinger equation. Totally deterministic and totally indistinguishable from standard QM. It's also very ugly and is riddled with conceptual difficulties, not least how you go about second-quantising to get something like a field theory, but even so, it's a totally deterministic formulation of standard QM.
Also, as basically a gravity theorist, it pains me to say it but GR is less reliable than the Standard Model. We can test it down to maybe 0.1mm and up to Pluto's orbit and beyond that, nah. It's because gravity's so damned weak. The other forces are much stronger and much better probed, and electromagnetism is the best known of all. QED is still one of our best theories. The problem with taking that too far is that then people have a bad habit of assuming gravitons must form the basis of a successful quantum theory of gravity and throw out geometry altogether -- knowing all the time of course (because they know this shit much better than I) that GR is non renormalisable and you're doomed to failure going at it in anything like a typical way. That way leads to string theories and supergravities and 30 years of intensive maths with absolutely zero solid predictions. A nice alternative is loop quantum gravity, which takes a much more conservative approach, takes geometry seriously, finds a way of quantising spacetime without gravitons and 7 extra dimensions, and has absolutely zero solid predictions for about fifteen years less intensive study. Other alternatives are brilliant, and all of them have absolutely zero predictions, to my knowledge.
As for the black hole question, it's not answerable within "big bang cosmology" because GR would have broken down before you get to the point that all the matter is within its own event horizon. Before GR is at all valid (even assuming we could use it that early) you can't extrapolate. So there's no way of saying that a black hole would inevitably be there because there might not *be* black holes in quantum gravity and on those scales probably aren't. (QG is expected to wash out singularities in some manner, and we have no idea what it's gonna do to event horizons on such small scales but if you've, say, granulised spacetime in some manner you've immediately got quantised areas and volumes as in loop quantum gravity, and that puts limits on horizons.) But it's all speculation... as is saying that there'd be a black hole back then because that extrapolates GR well beyond its regime of validity.
I wouldn't say "it didn't, we live inside still". I've heard people suggest that (I've even seen a Creationist cosmology that used external coordinates to chart the interior of a Schwarzschild hole, put us bang in the centre on top of a singularity and then used gravitational time dilation to prove that the universe is only 6000 years old. No shit, he did it. Of course, it relies on using the wrong time coordinate and the wrong coordinate patch in general, and then assuming that the singularity on the event horizon is real -- it's not -- and then that we can live in the centre of those coordinates, but it was still a stroke of genius.) But I wouldn't. The geometry is weird if you believe the extended Scwarzschild solutions and it's even weirder if you think we live in a charged or rotating hole. I don't believe any of it. I don't really have any firm beliefs one way or another, in all honesty, but definitely not that we live inside some big black hole.
I think I've made my opinions about dark matter and dark energy fairly clear somewhere in these comments, three times I think :) They are band-aids, and very ugly ones at that. MOND is an interesting one. No-one, including Milg
SUSY *is* a hack-job in my opinion. But what I meant is that in, say, MSSM, you've got well more than a hundred free parameters. Some of them are well-constrained; others aren't. Change a few of these and you can change the phenomenology. But as I've said a few times today, I'm not a particle physicist so I'm also happy to be told I'm wrong -- I talk with particle physicists sometimes, and the rest I get through the kind of academic osmosis that causes confusion worldwide... :)
1: Pure phenomenology. No-one constructing inflationary models that I know of actually seriously believes that it's fundamental physics (at least, not after the second or third year of their PhD). What they *do* believe, frequently, is that the phenomenology can help guide a more fundamental theory. Personally I don't always agree with that; I think it can shroud a fundamental theory (in a similar vein to how cosmology is built on phenomenology that basically shrouds a very serious and neglected underlying issue).
Unless you're using the Higgs itself to drive inflation -- Guth's first model did this but it ran into problems with a graceful exit; it's recently been reawakened and re-examinded, though -- you're going to have a massive problem identifying an inflaton. We've not observed *any scalar fields whatsoever*. Even the Higgs remains elusive, though that might change in the near future. (Don't hold your breath.) So you immediately have a problem that what you're doing is specious. You can then either ground your inflaton in a well-reasoned model of high-energy physics or, and this is the standard approach, just invent a scalar field, call it the inflaton, and give it an arbitrary potential. So long as you make the potential flat enough that scalar field is an inflaton.
Basically it's phenomenology. But the people who do it are convinced it gives *suggestions* about what lies underneath, and in some ways they've got a point. Inflation works extremely well and it's standard to assume there was an inflationary epoch. You solve the horizon problem, the flatness problem and (if you believe in various GUTs) the monopole problem. (Basically -- why does the CMB look identical in opposite directions when the universe is too young for them to have ever interacted; why is the universe so fucking SMOOTH; and why do we not see any of these magnetic monopoles that GUTs produce in abundance?) Even more importantly, though, the quantum fluctuations of a scalar field coupled to gravity in the early universe produce tiny seeds that are basically exactly right. You can make models that get them exactly wrong but actually you have to work a bit; basic inflation made a prediction of those seeds, and when WMAP came along and looked at the CMB in unprecedented detail, it was dramatically confirmed. Basically those ripples had to be almost exactly Gaussian random, and "scale-invariant" meaning that the extremely large wavelengths were massively more powerful than the shorter wavelengths. That maps through to the formation of the CMB, when electrons condensed into protons to form hydrogen and light rays could suddenly free-stream carrying with them a photograph of the early universe. And it maps through even further, to the large-scale structure of galaxy clusters where we can look at those very same wavelengths. Much of a shift from those early imprints and that distribution is changed actually quite dramatically.
2: Dark matter is a big issue (well, duh). Basically "dark matter" is a catch-all term for whatever is causing rotation curves to deviate from the Newtonian prediction. I get irritated when people immediately assume it's a new exotic species of particle. I've put a couple of rants on this thread aimed at this kind of thing. My feeling is that dark matter in galaxies (and galaxy clusters) is made up of five or six different effects, *all* of which act as "dark matter", ie to flatten rotation curves: exotic particles perhaps, if supersymmetry is true; massive neutrinos since we now know that they are massive even if we don't know the mass, and neutrinos are so abundant that with *any* mass they form at least a dark matter even if it can't be the full dark matter (attributing the entire dark matter to massive neutrinos badly washes out structure on galaxy cluster scales); relativistic corrections coming from our naive assumptions that galaxies inhabit Minkowski (ie normal flat) space, since they don't, and that may -- *may* -- be able to account for up to roughly a tenth or more of spiral galaxies' dark matter; i