Tuesday, April 08, 2014

Alan Guth's 1979 handwritten notes

First, let me mention that the CERN accelerator complex is in the middle of the Two Years' Vacation which is the right moment to begin its reawakening. See LHC begins long road to restart at the Symmetry Magazine.

They first restart the source, then the smaller rings and boosters etc. – the chronology of the restart pretty much mimics the 6-minute video above that describes the CERN rings etc. culminating in the LHC itself. In one year, e.g. in April 2015, the collisions at 13 TeV should begin. Because of the quantitative leap in energy, all "null results" may be instantly forgotten and there will be a completely new chance – but not certainty – to detect previously undetected physical phenomena.

But I want to mention another article at the Symmetry Magazine that was posted in 2005.

The article is called Inflation and it contains the photographs of some of the first pages at which Alan Guth made the realization that allowed cosmic inflation to be born.

Now, Alan Guth won the "messiest office" contest but that didn't prevent him from storing the first inflationary notebook pages from December 1979!

Click to zoom in.

If you can't read it (the notebook is a part of the permanent exhibition at the Adler museum in Chicago), it says:
EV 5
Dec 7, 1979

Spectacular realization

This kind of supercooling can explain why the universe today is so incredibly flat – and therefore resolve the fine-tuning paradox pointed out by Bob Dicke in his Einstein day lectures.

Let me first rederive the Dicke paradox. He relies on the empirical fact that the deacceleration parameter today \(q_0\) is of order 1.
Just to be sure, the Dicke paradox is the puzzling realization that \(\Omega=1\), the Universe seems to be flat without no good reason in the Big Bang cosmology. As far as I understand, the "supercooling" mechanism was just his original terminology for the old inflation – with a false vacuum state that ultimately tunnels into the true vacuum. Only weeks after these notes were written, Guth would hear about the horizon problem. But he would already be giving talks about "inflation" at the end of January 1980 (01/23/1980, three days after the first inauguration of Ronald Reagan).

Note that the inflationary expansion is a fancy example of "supercooling", i.e. so fast cooling without sufficiently good condensation nuclei that the liquid doesn't become solid during the cooling process. The temperature during inflation decreases 100,000 times (no, it is not the same googol-like huge numbers that describe the linear expansion of the space, but it is still large), from \(10^{27}\) to \(10^{22}\) kelvins. At the end of inflation, the temperature increases again – it's the "reheating" when the seeds of matter that later turns to galaxies and us is born out of the kinetic energy of the inflaton field.

A continuation of the page above. Click to zoom in.

"So, after a few of the most productive hours I had ever spent at my desk, I had learned something remarkable. Would the supercooled phase transition affect the expansion rate of the universe? By 1:00 a.m. I knew the answer: Yes, more than I could have ever imagined."

Alan Guth, The Inflationary Universe, Cambridge MA, 1998


  1. kashyap vasavadaApr 8, 2014, 4:41:00 PM

    Hi Lubos: It is nice to see Guth's notes.It shows how a genius works. But I have a question about BICEP2. I have heard that in BICEP2, you have to eliminate foreground rather than background as in the usual experiments. I sort of vaguely understand it. But can you elaborate on this?

  2. kashyap vasavadaApr 8, 2014, 5:54:00 PM

    I forgot to add that Guth has remarkably clear handwriting considering that he is a great scientist and got an award for messiest office in Boston!!

  3. You are using the words in two different ways, kashyap. In BICEP2 “foreground” simply means closer whereas “background” in any measurement refers to unwanted interference; it has nothing to do with proximity.

  4. kashyap vasavadaApr 8, 2014, 7:00:00 PM

    Thanks. But proximity(closer) to what?

    Can you elaborate?

  5. This is off topic--but an interesting update to a previous LM post on the death of the Palestinian ambassador in Praha:


    The letter killeth!

  6. I believe that here the foreground refers to galactic and extragalactic sources (Galactic synchrotron and polarized dust emission). This is the foreground with respect to the surface of last scattering,

  7. Hi Lubos! Though a layman I always enjoy reading articles like this on your blog about inflation and other topics like holography or ER=EPR. I found this paper on the latter at http://arxiv.org/abs/1404.0794 and thought you might like to take a look at it (and maybe comment ;)).

  8. Thank you so much for sharing the argument. Some people follow the dream while others don't. I do agree with your idea because i am currently following the money because it's hard to follow the dream without the money. I also quoted some parts of your ideas into my post here too: http://setprofitonline.com/follow-dream-or-money/

    Thank you,

  9. I saw the paper, it's pretty fascinating although I don't quite understand the relationship between these two very different quantities (entanglement entropy and scattering amplitude) yet.

  10. " Closer because they are low redshift compared to the surface of last scattering."

    Not that I disagree or even think Kashyap would have trouble understanding, but his has to be the most non-layman friendly explanation of "closer to us" I've ever seen :)

  11. Yes sorry about that :) I meant that these sources of foreground contamination are at redshift between 0 and 2 (closer to us), whereas the CMB photons that we observe today were emitted at a redshift Z = 1100 roughly (far away).

  12. kashyap vasavadaApr 9, 2014, 2:14:00 PM

    @scooby: Thanks. I did understand low vs high redshift sources. But it was not clear from Gene Day's answer if he meant these. From what you are saying it looks like there are lots of nearby objects which produce magnetic fields obscuring the B-mode effect which BICEP2 is looking for.It will be interesting if Planck results, which are being delayed for some reason, would have the same difficulty with analysis.Of course one can argue that background is anything which obscures what you are looking for!!

  13. kashyap vasavadaApr 9, 2014, 4:42:00 PM

    @scooby: Thanks. My first reply did not get through for some reason. Hopefully this is not duplicate. I understand the point about low z and
    high z sources. As you say, there are probably lots of galactic and extragalactic sources of magnetic fields which may obscure what they are looking for. Low z sources (nearby as compared to last scattering surface of
    CMB) are probably called foreground. Although, in principle, anything which you
    are not looking for can be called background!! Let us see if Planck group has
    similar difficulties. For some reason, their results are delayed.

  14. Alan Guth may have a lot of enthropy in his office, but not in his writings which seems more important ... ;-)

    His writings looks so nice and clear, as I have observed it of many physicists and mathematicians who exactly know what they are doing ...


  15. Gre exam is very important so students must keep in mind
    that only hardworking can help for passing it

    gre score

  16. GRE exam is a requirement for admission in such programs i.e. Masters, PhD etc. But no where in the world..., they use it to grade your PhD or your working during PhD.

    gre problem

  17. Hi Lubos,

    What evidence you have for String Theory?

  18. As it turns out...he was right about the primordial gravitational waves.

  19. Will you also kindly tell us what led you to this conclusion?

  20. I have heard rumors, but seen no numbers. The paper was accepted and published in physical review letters 112, 241101 (2014) and they stand by their analysis and numbers.

  21. Did you hear some new rumors, like concerning Planck which will probably speak within a month?

  22. see abstract http://journals.aps.org/prl/abstract/10.1103/PhysRevLett.113.031301. I have not read it but had heard about the "tension" .which for me is non existent. The numbers are compatible in my books. imo planck has to come up with a measurement and not a limit.

  23. LOL, is there any difference between the measurement r=0.03 +- 0.03 and the limit "r is less than 0.06", for example? ;-)

  24. not in this case. But an error allows to see how many standard deviations from zero the r measurement is (3 to 4 for BICEP2) whereas a 95% confidence limit says that it is 2 sigma from zero in a mouth filling way. So until the Planck measurement gets more precise I will go with BICEP2 . :)

  25. I think that this study is quite wrong . It's more likely that people with the highest IQ in the world are the string theorists from czech republic whereas people with lowest IQ levels are those loop quantum gravity specialists (the likes of lee smolin) and people working in Earth and climate sciences .

  26. It's actually pretty much implied by the results, too.

  27. "There is no schism in quantum mechanics. "

    Luboš, you are much more knowledgeable in this field than I am, so I believe you. However, there is a question you may be able to shed some light on. It is bugging me endlessly.

    If different mass distributions are involved in a quantum superposition, you have to add up several wave functions defined over non isometric domains. In order to do so, you need a mapping between their respective domains. How is it constructed?

  28. Dear Peter, it is simply not true that you need a "mapping between their respective domains" (I guess you mean the background spacetime geometries by the "domains"?).

    The superposition principle of quantum mechanics guarantees that for each states psi1, psi2, any complex linear superposition c1*psi1+c2*psi2 is an equally allowed state of the physical system.

    This includes superpositions of states that differ by spacetime geometries. The geometries may differ significantly so that no natural or canonical map between the geometries associated with each exists. The states psi1,psi2 in the superposition may even have a different topology in the bulk of the spacetime.

    Such superpositions may sometimes be "disfavored" if there are superselection sectors. Linear superpositions of wave functions from different superselection sectors are "less allowed".

    But superselection sectors must differ either by values of conserved quantities, or some behavior of the spacetime and other fields in the asymptotic region at infinity. If you only make a local regional modification of the state to get from psi1 to psi2, they are clearly in the same superselection sector, so the superposition is as allowed as psi1,psi itself even if the geometries are completely different and "incomparable" or "unmappable" to each other!