Skip to the Main Content

Note:These pages make extensive use of the latest XHTML and CSS Standards. They ought to look great in any standards-compliant modern browser. Unfortunately, they will probably look horrible in older browsers, like Netscape 4.x and IE 4.x. Moreover, many posts use MathML, which is, currently only supported in Mozilla. My best suggestion (and you will thank me when surfing an ever-increasing number of sites on the web which have been crafted to use the new standards) is to upgrade to the latest version of your browser. If that's not possible, consider moving to the Standards-compliant and open-source Mozilla browser.

July 19, 2004

Talking to ’t Hooft about tossing TVs

Posted by Urs Schreiber

Today Prof. Gerard ’t Hooft gave a talk at University Duisburg-Essen on Black Holes in Elementary Particle Physics. Maybe due to the media hype about Hawking’s announcement of his new idea about the black hole information ‘paradox’, ’t Hooft decided to throw his TV set away, and not only his but lots of them, in fact enough that they would form a spherical shell collapsing to a black hole.

Using this picture to emphasize the process in which ‘known physics’, represented by well understood TV sets, passes the horizon and hence a border beyond which all kinds of apparent paradoxes lurk, he talked about some standard facts of high energy physics and then briefly mentioned some of his intriguing observations and speculations concerning physics of the stretched horizon, the collision of infalling particles with outgoing Hawking radiation as well as the possibility of a deterministic hidden variable model of quantum theory, which, as he says, he develops as a hobby.

After the talk we went to a nearby Biergarten and I had the chance to ask some more detailed questions.

I have to admit that I haven’t read any of ’t Hooft’s papers concerning the above mentioned issues, so I learned for the first time about his calculation which indicates that, somehow, the scattering of Hawking radiation at infalling matter (one form - even though not the only one relevant I’d think - of back reaction which is not usually taken into account in related discussions, but which certainly should be) has some surprising resemblance to string scattering amplitudes - well, except for the curious fact that the analogy requires a imaginary string tension.

Very interesting are also his ideas about the foundations of quantum mechanics, holography and string theory.

He says that he expects that there is a deterministic and local (yes, local) hidden variable theory behind it all, which would be apparent if only we knew the correct degrees of freedom of nature. Since we don’t, we only see a statistical average of this deterministic process, and this translates in a non-local way to the quantum mechanical wavefunction, roughly.

To me this philosophy sounded a lot like approaches by Lee Smolin to get quantum mechanical dynamics from the classical statistics of ensembles of large matrices that encode the deterministic interrelation of all particles (well, probably, if at all, of all D0 branes) in the universe. But when I asked Prof. ’t Hooft about this he said he wasn’t fully familiar with Smolin’s approach.

Anyway, ’t Hooft’s idea now is that the full deterministic theory has no information loss, but that on the ‘coarse grained’ level of familiar quantum theory information is lost all the time in virtual black holes that are abundant in vacuum fluctuations. The point is that, he says, this way information about degrees of freedom in the bulk diasappears. The only information left is that at some holographic boundary! This way, I think, he tries to give a ‘dynamical’ explanation of holography.

I asked if and how he sees string theory fit into this picture, and he said that he thinks that since in string theory essentially only the S-matrix is a well defined observable, and since this means that only on-shell information at the ‘boundary’ is available while local physics in the bulk is fundamentally out of reach of present day string theory, this fits in perfectly with the above picture, where ordinary quantum mechanics is kind of an ‘effective theory’ on the boundary while the true bulk theory is a deterministic hidden-variable thingy.

I have to say that when first confronted with speculations like this some alarm bells go off - but then I realize that when ’t Hooft discovered holography a while back this idea must have sounded - before Maldacena came along and gave an explicit realization - just as weird, and now it is widely accepted and even standard lore.

So maybe in this little chat over a glass of beer I was actually shown a glimpse of the big physics picture of the future, without my poor mind being able to fully grasp it.

On the other hand, when asked what he thinks about how his ideas about string/gauge duality and holography have come to life in string theory, he answered, humbly and jokingly, that he almost fails to recognize his original ideas.

There was much more discussion, but that’s all I am going to report here. It was a big pleasure to talk to such an outstanding person as ’t Hooft is, and I have some things to think about now. First of all, I’ll toss away my TV set…

Posted at July 19, 2004 9:31 PM UTC

TrackBack URL for this Entry:

9 Comments & 1 Trackback

Re: Talking to ’t Hooft about tossing TVs

Hello Urs,

thanks for this interesting report.

For German reading people perhaps my popular review article about information loss in black holes might be interesting:

Vaas, R. (2002): Finales Fiasko. bild der wissenschaft Nr. 9, pp. 60-65.
(unfortunately not online)

It contains an interview with Gerard ‘t Hooft.

Regarding Lee Smolin’s work: Lee and Fotini are developing an interesting idea for a hidden variable quantum theory from quantum gravity at the moment, see

Sketchy, but worth looking.

Best wishes,

Posted by: Rüdiger Vaas on July 21, 2004 8:16 PM | Permalink | Reply to this

Re: Talking to ’t Hooft about tossing TVs

Hi Rüdiger -

nice to hear from you!

Yes, I am aware of Smolin’s Quantum Theory from Quantum Gravity. As far as I can see here he tries to apply idea to spin networks which he previously applied to Matrix Theory and motivated even earlier in the 80’s, I think, from very elementary assumptions.

All this is highly speculative, but I was always interested in following it. In particular I found interesting the idea of getting QM from statistics of large matrices in light of the fact that the IKKT Matrix Model has at least a superficial resemblance to the statistical version of the BFSS model.

But everybody not equipped with a Nobel Price or similarly renowned should probably better not talk and think about such things too much. ;-)

That’s why I would be interested in learning about Gerard ’t Hooft’s opinion on Smolin’s ideas. Did he say anything related in that interview which you mentioned? (I don’t have access to it right now.)

Posted by: Urs Schreiber on July 22, 2004 12:07 PM | Permalink | PGP Sig | Reply to this

Re: Talking to ’t Hooft about tossing TVs

Hi Urs,

I did not ask Gerard ‘t Hooft about Lee Smolin’s work, and I didn’t know that at that time (early 2002). Sorry.

By the way, another self-advertisement, if you allow: In the latest issue of bild der wissenschaft (8/2004) is a cover story about the measurement problem, covering three “objective” approaches (Bohmian mechanics, many histories, spontaneous localization). Quantum gravity is only briefly mentioned.

See the advertisement:
Einstein und die Quantenwelt

Best regards,

P.S. You are doing a great job here!

Posted by: Rüdiger Vaas on July 22, 2004 10:37 PM | Permalink | Reply to this

Re: Talking to ’t Hooft about tossing TVs

Hi Rüdiger -

By the way, another self-advertisement, if you allow

Yes, please. Nothing wrong with self-advertisements in science, I think.

Bohmian mechanics, many histories, spontaneous localization

In case anyone is interested in hearing my opinion on these matters:

There are two aspects to Bohm’s insight.

One is the undisputable observation that the complex Schroedinger equation (even the relativistic wave equation) is equivalent to a real current conservation together with a real equation which is the classical Hamilton-Jacobi equation corrected by a term containing Planck’s constant - the ‘quantum potential’.

This is just a mathematical fact and has nothing to do with interpretations of QM or whatnot. All this only comes about when people try to figure out what this fact might be telling them.

Yes, naively it looks like we have a current ‘guided’ by a ‘pilot wave’. On the other hand, when you remove the ‘quantum potential’ term you are just left with ordinary classical mechanics, and nobody would say that the Hamilton-Jacobi equation is a ‘pilot wave’ for classical trajectories.

It might be that this equivalent reformultation of the Schrödinger equation is just a curiosity without further meaning. In any case, writing this equation in Bohm’s way and using the words ‘pilot wave’ instead of ‘wave function’ alone does not change anything about quantum mechanics.

I know only one more sophisticated interpretation of the equaivalence of the Schrödinger equation to a certain pair of real equations, and that’s Edward Nelson’s result that these two equations are those governing stochastic processes that satisfy a certain constraint. (I have once written an overview over this ‘stochastical interpretation’ which can be found here.)

This I once found extremely intriguing. However, Nelson does not give any examples how stochastic processes with the required property might arise naturally. It is the fascinating claim by Smolin in a predecessor paper of hep-th/0201031 that the dynamics of eigenvalues of certain ensemles of large matrices is indeed described by a stochastic process of the required sort.

I should say that I haven’t checked this claim in detail, but assuming that it is true it is just a small step to think of places in theoretical physics where large matrices show up. Speculations concerning relations of this result to Matrix Models of string theory are therefore at least not too surprising, even though they may of course turn out to be fundamentally misguided, I don’t know.

In any case, that Prof. t’ Hooft thinks that the nature of the S-matrix and other aspects of string theory might be intimately related to a stochastic model of QM makes one wonder.

But I for one won’t spend my time thinking about this any more until I am retired! :-)

(I should mention, though, that there is another intriguing apparent relation between Bohm’s ‘quantum potential’ and terms that appear in supersymmetric quantum mechanics. I discuss that on pp. 295 and pp. 64 of my ‘master’ thesis.)

Next you mentioned many histories. This, too, is just an equivalent reformulation of the evaluation of expectation values in quantum mechanics. It again suggests certain realistic models of QM, but the formalism and the results obtained using it are independent of that.

This is in sharp contrast to spontaneous localization, if by that you mean models where we modify the Schrödinger equation so that ‘collaps’ is built in by hand.

You are doing a great job here!

Thanks. It is the more fun the more people participate. So thanks for all your messages!

BTW, there are some simple tricks to built in clickable hyperlinks into comments (which could increase the number of hits you get). Have a look at this SCT HowTo for more information.

Posted by: Urs Schreiber on July 23, 2004 10:56 AM | Permalink | PGP Sig | Reply to this

Re: Talking to ’t Hooft about tossing TVs

There was a paper that showed up this week

“Jensen-Shannon divergence, Fisher information, and Wootters’ hypothesis”


which attempts to assert how statistical fluctuations of measurement are responsible for the Hilbert space structure of quantum mechanics.

I don’t know what to think of this result, or whether it actually means anything. If somebody can show in a rigorous manner how statistical fluctuations due to measurements can actually “induce” a Hilbert space structure on a particular underlying theory (whether it’s physics, or anything else like economics or politics, etc …), then maybe it would be interesting. At this point I would be somewhat skeptical of this assertion. Though if this assertion is true, then perhaps “quantum theory” isn’t much more than a convolution of this “measurement theory induced Hilbert space geometry” with the underlying deterministic physical system?

(I wonder what the mathematical statistics folks would think about this sort of assertion).

Posted by: JC on July 24, 2004 12:39 AM | Permalink | Reply to this

Re: Talking to ’t Hooft about tossing TVs

My book “Quantum Theory as an Emergent Phenomenon” should
be available at the end of this month. The first chapter is a
general non-technical introduction to the ideas developed in
the remainder of the book. In contrast to ‘t Hooft’s approach,
the underlying hidden variables are even more non-local than
is standard quantum mechanics; the local canonical commutation
relations of quantum field theory emerge only after statistical
averaging. In a nutshell, I start from a non-commutative
extension of classical mechanics, in which non-commutativity of
factors is dealt with by using cyclic permutation under a trace.
I argue that the statistical thermodynamics of this trace dynamics
obeys the rules of quantum mechanics (or more generally, of
quantum field theory), and that the Brownian motion corrections
to this thermodynamics lead to the probability interpretation via
the Born rule, and to the nonunitary, nonlinear process of state
vector reduction.

Posted by: Steve Adler on September 28, 2004 7:22 PM | Permalink | Reply to this

Re: Talking to ’t Hooft about tossing TVs

I haven’t looked at your papers yet, but thanks for letting us know about your book.

Do you see any relation of your approach to the approaches by Nelson, Smolin, and maybe ’t Hooft (does it appear possible that your approach comes from or gives rise to these in a certain limit, say)?

Do you see any connection of your approach to known ‘ordinary’ concepts (like ’t Hooft relates his proposal to holography and Smolin his construction to Matrix Theory and recently to spin networks)?

Posted by: Urs Schreiber on September 28, 2004 7:33 PM | Permalink | PGP Sig | Reply to this

Re: Talking to ’t Hooft about tossing TVs

Readers of this thread will be interested in the recent work of Stephen Adler (of IAS/Princeton) which is about to appear in book form:


Quantum Theory as an Emergent Phenomenon (August 2004):

“Quantum mechanics is our most successful physical theory. However, it raises conceptual issues that have perplexed physicists and philosophers of science for decades. This book develops a new approach, based on the proposal that quantum theory is not a complete, final theory, but is in fact an emergent phenomenon arising from a deeper level of dynamics. The dynamics at this deeper level are taken to be an extension of classical dynamics to non-commuting matrix variables, with cyclic permutation inside a trace used as the basic calculational tool. With plausible assumptions, quantum theory is shown to emerge as the statistical thermodynamics of this underlying theory, with the canonical commutation/anticommutation relations derived from a generalized equipartition theorem. Brownian motion corrections to this thermodynamics are argued to lead to state vector reduction and to the probabilistic interpretation of quantum theory, making contact with recent phenomenological proposals for stochastic modifications to Schrödinger dynamics.”

[Urs: I would be most interested in hearing more from ‘t Hooft on Adler’s work. I would be very surprised if he hasn’t read the preprint.]

Posted by: Chris W. on August 28, 2004 7:07 PM | Permalink | Reply to this

Re: Talking to ’t Hooft about tossing TVs

Hi Chris -

thanks for mentioning this text. I haven’t read it and I probably won’t find the time to indulge in this business until when I’m sixty-four (or something ;-), but I am interested in being informed about the state of the art.

Urs: I would be most interested in hearing more from ’t Hooft on Adler’s work. I would be very surprised if he hasn’t read the preprint.

I don’t know if he has. Maybe not. When I talked to him he was not aware of Smolin’s ideas on how certain eigenvalue statistics follow Nelson’s stochastic process which is equivalent to the Schrödinger equation. Since Adler mentions Smolin’s approach as being ‘similar’ to his, I am guessing that ’t Hooft isn’t familiar with Adler’s ideas either.

(And maybe he is right about that, since crackpottery is just a step away, so it’s always hard to tell. ;-)

I would be interested in learning about hard results by Adler. Does he provide some ensemble of matrices which is described by Nelson’s stochastic process? I don’t see him citing Nelson, so apparently not.

If you want to summarize some ideas and results by Adler here I would very much appreciate it. (See here for details on how to include LaTeX math in your SCT comment.)

Posted by: Urs Schreiber on August 28, 2004 7:28 PM | Permalink | PGP Sig | Reply to this
Read the post Making AdS/CFT Precise
Weblog: The n-Category Café
Excerpt: The last session of Recent Developments in QFT in Leipzig was general discussion, which happened to be quite interesting for various reasons....
Tracked: July 22, 2007 7:34 PM

Post a New Comment