17 views

Skip to first unread message

Mar 24, 2003, 2:23:42 AM3/24/03

to sci-physic...@moderators.isc.org

In http://arxiv.org/abs/hep-th/0302219, Leonard Susskind declares that

the

recent attempt by Kachru et al,

http://arxiv.org/abs/hep-th/0301240,

to get hold of a positive cosmological constant in string theory,

involves

extremely unnatural, artificial models ["Rube Goldberg contraptions".]

However,

he doesn't mean this as criticism: he solemnly goes on to state that

this is

ok, because it is just what one would expect in an anthropic context;

and

he believes that string theory is anthropic. Of course he's right:

anthropicists

*should* expect successful physical theories to look really ugly and

contrived. And this is the most powerful argument against anthropicism

that I have ever seen. [I just hope that he is wrong about string

theory being anthropic!]

I can now justify my misanthropicism. Yay!

Mar 28, 2003, 2:54:18 AM3/28/03

to sci-physic...@moderators.isc.org

"Serenus Zeitblom" <serenusze...@yahoo.com> wrote in message

news:c7fd6c7a.03032...@posting.google.com...

I disagree.

I would expect that our universe to be a particular example selected from an

ensemble.

The theory describing the ensemble need not be ugly, and probably isn't.

However, curve fitting almost always is.

Dirk

Mar 29, 2003, 11:57:04 AM3/29/03

to

serenusze...@yahoo.com (Serenus Zeitblom) wrote in message news:<c7fd6c7a.03032...@posting.google.com>...

Lenny Susskind supports both string theory and the anthropic

principle, and says that string theory is anthropic, and anthropic

theories may be ugly but this is not a criticism. I don't why you

would interprete that as a reason to justify your misanthropicism. Let

me explain. He says the string theorists are looking for an extremely

beautiful simple underlying theory, and really there is no such

theory. You have the five superstring theories, and these along with

11d supergravity are all limiting points on the moduli space of an

underlying M-theory. String theories are hoping M-theory is the simple

beautiful underlying theory. Well, it may be underlying but it's

neither simple nor beautiful, at least in terms of how it reduces to

the real world. Susskind says that there are a googleplex or an

infinite number of other points in the M-theory moduli space, and our

universe must correspond to one of these other points, and not any of

the five known superstring theories, because they don't allow for

deSitter space. The likelihood of finding the point in the M-theory

moduli space that corresponds to the real world is infinitesimal, so

it's hopeless to use M-theory to actually calculate the details we

observe about the real world. In that sense, you can call it ugly.

However, there's nothing wrong with that. Try combining the Susskind

Landscape with inflationary cosmology. Each bubble in the inflating

universe would correspond to a different point in the moduli space of

the underlying M-theory, and give rise to totally different laws of

physics. Most would not be compatible with life. However, using the

anthropic principle, we would be in a bubble that would be condusive

to life, and so here we are. So, this seems to suggest that

superstring theory or M-theory is probably true, and the anthropic

principle is probably true, despite the "ugliness" of theory in the

sense that it encompasses a googleplex or an infinite number of

different point points in the moduli space, which are like different

theories, so we can't find which one is ours, so we can't calcuate the

laws of physics of our universe.

Jeffery Winkler

Mar 29, 2003, 8:14:47 PM3/29/03

to sci-physic...@moderators.isc.org

Serenus Zeitblom wrote:

> In http://arxiv.org/abs/hep-th/0302219, Leonard Susskind ...

> Of course he's right:

> anthropicists *should* expect successful physical theories to look really

> ugly and contrived.

Why ?

> And this is the most powerful argument against anthropicism

> that I have ever seen. [I just hope that he is wrong about string

> theory being anthropic!]

> I can now justify my misanthropicism. Yay!

I am really not clear about the status of the anthropic principle, my belief

is that there is an unspoken consensus about it that turns it into a form of

flag or boundary stone treated with scorn. It is difficult to debate because

it is unclear what positive value physicists admit for it; they usually allude

to it with negative undertones more than they tell what they think it is or

should be and why. The above is an example.

From my memory of a few sentences by Weinberg on AP - who clearly doesn't

like it either - my vigorous perception is that AP is grudgingly allowed a

position that's easier to explain as a consequence of the sociology and

genealogy of natural sciences than as the result of a properly reasoned

assessment. The metaphor of the boundary stone is I guess quite apt : at one

level, it is viewed as a boundary stone between natural science and religion

or faith, and the distinctively scornful conventional attitude appears at that

level to be something of a "patriotic obligation" of natural scientists not to

allow ground to what they view as the pathological anthropocentrism central to

revealed religion.

At another level, it is meant as a boundary stone internal to natural

sciences, a triple point between chemistry, physics and biology. That's the

position Weinberg ascribes to it when he quite directly reduces it to a

(pretty useless in terms of predictive power) apology of carbon chemistry.

This equation of AP with an apology of carbon chemistry has imo very little

rational justification and quite purely reflects the intention to reduce the

family of intuitions it represents, to practical insignificance - while

clothing the gesture as a tribute to darwinian reasoning in a minimal alliance

with biologists (studying the "ugly and contrived" forms living beings take)

against the common enemy of creationism.

To get at my point by a shortcut, the problem with equating AP to an apology

of carbon chemistry is imo made clear when one notes that carbon chemistry is

the widest imaginable encircling boundary one can draw around the known

observers of nature relevant to the anthropic argument, while the name of

"anthropic" itself alludes to a much closer boundary that's nevertheless much

wider than the one implied by the argument itself - if one means to take

it seriously. The question then becomes : having the more restricted anthropic

populations in view, can't we think of "conjugate" features just as reasonable

than a carbon-based biochemistry, and with distinct predictive power, to

abstract from them as candidate conditions of possibility of scientific

observers of a physical Universe ?

Otoh, one can't help noticing a correlation between this domination of

"anthropism-as-carbonism" in physics and the sadly observable fact that

physics-as-a-cultural-asset serves more often than (I presume) most physicist

would wish, to reduce human beings to (ugly) inanimate carbon.

Regards, Boris Borcic

--

"Archimedes did weapons research - while his life was irrefutably at risk"

Mar 29, 2003, 8:18:14 PM3/29/03

to sci-physic...@moderators.isc.org

"Dirk Bruere at Neopax" <di...@neopax.com> wrote in message news:<b5uko1$2di47v$2...@ID-120108.news.dfncis.de>...

> "Serenus Zeitblom" <serenusze...@yahoo.com> wrote in message

> news:c7fd6c7a.03032...@posting.google.com...

> >However,

> > he doesn't mean this as criticism: he solemnly goes on to state that

> > this is

> > ok, because it is just what one would expect in an anthropic context;

> > and

> > he believes that string theory is anthropic. Of course he's right:

> > anthropicists

> > *should* expect successful physical theories to look really ugly and

> > contrived. And this is the most powerful argument against anthropicism

> > that I have ever seen.

>

> > he doesn't mean this as criticism: he solemnly goes on to state that

> > this is

> > ok, because it is just what one would expect in an anthropic context;

> > and

> > he believes that string theory is anthropic. Of course he's right:

> > anthropicists

> > *should* expect successful physical theories to look really ugly and

> > contrived. And this is the most powerful argument against anthropicism

> > that I have ever seen.

>

> I disagree.

> I would expect that our universe to be a particular example selected from an

> ensemble.

> The theory describing the ensemble need not be ugly, and probably isn't.

> However, curve fitting almost always is.

> I would expect that our universe to be a particular example selected from an

> ensemble.

> The theory describing the ensemble need not be ugly, and probably isn't.

> However, curve fitting almost always is.

But think of what Susskind is opening the door to here. Basically, now anyone

can put forward a ridiculously artificial model "motivated" by string theory,

and claim that this is ok because of anthropic selection.

PROFESSOR NAIBA: TIME HAS INERTIA!

PROFESSOR ZEAB: What a ridiculous, Rube Goldberg contraption!

PROFESSOR NAIBA: Yes, I agree. Isn't it wonderful?

Apr 1, 2003, 2:24:16 AM4/1/03

to

Jeffery wrote:

>So, this seems to suggest that

>superstring theory or M-theory is probably true, and the anthropic

>principle is probably true, despite the "ugliness" of theory in the

>sense that it encompasses a googleplex or an infinite number of

>different point points in the moduli space, which are like different

>theories, so we can't find which one is ours, so we can't calcuate the

>laws of physics of our universe.

The reason no one can calculate anything in string/M theory has

nothing to with not knowing which vacuum state history has landed us

in amidst a googleplex of possibilities. No one understands what the

underlying M-theory is, and that is why no one can calculate anything.

If one had a well-defined theory and what it told us about the early

universe was that we could only get probabilistic information about

the late-time vacuum state, then talking about anthropic principles

might make sense. Right now this is just one more excuse for not

dealing with the fact that there is no theory.

Peter

Apr 1, 2003, 4:14:38 PM4/1/03

to

[I had posted the following on 03/24 but haven't seen it appear yet.]

Serenus Zeitblom schrieb:

>

> In http://arxiv.org/abs/hep-th/0302219, Leonard Susskind declares that

> the recent attempt by Kachru et al,

> http://arxiv.org/abs/hep-th/0301240,

> to get hold of a positive cosmological constant in string theory,

> involves extremely unnatural, artificial models ["Rube Goldberg contraptions".]

[...]

> [I just hope that he is wrong about string theory being anthropic!]

You have probably seen the long paper hep-th/0303194 by Michael R.

Douglas today on the server, which contains attempts to understand the

statistics of string vacua, so to say, motivated by the idea that

instead of studying individual vacua it might be more helpful to have an

idea about which volume in parameter space a class of vacua occupies.

On p. 63 it says:

"Not having a reliable estimate, we would still conjecture that the

qualitative structure of the Standard Model is the result of discrete

choices which are not hard to realize, and that the fraction of models

which meet the qualitative tests (ignoring values of couplings) is

closer to O(10^-10) than to O(10^-100)."

The last sentence is:

"We even argued that depending on what comes out [of the study of string

vacuum distribution], we might find that string/M theory has much less

predictive power than we thought, perhaps none. At present it is

reasonabe to think that string/M theory will have predictive power, but

we should admit that we do not really know, and try to find out."

Apr 1, 2003, 4:17:46 PM4/1/03

to

Boris Borcic <bor...@users.ch> wrote:

> I am really not clear about the status of the anthropic principle,

> I am really not clear about the status of the anthropic principle,

It's a theory of initial/boundary conditions which says that the

initial condition on that part of a suitable Cauchy surface that

intersects the Earth at January 1, 2001 is that there is carbon

based sentient life on that part of the Cauchy surface that

intersects the Earth; and that all theories are constrained

by this boundary condition.

It's perfectly valid to impose boundary conditions on a physical

system: even when the system is the entire universe.

Apr 1, 2003, 8:22:30 PM4/1/03

to sci-physic...@moderators.isc.org

serenusze...@yahoo.com (Serenus Zeitblom) wrote in message news:<c7fd6c7a.03032...@posting.google.com>...

> Of course he's right: anthropicists

> *should* expect successful physical theories to look really ugly and

> contrived. And this is the most powerful argument against anthropicism

> that I have ever seen. [I just hope that he is wrong about string

> theory being anthropic!]

> I can now justify my misanthropicism. Yay!

> *should* expect successful physical theories to look really ugly and

> contrived. And this is the most powerful argument against anthropicism

> that I have ever seen. [I just hope that he is wrong about string

> theory being anthropic!]

> I can now justify my misanthropicism. Yay!

Why string theory should not be anthropic - I had the impression that the

antropic principle was invented precisely because string theory failed to explain

certain experiments, notably a positive cosmological constant and the emergence

of the standard model.

However, not everyone seems to think that making predictions up to a googleplex

of alternatives is the best we can hope for. Recently (in the abstract to

http://www.arxiv.org/abs/hep-th/0212247), Witten wrote about string theory that

"It is also more predictive than conventional quantum field theory".

Apr 1, 2003, 10:05:43 PM4/1/03

to

In article <4b8cc0a6.03040...@posting.google.com>,

thomas....@hdd.se (Thomas Larsson) wrote:

thomas....@hdd.se (Thomas Larsson) wrote:

> serenusze...@yahoo.com (Serenus Zeitblom) wrote in message

> news:<c7fd6c7a.03032...@posting.google.com>...

> > Of course he's right: anthropicists

> > *should* expect successful physical theories to look really ugly and

> > contrived. And this is the most powerful argument against anthropicism

> > that I have ever seen. [I just hope that he is wrong about string

> > theory being anthropic!]

> > I can now justify my misanthropicism. Yay!

> Why string theory should not be anthropic - I had the impression that

> the anthropic principle was invented precisely because string theory

> failed to explain certain experiments, notably a positive

> cosmological constant and the emergence of the standard model.

Ack, no. (Yeesh.) The anthropic principle goes way back. I don't have a

cite, but it's a fairly obvious idea. It has probably been

re"discovered" a number of times.

Aaron

--

Aaron Bergman

<http://www.princeton.edu/~abergman/>

<http://aleph.blogspot.com>

[Moderator's note: The classic reference, which also contains a history

of earlier thinking about the anthropic principle, is

John D. Barrow and Frank J. Tipler, The anthropic cosmological principle

Oxford University Press, Oxford, 1986.

- jb]

Apr 4, 2003, 2:50:58 AM4/4/03

to sci-physic...@moderators.isc.org

thomas....@hdd.se (Thomas Larsson) wrote in message news:<4b8cc0a6.03040...@posting.google.com>...

> serenusze...@yahoo.com (Serenus Zeitblom) wrote in message news:<c7fd6c7a.03032...@posting.google.com>...

> > Of course he's right: anthropicists

> > *should* expect successful physical theories to look really ugly and

> > contrived. And this is the most powerful argument against anthropicism

> > that I have ever seen. [I just hope that he is wrong about string

> > theory being anthropic!]

> > I can now justify my misanthropicism. Yay!

>

> Why string theory should not be anthropic - I had the impression that the

> antropic principle was invented precisely because string theory failed to explain

> certain experiments, notably a positive cosmological constant and the emergence

> of the standard model.

I guess my main point is just this. Theories rarely get destroyed

overnight. The

evidence against them gradually piles up, and their adherents are

gradually forced to make more and more unnatural assumptions to get

out of trouble until finally everybody gives up because the theory

looks so artificial and ugly. Think of what happened to the Steady

State Cosmology for example---those guys were never *proved* wrong,

but the initially simple and beautiful theory had to be patched up so

many times that it ended up looking like a Rube Goldberg contraption.

That's how you know a theory is wrong---when too many excuses have to

be made. Now El Susskind comes along and says that no, the extreme

artificiality of the string models with positive cosmological constant

is ok, it is just what one would expect anthropically. My objection to

this is that you can get away with *anything* this way. The anthropic

principle is like a licence to kill. I'm *not* saying that string/M

theory is wrong. I'm just saying that things are looking bad right

now, but the Anthropic Principle is *not* the way to salvation.

And no, the AP was NOT invented to save string theory! Thank God.

Speaking of whom, Ed Witten has often come out in the past against

anthropic ideas in string theory, and I have not heard that he has

changed his mind.

Apr 3, 2003, 6:12:06 AM4/3/03

to

Peter Woit <wo...@cpw.math.columbia.edu> wrote in message

> The reason no one can calculate anything in string/M theory has

> nothing to with not knowing which vacuum state history has landed us

> in amidst a googleplex of possibilities. No one understands what the

> underlying M-theory is, and that is why no one can calculate anything.

> If one had a well-defined theory and what it told us about the early

> universe was that we could only get probabilistic information about

> the late-time vacuum state, then talking about anthropic principles

> might make sense. Right now this is just one more excuse for not

> dealing with the fact that there is no theory.

I think that's not really fair. Take a look, for example, at Bobby

Acharya's recent preprint and his papers cited there. He shows that

there are interpretations of M theory that make pretty definite

predictions about the cosmological constant. Unfortunately, those

predictions are almost certainly wrong. But they are predictions.

If you look at the recent papers of Acharya, Douglas, and Susskind,

three things become rather clear:

[1] These guys are even smarter than I thought they were.

[2] The discovery of a positive cosmological constant is getting

string theory into really serious trouble

[3] This last fact shows that string/M theory is real science.

Reading the literature now, you can detect a change of attitude.

People are more open to new ideas, and they are more willing to

accept that the theory could be in trouble; they don't have such a

tendency to weasel their way around criticism any more. The view of

string theory as a kind of catapult for launching bandwagons at

regular intervals is also dying down--see e.g. the notable lack of

enthusiasm for the latest candidate, the Dijkgraaf-Vafa stuff.

It's true that there is unfortunately a lot of anthropic talk, but I

guess that nobody really takes that very seriously---think of it as

a cry of pain.

Things are looking up for string theory! People are paying attention

to the serious stuff again.

Apr 4, 2003, 6:51:58 PM4/4/03

to sci-physic...@moderators.isc.org

serenusze...@yahoo.com (Serenus Zeitblom) wrote in message news:<c7fd6c7a.03040...@posting.google.com>...

>

> I guess my main point is just this. Theories rarely get destroyed

> overnight. The

> evidence against them gradually piles up, and their adherents are

> gradually forced to make more and more unnatural assumptions to get

> out of trouble until finally everybody gives up because the theory

> looks so artificial and ugly. Think of what happened to the Steady

> State Cosmology for example---those guys were never *proved* wrong,

> but the initially simple and beautiful theory had to be patched up so

> many times that it ended up looking like a Rube Goldberg contraption.

> That's how you know a theory is wrong---when too many excuses have to

> be made. Now El Susskind comes along and says that no, the extreme

> artificiality of the string models with positive cosmological constant

> is ok, it is just what one would expect anthropically. My objection to

> this is that you can get away with *anything* this way. The anthropic

> principle is like a licence to kill. I'm *not* saying that string/M

> theory is wrong. I'm just saying that things are looking bad right

> now, but the Anthropic Principle is *not* the way to salvation.

> And no, the AP was NOT invented to save string theory! Thank God.

> Speaking of whom, Ed Witten has often come out in the past against

> anthropic ideas in string theory, and I have not heard that he has

> changed his mind.

Most physicists have a very negative reaction to the anthropic

principle because they think it's a cop out, like "I can't figure out

the theory so we'll slap on the anthropic principle". Some people

think that physicists are using the anthropic principle as a crutch,

that they don't have to work as hard to come up with a theory because

they can just fall back on the anthropic principle. I don't think we

have a problem of physicists misusing it in this way. I think string

theorists are trying just as hard to figure out M-theory as if there

was no anthropic principle. Also, there is nothing wrong with the

anthropic principle itself. A point in the Universe chosen at random

would be very unlikely to be as close to a star as we are. How do you

explain the apparent unlikelihood of us being so close to a star? You

explain it with the enthropic principle, meaning life would be much

more likely to arise near a star. We are not located at a totally

random point in the Universe because we are here to ask the question.

Similarly, the anthropic principle applied to M-theory, might actually

be (or may not be) the best explanation for why we are in the point in

the quantum vacua, or in the type of Universe, we are in.

Jeffery Winkler

Apr 6, 2003, 4:23:01 PM4/6/03

to sci-physic...@moderators.isc.org

Serenus Zeitblom wrote:

>>The reason no one can calculate anything in string/M theory has

>>nothing to with not knowing which vacuum state history has landed us

>>in amidst a googleplex of possibilities. No one understands what the

>>underlying M-theory is, and that is why no one can calculate anything.

>>If one had a well-defined theory and what it told us about the early

>>universe was that we could only get probabilistic information about

>>the late-time vacuum state, then talking about anthropic principles

>>might make sense. Right now this is just one more excuse for not

>>dealing with the fact that there is no theory.

>>

>>

>

>I think that's not really fair. Take a look, for example, at Bobby

>Acharya's recent preprint and his papers cited there. He shows that

>there are interpretations of M theory that make pretty definite

>predictions about the cosmological constant. Unfortunately, those

>predictions are almost certainly wrong. But they are predictions.

>

>

The word "prediction" seems to mean something different to me and to

a lot of string theorists. I don't see anything I would call a

prediction of the cosmological constant in Acharya's recent paper.

He's studying

compactifications on singular G2 manifolds, and claiming that for some

choice of what to put at the singularities, he can get a superpotential

whose minima fix moduli. He notes that he gets supersymmetric

vacua with negative cosmological constant this way, and that if

one somehow adds in supersymmetry breaking to make this all realistic,

"In principle, therefore, these models could have positive cosmological

constants."

He also notes that depending on how one chooses the singularities

and what one does at them, one can get a "vast number of possibilities

for the low energy physics" and he suggests that there are lots of

M-theory vacua that look like the standard model.

>If you look at the recent papers of Acharya, Douglas, and Susskind,

>three things become rather clear:

>

>[1] These guys are even smarter than I thought they were.

>

I don't know much about Acharya, but personally I found the Susskind

and Douglas recent "anthropic" papers sad and disturbing. Both of

them have a history of real accomplishments but to me these two

papers seem to represent an explicit abandonment of the idea that

theoretical physicsists should try and predict anything about the

real world.

>

>[2] The discovery of a positive cosmological constant is getting

>string theory into really serious trouble

>

>[3] This last fact shows that string/M theory is real science.

>

>

String/M-theory has always had a huge problem with the cosmological

constant and nothing has really changed on that front, although the

fact that it is non-zero may cause some people to take the problem

more seriously. I'd like to believe you that string theory is getting

into really serious trouble.

>Things are looking up for string theory! People are paying attention

>to the serious stuff again.

>

>

The Susskind/Douglas papers seem to me to be part of the final death-spiral

of the theory, as they and others finally recognize that the framework they

are working in can only lead two places:

1. There are no stable non-supersymmetric M-theory vacua that look anything

like the real world, so M-theory is wrong.

2. There are an infinity of M-theory vacua that look like the real

world, as well

as like anything else you might imagine, so M-theory is vacuous.

Peter

Apr 7, 2003, 1:04:17 AM4/7/03

to sci-physic...@moderators.isc.org

Peter Woit schrieb:

[...]

> 2. There are an infinity of M-theory vacua that look like the real

> world, as well as like anything else you might imagine, so M-theory

> is vacuous.

I do not quite understand this reasoning: Why should a theory be vacuous

that does not predict its own initial conditions?

Why should we expect a theory of quantum gravity to predict not only how

the world works but in addition why it works that way (e.g. by

predicting that it's the unique stable vacuum)?

I recall that in his recent paper (hep-th/0303185) Lee Smolin advertises

the fact that loop quantum gravity can consistentently be formulated for

many kinds of matter couplings, for arbitrary dimensions and for

supersymmetric as well as for non-supersymmetric scenarios. Why should

versatility be a virtue for LQG but be a flaw for strings?

What should we demand a quantum gravity theory to predict? If it

correctly predicts Plack scale physics but does not uniquely fix the

low-energy gauge group and particle spectrum, does that mean its

worthless (or vacuous)?

Apr 7, 2003, 9:46:40 PM4/7/03

to

In article <3E90A600...@uni-essen.de>,

Urs Schreiber <Urs.Sc...@uni-essen.de> wrote:

Urs Schreiber <Urs.Sc...@uni-essen.de> wrote:

>Peter Woit schrieb:

>[...]

>> 2. There are an infinity of M-theory vacua that look like the real

>> world, as well as like anything else you might imagine, so M-theory

>> is vacuous.

...

>What should we demand a quantum gravity theory to predict? If it

>correctly predicts Plack scale physics but does not uniquely fix the

>low-energy gauge group and particle spectrum, does that mean its

>worthless (or vacuous)?

A QG theory that did this would be unsatifying, but would have value.

However, I think the problem is if you have a large number of vacua,

which all reproduce the low-energy physics, then you don't have any

predictive power, because you can't fill in the gap between low-energy

and the Planck scale.

--

======================================================================

Kevin Scaldeferri Calif. Institute of Technology

The INTJ's Prayer:

Lord keep me open to others' ideas, WRONG though they may be.

Apr 7, 2003, 9:46:30 PM4/7/03

to

Urs Schreiber <Urs.Sc...@uni-essen.de> wrote in message

news:<3E90A600...@uni-essen.de>...

news:<3E90A600...@uni-essen.de>...

> What should we demand a quantum gravity theory to predict? If it

> correctly predicts Plack scale physics but does not uniquely fix the

> low-energy gauge group and particle spectrum, does that mean its

> worthless (or vacuous)?

It might be somewhat difficult to evaluate whether Planck scale physics is

correctly predicted or not, given the difficulty to conduct Planck scale

experiments. And it is hard to tell what M-theory's Planck scale predictions

are, since it lacks a mathematical formulation.

Apr 8, 2003, 7:20:42 PM4/8/03

to

Urs Schreiber wrote:

>Peter Woit schrieb:

>

>[...]

>

>

>>2. There are an infinity of M-theory vacua that look like the real

>>world, as well as like anything else you might imagine, so M-theory

>>is vacuous.

>>

>>

>

>I recall that in his recent paper (hep-th/0303185) Lee Smolin advertised

>the fact that loop quantum gravity can consistentently be formulated for

>many kinds of matter couplings, for arbitrary dimensions and for

>supersymmetric as well as for non-supersymmetric scenarios. Why should

>versatility be a virtue for LQG but be a flaw for strings?

>

>

I don't think the people who do LQG think of this versatility as a

virtue, but they

have never claimed that they have a theory which can explain the

standard model.

String/M-theory has been claiming this for nearly twenty years, but not

actually

ever coming up with a theory that actually explains anything at all

about the

standard model.

>What should we demand a quantum gravity theory to predict? If it

>correctly predicts Plack scale physics but does not uniquely fix the

>low-energy gauge group and particle spectrum, does that mean its

>worthless (or vacuous)?

>

>

If string/M-theory actually did this, it would be unsatisfying, but not

vacuous.

The fact of the matter is that it not only doesn't predict low energy

physics,

it also doesn't predict Planck scale physics. Ask a string/M-theorist

to write

down the spectrum of excitations of the theory at or above the Planck scale.

They can't do this. Despite claims one often sees, the

problem with string/M-theory is not that one

can't compute its vacuum state either because the calculation is too

difficult or because this is determined by an initial condition at the

big bang.

The problem is that there is no theory, just some fragmentary ideas which

many take as evidence for the existence of a wonderful unknown theory.

The hope of string/M-theorists recently seems to have been that it doesn't

matter that they don't have a theory, that perhaps the ideas about

string/M-theory they do have are

enough to constrain the possible vacuum states of the unknown theory to

a finite number, one of which would be the standard model. All the evidence

I've seen so far is that this doesn't work. It appears that if you can

write down a vacuum state

consistent with the Standard model, you can write down an infinity of

them, with whatever couplings, masses, etc. you want, predicting nothing.

Apr 8, 2003, 7:32:47 PM4/8/03

to sci-physic...@moderators.isc.org

In article <325dbaf1.03040...@posting.google.com>,

Jeffery <jeffery...@mail.com> wrote:

>

>Most physicists have a very negative reaction to the anthropic

>principle because they think it's a cop out, like "I can't figure out

>the theory so we'll slap on the anthropic principle". Some people

>think that physicists are using the anthropic principle as a crutch,

>that they don't have to work as hard to come up with a theory because

>they can just fall back on the anthropic principle. I don't think we

>have a problem of physicists misusing it in this way. I think string

>theorists are trying just as hard to figure out M-theory as if there

>was no anthropic principle. Also, there is nothing wrong with the

>anthropic principle itself. A point in the Universe chosen at random

>would be very unlikely to be as close to a star as we are. How do you

>explain the apparent unlikelihood of us being so close to a star? You

>explain it with the enthropic principle, meaning life would be much

>more likely to arise near a star. We are not located at a totally

>random point in the Universe because we are here to ask the question.

>Similarly, the anthropic principle applied to M-theory, might actually

>be (or may not be) the best explanation for why we are in the point in

>the quantum vacua, or in the type of Universe, we are in.

Jeffery <jeffery...@mail.com> wrote:

>

>Most physicists have a very negative reaction to the anthropic

>principle because they think it's a cop out, like "I can't figure out

>the theory so we'll slap on the anthropic principle". Some people

>think that physicists are using the anthropic principle as a crutch,

>that they don't have to work as hard to come up with a theory because

>they can just fall back on the anthropic principle. I don't think we

>have a problem of physicists misusing it in this way. I think string

>theorists are trying just as hard to figure out M-theory as if there

>was no anthropic principle. Also, there is nothing wrong with the

>anthropic principle itself. A point in the Universe chosen at random

>would be very unlikely to be as close to a star as we are. How do you

>explain the apparent unlikelihood of us being so close to a star? You

>explain it with the enthropic principle, meaning life would be much

>more likely to arise near a star. We are not located at a totally

>random point in the Universe because we are here to ask the question.

>Similarly, the anthropic principle applied to M-theory, might actually

>be (or may not be) the best explanation for why we are in the point in

>the quantum vacua, or in the type of Universe, we are in.

Would it make sense to say the anthropic principle says one of

several things:

1) We are where we are because we can't be anywhere else. In this case

where refering not only to position is space, but of one of many possible

universes.

2) Any combination of flexible parameters will give rise to a universe with

life of some sort. This seems not to be the case in general, but might be

the case for atleast some variation.

3) The weirdest possibility, bordering on religion and requiring time

travel :) That intelligent life somehow plays a role in the existance

of the universe. I'm thinking of some sort of odd bootstrapping,

where adjustments influencing the past made by intelligent life

allow the universe to exist.

The interesting thing is all of the possibilities I can think of have

implications which run beyond whats currently observable. I believe the

fact that we exist is infact an important piece of data, but one which

is hard to interpret.

--

Be a counter terrorist perpetrate random senseless acts of kindness

Rave: Immanentization of the Eschaton in a Temporary Autonomous Zone.

"Anyone who trades liberty for security deserves neither liberty nor security"

-Benjamin Franklin

Apr 9, 2003, 6:14:00 PM4/9/03

to sci-physic...@moderators.isc.org

Peter Woit schrieb:

>

> Urs Schreiber wrote:

> >I recall that in his recent paper (hep-th/0303185) Lee Smolin advertised

> >the fact that loop quantum gravity can consistentently be formulated for

> >many kinds of matter couplings, for arbitrary dimensions and for

> >supersymmetric as well as for non-supersymmetric scenarios. Why should

> >versatility be a virtue for LQG but be a flaw for strings?

> >

> I don't think the people who do LQG think of this versatility as a

> virtue,

Lee Smolin emphasizes it because it means that LQG would not be

invalidated by possible future discovery of such phenomena. It's not

regarded as a "problem" because for some reason (that still escapes me)

people do not demand of LQG (or any other approach to QG) what they

demand of strings: Namely (among other things) that it fixes the why and

not just the how of the world that we perceive.

The point seems to be that in string theory it looks as if there are, or

might be, mechanisms to actually say something about why-questions,

which is something, I'd think, completely beyond of what we could

reasonably hope a theory does for us. This looks like a bonus for me,

not like something that makes the theory vacuous if it does not exist.

> but they

> have never claimed that they have a theory which can explain the

> standard model. String/M-theory has been claiming this for nearly

> twenty years, but not actually ever coming up with a theory that

> actually explains anything at all about the standard model.

OK. So what should then be criticized in this regard are the claims

about the theory, not the theory itself. It's not a problem of string

theory, as a theory of quantum gravity, not to uniquely predict the

standard model (partly because it shares this characteristic with all

other approaches to QG), but a failure that people have argued

otherwise. Right?

> >What should we demand a quantum gravity theory to predict? If it

> >correctly predicts Plack scale physics but does not uniquely fix the

> >low-energy gauge group and particle spectrum, does that mean its

> >worthless (or vacuous)?

> >

> If string/M-theory actually did this, it would be unsatisfying, but not

> vacuous. The fact of the matter is that it not only doesn't predict low

> energy physics, it also doesn't predict Planck scale physics. Ask a

> string/M-theorist to write down the spectrum of excitations of the theory

> at or above the Planck scale. They can't do this.

I am not sure what you mean by this. At the Planck scale string theory

predicts that you see the massive modes of the string, the string's

spectrum. Is that not what you are referring to here? There are lots of

papers on what signatures of Planck scale physics to expect if string

theory applies, as Lubos Motl has recently emphasized in another thread.

Here are two examples that predict stringy signatures that could be seen

in coming experiments:

Kingman Cheung: Black hole, string ball, and p-brane production at

hadronic supercolliders, hep-ph/0205033

Abstract:

In models of large extra dimensions, the string and Planck scales become

accessible at future colliders. When the energy scale is above the

string scale or Planck scale a number of interesting phenomena occur,

namely, production of stringy states, p-branes, string balls, black

hole, etc. In this work, we systematically study the production cross

sections of black holes, string balls, and p-branes at hadronic

supercolliders. We also discuss their signatures. At the energy scale

between the string scale M_s and M_s/g_s^2, where g_s is the string

coupling, the production is dominated by string balls, while beyond

M_s/g_s^2 it is dominated by black holes. The production of a p-brane is

only comparable to black holes when the p-brane wraps entirely on small

extra dimensions. Rough estimates on the sensitivity reaches on the

fundamental Planck scale M_D are also obtained, based on the number of

raw events.

Kin-ya Oda: Alternative Signature of TeV Strings, hep-ph/0111298

Abstract:

In string theory, it is well known that any hard scattering amplitude

inevitably suffers exponential suppression. We demonstrate that, if the

string scale is M_s < 2TeV, this intrinsically stringy behavior leads to

a dramatic reduction in the QCD jet production rate with very high

transverse momenta p_T > 2TeV at LHC. This suppression is sufficient to

be observed in the first year of low-luminosity running. Our prediction

is based on the universal behavior of string theory, and therefore is

qualitatively model-independent. This signature is alternative and

complementary to conventional ones such as Regge resonance (or string

ball/black hole) production.

(end of abstract)

These papers of course assume large extra dimensions. If extra

dimensions are small this simply moves the signatures beyond the reach

of present and upcoming technology. But they still can be tested in

principle. The problem of actually measuring Plack scale physics is

inherent to QG in general. Once it was also hard to test classical GR.

In fact it still is hard to check for gravitational waves.

Apr 9, 2003, 8:40:23 PM4/9/03

to

news:b64t3b$4nm$1...@newsmaster.cc.columbia.edu...

> Jeffery wrote:

Isn't the real situation one where the competing trial theory to the

one running on the Anthropic Principle asserts that, in theory, given

complete knowledge of all initial conditions (hmmm, where've we heard

of that ambitious ability), then in theory all downstream stuff

including creative acts can be calculated, er, except for the

non-computable stuff? However, that program of perfect scientific

omniscence has not materialized on the Earthly plane and the calcs are

chaotic so while folks assert there IS a theory, it is a

non-functional one leaving us to still be guided by the functional AP.

Apr 10, 2003, 3:57:13 PM4/10/03

to

Urs Schreiber <Urs.Sc...@uni-essen.de> wrote in message

news:3E93DBB3...@uni-essen.de...> Lee Smolin emphasizes it because it means that LQG would not be

> invalidated by possible future discovery of such phenomena. It's not

> regarded as a "problem" because for some reason (that still escapes me)

> people do not demand of LQG (or any other approach to QG) what they

> demand of strings: Namely (among other things) that it fixes the why and

> not just the how of the world that we perceive.

It would be truly great if string theory could explain the why, and even if

it only could explain the how. However, most people would be quite happy

with a much more modest achievement, namely that it would make *some*

testable prediction, and that this prediction is confirmed experimentally

within a reasonable amount of time. Witten recently claimed that string

theory is more predictive that conventional QFT, e.g. because it predicts

supersymmetry. However, few string theorists seem willing to stand up and

unambigously predict that sparticles will be seen at the LHC. (To me the

word "prediction" means that non-confirmation proves the theory wrong).

For a very clear discussion on why physics must produce reliable knowledge

about the real world, based on experiments, see Section 1.4 of

http://www.arxiv.org/abs/hep-th/0204131. Being a string theory pioneer, the

author should know what he talks about.

> These papers of course assume large extra dimensions. If extra

> dimensions are small this simply moves the signatures beyond the reach

> of present and upcoming technology.

For tests of large (> 0.01 mm) extra dimensions, see

Nature 421 922 (2003)

http://arXiv.org/abs/hep-ph/0210004

http://arxiv.org/abs/hep-ph/0303057

(thanks to Uncle Al).

Apr 10, 2003, 5:15:13 PM4/10/03

to

Urs Schreiber wrote:

>OK. So what should then be criticized in this regard are the claims

>about the theory, not the theory itself. It's not a problem of string

>theory, as a theory of quantum gravity, not to uniquely predict the

>standard model (partly because it shares this characteristic with all

>other approaches to QG), but a failure that people have argued

>otherwise. Right?

>

>

String/M-theory doesn't just not uniquely predict the standard

model. It doesn't predict anything. If one wants to evaluate it

purely as an explanation of how to reconcile quantum mechanics

with general relativity, see Smolin for details of why this is an

unfinished project, one that LQG arguably does better at.

>

>

>>>What should we demand a quantum gravity theory to predict? If it

>>>correctly predicts Plack scale physics but does not uniquely fix the

>>>low-energy gauge group and particle spectrum, does that mean its

>>>worthless (or vacuous)?

>>>

>>>

>>>

>>If string/M-theory actually did this, it would be unsatisfying, but not

>>vacuous. The fact of the matter is that it not only doesn't predict low

>>energy physics, it also doesn't predict Planck scale physics. Ask a

>>string/M-theorist to write down the spectrum of excitations of the theory

>>at or above the Planck scale. They can't do this.

>>

>>

>

>I am not sure what you mean by this. At the Planck scale string theory

>predicts that you see the massive modes of the string, the string's

>spectrum. Is that not what you are referring to here? There are lots of

>papers on what signatures of Planck scale physics to expect if string

>theory applies, as Lubos Motl has recently emphasized in another thread.

>

What exactly is the string spectrum you expect to see at high

energy? Remember the diagram with six cusps that M-theorists like

to draw at the start of their talks. The only thing you know how

to calculate involves what happens at the points of the cusps and

even there you have the compactification problem. Where are we

in this diagram and what is the theory at any point in this

diagram? M-theorists seem to claim that the underlying theory is

some sort of eleven dimensional supersymmetric theory whose

fundamental excitations are higher-dimensional branes, but have

no idea what the dynamics of these is supposed to be.

>Here are two examples that predict stringy signatures that could be seen

>in coming experiments:

>

>

A prediction is not something that tells you what "could" be seen

in coming experiments, it is something that tells you what "will"

be seen in coming experiments. A scientific prediction is

supposed to be something that can be used to falsify a theory.

If black holes don't come out of the interaction regions at the

LHC, no one is going to think this says anything one way or

another about string/M-theory.

The large extra dimension scenario has all the problems of

string/M-theory, together with the added one of not even knowing

what the overall scale of the theory is. It also throws out the

only two arguments for supersymmetry (hierarchy problem and

unification of coupling constants).

Have to run now and get ready to watch the TV program "Einstein's

Dream" tonight. The advertisement indicates that the physics

faculty of the Institute for Advanced Study (Seiberg, Witten,

Maldacena) will be explaining their pioneering work on a theory

to explain all the forces of nature in the same terms. Freeman

Dyson is also supposed to appear in the role of the skeptic.

Peter

Apr 11, 2003, 8:41:10 PM4/11/03

to sci-physic...@moderators.isc.org

> However, few string theorists seem willing to stand up and

> unambigously predict that sparticles will be seen at the LHC. (To me the

> word "prediction" means that non-confirmation proves the theory wrong).

The reason is that TeV scale symmetry is nice for solving the hierarchy

problem, but one can easily imagine breaking SUSY at a higher scale.

Still, the lack of low energy SUSY would be a fair sized blow, IMO.

Apr 11, 2003, 10:47:16 PM4/11/03

to sci-physic...@moderators.isc.org

Peter Woit schrieb:

> String/M-theory doesn't just not uniquely predict the standard

> model. It doesn't predict anything.

I can well understand dissatisfaction with the achievements of string

theory. But annoyance about the habit of some string proponents to

overemphasize the success of the theory should probably not justify

copying this habit under reversed sign and deliberately underestimating

the success to the point of claiming that "there is no theory" and that

"it does not predict anything". To take the most banal counter example

(and there are less banal ones as you know better than I do): As soon as

accelerators see a total of 12 or more spacetime dimensions (yes, that's

not likely to happen anywhere soon, but that's how it goes with quantum

gravity) string theory is physically falsified. On the other hand, LQG

would apparently still be a possible candidate.

> If one wants to evaluate it

> purely as an explanation of how to reconcile quantum mechanics

> with general relativity, see Smolin for details of why this is an

> unfinished project,

Without any doubt is it unfinished. It seems that this is one point that

really everybody agrees on.

> one that LQG arguably does better at.

In the present state it is certainly a matter of individual opinion on

which approach to put more personal hopes, if on any. I am prepared to

seriously consider your and Thomas Larsson's point of view, that string

theory has "failed", as worthy of consideration. But to be useful this

claim must be backed up with good examples of such failures. So far we

have been arguing about whether or not the non-uniqueness of the vacuum

is really such a failure. To me it seems, and please correct me if this

is wrong, that all the objections that you have brought forth against

string theory rest on the single assumption that it needs to be

considered a failure that the theory does not predict where in moduli

space we are, i.e. which state the world is in:

> A prediction is not something that tells you what "could" be seen

> in coming experiments, it is something that tells you what "will"

> be seen in coming experiments.

That's true as soon as the state is known, which the experiment is

probing. Given some state, theory predicts its properties. Without

knowledge of the state being measured theory can only say what could be

seen, namely anything compatible with some state.

For instance, in classical celestial mechanics, where (bound) states are

Kepler ellipses, let E be any such state. Then the theory says that we

could find a planet in the solar system with trajectroy E. The theory

does not predict that beyond Neptun we will find Pluto (ignoring

planet-planet interactions). But as soon as position and velocity of

Pluto are determined the state is known and theory predicts all the

future trajectory of the object.

The analogue applies of course to string theory. As long as we do not

know the state the world is in (the "vacuum") we don't know the string

scale, the type of string (I,II, etc.), the compactification geometry,

for instance. As soon as we know these, string theory predicts for

instance amplitudes of string scattering events.

I don't think this is unusual or even diconcerting. As another example,

classical cosmology predicts that we could inhabit an open FRW universe.

It does not predict that we necessarily do, because there are other

possible states in the theory, like flat and closed FRW universes. But

as soon as the state we observe is known (the value of the curvature

parameter k in this example) the theory predicts all properties of this

state that can be measured.

Also, as with any other theory, some predictions should be possible

without a full knowledge of the state. That's why the author of the

hep-ph/0111298 paper, that I had cited, says (in the abstract):

Apr 13, 2003, 10:30:21 PM4/13/03

to

Urs Schreiber wrote:

>To me it seems, and please correct me if this

>is wrong, that all the objections that you have brought forth against

>string theory rest on the single assumption that it needs to be

>considered a failure that the theory does not predict where in moduli

>space we are, i.e. which state the world is in:

My objections to string/M-theory are not just based on the

vacuum selection problem. The vacuum selection problem is

just a symptom of more serious underlying problems.

The minute one tries to calculate anything in

string/M-theory one runs into trouble. The trouble

with vacuum selection occurs first because that's the first

thing one needs to calculate to use the theory to do physics.

It's not that we don't know where in moduli

space we are, more to the point we don't know what the

moduli space is, or what the theory is that corresponds

to a generic point in this supposed moduli space.

The more serious problems of string/M-theory are:

1. There is no theory, just a few fragments of a theory

which some people have been hoping for twenty years to turn

into a real theory, with no success. If you don't believe

me that there is no theory, consult the following web page

http://www.physics.ucsb.edu/%7Egiddings/Mquest.html

which has a list of unanswered questions in M-theory,

including "What are the fundamental degrees of freedom

of M-theory?", "What are the dynamical laws governing

their interaction?" "What are the observables of M-theory?"

If you don't know what the fundamental degrees of freedom,

dynamical laws or observables of your theory are, how can

you claim to have a theory?

2. The fragments of a theory that exist are ugly and

mathematically unconvincing. All attempts to relate

string/M-theory to the real world involve

very complicated and ugly constructions (e.g. an

11-dim space, R^4 X a seven dimensional singular

space with a complicated and random-looking set of

singularities). The standard model is based on two

of the deepest notions in modern geometry (the Dirac

operator and the notion of a connection). An

aesthetically and mathematically convincing extension

of the standard model should involve mathematics at

this depth or greater, not random complicated geometrical

constructions obviously set up to get the answer one

wants.

3. The effort to get beyond the standard model using

string/M-theory has now occupied most of the smartest

people in the field for nearly twenty years, producing

somewhere between 10K-100K scientific papers. Most

importantly, this has used up about twenty years of intense

intellectual effort by Witten. The end result is

not only no progress towards the goal, but the goal

is getting further and further away, with the

Susskind/Douglas papers just the latest example of how

little string/M-theorists now think their theory can ever

explain. In 84-85, many were convinced that they would

be calculating standard model parameters within months. This

now seems impossibly far away. There has been real

progress in understanding what string/M-theory might be,

but all this new understanding has just made it more and

more clear that the original hopes for the theory don't

work.

If this isn't "failure", then nothing is.

String/M-theory is still being heavily promoted

these days (as in the TV program I saw last night), but

it is not being held either to a standard of

consistency or to a standard of making falsifiable

predictions that can be checked. From the point of

view of its promoters, the theory can never be said

to have failed.

Apr 13, 2003, 10:30:22 PM4/13/03

to

Urs Schreiber <Urs.Sc...@uni-essen.de> wrote in message

news:3E96A902...@uni-essen.de...> In the present state it is certainly a matter of individual opinion on

> which approach to put more personal hopes, if on any. I am prepared to

> seriously consider your and Thomas Larsson's point of view, that string

> theory has "failed", as worthy of consideration. But to be useful this

> claim must be backed up with good examples of such failures. So far we

> have been arguing about whether or not the non-uniqueness of the vacuum

> is really such a failure. To me it seems, and please correct me if this

> is wrong, that all the objections that you have brought forth against

> string theory rest on the single assumption that it needs to be

> considered a failure that the theory does not predict where in moduli

> space we are, i.e. which state the world is in:

That string theory fails to make unique predictions is embarrassing,

but it has never been my main objection. LQG does not make terribly

many predictions either (besides, Smolin's claim that special

relativity may break down makes me confused - it does not sound

right).

My main point, as expressed in

http://www.arxiv.org/abs/math-ph/0103013

is the existence of new mathematics, which both generalizes important

structures appearing in perturbative string theory, and is closely

related to theories describing real physics. You can also see that

although the e-print is a political pamphlet intended to provoke, I do

emphasize the successes of stringy mathematics, both as pure

mathematics and in statistical physics. It has rather been (some)

string theorists here that have described perturbative string theory

as boring and irrelevant, when I have pointed out that its mathematics

is less distinguished than previously thought.

This is IMO a very strong point, because string theory evolved

historically using mathematical beauty and consistency as sole

guidelines. In particular, string theory's pioneers were obsessed with

*central* extensions, as witnessed by terms like Virasoro,

Neveu-Schwartz and Ramond algebras. If they had known that similar

(but non-central) extensions exist and admit representations for all

algebras of vector fields, string theory might have taken a different

route during its formative years.

As you and other regulars on this newsgroup know (but others may read

this too), I refer of course to two algebraic structures:

The higher-dimensional Virasoro algebra, which combines key aspects of

general relativity (diffeomorphisms) and quantum mechanics (projective

representations).

mb(3|8), which is an exceptional simple Lie superalgebra closely related

to the symmetries of the standard model.

I believe that my point is both original and constructive, and I am

sad that my position has been construed as mere negativist

string-hating. It seems to me that investigating these structures, and

transforming them into physics, would be a much better use for all the

talent in string theory, especially since recent anthropic ideas seem

to be the alternative. The worst thing that can happen is that this

turns out to be cool mathematics only.

Sadly, my marketing efforts have obviously failed, but I don't really

see what I could have done differently. The chances that people look

into these structures must be better if they dislike them than if they

have never heard about them at all.

Apr 14, 2003, 11:58:16 AM4/14/03

to

ent...@farviolet.com (Lawrence Foard) wrote in message news:<b6nugl$fr5>

1) We are where we are because we can't be anywhere else. In this case

> where refering not only to position is space, but of one of many possible

> universes.

Not that it would be impossible, but you can point to characteristics

of our universe that would make life more likely.

>

> 2) Any combination of flexible parameters will give rise to a universe with

> life of some sort. This seems not to be the case in general, but might be

> the case for atleast some variation.

You can certainly come up with hypothetical universes in which life

was impossible. What if all matter in the Universe was in a giant

black hole? What if we had only bosons with no fermions? What if the

proton had a tiny lifetime like the other hadrons? What if there were

one, or none, extended spatial dimensions?

However, you are right in that there are probably universes that

contain life totally different from us, totally different from

anything we could imagine?

>

> 3) The weirdest possibility, bordering on religion and requiring time

> travel :) That intelligent life somehow plays a role in the existance

> of the universe. I'm thinking of some sort of odd bootstrapping,

> where adjustments influencing the past made by intelligent life

> allow the universe to exist.

I don't think this makes any sense at all.

Jeffery Winkler

Apr 14, 2003, 3:34:01 PM4/14/03

to sci-physic...@moderators.isc.org

ent...@farviolet.com (Lawrence Foard) wrote in message news:<b6nugl$fr5$1...@farviolet.com>...

No it doesn't say that. It is a constraint on the fundamental

physical parameters by the evidence of our existence. That is why it

is troubling. Its statement is causally incorrect.

>

> 2) Any combination of flexible parameters will give rise to a universe with

> life of some sort. This seems not to be the case in general, but might be

> the case for atleast some variation.

No it doesn't say that either although the strong version requires

carbon based life to emerge from the particular set of parameters that

govern our universe.

>

> 3) The weirdest possibility, bordering on religion and requiring time

> travel :) That intelligent life somehow plays a role in the existance

> of the universe. I'm thinking of some sort of odd bootstrapping,

> where adjustments influencing the past made by intelligent life

> allow the universe to exist.

This is where you are getting caught in the causal cul de sac of #1.

There is no evidence that time is reversable. Now if you want to

consider some type of cyclical universe model (which requires a closed

universe) then you might argue that there might be a "reprocessing" of

the parameters in which life in the collapsing universe might

influence the next cycle. (see Tiplers "The Physics of Immortality and

get rid of all the infinities and base it on a quantum theory of

gravity and it might come close to what you suggest.

>

> The interesting thing is all of the possibilities I can think of have

> implications which run beyond whats currently observable. I believe the

> fact that we exist is infact an important piece of data, but one which

> is hard to interpret.

If you can't falsify it (even in principle), it's not science, its

religion.

e.

Apr 14, 2003, 3:37:02 PM4/14/03

to

In article <3E96A902...@uni-essen.de>,

Urs Schreiber <Urs.Sc...@uni-essen.de> wrote:

Urs Schreiber <Urs.Sc...@uni-essen.de> wrote:

>Peter Woit schrieb:

>

>> String/M-theory doesn't just not uniquely predict the standard

>> model. It doesn't predict anything.

>I can well understand dissatisfaction with the achievements of string

>theory. But annoyance about the habit of some string proponents to

>overemphasize the success of the theory should probably not justify

>copying this habit under reversed sign and deliberately underestimating

>the success to the point of claiming that "there is no theory" and that

>"it does not predict anything". To take the most banal counter example

>(and there are less banal ones as you know better than I do): As soon as

>accelerators see a total of 12 or more spacetime dimensions (yes, that's

>not likely to happen anywhere soon, but that's how it goes with quantum

>gravity) string theory is physically falsified.

Do you truly beleive that? 10 years ago, the same claim would have

been made, but the number was 11 or more.

Apr 14, 2003, 3:50:57 PM4/14/03

to sci-physic...@moderators.isc.org, jeffery...@mail.com

jeffery...@mail.com (Jeffery) writes:

[...]

> Also, there is nothing wrong with the anthropic principle itself.

> A point in the Universe chosen at random would be very unlikely to be as

> close to a star as we are.

That depends on what probability measure you use to choose your point

"at random." Since matter is clumpy, and tends to get concentrated

in and around stars, drawing points "at random" in proportion to

mass density may not only land your "near" a star, but _in_ a star !!!

> How do you explain the apparent unlikelihood of us being so close to a

> star? You explain it with the enthropic principle, meaning life would be

> much more likely to arise near a star.

Not necessarily. There is growing evide nce that large, tidally-heated

icy moons of gas giant planets and "brown dwarves" will have liquid water

"oceans" under their ice crusts, as Europa, Ganymede, and perhaps Callisto

do, and that if one also considers Water/Ammonia mixtures, then Titan,

Triton, and even Pluto may have liquid "oceans" under their ice crusts.

Even the larger Kuiper Belt Objects may have liquid Water/Ammonia "oceans"

that are kept liquid and supplied with energy from the decay of radioactive

isotopes in their cores. If so, then the majority of potential abodes for

water-based carbon life in the galaxy may not be "near" stars at all, but

rather in modest-sized iceballs floating in the outer darkness. Hence,

humanity's existence on a hot, oversized rock far too close to its sun

starts looking very unlikely indeed, and one should start talking about the

"Water/Ammonia-based Cephalopodic Principle," not the "Anthropic principle,"

since the squid-populated iceballs will outnumber the earthlike worlds by

one to three orders of magnitude... :-T

(The above, BTW, is a prime example of why the "Anthropic Principle"

has virtually ZERO predictive power --- one can make it fit just about

any scenario one pleases...)

-- Gordon D. Pusch

perl -e '$_ = "gdpusch\@NO.xnet.SPAM.com\n"; s/NO\.//; s/SPAM\.//; print;'

Apr 16, 2003, 5:29:45 PM4/16/03

to sci-physic...@moderators.isc.org

Thomas Larsson schrieb:

[...]

> The higher-dimensional Virasoro algebra, which combines key aspects of

> general relativity (diffeomorphisms) and quantum mechanics (projective

> representations).

What I don't understand is why the higher-dimensional Virasoro algebra

should be the constraint algebra of (quantum) GR: Isn't there more to GR

than diffeo invariance? When starting from the Einstein-Hilbert action

one gets the constraint algebra that is used in LQG to quantize. Which

action would give the constraint algebra that you are proposing? Or

would it follow from the same action (the EH action) but using a

different quantization scheme?

In your paper you are saying that string theory used the Virasoro

algebra at a point when no (few?) "deeper structures" were known. I

don't know about that. But I know that the Virasor algebra turns up in

string theory not because somebody considered it deep enough to be

worthy of a TOE, but because it comes up all by itself when quantizing

the Polyakov action, which is the simple generalization of the point

particle action. Actuallly, this is where all of string theory follows

from by consistency requirements. In this sense I always find it a

little odd to complain about the structures in string theory not being

"deep enough": They are not put in by hand but follow from dealing with

the Polyakov action, in the end. This is where all these concepts are

rooted in something that is arguably physics (even if not necessarily

the physics of our world). But what is the physics behind your concepts,

in this sense. From which physical assumptions do we arrive at the

multi-dimensional Virasoro algebra? What if tomorrow somebody else comes

along and promotes that not depth is the ultimate indicator of

algebraic profoundness but some other algebraic characterization. How do

I know who's right?

Can you say anything about the classical limit of states that satisfy

your constraints, can you recover classical GR in some sense or at least

speculate about how that would have to be approached? In other words:

The higher-dim Vir algebra is nice mathematics: What does it have to do

with physics, with general relativity in particular? It is an extension

of the diffeomorphism group (right?) but that alone does not give GR,

does it? You say it incorporates diff-invariance and projective

representations and thus must be GR and must be quantum. This doen't

sound implausible (though I cannot really judge this statement) but

gives me much less information than I usually get with a (new) theory of

physics. I am left to trust your intuition about how QG should look

like. It's like somebody writing down just E8 and then claiming that

this must be the TOE. I wouldn't see why he could even claim so. But

since the appearance of E8 can actually be derived from an assumption

about physics (strings instead of points), I can see why this claim is

at least not without some basis.

> mb(3|8), which is an exceptional simple Lie superalgebra closely related

> to the symmetries of the standard model.

I read in your paper that the 0-grade subalgeba is sl(3) x sl(2) x

gl(1), which you call the standard model algebra. But of course that is

really su(3) x su(2) x u(1). You say the former is the non-compact form

of the latter. Well possible that I am ignorant of a basic fact of gauge

theory: How do you get from sl(3) x sl(2) x gl(1) to the observed gauge

group?

Apr 18, 2003, 2:15:58 AM4/18/03

to sci-physic...@moderators.isc.org

In article <3E93DBB3...@uni-essen.de>,

Urs Schreiber <Urs.Sc...@uni-essen.de> wrote:

>Peter Woit schrieb:

>> Urs Schreiber wrote:

>> >I recall that in his recent paper (hep-th/0303185) Lee Smolin advertised

>> >the fact that loop quantum gravity can consistentently be formulated for

>> >many kinds of matter couplings, for arbitrary dimensions and for

>> >supersymmetric as well as for non-supersymmetric scenarios. Why should

>> >versatility be a virtue for LQG but be a flaw for strings?

>> I don't think the people who do LQG think of this versatility as a

>> virtue,

>Lee Smolin emphasizes it because it means that LQG would not be

>invalidated by possible future discovery of such phenomena. It's not

>regarded as a "problem" because for some reason (that still escapes me)

>people do not demand of LQG (or any other approach to QG) what they

>demand of strings: Namely (among other things) that it fixes the why and

>not just the how of the world that we perceive.

People place different demands on loop quantum gravity and string theory,

because the supporters of these theories have made quite different

claims about these theories.

String theory has often been touted as a "theory of everything" which

could explain all the forces and particles of nature, if only the details

could be worked out. It's often been suggested that there's a more or less

unique consistent version of the theory. Claims like this help string

theorists get lots of money. They also raise big expectations - and lead

to criticism when these expectations are not quickly met.

Loop quantum gravity has always been a more humble, low-budget affair.

Its goal is to find a background-free theory of gravity that takes

quantum mechanics into account. Its supporters never seem to claim

that there's a unique way to incorporate matter into this theory, or

that it's a "theory of everything". So it gets less funding, and

people demand less of it.

Of course, one reason for this is that string theory came from particle

physics, which got lots of money in the heyday of particle accelerators,

before the debacle of the superconducting supercollider. Loop quantum

gravity came from general relativity, which got very little money in the

days before LIGO.

These issues are almost more sociological than scientific, but the

sociology affects the progress of science. People would reduce

their expectations of string theory, become less critical of it,

give it less funding, but give it more time to work, if string

theorists lowered their claims for it. I think this would be a

good thing. It may be in the process of happening right now, thanks

to the work of Susskind and others.

I should emphasize that both in superstring theory and loop quantum

gravity, the strength of the claims made varies drastically from

researcher to researcher. In loop quantum gravity, Smolin probably

makes more enthusiastic claims than almost anyone else.

Apr 18, 2003, 2:16:37 AM4/18/03

to sci-physic...@moderators.isc.org

gdp...@NO.xnet.SPAM.com (Gordon D. Pusch) wrote in message news:<giof3aq...@pusch.xnet.com>...

> jeffery...@mail.com (Jeffery) writes:

<snip>

> > How do you explain the apparent unlikelihood of us being so close to a

> > star? You explain it with the enthropic principle, meaning life would be

> > much more likely to arise near a star.

>

> Not necessarily. There is growing evide nce that large, tidally-heated

> icy moons of gas giant planets and "brown dwarves" will have liquid water

> "oceans" under their ice crusts, as Europa, Ganymede, and perhaps Callisto

> do, and that if one also considers Water/Ammonia mixtures, then Titan,

> Triton, and even Pluto may have liquid "oceans" under their ice crusts.

> Even the larger Kuiper Belt Objects may have liquid Water/Ammonia "oceans"

> that are kept liquid and supplied with energy from the decay of radioactive

> isotopes in their cores. If so, then the majority of potential abodes for

> water-based carbon life in the galaxy may not be "near" stars at all, but

> rather in modest-sized iceballs floating in the outer darkness. Hence,

> humanity's existence on a hot, oversized rock far too close to its sun

> starts looking very unlikely indeed, and one should start talking about the

> "Water/Ammonia-based Cephalopodic Principle," not the "Anthropic principle,"

> since the squid-populated iceballs will outnumber the earthlike worlds by

> one to three orders of magnitude... :-T

>

> (The above, BTW, is a prime example of why the "Anthropic Principle"

> has virtually ZERO predictive power --- one can make it fit just about

> any scenario one pleases...)

All very interesting speculation. However, the Anthropic Principle

does make a prediction - given that we are observers, and that we

exist when and where we do, and given no information as to whether

observers exist elsewhere, then observers *will* tend to be located

close to a star. Of course, one may be wrong, and self-aware entities

may be found to exist orbiting Saturn, but of course you adapt your

reasoning to fit the facts. All the Anthropic Principle says is that

observers are typically found in conditions no more unusual than those

required for their existence. It doesn't say much, true, but it does

indicate where to place your bet.

-I

Apr 18, 2003, 10:51:47 AM4/18/03

to

Urs Schreiber <Urs.Sc...@uni-essen.de> wrote in message news:<3E9A6EEF...@uni-essen.de>...

> Thomas Larsson schrieb:

>

> [...]

> > The higher-dimensional Virasoro algebra, which combines key aspects of

> > general relativity (diffeomorphisms) and quantum mechanics (projective

> > representations).

>

> What I don't understand is why the higher-dimensional Virasoro algebra

> should be the constraint algebra of (quantum) GR: Isn't there more to GR

> than diffeo invariance?

> Thomas Larsson schrieb:

>

> [...]

> > The higher-dimensional Virasoro algebra, which combines key aspects of

> > general relativity (diffeomorphisms) and quantum mechanics (projective

> > representations).

>

> What I don't understand is why the higher-dimensional Virasoro algebra

> should be the constraint algebra of (quantum) GR: Isn't there more to GR

> than diffeo invariance?

Sure there is, but diff invariance is a crucial aspect (and one that

string theory misses, lacking a background-independent formulation). In

fact, tensor calculus is secretly the classical representation theory of

the diff group - think primary fields in CFT. There is certainly more to

GR than tensor calculus, but it is hard to imagine GR without it.

Similarly, the Fock modules of Vir(n) can be considered as the quantum

analogue of tensor calculus. This is not QG by itself, but probably a

necessary piece of mathematical infrastructure, which does for QG what

tensor calculus does for GR.

One reason why I believe very strongly in this has to do with how it was

discovered. The QG argument came first, at a time when there even was a

no-go theorem stating that a multi-dimensional Virasoro algebra could

not exist: vect(n) has no central extension when n > 1. That a

generalization of the Virasoro algebra nevertheless exists and has a

natural representation theory (something that other extensions don't

seem to have) confirmed to me that my instincts were right.

> When starting from the Einstein-Hilbert action

> one gets the constraint algebra that is used in LQG to quantize. Which

> action would give the constraint algebra that you are proposing? Or

> would it follow from the same action (the EH action) but using a

> different quantization scheme?

The latter. Now when the theory of Fock modules seems essentially

complete, I constructed in http://www.arxiv.org/abs/math-ph/0210023

more complicated modules that depend on dynamics in the

form of Euler-Lagrange (EL) equations. The idea is as follows:

Phase space is the space of solutions to the EL equations modulo gauges;

one may coordinatize phase space by specifying the values on a spacelike

surface, but this is not intrinsic. I have read this somewhere, probably

in

http://www.arxiv.org/abs/gr-qc/9910079

Loop Quantum Gravity and the Meaning of Diffeomorphism Invariance

Marcus Gaul, Carlo Rovelli

but now I am unable to find the quote (the paper is worth reading

anyway). Anyway, I consider a "large phase space" where gauges have not

been eliminated, but where the gauge group instead acts on (the space of

functions over) this large phase space. Classically, and in the absense

of anomalies, there is no difference; you can always mod out gauges by

passing to the orbit space.

Since the large phase space is built from tensor fields, the Fock

construction applies, and one winds up with Fock modules of Vir(n) which

to me seem like a quantization of the dynamical system - the modules

depend on the EL equations and carry a representation of the

diffeomorphism algebra, where some brackets have quantum corrections.

If you look at things more closely there are some consistency conditions

which only seem to work out if spacetime has four dimensions and there

are two bosons for every three fermions with the naive counting of

degrees of freedom; this relation holds in the standard model coupled to

gravity. However, the same argument also requires new gauge symmetries,

including fermionic ones, perhaps indicating the need for some new physics.

> What if tomorrow somebody else comes

> along and promotes that not depth is the ultimate indicator of

> algebraic profoundness but some other algebraic characterization. How do

> I know who's right?

You don't! If you want fool-proof guarantees in advance, you shouldn't

be doing research. Mathematical beauty alone can never be the ultimate

arbiter, since beauty lies in the eyes of the beholder. It is eventually

the job of the experimentalist to select between candidate theories.

Besides, the coolest thing about maximal depth is that it happens to

lead to a symmetry group which is closely related to the standard model,

i.e. experiments.

> I read in your paper that the 0-grade subalgeba is sl(3) x sl(2) x

> gl(1), which you call the standard model algebra. But of course that is

> really su(3) x su(2) x u(1). You say the former is the non-compact form

> of the latter. Well possible that I am ignorant of a basic fact of gauge

> theory: How do you get from sl(3) x sl(2) x gl(1) to the observed gauge

> group?

You go from compact to non-compact by complexification, and the other

way by picking a compact real form. One must eventually work with the

real form, but working with the complexification seems like a natural

first step.

Apr 18, 2003, 11:17:45 AM4/18/03

to

Peter Woit schrieb:

> 1. There is no theory, just a few fragments of a theory

> which some people have been hoping for twenty years to turn

> into a real theory, with no success. If you don't believe

> me that there is no theory, consult the following web page

>

> http://www.physics.ucsb.edu/%7Egiddings/Mquest.html

>

> which has a list of unanswered questions in M-theory,

> including "What are the fundamental degrees of freedom

> of M-theory?", "What are the dynamical laws governing

> their interaction?" "What are the observables of M-theory?"

> If you don't know what the fundamental degrees of freedom,

> dynamical laws or observables of your theory are, how can

> you claim to have a theory?

> 1. There is no theory, just a few fragments of a theory

> which some people have been hoping for twenty years to turn

> into a real theory, with no success. If you don't believe

> me that there is no theory, consult the following web page

>

> http://www.physics.ucsb.edu/%7Egiddings/Mquest.html

>

> which has a list of unanswered questions in M-theory,

> including "What are the fundamental degrees of freedom

> of M-theory?", "What are the dynamical laws governing

> their interaction?" "What are the observables of M-theory?"

> If you don't know what the fundamental degrees of freedom,

> dynamical laws or observables of your theory are, how can

> you claim to have a theory?

Since this applies to M-theory only, I propose the following summary of

this discussion:

1) Perturbative string theory exists and makes predictions.

2) M-theory hardly exists and hence hardly makes predictions.

3) Many popular overly enthusiastic claims about string/M-theory are

problematic.

[...]

> The standard model is based on two

> of the deepest notions in modern geometry (the Dirac

> operator and the notion of a connection). An

> aesthetically and mathematically convincing extension

> of the standard model should involve mathematics at

> this depth or greater,

Isn't it fascinating how much string theory has to say about Dirac

operators, gauge theory and modern geometry?

Apr 20, 2003, 8:22:14 PM4/20/03

to

"Kevin A. Scaldeferri" schrieb:

> In article <3E96A902...@uni-essen.de>,

> Urs Schreiber <Urs.Sc...@uni-essen.de> wrote:

> >Peter Woit schrieb:

> >> String/M-theory doesn't just not uniquely predict the standard

> >> model. It doesn't predict anything.

[...]

> >As soon as

> >accelerators see a total of 12 or more spacetime dimensions (yes, that's

> >not likely to happen anywhere soon, but that's how it goes with quantum

> >gravity) string theory is physically falsified.

>

> Do you truly beleive that? 10 years ago, the same claim would have

> been made, but the number was 11 or more.

Though the dimension did not just go up by one. The prediction is that

up to ten dimensions you see stringy signatures, e.g. Regge behaviour

and Hagedorn statistics. As soon as you begin to see an 11th dimension

this would gradually be replaced by the corresponding signatures of

full membranes.

Superstrings as such live in ten dimensions, and that cannot change.

They are limits of supermembranes which necessarily live in 11

dimensions, and that cannot change either. I don't know if membranes

could in priciple possibly some day be found out to be limits of

something even higher dimensional, as you suggest. Actually there is

F-theory, which lives in 12 dimensions. But that's usually considered

as a mere computational tool to get 11d physics. (Quite similar to how

the Wess-Zumino term in a 2-d WZ model is naturally written as a

topological term integrated over a 3-fold.)

So I think there is reason to believe in my claim above, but you are

right that it should be made more precise. Let's see: As soon as

accelerators see 11 dimensions and still no deviation from

Regge/Hagedorn signatures string theory is physically falsified. If no

Regge/Hagedorn behaviour (stringy signatures) have been seen before

that string theory is already physically falsified. As soon as

accelerators see 12 dimensions and spacetime is still described by

higher-order corrected gravity the theory is physically falsified

(since no sugra possible in d>11). If no signs that higher-order

corrected sugra applies before that have been seen the theory is

already physically falsified.

But I also think that this dimension-focused question is too narrow.

Long before accelerators (or any other experiment) can probe 10 or 11

dimensions all kinds of generic stringy signatures need to be

observed, as discussed in the papers that I had cited.

Apr 21, 2003, 6:21:12 PM4/21/03

to

iainm...@yahoo.com (Iain) writes:

> All very interesting speculation. However, the Anthropic Principle

> does make a prediction - given that we are observers, and that we

> exist when and where we do, and given no information as to whether

> observers exist elsewhere, then observers *will* tend to be located

> close to a star.

I'm sorry, but I don't consider "we're here, because we're here,

because we're here, because we're here" to be a very interesting

or useful "prediction;" in fact, I don't consider it to be a

"prediction" in any meaningful sense of the word --- at best,

it is a particularly banal and useless tautology.

> All the Anthropic Principle says is that observers are typically found

> in conditions no more unusual than those required for their existence.

> It doesn't say much, true, but it does indicate where to place your bet.

You are confusing the "Anthropic Principle" with the "Generalized

Copernican Principle:" The former is the belief that the Universe must

have exactly the properties it does or we would not be here to observe it;

the latter is the belief that we must live in the most mediocre of all

possible worlds. The only thing these two beliefs share is that they are

both tautological in nature and both have ZERO predictive power.

Apr 22, 2003, 4:04:43 PM4/22/03

to

Urs Schreiber wrote:

>Since this applies to M-theory only, I propose the following summary of

>this discussion:

>

>1) Perturbative string theory exists and makes predictions.

Sorry, but I can't agree with either of these statements.

The string theory perturbation series is well-defined to

two loops and may or may not have finite, well-defined

terms for higher loops. Even if the terms are finite,

the series is almost certainly a divergent series. So

I still have a problem with the statement that the theory

"exists", or string theorist's claims that they have a

consistent, finite theory of quantum gravity.

If you just define the theory by ignoring everything

after two loops, you do have a well-defined framework

in which to calculate things. Unfortunately it has lots

of problems including:

1. It's not unitary (since you are not including higher loops)

2. The theory has unbroken space-time supersymmetry

3. Before you can calculate anything you have to

choose a background geometry. There are an infinite

number of ways to do this.

So, the only thing that I see that really exists is

a calculational scheme that is not really consistent

and predicts just about nothing. One of the few

things you could possibly argue is a prediction

(space-time supersymmetry) disagrees with

experiment.

>2) M-theory hardly exists and hence hardly makes predictions.

>

>3) Many popular overly enthusiastic claims about string/M-theory are

>problematic.

Can't disagree with either or these. But when will leading figures

in the string/M-theory community stop making these problematic

claims?

Peter

Apr 23, 2003, 2:06:50 AM4/23/03

to sci-physic...@moderators.isc.org

Urs Schreiber wrote:

>Peter Woit schrieb:

>

>

>

>>String/M-theory doesn't just not uniquely predict the standard

>>model. It doesn't predict anything.

>>

>>

>As soon as

>accelerators see a total of 12 or more spacetime dimensions (yes, that's

>not likely to happen anywhere soon, but that's how it goes with quantum

>gravity) string theory is physically falsified.

>

>

I'm still waiting to hear a single prediction from string theorists

about how the

physical world will behave at any energy scale. A prediction for what won't

happen isn't at all the same thing.

An example: suppose I announce that I have a wonderful theory of the

universe

that says that it is built out of golf balls. It might be true that if

tennis

balls start coming out of the interaction region at the LHC, that would be

evidence towards falsifying my theory. But calling "no tennis

balls at the LHC" a prediction of the theory is stretching the notion

of what a scientific prediction is past the breaking point.

Apr 25, 2003, 2:14:10 AM4/25/03

to sci-physic...@moderators.isc.org

Peter Woit <wo...@cpw.math.columbia.edu> wrote in message news:

> I'm still waiting to hear a single prediction from string theorists

> about how the

> physical world will behave at any energy scale.

Since you are a mathematician perhaps you have not read about the

history of physics - I don't know if you have or not. Physicists did

not study the theory of black holes before the advent of GTR, however,

they did study various quantum phenomena before the advent of the

full-fledged quantum theory in the 1920s.

E. Witten once said something ala string theory is like 21st century

mathematics that fell into the 20th century. Physicists are not trying

to be rigorous logicians and if they happen to get

mathematical-physical insights before getting any clear experimental

hints then that is what they have to work with.

You can be assured that there is no good current reason for rejecting

either string theory or loop QG because if there was then the

discoverer would not keep it a secret. Also, you should probably

ignore anthropic speculations and the like in string theory because

such ideas are most likely premature and probably meaningless if they

have no a priori basis.

You keep wanting to emphasize the importance of Dirac operators but

these are also used in string theory, e.g. Pierre Ramond has written

about the role of the algebraic Kostant-Dirac operator in SUGRA.

In summary, ugly theories would only be good if they were actually the

truth but it is too early to give up on string theory, especially

since string theory is still smarter than you, me or any other

individual because no one person has produced as many new ideas in

math or (potentially) physics as string theory has.

Apr 25, 2003, 8:10:56 PM4/25/03

to sci-physic...@moderators.isc.org

zir...@hotmail.com (zirkus) wrote in message news:<8c7d34cb.03042...@posting.google.com>...

> Peter Woit <wo...@cpw.math.columbia.edu> wrote in message news:

>

> > I'm still waiting to hear a single prediction from string theorists

> > about how the

> > physical world will behave at any energy scale.

>

> Since you are a mathematician perhaps you have not read about the

> history of physics - I don't know if you have or not. Physicists did

> not study the theory of black holes before the advent of GTR, however,

> they did study various quantum phenomena before the advent of the

> full-fledged quantum theory in the 1920s.

>

In 1784, the English geologist John Michell realized that it would be

theoretically possible for gravity to be so overwhelmingly strong that

nothing, not even light could escape. So yeah, theory of black holes

existed before the advent of GTR.

Jeffery Winkler

Apr 29, 2003, 4:33:56 AM4/29/03