251 internautes sur 266 ont trouvé ce commentaire utile
A slick sales job with a large side of information
5 janvier 2006
Par
Royce E. Buehler
- Publié sur Amazon.com

Format: Relié

Unlike the physicists who wrote the first two reviews, I don't know much 'bout string theory. Which is why I turn to books like this, or Greene's _The Elegant Universe_. Let me try to explain what this book is trying to do, and how, for one proverbial intelligent layman, it stacks up.

Susskind is a man with a mission. What he's describing here is not settled science, but his own view of the direction fundamental physics should be trying to go. In order to describe that properly, of course, he has to explain a good deal of settled physics along the way. He does this engagingly and fairly clearly, though he doesn't have the truly remarkable expository gifts of Brian Greene, and I strongly recommend that anyone who wants to tackle this book should read _Elegant Universe_ first.

The book has two tightly intertwined main theses. The first has to do with the Anthropic Principle: the observation that a large number of physical constants are required to fall within a surprisingly narrow range of values, in order for the apparatus of biology ever to appear. Slight tweaks to any of them would make galaxies, stars, atoms, chemical elements heavier than helium, to say nothing of carbon based life forms, impossible. Susskind's thesis here is that the AP is neither, as many theists would like to claim, evidence for a Designer who tailored the universe to make us possible; nor, as secular physicists would like to claim, an uninteresting tautology requiring no explanation. Rather, its explanation is to be found in the last decades' developments in string theory.

His second thesis is that these developments - especially the way in which string theory, which was originally hoped could prove mathematically that the various physical constants could have only one uniquely determined set of values, turned out to be a family of five, then many, then mind bogglingly many, distinct theories - are not the intellectual catastrophe some have felt them to be. Rather they are an argument in favor of the truth of string theory, because the innumerable variations in the laws of physics permitted by the various string theories provides a naturalistic explanation for the Anthropic Principle. To wit: cosmic inflation creates innumerable new universes all the time, each with its own set of physical constants, and it is not surprising that some of them should have laws (and in particular a value for Einstein's cosmological constant, which is more extremely constrained than any of the others) which permit life to arise. The collection of all these possible universes, by analogy with the "fitness landscapes" of evolutionary theory in biology, is what Susskind designates as "the cosmic landscape" of the title.

There are a lot of problems with this point of view. Susskind considers them, and argues enthusiastically, subtly and fairly that none of them is a show stopper. In the end, I felt he failed to close the sale. Until someone solves what he calls the "measure problem", the whole scheme is dead in the water. Further, we are never given a positive reason to believe in the truth of string theory, other than the fact that no other consistent theory unifying gravity and quantum mechanics has surfaced yet.

In his final chapter, Susskind tries to summarize the disparate attitudes of a dozen major living theorists toward this emerging Landscape picture. The two most telling criticisms come from physicist David Gross. The book gives them a pretty fair hearing , and doesn't claim to dispose of them.First, if we adopt the idea that the physical constants have randomly created values, the enterprise of trying to explain why they have the particular values they do comes to a dead halt - perhaps prematurely. "Quantum fluctuations did it" puts the kibosh on further inquiry as surely as "God did it" would. Second, we don't really know how wide a range of physical constants could produce life and intelligence in *some* form.

Those who are looking for a primer on string theory, or on the latest truths that scientists have learned and agreed on, won't find it here. But if you are interested in the Anthropic Principle, or in the ferment of controversies at the edge of the presently knowable, you won't have to agree with Susskind to take delight, as I did, in colorfully articulated, intriguing, and frequently illuminating read.

59 internautes sur 68 ont trouvé ce commentaire utile
A very interesting book about the properties of reality
25 décembre 2005
Par
Jill Malter
- Publié sur Amazon.com

Format: Relié

Is this book just what I wanted? Well, what I think I really wanted was for Einstein to return to us today and write a book on the philosophy of modern physics based on today's understanding of things. Yes, that would have been just great! But Einstein is dead. Luckily, of course, there are some excellent physicists around, such as the author of this book.

This book, written by an eminent String Theorist, has some fine explanations for the layman of some topics in modern physics, including String Theory. But the most interesting part is advertised in the title, namely the nature of the cosmic landscape.

The cosmic landscape refers to the mathematical space which has as its elements the values of the "fields" that constitute the physical laws and constants which apply to a particular "universe" (with a small u) or "pocket universe" if one prefers that term. The idea is that there may be many possible sets of physical laws and constants. The more we discover about physics, the more it seems that there are plenty of possible universes. But do they really exist? That is, is the landscape populated by more than our known universe? Is it heavily populated? The author argues that it is. And that certainly makes sense to me.

We're told about the anthropic principle. At its simplest, this principle merely states that we have to live in a universe that permits intelligent life. That's not very profound. But this principle also suggests that there is indeed a landscape of possible universes, and it encourages us to verify that only a very small fraction of them would permit the kind of complexity required for intelligent life. And in fact, Susskind gives us a good example of this. It turns out that the Cosmological Constant (which causes a universal repulsive force, sort of the opposite of gravity) we observe is about 120 orders of magnitude less than an unsophisticated theory might predict. That leaves us with a very tiny number to explain, a number which is fundamental to the laws of our universe. Well, sure enough, there appears to be an anthropic reason. Susskind tells us that Steven Weinberg calculated that a constant even ten times bigger would result in enough repulsion so that the clumpiness of the early universe would have been reduced so much that no galaxies, stars, or planets would have formed. Similarly, a large cosmological constant with the opposite sign would have caused our early universe to collapse.

Given that there may be a large multiplicity of actual universes, how do they originate? Since we appear to be in a rather unusual universe, it is tempting to hypothesize that there's some sort of selection principle at work. But what? We get into questions about a possible evolution of universes. But this could be quite different than evolution in biology. Are there incremental changes between generations of universes, and if so, how? Susskind doubts that changes are incremental. Are there universes that produce large numbers of other universes, given that there are no obvious selection benefits to having fewer offspring? As Susskind says, we see no mechanism to cause competition for resources. Are we in a produced universe, a producing universe, or both? If our universe is a producer, do miscellaneous black holes each produce new universes? Well, according to Susskind, they don't: the production mechanism he suspects is dominant is the cloning of space due to the metastability of the vacuum.

I think it helps to remind ourselves that if there is a large system of pretty much random entitites, that system will be dominated by the biggest ones. If time exists, that system will soon be dominated by not just the biggest, but the most stable and longest-lived ones. And those objects that produce, get produced, or are reproducing (or even better, reproducing in a manner than permits improvements) have a huge advantage. That applies to the Earth's biosphere, and presumably it applies to Reality as a whole.

Susskind discusses the "many-worlds" interpretation of quantum mechanics. In this version, when there is a random choice of polarizations for a particle (or a random choice of which slit a photon traveled through), both possibilities actually occur, but in different aspects of reality! That works mathematically, although it does produce a truly huge number of alternate realities. Susskind does say that this is a little like the huge number of alternate realities of the cosmic landscape. But there is a difference which he admits. The many-worlds alternatives all have the same physical laws, while the cosmic landscape does not.

The book includes a very good discussion of the question about possible information loss in a black hole. Stephen Hawking used to argue that such an information loss actually occurs. Susskind and `t Hooft argued that information has to be conserved, and that it must somehow be stored outside the black hole, and Hawking now has conceded this point. And we also learn about black hole complimentarity, which helps explain all this by making it clear that information has no definite location in space.

I truly enjoyed this book, and I strongly recommend it.

21 internautes sur 22 ont trouvé ce commentaire utile
Bold speculation, but refuted as science
10 juillet 2011
Par
dcleve
- Publié sur Amazon.com

Format: Relié
Achat vérifié

This is an excellent book, by a clear writer, who has big ideas, and expresses them well. I recommend reading this book, then reading a critic of String theory, Lee Smolin's The Trouble With Physics.

Leonard Susskind is the original inventor of String Theory over 25 years ago, and has now merged its latest permutations with Inflationary cosmology to produce a theory of everything. Susskind throughout the book talks about how he spends much of his time trying to figure out how to explain esoteric ideas in physics to laymen. I have an undergraduate degree in physics, and have read perhaps a dozen books on cosmology, and was able to follow this book fairly well. I am a "layman" compared to Susskind, but have a much stronger astrophysics background than most "laymen", so take warning. Understanding this book is not as easy as falling off a log.

His starting point is that Fine Tuning is a legitimate argument - that our universe is fine tuned to create life to a bizarre degree. I will not present his rationale for Fine Tuning. He for many years rejected the Fine Tuning claims, but became convinced himself when the Cosmological Constant was shown to be very small but positive. Since the possible range of the CC is huge, and anything but zero or very close to zero values will lead to very short-lived universes, or ones with no matter concentrations, many physicists assumed that the CC was somehow forced to be zero, by unknown physics laws. That it is not was startling to astrophysicists, and for Susskind this was the last straw to support a Fine Tuning argument.

There are two ways we know of to get Fine Tuning, design or evolutionary selection. Selection requires multiple random options to select between, and because he rejects Design, this is where Susskind goes. He proposes the Anthropic Principle, in which virtually infinite universes are created, and only the very, very, very few which are habitable are ever observed by sentient life.

To take you through his reasoning requires a peculiar journey through speculative physics. Inflation is what happens when the vacuum fluctuations of virtual particles in empty space sum up to a positive value. This positive term ends up producing a repulsion force against anything which has mass, meanwhile the virtual particles themselves DO have a mass, and the repulsion exceeds their gravitational attraction - which causes space to grow. The Cosmological Constant is the bulk term which describes the speed at which growth occurs. Since this new space is also filled with the same virtual particle field, mass is basically being created from nothing. The current Big Bang model is based on an initial high value of the CC, and then a drop down to our current very, very, very small but positive CC.

String theory has an explanation for the CC, as well as for everything else. String theory started with an attempt to describe the behavior of quarks and other similar elementary particles. Some behave like elastic strings with masses at each end, and others like three strings linked like a Y, each with a mass at the end. This part of the theory is very solid. It was then extended to all the rest of physics. When one tries to explain the behavior of all the other elementary particles and forces using a similar model, a model CAN describe approximately our suite of particles and forces, but it requires 11 dimensions plus time (8 of these dimensions are rolled up rather than extended, which is why our universe looks 3-dimentionsla rather than 11-dimensional). Currently none of the models produced quite work out, so String Theorists have kept coming up with variations on the equations which can add features which might let them actually match our universe. As a minimum, they add things like shape, twist, flux, branes (string localizers) of various dimensions, shape for the branes, and singularities to the equations for each dimension. This produces a near infinite number of possible "String Theories", and a near infinite number of variables to dial to fit our world. Within the theory, the CC, fundamental forces, and elementary particles all could be different with a different set of dimensions, shapes, fluxes, branes, etc.

One way to bring order to this is that not all possible combinations are stable. Basically, if there is an adjacent state which has a lower energy level (translates to a lower CC), frictional effects will lead a possible universe to drop to the lower state. What this means is that only local energy minima (valleys in a rough surface) are stable, and any universe that exists for any length of time must be in one of these en43ergy minima valleys.

Susskind adopts the Eternal Inflation hypothesis of Alan Guth and Andrei Linde, and assumes that there is some initial stable high CC state. Quantum effects cause fluctuations in these String Theory properties, and if they lead to a lower energy state than the locally stable one, this fluxuation acts as a crystal seed for the spreading transition of the high CC region to the lower one. BUT, since the high CC region is inflating faster than the new lower one, the transition to the new lower one is slower than the growth of the old higher CC space, and the high CC universe never goes away, it just has infinite bubbles of lower CC inside it. Each of these likewise can spontaneously create new lower CC regions within them, each of which will continue creating infinite space a well.

This view of the world - a inflating field of extremely high energy and high CC, with an infinite number of mini universes inside it that have bubbled out, is what Susskind calls his Cosmic Landscape. He proposes that our universe is one of these bubbles. To match the history of our universe, our bubble would have had to initially form near a trough leading to our valley, because inflation was initially high for a while (moving down a fairly level trough), then quickly dropped to our current rate.

String Theory, and this Infinite Multiverse have been criticized as non-scientific. A recently published critical book on string theory is titled Not Even Wrong which is the ultimate dismissal in science. A claim which is untestable in principle is worse than wrong, it is not science, and useless. Susskind is defensive about this, and challenges what the definition of science is. He asserts that science is whatever scientists are doing - and in his world scientists engage in extensive speculative mathematics, with virtually no possibility of experimental tests. When people try to redefine science because what they are doing does not fit the definition is generally considered a fairly definitive failing, and was cited in the Intelligent Design court case as definitive proof that Intelligent Design is not science. It is not good news for Susskind that he is sharing the tactics of the ID movement.

Scientific ideas to go through three steps: speculation, hypothesis, and theory. I don't think String Theory has quite mastered the hypothesis category yet, and the Landscape and Multiverse are still only speculations. A very legitimate critique is that the concepts just bring in too many free variables. Each one is another assumption, and anyone who has ever tried to fit a curve to data knows that once your free variables equal or exceed your data, then you have ZERO confidence in the predictive ability of the curve-fit.

[Unfortunately only physical scientists and engineers ever fit anything other than straight lines to data. If anyone is interested, try putting three points not in a line on a piece of paper. A straight line will at best approximate them. But one can hit all the points with a wavy curve of almost any shape. But since a curve which comes in and goes off the top of the paper, and hits all the points fits the data as well as one that comes in off the bottom or either edge, there is no confidence in the predictive power of these curves. With the straight line, you have some confidence, and it is related to the number of data points minus the constants used to create the curve (2 for a line), divided by the number of data points. So the near infinite variables of string theory vs. only 100 or so fundamental constants and properties of our universe that the theory deals with gives zero confidence in the predictive power of that theory.]

For my own interest, I have tried to spell out the major assumptions of Susskind in his Landscape fusion of String Theory and Inflation.

1-8 There are 8 additional dimensions to space.

9 The properties of everything are dependent on features of these 11 dimensions,

10 The features include whether a dimension unrolls

11 Also the shape of the dimension (number of holes)

12 Also number of twists

13 Also the flux value (integer) in each hole

14 Also number of 1-D branes (termination surfaces for strings)

15 Also number of 2-D branes

16 Also number of 3-D branes

17 Also number of 4-D branes

18 Also number of 5-D branes (I don't think any go higher than 5 now, not sure)

19 Whether any branes are negative (mentioned but not explained by Susskind)

20 Shape of the branes

21 Twist properties of the branes

22 Number of singularities (different from black holes)

23 TBD other mathematical structures

24 Specific string properties which create the elementary particles and fundamental forces (not sure if these are set by all of the above or not)

25 All of the above define the vacuum energy, and thus the cosmological constant (CC)

26 A high CC creates an inflating universe

27 Inflating universes can create mass and energy out of nothing, but it is all in balance, since gravitational energy is negative and exactly balances all the positive energy.

28 Fluxuations in the vacuum energy create transitions to lower CC regions within the inflating space

29 The "friction" transitioning from a higher to lower CC is released in the form of an elementary particle plasma (read hot Big Bang)

30 There is a texture to the Landscape of vacuum energy states, and new bubbles of a CC will transition down the slope of this landscape until they reach a local minima, or valley.

31 Only lower CC states can appear, not higher (not sure why)

32 There is a continuing cascade to a zero CC level (string theory models can easily create negative CC, so not sure why things stop at zero, negative CC would be lower energy than zero, so I think the model should not stop. This is a problem for the Landscape, because if high negative CCs are the innate end point of the Landscape, all the universes will eventually just puff out)

33 Lot and lots of local minima exist (close to infinite, I think it depend on when you give up speculation on new arbitrary structures in the String Theory model)

34 All these local minima are actually reachable from the particular high CC local minima of the Eternal Inflation field. (this assumption is one where the non-science accusation may be valid. Most of the minima will be vary far in properties from the starting condition, and the probability of a quantum fluxuation or tunneling effect reaching a distant state rather than the closest alternate state is vanishingly small. But by postulating an infinite multiverse of Eternal Inflation, it would not matter if one state had 200 orders of magnitude greater probability of being reached - all possible states will be populated with infinite numbers of bubble universes.)

35 Anthropic Principle - life is rare and will only arise in an exotic very, very, few local minima which can support complex chemistry, and have long term stability. That is why our universe is exotic and unusually stable.

36 Our universe had a CC that started in a nearly flat-bottomed trough that lead to our steep-sided valley

As you can see, at just the major assumption level, this speculation/hypothesis is quite a doozy.

It is science however, and there are a number of predictions made by these theories. The confirmed ones include:

* Flatness of the universe

* Homogeneity of the universe

* Size of galactic and galactic cluster mass concentrations

The unconfirmed ones include:

* At the very largest mass structure level, we will see the effect of a higher CC as the universe settled into the bottom of the trough

* A string theory model corresponds to our universe

* A series of string theory models correspond to the trough leading to our current valley

* Our universe has an edge with different properties, since it is continually transitioning high CC space around us into our low CC space (and we may be able to see a property variation, depending on how close we are to that edge. Note this is my own conclusion, not Susskind's)

* Portions of our universe may or have already spontaneously dropped to a lower CC level. These regions would grow within our universe, and be observable.

* Gravitons have zero mass, and behave like closed loop strings rather than strings with ends.

There are also two predictions which currently are falsified:

* Our CC is calculable and small based on string theory (the math sums up to infinity - in order to get a finite CC they throw out all numbers larger than a certain arbitrary value)

* All the particles for which no rotational inertia has been observed to date (electrons, neutrinos, photons), actually have rotational inertia

Since the confirmations to date all are just part of the Big Bang with Inflation model, and really have nothing to do with either Eternal Inflation, String Theory, or the Landscape, all of these ideas are currently unsupported by any observations. Having two contradicting observations is also a problem - but a common one for hypotheses as they work out their bugs.

I consider this hypothesis, however, to be the most coherent and complete materialist hypothesis I have ever seen for the formation of the universe. But please note what it includes:

* by my count 36 major assumptions - most of which are not confirmable in any way

* Another substrate to the universe other than the material, which is more fundamental, and from which the material emerged.

* Assumptions with infinite properties

* No explanation for where the original high CC Eternal Inflating space came from

Compare these features with the critiques of religious theories of the origin of the universe. The ill-defined infinite properties of God are criticized, as is the inability to explain God's origin, the complexity of assuming another substrate to the universe, and lots and lots of unconfirmable assumptions.

I really liked Susskind's book. He is a clear writer, and both a clear thinker and a very big thinker. I suspect Susskind has gone wrong early in String Theory, and strings are not really applicable to electrons, neutrinos, and photons, based on their not having rotational inertia. The reason he thinks they must be strings is that otherwise the interactions between these particles and the string-based particles run into a summing-to-infinity problem. Since he is just ignoring a similar problem in calculating the CC, I didn't see why one was a convincing argument to him that that everything was strings, and the other summing to infinity problem is dismissible. Once this summing problem is readdressed, and an answer to it found, I suspect a simpler partial string and partial particle theory will replace String Theory, and I don't know if this Landscape concept would survive.

Lee Smolin's The Trouble With Physics The Rise of String Theory, the Fall of a Science, and What Comes Next, provides a useful set of objections to counter Susskind's optimism over his Landscape theory. Smolin points out three important facts that Susskind left out. The first is that before String Theory, there were about a dozen attempts to unify quantum gravity and general relativity -- all of them were mathematically consistent, but every one made predictions that were tested and falsified. Susskind is advocating that Physics be judged by mathematical consistency rather than experiment, but if we had not tested previosly, we would now be using one of those falsified theories rather than String Theory.

The second point is that the hidden dimensions of String Theory are unstable. Quantum fluxuations cause them to spontaneously collapse, or inflate to infinity like our three conventional dimensions. The ONLY way String Theory can lock them down so they are stable AND hidden is if they have a more complex shape than just being rolled into a tube. String Theory has a prediction -- hidden dimensions will unroll -- that is disproven by observation. What Susskind did not describe was the special case for these extra dimensions, which is the only way they could be stable, is that each must have at least two holes in it, and that each hole must have one or several charged Branes wrapped around the surface (basically a pretzel shape is the least complex stable shape). Only then will they stay stable and hidden. This is a bizarre kluge to the theory that makes it effectively absurd.

The third point is that String Theory for most of its history predicted a negative or zero cosmological constant. Theorists worked with the zero CC versions until a small positive CC was discovered. Then they found a way to further kluge the theory. That is what the negative branes were invented for. Branes are an imagined surface or shape (can be from 1 to 5 D) that the ends of strings must stay connected to. They were first imagined in 2-D as membranes, then expanded into more Ds, and called D-branes, or just Branes. A negative brane was never explained by Susskind, and I suspect it is just a mathematical artifact to somehow bring the vacuum energy to a positive value. The addition of them to String theory is another patch to deal with a contrary observation.

Smolin points out that good theories are usually quickly confirmed by observations, and they suggest all sorts of new and surprising things which then advance physics in other areas. The theories that have been discarded in physics are the ones which try to be kluged up with special cases to deal with one embarrassing observation after another. This looks to me to be the case with String Theory. Smolin started his book with 5 big questions that faced physics 30 years ago, when String Theory first became the dominant idea in theoretical physics, and that NONE of these have been solved in the last 30 years. He thinks it is due to physics departing from testability, and embracing a theory which is infinitely klugable and non-falsifiable -- and as a result the experimentalists have been starved of viable theories to test to advance our understanding of these issues.

Smolin is as good and clear a writer as Susskind, and he has a better case, since he does not try to hide embarrassing facts, nor attack the principle of experimental confirmability at the core of science. Reading the two together, Smolin blows Susskind out of the water -- String Theory is a dead end. Despite this judgement, I still reccommend the book as very enjoyable and thought provoking.