Monday, May 01, 2006

Heisenberg + LSD = Stephen Hawking's Flexiverse

The article was in New Scientist this month, which you need a subscription to for lack of a better way to link to the article I've just posted it right here. (If anyone knows how to put the article behind a link...feel free.)

He's saying he can walk, talk, and kick your ass at Horse (all while not drooling on himself)...depending on how you look at it:

Exploring Stephen Hawking's Flexiverse
20 April 2006
Amanda Gefter

Here's how to build a universe. Step one: start at the beginning of time. Step two: apply the laws of physics. Step three: sit back and watch the universe evolve. Step four: cross your fingers and hope that it comes out looking something like the one we live in.

That's the basic prescription for cosmology, the one physicists use to decipher the history of the universe. But according to Stephen Hawking of the University of Cambridge and Thomas Hertog of the European Organization for Nuclear Research (CERN), the steps are all backward. According to these physicists, there is no history of the universe. There is no immutable past, no 13.7 billion years of evolution for cosmologists to retrace. Instead, there are many possible histories, and the universe has lived them all. And if that's not strange enough, you and I get to play a role in determining the universe's history. Like a reverse choose-your-own-adventure story, we, the observers, can choose the past.

This bizarre state of affairs has its roots in Hawking's work in the 1970s. Early in his career, Hawking, along with physicist Roger Penrose, proved a theorem showing that our expanding universe must have emerged from a singularity - a place where gravity becomes so strong that space and time are curved beyond recognition. In this situation, general relativity - our best description of how space, time and matter interact - no longer applied.

So what rules did apply? Hawking and Hertog suggest that the universe was so small at this time that quantum effects must have been important. We don't yet have a quantum theory of gravity, so we can't be sure exactly what the rules were, but the principle still stands, they say. "The real lesson of these so-called singularity theorems is that the origin of the universe is a quantum event," Hertog claims. And that, of course, opens the whole universe up to some very strange phenomena.

The famous double-slit experiment highlights the bizarre reality of how a universe born in quantum mode might behave. In the experiment, a screen with two open slits faces a sheet of photographic film. When light is shone through the slits the film registers where it lands. If the light goes through both slits the film shows an "interference pattern" of light and dark bands. Such a pattern is typically produced by interfering waves - one from each slit. What's spooky is that even when a lone photon is fired at the slits it still creates a pattern of light and dark bands - as if it were two waves.

In 1983, Hawking and James Hartle of the University of California at Santa Barbara, took up this picture and applied it to the evolution of the whole universe. They did that using the "sum over histories" interpretation of quantum theory, first set out by the late Richard Feynman. Feynman suggested that the way to interpret quantum phenomena such as the double-slit experiment was to assume that when a particle travels from point A to point B, it doesn't simply take one path - it takes every possible path simultaneously; the photon travels through both slits at the same time and interferes with itself, for example.

In this scheme, when a photon travels from a lamp to your eye it moves in a straight line, but it also dances about in twists and swirls, travels to Jupiter and back, and ricochets off the Great Wall of China. The obvious question, then, is why do we see only ever see one path, straight and simple? Feynman's answer was, because all the other paths cancel each other out. In the sum-over-histories interpretation, each path can be mapped out as a wave. Each wave has a different phase (effectively a starting time), and all the waves added together create an "interference pattern", building upon one another where their phases align and cancelling each other out where their phases are mismatched. The sum of all the waves is one single wave, which describes the path we observe.

Applied to the universe, this idea has an obvious implication. Just as a particle travelling from point A to point B takes every possible path in between, so too must the history of the universe. In one history, the Earth never formed. In another, Al Gore is president. And in yet another, Elvis is still - well, you get the idea. "The universe doesn't have a single history, but every possible history, each with its own probability," Hertog says.

But there is a twist: the history that we see depends on the experimental setup. In the double-slit experiment, it has been shown time and again that if we use a photon detector to find which of the two slits the photon went through, it no longer creates an interference pattern, just a single spot on the film. In other words, the way you look at the photon changes the nature of its journey. The same thing happens in Hawking and Hertog's universe: our observations of the cosmos today are determining the outcome - in this case, the entire history of the universe. A measurement made in the present is deciding what happened 13.7 billion years ago; by looking out at the universe, we assign ourselves a particular, concrete history.

If true, this is no mere curiosity; Hawking and Hertog have tossed the notion of a unique, observer-independent cosmology out the window and thrown the sacred laws of cause and effect into question. But they're not exactly being violated, Hawking says - it's all to do with perspective. If we could stand outside the world, we would be able to see the present affecting the past, as when an observer affects a photon's path through the universe. From inside the universe, though - from the only place we can possibly be - no observer sees causality violated. What we observe in the present, the "final" state, is one entire, causally consistent history or another: from within any given history, cause and effect proceed in the usual manner.

"Observations of final states determine different histories of the universe," says Hawking. "A worm's-eye view from inside the universe would have the normal causality. Backwards causality is an angel's-eye view from outside the universe."

So the idea is that to unravel the past, we must sum together all possible histories of the universe. What does that mean?

Hawking and Hertog equate the cosmic histories with how the geometry of the universe evolves in each possible case of going from point A (the beginning of time) to point B (now). To start with, this seems straightforward enough. We can specify the state of the universe at point B by making certain observations of the world around us - the universe has three large spatial dimensions, its geometry is close to flat, it is expanding, and so on.

What about point A, though? Mapping out the paths of a photon from a lamp to our eye is not too hard because we know the beginning point - the lamp - and the final point: our eye. We know nothing about the universe at the beginning of time, however. After all, that's what cosmology is supposed to tell us.

This is where the sum-over-histories interpretation comes into its own. The mathematics behind this approach to quantum theory contains an oddity: the answers only come out right when the calculation is done in imaginary time. That doesn't mean make-believe time, but rather a time dimension that is expressed using complex numbers. This is not an entirely esoteric idea: electrical engineers routinely use complex numbers, which are split into "real" and "imaginary" parts, to design electrical circuits. In the hands of cosmological engineers, imaginary numbers turn out to have profound consequences.

Hawking and Hartle's original work on the quantum properties of the cosmos suggested that imaginary time, which seemed like a mathematical curiosity in the sum-over-histories approach, held the answer to understanding the origin of the universe.

Add up the histories of the universe in imaginary time, and time is transformed into space. The result is that, when the universe was small enough to be governed by quantum mechanics, it had four spatial dimensions and no dimension of time: where time would usually come to an end at a singularity, a new dimension of space appears, and, poof! The singularity vanishes.

In terms of the universe's history, that means there is no point A. Like the surface of a sphere, the universe is finite but has no definable starting point, or "boundary". Hence the idea's name: the no-boundary proposal.

This has led Hawking to define a new kind of cosmology. The traditional approach, which Hawking calls "bottom-up" cosmology, tries to specify the initial state of the universe and work from there. This is doomed to fail, Hawking says, because we know nothing about the starting conditions. Instead, he suggests, we should use the no-boundary proposal to do "top-down" cosmology, where the only input into our models of the universe comes from what we observe now - together with the idea that our universe has no boundary in the past.

Improbable tuning

The result of this process, he says, solves a long-standing problem of cosmology: fine-tuning. Most cosmologists think, for example, that the universe went through an early burst of rapid expansion, or "inflation". There is some evidence to support the claim, but there's also a problem. Standard inflationary models require a very improbable initial state, one that must have "finely tuned" values that cause inflation to start, then stop in a certain way after a certain time: a complicated prescription whose only justification is to produce a flat universe without any strange topology, and so on - a universe like ours.

Such a prescriptive method makes hard and unsatisfying work of producing the universe we see today. While a cosmologist can put these values into the equations "by hand", it is not exactly a satisfactory way to develop our model of how the universe works. In the no-boundary theory, however, there simply is no defined initial state. "In the usual approach it is difficult to explain how inflation began," says Hawking. "But it occurs naturally in top-down with the no-boundary condition. It doesn't need fine tuning."

To do top-down cosmology, Hawking and Hertog first take a whole raft of possible histories, all of which would result in a universe with features familiar to us. "We then calculate the probability for other features of the universe, given the constraints," Hertog says. Specify a universe that is three-dimensional and flat, for instance, and you can have histories that involve inflation and histories that don't. "Top-down cosmology does not predict that all possible universes have to begin with a period of inflation, but that inflation occurs naturally within a certain subclass of universes," Hertog says. The process creates a probability for each scenario, and so Hertog can see which kind of history is most likely. "What we find is that the inflating histories generally have the largest probability."

In many ways, top-down cosmology is an unsettling idea. Usually, science demands that our observations come out as output - we certainly don't expect them to be the input. That, after all, denies us the chance to see if the theory matches up with observations. What's more, the sum over histories is formed by calculating the various probabilities for a universe like ours to arise out of literally nothing: that means we can never know anything for certain about how our universe got to be as it is.

We shouldn't be surprised, Hertog says: quantum theory has long shown us that it is impossible for us to know everything about the world around us. In "classical" physics, we can predict both the exact momentum and position of a particle at any time, but quantum mechanics doesn't allow it. No one suggests that quantum mechanics is wrong because of this, Hertog points out - and experiments have shown that it is not. What quantum theory has given us now, Hertog says, is some indication about the nature of inflation, where before we had none. "Before, we had no prediction at all - and indeed no notion of likeliness - on this issue."

For many, it remains a difficult argument to swallow. Science since Copernicus has aimed to model a universe in which we are mere by-products, but top-down cosmology turns that on its head, rendering the history of the universe a by-product of our observations. All in all, it is very like the "anthropic landscape" argument that is causing controversy among string theorists (see "Putting the you into universe").

Princeton University physicist Paul Steinhardt is certainly unimpressed by Hawking and Hertog's scheme. "It's kind of giving up on the problem," he says. "We've all been hoping to calculate things from first principles. Stephen doesn't think that's possible, but I'm not convinced of that. They might be right, but it's much too early to take this approach; it looks to me like throwing in the towel."

Stanford University's Andrei Linde is similarly unconvinced. There are a number of technical assumptions that make him sceptical. "I don't buy it," he says.

The past is out there.

The merits of Hawking and Hertog's new approach to cosmology might be decided by experiment. The theory predicts specific kinds of fluctuations in two cosmological phenomena: the cosmic microwave background radiation produced just after the big bang, and the spectrum of primordial gravitational waves. These fluctuations arise from applying the uncertainty principle of quantum mechanics to Hawking and Hertog's scheme: in this scenario, the universe's shape is never precisely determined, but is influenced by other histories with similar geometries.

If Hawking and Hertog are right, quantum uncertainty will manifest as slight differences from what standard inflationary theory predicts for the CMB. The top-down predictions only differ from the standard cosmological model at a level of precision that has not yet been reached in observations, however. The top-down signature in the gravitational wave spectrum should be easier to differentiate, but since we haven't yet detected any gravitational waves, we'll have to wait for that proof too.

For Hawking and Hertog, there's simply no doubt that top-down cosmology is the only answer. It's simple: if you can't know the initial state of the universe, you can't work forwards from the beginning: the top-down approach is the only one that works.

Hartle agrees. Hawking and Hertog's scheme may seem strange, but it is the only way forward because we are part of the experiment we are trying to observe. "It's a different viewpoint, but it's sort of inevitable," he says. "Colsmologists certainly should be paying attention to this work."

The trouble, of course, is that if they are right, we're involved in the making of that history. In that case, we have a new set of instructions for building a universe. Step one: look around you. Step two: find the set of all possible histories that end up as a universe like the one you see. Step three: add them together and create a history for yourself.

It seems there are two ideas...the past is defined by how you look at / measure it and that every variation of the universe has the same probability of existing. I'm a little lost on how they're putting those ideas together though. Either way, if he's right the immortalized face of the prophet who first told us we are all in fact God is hilarious.


island said...

This just gets funnier by the minute. I'm going to reference a section from Amanda's article from here, for my blog, science in crisis

Physics used to be causally accountable. Now it's a free-for-all for whomever can abuse infiity and uncertainty the best to create the most ludicrous reality that they can come up with.

Mike said...

That's pretty mind-blowing stuff. Also, utterly incomprehensible by me except on the most base and primitive level of human thought. When Hawking dies he'll probably just dissolve into thin air like Darth Vader.

But if this theory's right, that means in some other dimension I really did hit a home run at the state championship and go on to marry the head cheerleader, Jane Madsen. Boy, what gams. Now that's a universe.

Horse said...

awesome post, gibbo. and to that last sentence, all i have to say is, LOL!

maria said...

mike is that a references to that jim belushi movie? good show.

and i agree--i too merely comprehend on a base and primitive level. the article was about time travel right?

island said...

heh... it's a cosmological model that uses the same kind of interpretation of time-warping quantum weirdness as John Wheeler's Final Anthropic Principle to account for fine tuning.

Get rid of the infinities and uncertainty and they've got themselves a valid model of our universe, (minus the mechanism for it's evolution)... in my humble but supported opinion.

Slac said...

It seems there are two ideas...the past is defined by how you look at / measure it and that every variation of the universe has the same probability of existing.

Umm... didn't they say that some universes are more probable than others, namely the ones that result in inflation?

JasonGetsDown said...

Island, its easy to take an anti-intellectual approach and call it all melarky, but think about it. the double slit experiment shows us that things are indeed incredibly wierd at the quantum level. Particles do in fact travel multiple paths with varying probabilities until we observe them and the system collapses into one percieved reality. If the particles that make up the universe all have this behavior why shouldn't the universe itself behave this way?
If you aren't familiar with the double slit experiment I suggest watching this:
ignore the pseudo superhero.