What's the trouble with anthropic reasoning?

[Chronos]

I couldn't help but notice this interesting paper. I'm guessing some lively discussion will be emergent on the blogs.

http://arxiv.org/abs/astro-ph/0610330
What's the trouble with anthropic reasoning?
Authors: Roberto Trotta (Oxford University), Glenn D. Starkman (Oxford University & Case Western Reserve University)
Comments: 8 pages, no figures. Contribution to the proceedings of the conference "The Dark Side of the Universe", Madrid, June 2006

Selection effects in cosmology are often invoked to "explain" why some of the fundamental constant of Nature, and in particular the cosmological constant, take on the value they do in our Universe. We briefly review this probabilistic "anthropic reasoning" and we argue that different (equally plausible) ways of assigning probabilities to candidate universes lead to totally different anthropic predictions, presenting an explicit example based on the total number of possible observations observers can carry out. We conclude that in absence of a fundamental motivation for selecting one weighting scheme over another the anthropic principle cannot be used to explain the value of Lambda. .

 

[Kea]

Quote: We conclude that in absence of a fundamental motivation for selecting one weighting scheme over another, the anthropic principle cannot be used to explain the value of Lambda.

Who ever thought the anthropic principle could be used to explain any fundamental parameters?

 

[marcus]

Who ever thought the anthropic principle could be used to explain any fundamental parameters?

sometimes I'm a little slow to catch on, Kea
it sounds like a rhetorical question suggesting that NOBODY (or nobody except goofies) ever thought anthropix could explain parameterz.

I can make sense of that. If I understand you, the anthropic principle isn't a way to explain, but rather a way to DISTRACT people from trying to explain, and mainly it is a kind of DISCOURAGING NOISE that one makes to kind of demoralize people and get them to GIVE UP trying to explain.

AN ACCURATE PARAPHRASE OF THE ANTHROPIC PRINCIPLE
[Jeers]: hiss! boo! you incompetent morons, you pathetic losers! you will never explain those parameters. they are just however they are, you donkey's bottom! and if they weren't you wouldn't be here...In this light, Kea, I can sort of understand what you said. Let me know if I am missing some subtle point.

 

[Chronos]

I have always felt the anthropic principle is a useful to rule out the impossible, but otherwise has no predictive prowess. It reminds me of Occam's razor, which is strictly utilitarian: the simplest model that works is preferred. It has even less predictive power, but, is also useful. I have a hammer, a superb pounding tool. I have a hatchet, a superb chopping tool. But neither tool is very useful to a watchmaker.

 

[Kea]

Let me know if I am missing some subtle point...

Dear Marcus, no. There isn't anything subtle here (after all, the WMAP pictures are there for all to see for themselves)...and I must say I usually do enjoy your paraphrasing.

 

[Kalimaa23]

Anthropic reasoning is a wonderful, well develloped piece of scientific pursuit, if not for one tiny little flaw...

It's bollocks!

 

[selfAdjoint]

Anthropic reasoning is a wonderful, well develloped piece of scientific pursuit, if not for one tiny little flaw...

It's bollocks!

ROTFLOL! Thanks Dimitri!

 

[Kea]

Let's get this straight

Since a large number of people don't seem to have figured this out yet, allow me to emphasise the point:

a. The Anthropic Principle is garbage
b. There is no real Dark Energy because cosmology is quantum gravitational

Unfortunately, from CosmicVariance we have the comment

NASA and DOE have appointed a National Academies panel to recommend whether LISA (space-based gravity wave probe), CON-X (space-based X-ray telescopes to study black holes) or JDEM (joint dark energy mission) should be launched. NASA has the funds to do exactly *one* of these projects. The panel is just being formed now and will report a year from now.

They need a panel to figure out the bleeding obvious?

 

[Chronos]

I vote LISA as the percentage play for most likely to succeed. At least in principle, LISA has the most utility, IMO. I am curious, however, kea. Why do you think DE is not a 'player'?

 

[Kea]

LISA! What?! No way. The NASA site http://lisa.gsfc.nasa.gov/ says

LISA is a constellation of three spacecraft that uses laser interferometry to precisely measure distance changes between widely separated freely falling test masses housed in each spacecraft .

Now, I don't know much about the details of the design, but that says to me that they'll be looking for gravitational waves from distant sources, and pretty well nothing else. In other words, on what grounds do they actually expect to see anything? Do they have a working theory of QG?

Why do you think DE is not a 'player'?

To quote NASA again:

DE...calls for accurately measuring how that expansion rate is increasing with time.

Oh, really? Well, that would be nice if the expansion of 'space' (the stuff we wanted to get rid of, remember?) actually was increasing, which it isn't. A number of respectable cosmologists now agree with this statement and, more importantly, some unrespectable cosmologists also agree (eg. Riofrio at http://riofriospacetime.blogspot.com/ ...or myself). The best way to see this is as a varying-speed-of-light cosmology.

Now, if I'm counting right, that only leaves one project which can reasonably be expected to produce results.

 

[Severian]

I agree with Chronos. LISA is most interesting. Second most interesting is JDEM. And last on my list would be CON-X.

The sensitivity of LISA is in the right ballpark to see something. If they don't see anything, then that too will be an interesting result.

Irrespective of whether or not you believe the DE to be a cosmological constant or not, figuring out what is going on should be a top priority. After all, does anyone have a working theory for why the speed of light should change?

CON-X is also interesting for its secondary goals, but it's primary goal of investigating black holes seems less so.

 

[Kea]

The sensitivity of LISA is in the right ballpark to see something. If they don't see anything, then that too will be an interesting result.

Well, I agree that a LISA would be much better than no LISA. But if there's a good reason to expect a null result, it shouldn't get top priority. And there might be.

After all, does anyone have a working theory for why the speed of light should change?

Do you mean a working cosmology or a full theory of QG? If the former, you should probably take a look at Riofrio's WMAP graphs. If the latter...well, we're working on it.

 

[Severian]

Do you mean a working cosmology or a full theory of QG? If the former, you should probably take a look at Riofrio's WMAP graphs. If the latter...well, we're working on it.

How can you be working on an explanation for the apparent acceleration and not be interested in a Dark Energy experiment? It is clearly something we need to understand better. However, aesthetically, a cosmological constant is much much simpler and more pleasing than a varying speed of light.

Which plots are you referring to? I can't find any papers by him on SPIRES. Are you referring to the plot on his blog? ie. http://riofriospacetime.blogspot.com/2006/09/funny-stuff-from-wmap.html

He says:

This is one of the graphs that "proves" a Concorde cosmology. Look at the left side. The data points only follow the prediction line for angles less than 30 degrees. Starting at l = 3 the points depart significantly from the prediction. Presented here in logarithmic form, the departure does not seem obvious. However, most of the sky is greater than 30 degrees!

However, the data on the left side of the plot are pefectly consistant with the prediction. The end point (the one he claims they are hiding) is only 2 standard deviations away from the line. It appears Mr Riofrio does not understand statistics. (In actuality, I suspect their error estimates on the left hand measurements are a little low.) One should also be aware that the prediction has large errors for large angles.

Or were you referring to something else?

 

[Kea]

How can you be working on an explanation for the apparent acceleration and not be interested in a Dark Energy experiment?

Of course I'm interested in all the experiments. I'm just claiming that it's obvious which one should be given top priority. And what makes you think that CON-X can't also be viewed as a DE experiment?

A cosmological constant is much much simpler and more pleasing than a varying speed of light.

I've heard that one before.

Are you referring to the plot on his blog?

That would be her blog. And yes.

Or were you referring to something else?

I was referring to the conceptual basis behind this and other blog entries.

 

[francesca]

a. The Anthropic Principle is garbage
b. There is no real Dark Energy because cosmology is quantum gravitational

Kea rules!

 

[Dcase]

Alternative perspective

I have never really understood the Anthropic Principle.

I think that I understand the difference between an anthropic perspective and, for example, a neutrino perspective.

Anthropic perspective:
Earth is relatively solid [especially if in an ice age] with tools [helical auger?] required for tunneling [fingernails generally inadequate] taking a significant amount of measurable anthropic time.

Possible neutrino perspective:
Earth is relatively porous [even if in an ice age] with only a flight path [helical EM?] required for tunneling taking an insignificant amount of measurable anthropic time.

If Einstein is correct, time slows down for the neutrino relative to man.

QLG appears to use an anthropic perspective as do strings.
QLG uses an anthropic cubed Planck length for volume although volume is more likely curved as a sphere or ellipse [how many cubed stars are there?]
QLG wants to ignore negative volumes when volumes are either contained [such as a star] or uncontained [such as a supernova].

 

[Ontoplankton]

It's bollocks!

Since a large number of people don't seem to have figured this out yet, allow me to emphasise the point:

a. The Anthropic Principle is garbage

You're being unreasonable. "Anthropic reasoning" just means taking into account observational selection effects. Without anthropic reasoning, there is no reply to silly arguments like "Isn't it a marvellous coincidence that the planet we live on is suited to the evolution and survival of intelligent life?"

stuff to read

 

[selfAdjoint]

You're being unreasonable. "Anthropic reasoning" just means taking into account observational selection effects. Without anthropic reasoning, there is no reply to silly arguments like "Isn't it a marvellous coincidence that the planet we live on is suited to the evolution and survival of intelligent life?"

stuff to read

Anthropic reasoning at level you cite is quite different from the "Anthropic Principle" used by landscape physicists, although interested parties certainly work hard to confuse them. The physics Anthropic Principle is BUNK.

 

[Ontoplankton]

I've seen a lot of different things being called the "anthropic principle" (both before and during the current string controversy), so I think it would be helpful if people made sure to be very clear about exactly what they were calling BUNK. "Anthropic reasoning" is a less ambiguous term that refers to reasoning having to do with observer selection effects, and I'm not sure that's something you can avoid in your thinking.

 

[Kea]

"Anthropic reasoning" is a less ambiguous term that refers to reasoning having to do with observer selection effects, and I'm not sure that's something you can avoid in your thinking.

It is understood here, as selfAdjoint has pointed out, that we are referring to the landscapology application of a (not very well thought out) principle, and in particular the conclusion thereof that fundamental parameters of the SM may not be computable.

 

[Chronos]

The anthropic principle is a very useful tool for ruling out the impossible. Beyond that, it is useless, IMO.

 

[Ontoplankton]

I've seen a lot of different things being called the "anthropic principle" even in landscapology. If the anthropic principle just means the fundamental SM parameters may not be computable (because supposedly multiple possibilities are realized and we live in one of those in which you can live), how is that something you can declare false from your armchair? Again, it may be useful to look at the analogy with planets; some people in the middle ages thought they could explain the Earth's properties (like distance from the sun) from first principles, but we now think there's a huge number of planets, and they vary in their properties, and we happen to live on one of them that supports life. The universe may or may not be like that, but without more information, how are you so sure that it isn't?

 

[Severian]

The life-supporting planets argument is not comparable to the landscape argument simply because we can investigate other planets. So we can look to see if they support life and reason that we must be on one that does. If the Earth had turned out to be the only planet in the universe (or one of very few) then the anthropic reasoning wouldn't have worked.

If the landscape idea is true, then our horizon is inside one of 'bubbles' of a particular vacuum state. But since we can't see beyond our horizon this will always be speculation - we can never look at the other bubbles to see what their properties are (or even if they exist!).

Science is all about testability. Since we can never test the landscape notion of bubbles of different vacua, it is not a scientific notion - it is religion. (To be fair, I have heard Linde and Susskind say that there are testable predictions of the landscape, but I didn't understand how this could be - if someone could explain how this could come about, I would be happy to amend my view that the landscape is religion.)

 

[Kea]

If the anthropic principle just means the fundamental SM parameters may not be computable ... how is that something you can declare false from your armchair?

 

[Ontoplankton]

The life-supporting planets argument is not comparable to the landscape argument simply because we can investigate other planets.

Correct; to make the two comparable, we would have to go back to the time when we didn't have any data on planets outside the solar system. I claim that, given the knowledge we had then, it was not reasonable to dismiss the possibility of there being a huge number of planets with most of them probably being uninhabitable. After all, that's what turned out to be the case later.

If the Earth had turned out to be the only planet in the universe (or one of very few) then the anthropic reasoning wouldn't have worked.

But that's not the situation we're in with respect to the landscape; we don't have any hard data saying there are multiple universes, but nor do we have any hard data saying there's only one universe.

Science is all about testability. Since we can never test the landscape notion of bubbles of different vacua, it is not a scientific notion - it is religion.

First, just because you can't see the other bubbles doesn't mean the theory isn't testable; it may just be testable in an indirect way only. Sometimes the best explanation for the phenomena that we can see implies that there are also things we can't see. For example, we can't interact with anything beyond the observable universe (and maybe we will never be able to), but we have a well-confirmed theory according to which the universe beyond the observable universe looks just the same as it does here. No one believes the universe is a ball with the Earth in the center.

Second, just because something isn't testable doesn't mean it's a religion. The idea that there do not exist multiple bubbles is just as untestable as the idea that there do exist multiple bubbles; if there is no information that can help us decide either way (and it's not a given that this will always be the case), what we should do is suspend judgment, possibly forever. I don't see people here suspending judgment; I see them dismissing possibilities out of hand.

(Sometimes an idea is untestable because it's meaningless, and sometimes an idea seems untestable because its proponents "cheat" and count any possible evidence as evidence for (or at least not against) the idea (e.g. Freud); but sometimes an idea is just untestable because of physical limitations to our testing abilities even though it says something meaningful about our world if it's true. I think this is one of those cases.)

 

[Dcase]

Concinnity of Perspective

I learned a new word today:
“concinnity \kuhn-SIN-uh-tee\, noun:
1. Internal harmony or fitness in the adaptation of parts to a whole or to each other.
2. Studied elegance of design or arrangement -- used chiefly of literary style.
3. An instance of concinnity“.
http://dictionary.reference.com/wordoftheday/archive/2006/10/12.html

Consider the Strong Interaction:
“The strong interaction or strong force is today understood to represent the interactions between quarks and gluons as detailed by the theory of quantum chromodynamics (QCD). The strong force is the fundamental force mediated by gluons, acting upon quarks, antiquarks, and the gluons themselves.
“Although the strong force only acts upon elementary particles directly, the force is observed between hadrons as the nuclear force. As has been shown by many failed free quark searches, the elementary particles affected are unobservable directly. This phenomenon is called confinement, a theory which allows only hadrons to be seen.”
http://en.wikipedia.org/wiki/Strong_nuclear_force

Concinnity seems particularly appropriate to the discussion of perspective.

When the strong interaction is viewed from the quark perspective, then this interaction is ionic-like among oppositely and unequally charged up and down quarks almost in the same category as H2O.

No wonder U(1) x SU(2) x SU(3) tends to unify forces as though all can be translated into EM.

Can the same be done for gravity?

 

[Loren Booda]

Might the Anthropic principle in general be just a self-fulfilling prophesy? Can one say how would it differ in a universe without [other] observers? Is it any more than Darwinism in a space suit?