Facts against BB Theory

In article , Steve Willner
writes:

I strongly object to the premise of most of this discussion.

Presumably to the premise of one camp in this discussion; I'm on your
side. :-)

In
science, the word "predict" has nothing to do with foretelling the
future. Instead it means deriving the consequences of a
theory. That's what we mean, for example, when we say Kepler's Laws
_predict_ Tycho Brahe's observations, which were done prior to
Kepler's derivation of his laws. In at least one way it's better
when the data come before the theory; the danger of "confirmation
bias" is thereby avoided.

That hits the nail on the head.

The real question is not the time order of theory and observation but
rather how many adjustable parameters the theory has, how well or
badly its predictions match the data, and how extensive the data are.

Right.

I see there's a new simulation that goes quite a way to predicting
cosmological observations:
http://www.cfa.harvard.edu/news/2014-10
I haven't read the article yet (and had nothing to do with the
research itself). If the work really has the galaxy morphologies and
chemical abundances right, that will be a huge achievement.

Indeed. I think there is a serious misunderstanding on the part of
people outside the field about how difficult such simulations are. It
is not a mark against theory if a simulation performed in the 1970s on a
computer which was orders of magnitude more expensive but also had
orders of magnitude less processing power and memory than a typical
modern mobile phone didn't get all the details right (mainly due to lack
of temporal and spatial resolution). The CMB, however, is different:
here the theoretical predictions were ahead of the data. Nature is
independent of how good our computing power is and of when we discover
something (which might depend on politically motivated funding decisions
and so on).

On 5/14/2014 9:36 AM, Phillip Helbig---undress to reply wrote:
In , Steve Willner

I strongly object to the premise of most of this discussion.

Presumably to the premise of one camp in this discussion; I'm on your
side. :-)

Steve seems to think "most of" the posters here are
in the other camp (or could it be a minority, but a
rather vocal one?)

In
science, the word "predict" has nothing to do with foretelling the
future. Instead it means deriving the consequences of a
theory.

Although "prediction" can indeed have this special
meaning of "observational consequence" or "observable
implication", I still think many scientists when they
use "predicted" simply mean predicted.

.. That's what we mean, for example, when we say Kepler's Laws
_predict_ Tycho Brahe's observations,

Then you use the *present tense*, which is strange to
begin with because the observations have been made
already and therefore cannot be predicted any more.
So now you must be using the "observable implication"
meaning.

...
... The CMB, however, is different:
here the theoretical predictions were ahead of the data.

That way of expressing it makes the meaning explicit
(of "prediction", I mean), but still, if you omit the
added clarification and just use the *past tense*
"theory predicted the CMB" or
"the CMB was predicted by theoretical models"
then most scientists would immediately say that isn't
true! Which proves that, also for them, "predicted"
has its ordinary meaning.

--
Jos

On Tuesday, May 13, 2014 12:05:15 PM UTC-4, Steve Willner wrote:
I strongly object to the premise of most of this discussion. In
science, the word "predict" has nothing to do with foretelling the
future. Instead it means deriving the consequences of a
theory. That's what we mean, for example, when we say Kepler's Laws
_predict_ Tycho Brahe's observations, which were done prior to
Kepler's derivation of his laws. In at least one way it's better
when the data come before the theory; the danger of "confirmation
bias" is thereby avoided.

You have just demonstrated the problem
that troubles me: the attempt to treat
retrodictions as true predictions.

The advance of the perihelion of Mercury
was retrodicted by General Relativity, but
the results of the 1919 eclipse experiment
were predicted by GR. There is a huge
difference between retrodictions and
predictions. Those scientists of the past
who we hold in high esteem understood this
distinction. Over the last few decades the
distinction has eroded badly, and I think
this is a threat to science.

I see there's a new simulation that goes quite a way to predicting
cosmological observations:
http://www.cfa.harvard.edu/news/2014-10
I haven't read the article yet (and had nothing to do with the
research itself). If the work really has the galaxy morphologies and
chemical abundances right, that will be a huge achievement.

I am preparing a post on the Illustris simulation.
Here is a preview.

Of course it models observations! It has been
adjusted and tweaked for decades for just this
purpose.

Joel Primack comments that the simulation has been
"oversold". "The people who are not in the field
are reading the hype and being misled into thinking
what they are doing is revolutionary ... And that's
just wrong."

Those are Primack's comments to Science.

RLO
http://www3.amherst.edu/~rloldershaw

Sometimes a "prediction" can be numerically wrong but still provide a
triumph for a theory. For example, Peebles (1966
http://adsabs.harvard.edu/abs/1966ApJ...146..542P) predicted the
primordial helium and deuterium abundances. His values (Yp =
0.26-0.28) disagree with modern values (Yp = 0.246) for reasons that
are now clear: his assumed lifetime of the free neutron was wildly
wrong, he did not know about the tau neutrino, many nuclear
cross-section had not yet been measured.

In the 48 years since this paper, there has sometimes been tension
between BBN models and observations of primordial abundances. There
must be over 100 papers proposing modifications to BBN such as
inhomogenous nucleosynthesis or exotic neutrinos to provide better
agreement. However these modifications have not been needed, and the
same 1966 prescription of Peebles works with zero adjustment or tweaks
but only updated nuclear physics (and better computers). This is one
of the great predictions of modern science -- that the observed
primordial light element abundances can be derived from calculations
of a cooling nuclear plasma in an expanding universe. --Wayne

On Friday, May 2, 2014 4:31:55 PM UTC-4, Dan Riley wrote:

See Dan's post above.
----------------------------------

Four related questions have been on
my mind recently.

1. Are we confident in the assumption
of "very cold" initial conditions in the
inflationary blip before the Hot Big Bang?

2. What would cause the temperature to go
from "very cold" to "hot"?

3. How confident can we be about events
that take place long before the CMB was
generated?

4. Is there an elephant in the room, i.e.,
a conflict with the basic laws of thermodynamics?

In article , "Robert L.
Oldershaw" writes:

1. Are we confident in the assumption
of "very cold" initial conditions in the
inflationary blip before the Hot Big Bang?

Do you doubt them and if so, why?

2. What would cause the temperature to go
from "very cold" to "hot"?

Reheating (look it up).

3. How confident can we be about events
that take place long before the CMB was
generated?

There are hundreds if not thousands of papers concerned with calculating
observable aspects of the CMB from generic-inflation initial conditions
assuming only known physics. The fact that terabytes of CMB information
can be fit with just 6 parameters shows that this is very well
understood.

4. Is there an elephant in the room, i.e.,
a conflict with the basic laws of thermodynamics?

Do you think so and if so, why?

In article ,
Jos Bergervoet writes:
"theory predicted the CMB" or
"the CMB was predicted by theoretical models"
then most scientists would immediately say that isn't
true!

I don't think most scientists would say that, and they'd be wrong if
they did. The CMB was predicted (in both the popular and scientific
senses) by Gamow and others in the 1940s (I think) and observed in
1965. Dicke and others knew of the prediction and had built an
antenna to try to observe the CMB, but they were "scooped" by the
Bell Labs group.

To the general point, most scientists I know, when speaking of
scientific questions, use "predict" in the sense I wrote, not in the
popular sense. That sense is what the scientific method is concerned
with, despite what some people may believe.

--
Help keep our newsgroup healthy; please don't feed the trolls.
Steve Willner Phone 617-495-7123
Cambridge, MA 02138 USA

On Thursday, May 15, 2014 2:02:19 AM UTC-4, wlandsman wrote:
same 1966 prescription of Peebles works with zero adjustment or tweaks
but only updated nuclear physics (and better computers). This is one
of the great predictions of modern science -- that the observed
primordial light element abundances can be derived from calculations
of a cooling nuclear plasma in an expanding universe. --Wayne

But an important question is whether these
were true definitive predictions or
whether they should be more accurately
called retrodictions?

Were the approximate abundances of various
nuclei already well known? In the cases
of H and He, I think they were known fairly
well from stellar spectra and other data BEFORE
1966.

I think Peebles knew the answers he was aiming for.
In that case, you do not have a prediction at all.

Also, there have also been problems with He and Li
that have cropped up over the decades, but
it has been relatively easy for model-builders
to offer multiple solutions for 'saving the
phenomenon'.

[Mod. note: reformatted -- mjh]

In article , Steve Willner
writes:

In article ,
Jos Bergervoet writes:
"theory predicted the CMB" or
"the CMB was predicted by theoretical models"
then most scientists would immediately say that isn't
true!

I don't think most scientists would say that, and they'd be wrong if
they did. The CMB was predicted (in both the popular and scientific
senses) by Gamow and others in the 1940s (I think) and observed in
1965. Dicke and others knew of the prediction and had built an
antenna to try to observe the CMB, but they were "scooped" by the
Bell Labs group.

Right. Not only the existence of the CMB itself, but details. I don't
know how many plots of the CMB power spectrum I had seen, in very much
detail, even before the first peak was observed. Planck is now
approaching the details of theoretical predictions. Yes, there are some
free parameters, but a) most of them are known from other sources, so
not really free, and b) fitting all the data in a CMB map with just a
handful of parameters illustrates that people actually understood this
in detail before the observations.

In article , "Robert L.
Oldershaw" writes:

On Thursday, May 15, 2014 2:02:19 AM UTC-4, wlandsman wrote:
same 1966 prescription of Peebles works with zero adjustment or tweaks
but only updated nuclear physics (and better computers). This is one
of the great predictions of modern science -- that the observed
primordial light element abundances can be derived from calculations
of a cooling nuclear plasma in an expanding universe. --Wayne

But an important question is whether these
were true definitive predictions or
whether they should be more accurately
called retrodictions?

I think Peebles knew the answers he was aiming for.
In that case, you do not have a prediction at all.

Perhaps, but show me where, in his chain of reason, he has anything
adjustable which would allow him to fit arbitrary observations.

Also, there have also been problems with He and Li
that have cropped up over the decades, but
it has been relatively easy for model-builders
to offer multiple solutions for 'saving the
phenomenon'.

So you doubt big-bang nucleosynthesis as well. OK.

As Steve and I have pointed out, you are making a mountain out of a
molehill here. Yes, a prediction is more spectacular than a
postdiction, especially from a new theory, but as long as the theory
doesn't have any free parameters which one can tune to fit observations
the two are essentially equivalent. Also, your distinction would
indicate that GR would be a worse theory if the bending angle of light
had been known beforehand. But it would be the same theory, and there
is no fudge factor to get the bending angle correct.