On May 27, 8:26 pm, Tom Roberts wrote:
> On 5/27/13 5/27/13,
gehan.ameresek...@gmail.com wrote:
> > The problem I have is this: in scientific discovery we rely on observations
> > as well as computations. Are only observations valid? Are computations not
> > valid? Computations do not 'describe reality?'. All scientific knowledge
> > based on computations should be immediately discounted?
>
> No. You are VERY confused. You are missing the forest by concentrating on the
> veins in a leaf.
>
> Science is the human process by which we develop models of the world we inhabit.
> Hopefully over time, as we learn more and develop better technology, these
> models get better and better, with wider domains of applicability.
Yes, what Tom is saying is correct. Let Koobee Wublee reiterate what
science is.
** Start with a hypothesis.
** Compute the predictions to this hypothesis.
** Design experiment to test these preditions.
If any experiment shows a disagreement in these predictions, the
hypothesis is wrong.
Furthermore, if the hypothesis and its antitheses all satisfy the same
observations, you end up having nothing. You must show experiments
that verify your hypothesis and disprove its antitheses. <shrug>
> And that has
> indeed been the case: Newtonian mechanics gave way to SR,
This is just not true. SR is a hypothesis over Newtonian mechanics
that is capable of explaining the null results of the MMX. SR must be
subjected to experimental verifications before deciding that SR is
indeed a model built on top of the Newtonian mechanics. <shrug>
> which gave way to GR;
Ditto, Tom. <shrug>
> in each case the new theory provided higher accuracy in predicting experimental
> results, and a wider domain of applicability. A more remarkable progression has
> occurred in quantum theory....
It turns out that SR and GR manifest all sorts of mathematical
inconsistencies in which Koobee Wublee has addressed one by one in the
past. <shrgu>
> So what matters is how well a give model works to describe and predict physical
> phenomena in the world, not whether it involves "observation or computation".
All news models must be subjected to experimental verifications. This
is what scientific method is all about. This is what separate
scientists from crackpots. The question is if the self-styled
physicists in which Tom is one of them is indeed a scientist or not.
<shrug>
> Note that nature takes no notice of how humans happen to observe her behavior,
> so HOW humans make observations cannot possibly affect any physical phenomena
> (it can only affect how they are described).
Yes, this is true. So, why is the momentum term showing up in the
field equations? Momentum is an observer dependent quantity, no?
<shrug>
> In physics a primary means of
> observation is by using a coordinate system,
> which is an ARBITRARY choice of the humans using it.
Yes, this is true, but Tom is trying to downplay this important
concept. To describe any physical phenomena, you must choose a
coordinate to do so, and there are no exceptions. If you don’t have a
coordinate system in mind, there is no bloody way you can describe any
invariant geometry. <shrug>
> So no coordinate-dependent quantity can possibly be a valid
> model of any physical phenomenon.
Tom is grossly confusing observer and coordinate dependencies. No
observer can affect the invariant geometry that he is observing, but
that does not mean the description of that invariant geometry is the
same using different coordinate systems. <shrug>
> Closing speed is clearly coordinate dependent,
> and this is why no physical theory has ever been phrased in terms of closing
> speed.
There is no such thing as the closing speed. <shrug>
> Relative speed is likewise coordinate dependent, and no current
> fundamental theory involves it, either.
The whole point is that it is very stupid to talk about coordinate
dependency because it depends on the observer. What makes sense is
observer dependency. <shrug>
> Our current theories are all described
> by manifestly coordinate-independent equations among tensor or spinor quantities.
Nonsense. Tensor calculus is nothing but applying Leibniz’s chain
rules on derivatives to matrices. They are not coordinate independent
since it is just impossible to describe any invariant geometry without
using a set of coordinate system. You must supply a coordinate system
to begin solving for the field equations, and by the grace of God, you
shall stick with that choice of coordinate system to the very end of
derivations. <shrug>
> Whatever role(s) you might think that velocity, speed, or energy
> could play are taken over by 4-momentum (being a 4-vector, it is
> manifestly independent of coordinates). The relative and closing
> speeds of colliding objects are irrelevant, what matters are the
> Mandelstam variables, all of which are invariants.
Tom is spinning more myths to salvage his already weak case in support
of SR and GR. The reason under SR and GR why the 4-whatever is
invariant is because of the local event. For the simplest case, say
the Minkowski spacetime geometry which is generally described by the
following with two different observers observing the same observed
(the target).
** c^2 dt1^2 – ds1^2 = c^2 dt2^2 - ds2^2
Where
** dt1 = Observed flow rate of time by #1
** dt2 = Observed flow rate of time by #2
** ds1 = Observed local space at the target by #1
** ds2 = Observed local space at the target by #2
With all these observers, you can also find the target being the
observer observing itself. In doing so, the equation above simplifies
into the following.
** c^2 dTau^2 = c^2 dt1^2 – ds1^2 = c^2 dt2^2 - ds2^2
Where
** dTau = Local flow rate of time
In an abstract fashion where it can be mathematically proven, the
following shows the energy transform in the 4-vector domain.
** E1^2 – p1^2 c^2 = E2^2 – p2^2 c^2
Notice the equation above is invariant because it describes the energy
of the rest mass which must be invariant to any observers. Duh!
<shrug>