Causality
Peter
In order to avoid misunderstandings in several (sub)threads, I would like to
propose to find a common understanding of certain basic notions, such as
'causality'. (For 'locality', see the foregoing new thread.)
To begin with, my understanding of 'causality' is, that an event has got a
cause.
The principle of causality states that *all* events have got a cause ('nihil
sine ratione', as Leibnitz said).
This 'obvious truth' or 'necessity of thinking', as some may said, was
obscured by unlucky historical developments. It was incorrectly connected
with classical-mechanical motion. Quantum mechanics has revealed that such
motion does not take place within atoms. Unfortunately, not only the
classical-mechanical motion was rejected, but also causality in general. This
confusion seems to endure to our days...
Thus, I would like to agree with you, that causality is a universal
principle, where the concrete form of realization depends on the specific
area of application.
Of course, this is just an initial sketch -
Looking forward,
Peter
Igor
I'm not so sure that quantum mechanics denies causality as much as it
muddles the distinction between cause and effect.
Mike
> I'm not so sure that quantum mechanics denies causality as much as it
> muddles the distinction between cause and effect.- Hide quoted text -
>
Actually, I've shown that QM is derived by having to consider every
fact as the potential cause of every other fact. See my webpage at:
Oh No
becomes difficult to state in quantum theory. For example
"microcausality" is simply the locality condition in relativistic qft.
If causality is not to mean determist causality, how can we put it into
words? If it just refers to probabilistic results, does that reflect our
usual understanding of causality? Do we need a different word for
quantum causality, and how are we to define it?
Regards
--
Charles Francis
moderator sci.physics.foundations.
charles (dot) e (dot) h (dot) francis (at) googlemail.com (remove spaces and
braces)
Ken S. Tucker
On Aug 15, 1:51 pm, Peter <end...@dekasges.de> wrote:
> In order to avoid misunderstandings in several (sub)threads, I would like to
> propose to find a common understanding of certain basic notions, such as
> 'causality'. (For 'locality', see the foregoing new thread.)
>
> To begin with, my understanding of 'causality' is, that an event has got a
> cause.
Quote,"an event has got a cause." needs clarification.
Personally I have been irritated by the lack of a good
definition as to the meaning of "event" in relativity, so
in place of the word "event" I'll use "occurance".
((I've replaced the usual spelling of occurence)).
Within a finite volume DV = DxDyDz, is a detector or
an emitter, able to measure an incremental change of
Energy in a finite interval of time, DE/Dt = Power.
I think that the term "occurance" is a quantization
of the word "event", where event is presupposed to
be able to occur at a point, which is unrealistic.
Mathematically, let Occurance (o) be defined,
(o) = DV Dt DE.
((I'm using bracket's around "o" so we don't confuse it
with zero and because it is has a finite spacetime
volume)).
With OCCURANCE (o) defined we can examine and
define "causality", as a relation between two (o)'s,
such as the emission and subsequent absorption
of a photon, "~~~~",
(o) ~~~~~~> (o')
and (o) = -(o') defining causality.
I think "casuality" is another term for Conservation
of Energy, but I'm having problems with understanding
Power, which appears to be Energy Flow or IOW's
Energy Flux.
I'll stop here pending suggestions and/or approval.
Regards
Ken S. Tucker
Oh No
>Hi Dr. Enders
>
>On Aug 15, 1:51 pm, Peter <end...@dekasges.de> wrote:
>> In order to avoid misunderstandings in several (sub)threads, I would like to
>> propose to find a common understanding of certain basic notions, such as
>> 'causality'. (For 'locality', see the foregoing new thread.)
>>
>> To begin with, my understanding of 'causality' is, that an event has got a
>> cause.
>
>Quote,"an event has got a cause." needs clarification.
>
>Personally I have been irritated by the lack of a good definition as to
>the meaning of "event" in relativity, so in place of the word "event"
>I'll use "occurance". ((I've replaced the usual spelling of
>occurence)).
>
>Within a finite volume DV = DxDyDz, is a detector or an emitter, able
>to measure an incremental change of Energy in a finite interval of
>time, DE/Dt = Power.
>
>I think that the term "occurance" is a quantization of the word
>"event", where event is presupposed to be able to occur at a point,
>which is unrealistic. Mathematically, let Occurance (o) be defined,
>
>(o) = DV Dt DE.
Event just means a space-time coordinate. It should be implicit that a
space time coordinate is something measured, and equally it should be
implicit that measurement is to within error bounds. In relativity, I
don't see a problem with modelling coordinates using a continuum. The
problem starts to arise in quantum theory, when we cannot do the
measurement even in principle (without fundamentally changing the
situation).
Ken S. Tucker
> Thus spake Ken S. Tucker <dynam...@vianet.on.ca>
>
>
>
> >Hi Dr. Enders
>
> >On Aug 15, 1:51 pm, Peter <end...@dekasges.de> wrote:
> >> In order to avoid misunderstandings in several (sub)threads, I would like to
> >> propose to find a common understanding of certain basic notions, such as
> >> 'causality'. (For 'locality', see the foregoing new thread.)
>
> >> To begin with, my understanding of 'causality' is, that an event has got a
> >> cause.
>
> >Quote,"an event has got a cause." needs clarification.
>
> >Personally I have been irritated by the lack of a good definition as to
> >the meaning of "event" in relativity, so in place of the word "event"
> >I'll use "occurance". ((I've replaced the usual spelling of
> >occurence)).
>
> >Within a finite volume DV = DxDyDz, is a detector or an emitter, able
> >to measure an incremental change of Energy in a finite interval of
> >time, DE/Dt = Power.
>
> >I think that the term "occurance" is a quantization of the word
> >"event", where event is presupposed to be able to occur at a point,
> >which is unrealistic. Mathematically, let Occurance (o) be defined,
>
> >(o) = DV Dt DE.
>
> Event just means a space-time coordinate. It should be implicit that a
> space time coordinate is something measured, and equally it should be
> implicit that measurement is to within error bounds.
Exactly, those "error bounds" import HU,
(Heisenberg Uncertainty). I purposefully built
that in by using DV3=DxDyDx and DV4=DV3*Dt
to account for those boundaries, but my entry
into the subject was gentle.
For example, suppose Planck's "h" expressed in
Ergs Seconds is the eqivalent of (h= DE Dt) and
DV is invariant, then (o) is invariant.
((Fred, please see Spiegel, pg.203..66 and find
dV=dV' as invariant)).
((I'm uncertain my reasoning is entirely valid))
Next, DV3*h = invariant = (o).
Once the "occurance's" themselves are defined
invariantly, then the differences of the (o)'s above
are mathematically invariant.
Let's presume Tucker's reasoning is close.
If so, we find a means to define "casuality"
Generally Covariantly.
That is why I think Dr. Ender's question is important,
that I'll rephrase as, "Can Casuality be expressed as
a law of physics", true w.r.t all CS's?
> In relativity, I
> don't see a problem with modelling coordinates using a continuum.
Well I certainly do, and cared enough to formulate
and provide a rework of the sematic "event" to give
a (fairly) clear invariant definition of "occurance",
compatible with QT, QM.
> The
> problem starts to arise in quantum theory, when we cannot do the
> measurement even in principle (without fundamentally changing the
> situation).
Well of course, that's why we need to take fuzzy
semantics to hard core paper definitions.
> Regards
> Charles Francis
> http://www.teleconnection.info/rqg/MainIndex
Cheers
Ken S. Tucker
Hendrik van Hees
> The meaning of causality was clear with a determinist universe, but
> becomes difficult to state in quantum theory. For example
> "microcausality" is simply the locality condition in relativistic qft.
> If causality is not to mean determist causality, how can we put it
> into words? If it just refers to probabilistic results, does that
> reflect our usual understanding of causality? Do we need a different
> word for quantum causality, and how are we to define it?
I think it is very important, particularly in the context with quantum
theory (QT), to distinguish between causality and determinism.
I'd define a theory or model as causal as follows: A theory is called
causal if one knows the state of the system exactly at a certain time
t0, if its state is known exactly for all earlier times t<t0.
All our physical theories, including quantum mechanics, are causal in
that sense. They are even a little bit more than that, namely they are
local in time. It is sufficient to know the state of the system at one
time t0<t to know it at time t.
Further, I call a theory deterministic, if the exact knowledge of the
state of the system implies to know the values for all possible
observables of this system. Classical physics (including classical
relatvistic theories of course) is causal and deterministic. QT is of
course causal, but it's not deterministic since in principle if you
know the state of a system (represented by a ray in Hilbert space)
exactly at one instant of time, t0, you know it at any later time
either. It is not deterministic, since the state tells you which
observables have a certain value, namely if any of the ray's
representant Hilbert-space vectors is an eigenvector of the
self-adjoint operator, representing that observable. All other
observables are have no certain values, and thus even if one knows the
exact state of the system one does not know the values of all
observables but only of a certain set of observables.
A very concise and clear elaboration of all this, you can find as the
introductory chapter in
J. Schwinger, Quantum Mechanics, Symbolism for atomistic measurements,
Springer.
--
Hendrik van Hees Institut für Theoretische Physik
Phone: +49 641 99-33342 Justus-Liebig-Universität Gießen
Fax: +49 641 99-33309 D-35392 Gießen
http://theory.gsi.de/~vanhees/faq/
Peter
> Thus spake Peter <end...@dekasges.de>
> >
> >In order to avoid misunderstandings in several (sub)threads, I would like to
> >propose to find a common understanding of certain basic notions, such as
> >'causality'. (For 'locality', see the foregoing new thread.)
> >
> >To begin with, my understanding of 'causality' is, that an event has got a
> >cause.
> >
> >The principle of causality states that *all* events have got a cause
> >('nihil sine ratione', as Leibnitz said).
> >
> >This 'obvious truth' or 'necessity of thinking', as some may said, was
> >obscured by unlucky historical developments. It was incorrectly connected
> >with classical-mechanical motion. Quantum mechanics has revealed that such
> >motion does not take place within atoms. Unfortunately, not only the
> >classical-mechanical motion was rejected, but also causality in general.
> >This confusion seems to endure to our days...
> >
> >Thus, I would like to agree with you, that causality is a universal
> >principle, where the concrete form of realization depends on the specific
> >area of application.
> The meaning of causality was clear with a determinist universe, but
> becomes difficult to state in quantum theory. For example
> "microcausality" is simply the locality condition in relativistic qft.
This is an example for the fact that the historical development of quantum
theory has rather blurred this issue
Quantum theory was found by huge jumps of both formalism and notions. This
made the novel terrain extraordinarily unsurely. Accounting for the incomplete
philosophical education of the pioneers (absence of dialectics, overweight of
idealism), it may be even not surprising that many speculations arose.
> If causality is not to mean determist causality,...
of course, not :-)
> ...how can we put it into words?
See above - where Leibnitz's 'nihil sine ratione' refers to determinism?
> If it just refers to probabilistic results, does that reflect our
> usual understanding of causality?
Is the Langevin equation acausal?
> Do we need a different word for
> quantum causality, and how are we to define it?
I don't think so. However, I would agree that there are various *forms* of
*realization* of causality :-)
Thank you,
Peter
Salviati
Newsbeitrag news:g89ufv$r6a$1...@news2.open-news-network.org...
> I think it is very important, particularly in the context with quantum
> theory (QT), to distinguish between causality and determinism.
While I agree on that, my reason is quite different:
>
> I'd define a theory or model as causal as follows: A theory is called
> causal if one knows the state of the system exactly at a certain time
> t0, if its state is known exactly for all earlier times t<t0.
I consider this the elusive belief that anything can be reduced to states
of a closed and finite system. Isn't this determinism?
>
> All our physical theories, including quantum mechanics, are causal in
> that sense. They are even a little bit more than that, namely they are
> local in time. It is sufficient to know the state of the system at one
> time t0<t to know it at time t.
I learned that there are no closed system in reality but a largely unseen
variety of influences. Accordingly, reality is bound to causality which
simply means that any effect bundles many influencing causes even if
just one might be considered the decisive one. No effect precedes any
influence that caused it. Therefore, genuine causality is restricted to the
past.
While future "causality" is the guess that reality will behave as predicted,
we may be sure that the principle causality will also hold in future. For
instance, every mammal has a father and a mother, so far. However it is
uncertain what children it will have. A theory may not consider the
possibility of cloning.
What is called a causal signal is actually a deterministically calculable
one.
It equals zero for t<0.
> Further, I call a theory deterministic, if the exact knowledge of the
> state of the system implies to know the values for all possible
> observables of this system. Classical physics (including classical
> relatvistic theories of course) is causal and deterministic. QT is of
> course causal, but it's not deterministic since in principle if you
> know the state of a system (represented by a ray in Hilbert space)
> exactly at one instant of time, t0, you know it at any later time
> either.
Isn't this rather an indication for lost realism?
Claude Shannon: The past is known, in principle, but we cannot control it.
The future is unknown but we can control it, in principle.
Regards,
Salviati
Oh No
What is the cause by which a radioactive decay takes place in a given
time interval?
>
>> If it just refers to probabilistic results, does that reflect our
>> usual understanding of causality?
>
>Is the Langevin equation acausal?
My point is that to answer such a question we must first define what we
mean by causal.
>
>> Do we need a different word for
>> quantum causality, and how are we to define it?
>
>I don't think so. However, I would agree that there are various *forms* of
>*realization* of causality :-)
>
>Thank you,
>Peter
>
Oh No
giessen.de>
>Oh No wrote:
>
>> The meaning of causality was clear with a determinist universe, but
>> becomes difficult to state in quantum theory. For example
>> "microcausality" is simply the locality condition in relativistic qft.
>> If causality is not to mean determist causality, how can we put it
>> into words? If it just refers to probabilistic results, does that
>> reflect our usual understanding of causality? Do we need a different
>> word for quantum causality, and how are we to define it?
>
>I think it is very important, particularly in the context with quantum
>theory (QT), to distinguish between causality and determinism.
>
>I'd define a theory or model as causal as follows: A theory is called
>causal if one knows the state of the system exactly at a certain time
>t0, if its state is known exactly for all earlier times t<t0.
>
>All our physical theories, including quantum mechanics, are causal in
>that sense. They are even a little bit more than that, namely they are
>local in time. It is sufficient to know the state of the system at one
>time t0<t to know it at time t.
What about the state of one of the particles in an EPR experiment?
Presumably, if we cannot separately define the state of an individual
particle in an entangled pair, this must force on us non-locality?
>Further, I call a theory deterministic, if the exact knowledge of the
>state of the system implies to know the values for all possible
>observables of this system. Classical physics (including classical
>relatvistic theories of course) is causal and deterministic. QT is of
>course causal, but it's not deterministic since in principle if you
>know the state of a system (represented by a ray in Hilbert space)
>exactly at one instant of time, t0, you know it at any later time
>either. It is not deterministic, since the state tells you which
>observables have a certain value, namely if any of the ray's
>representant Hilbert-space vectors is an eigenvector of the
>self-adjoint operator, representing that observable. All other
>observables are have no certain values, and thus even if one knows the
>exact state of the system one does not know the values of all
>observables but only of a certain set of observables.
>
I think this takes us forward, but I would grumble that to define
causality in terms of "state" only resolves the issue if we have a clear
interpretation of "state", so that it reduces the issue of causality to
the perennial issue of interpretation. Since I use an interpretation
according to which "state" is defined in terms of the probabilities of
measurement results, I would like to modify this, along the lines of
saying:
There is a causal relation between two events (situations?) if the
outcome of a measurement at one event alters the probability of the
outcome of some measurement at the other.
Rich L.
thoughts may be a bit naive yet.
I think it is good to separate the ideas of causality and
determinism. I think too we need to clarify our ideas about the
nature of the past and future. As stated more eloquently above, we
assume the past is definite and fixed, whether we know what it is or
not. We all assume (as do I) that the sequence of events in the past
is fixed and, in an appropriate sense, relativistically covarient (I
think that is the right term). Likewise the future is often assumed
to be indeterminate, but predictable within certain limits. This,
ultimately, is the origin of our common sense notion of causality. At
least at a philosophical level, these assumptions should be made
explicit and questioned.
There are two possibilities for the sequence of events: they can be
definite(fixed) or indefinite(not fixed) (I'm not sure I like these
terms for the idea I'm trying to express...). We assume the past is
definite in that the events have happened and they will not change.
We can imagine a universe where the past also changes, but we don't
seem to be in such a universe. Likewise we assume the future is
changable by choices we make in the present. This may or may not be
true. We like to think it is true because it gives us some sense that
we are in control of our destiny, but I don't know any conclusive
argument that this is neccessarily the case. I consider this an open
question.
There is also the issue of knowability. We believe we can know the
past, that it is knowable, even if we don't actually know it
perfectly. Quantum Mechanics, and even statistical mechanics, weakens
this belief somewhat. I think the common sense idea of causality
depends critically on to what degree can we really know the past, or
even the current state, so that we can make a causal connection
between the present state and some future state. Bells' Inequality
experiments based on it I find very illuminating on this point, but
I'm not sure people have considered all the possible conclusions from
these observations.
I think more consideration should be given to the possiblity that the
future is itself fixed, if unknown and unknowable. One could
certainly argue that if it is unknown and unknowable, does it make any
difference if it is "fixed" or not? It could in the interpretation of
the Aspect experiments, and the philosophical question of what is the
true causal/deterministic/knowable nature of the universe.
Einstein famously argued that the outcome of events is not random, and
that those events (and theories) that appeared to be so merely
indicated that there were variables that we could not observe that
determined the actual outcome. The experiments based on Bell's
Inequality have generally been taken to refute this concept, but in
fact, they are in perfect agreement with it, as well as the non-causal
QM interpretation. They don't really settle the question. The reason
I say these experiments are consistent with deterministic concepts is
that a perfectly deterministic universe has a definite future as well
as past. We may not know the future yet, but it is as definite and
fixed as the past appears to be. In such a universe of course the
emitted photons alwyas show perfect correlations with each other
because the future is as fixed as the past. Of course humans, as the
control freaks that we are, don't like to think about this possibility
too much.
The universe is likely even stranger than we currently imagine, and
these ideas may all be too simple yet by a large margin.
For what it's worth,
Rich L.
Hendrik van Hees
> What about the state of one of the particles in an EPR experiment?
> Presumably, if we cannot separately define the state of an
> individual particle in an entangled pair, this must force on us
> non-locality?
Within the usual concept of QT that's very easy. If you have an
entangled particle pair, the single-particle state is (by definition)
not known and to be described by a probability operator which is not
of the form |psi><psi|. In the usual slang, one says it's not in a
pure but a mixed state. In other words, the single particle is not in
a definite quantum state.
What do you mean by "non-locality"? Now we enter the issue of defining
locality which has been put into another thread. In the usual
business of, e.g., the entangled-twophoton experiments ("Aspect
like") we have long-ranged correlations. You have to specify clearly,
what you mean with "non-locality" here.
> I think this takes us forward, but I would grumble that to define
> causality in terms of "state" only resolves the issue if we have a
> clear interpretation of "state", so that it reduces the issue of
> causality to the perennial issue of interpretation. Since I use an
> interpretation according to which "state" is defined in terms of the
> probabilities of measurement results, I would like to modify this,
> along the lines of saying:
>
> There is a causal relation between two events (situations?) if the
> outcome of a measurement at one event alters the probability of the
> outcome of some measurement at the other.
What I meant with "states" were "pure states" in the sense of usual
quantum-mechanics slang.
Chalky
> In order to avoid misunderstandings in several (sub)threads, I would like to
> propose to find a common understanding of certain basic notions, such as
> 'causality'. (For 'locality', see the foregoing new thread.)
>
> To begin with, my understanding of 'causality' is, that an event has got a
> cause.
>
> The principle of causality states that *all* events have got a cause ('nihil
> sine ratione', as Leibnitz said).
Within the context of physics, I have usually understood the principle
of causality to mean pre-relativistically that the cause precedes the
effect, or relativistically that the effect is contained within the
light cone constructed around the cause.
Whether this assumption is valid or not is another matter entirely. It
certainly seems to be inconsistent with the concept of temporal
entanglement in QM. Similarly, the above quoted Liebnitz definition
seems to be inconsistent with the concept of randomness.
Salviati
news:2c9a44f4-8217-433c...@m45g2000hsb.googlegroups.com...
> I think the common sense idea of causality
> depends critically on to what degree can we really know the past, or
> even the current state, so that we can make a causal connection
> between the present state and some future state.
To my understanding of reality, there is no causal connection between "the"
present state and any future state but a future state will result from
indefinitely many present states altogether.
> I think more consideration should be given to the possiblity that the
> future is itself fixed, if unknown and unknowable. One could
> certainly argue that if it is unknown and unknowable, does it make any
> difference if it is "fixed" or not? It could in the interpretation of
> the Aspect experiments, and the philosophical question of what is the
> true causal/deterministic/knowable nature of the universe.
Is spf a place for fatalism?
> The universe is likely even stranger than we currently imagine, and
> these ideas may all be too simple yet by a large margin.
I refuse to speculate on an obviously elusive basis. So far nobody proved
me wrong when I pointed to what I consider unjustified interpretation of
arbitrary mathematical constructs.
Regards,
Salviati
======================================= MODERATOR'S COMMENT:
'fatalism' is not a term of physics
Peter
> Oh No wrote:
>
> > The meaning of causality was clear with a determinist universe, but
> > becomes difficult to state in quantum theory. For example
> > "microcausality" is simply the locality condition in relativistic qft.
> > If causality is not to mean determist causality, how can we put it
> > into words? If it just refers to probabilistic results, does that
> > reflect our usual understanding of causality? Do we need a different
> > word for quantum causality, and how are we to define it?
>
> I think it is very important, particularly in the context with quantum
> theory (QT), to distinguish between causality and determinism.
Yes, indeed
> I'd define a theory or model as causal as follows: A theory is called
> causal if one knows the state of the system exactly at a certain time
> t0, if its state is known exactly for all earlier times t<t0.
This is a definition of determinism, not of causality; for this, it agrees
with your definition of determinism below
> All our physical theories, including quantum mechanics, are causal in
> that sense. They are even a little bit more than that, namely they are
> local in time. It is sufficient to know the state of the system at one
> time t0<t to know it at time t.
This is correct for (the so-called) Newton's equation of motion, for the
Schrödinger and Dirac equations, but not for Langevin's equation of motion
> Further, I call a theory deterministic, if the exact knowledge of the
> state of the system...
This invokes the notion of state to be defined first :-)
> ...implies to know the values for all possible
> observables of this system.
This is new to me, that determinism connects 'state' with 'observables'
> Classical physics (including classical
> relatvistic theories of course) is causal and deterministic. QT is of
> course causal, but it's not deterministic since in principle if you
> know the state of a system (represented by a ray in Hilbert space)
> exactly at one instant of time, t0, you know it at any later time
> either. It is not deterministic, since the state tells you which
> observables have a certain value, namely if any of the ray's
> representant Hilbert-space vectors is an eigenvector of the
> self-adjoint operator, representing that observable. All other
> observables are have no certain values, and thus even if one knows the
> exact state of the system one does not know the values of all
> observables but only of a certain set of observables.
>
> A very concise and clear elaboration of all this, you can find as the
> introductory chapter in
>
> J. Schwinger, Quantum Mechanics, Symbolism for atomistic measurements,
> Springer.
Hope to get this affordably ;-)
Thank you,
Peter
Salviati
cause>
> Similarly, the above quoted Liebnitz definition
> seems to be inconsistent with the concept of randomness.
Leibniz, neither Liebnitz nor Leibnitz.
Nihil sine ratione(s) does not contradict to the method of treating
processes
as if they were random because a huge variety of influences evades any
deterministic description. Playing dice does not violate causality.
Regards,
Salviati
Rich L.
> > Oh No wrote:
...
> > I think it is very important, particularly in the context with quantum
> > theory (QT), to distinguish between causality and determinism.
>
> Yes, indeed
>
> > I'd define a theory or model as causal as follows: A theory is called
> > causal if one knows the state of the system exactly at a certain time
> > t0, if its state is known exactly for all earlier times t<t0.
>
> This is a definition of determinism, not of causality; for this, it agrees
> with your definition of determinism below
>
> > Further, I call a theory deterministic, if the exact knowledge of the
> > state of the system...
> > observables of this system.
There are degrees of the causality concept. In classical mechanics it
was assumed that, in principle at least, the current state can be
determined exactly and that the future can be calculated exactly.
Whether you interpret this as fatalism or not is not a question for
physics. I would call this "strong causality", although "determinism"
would also be a good term. Quantum mechanics softened this concept
such that the current state can only be known as a wavefunction with
somewhat indeterminate position/momentum and with only predictions of
the probabilities for various possible future development. This is
still causal, but is much weaker. Clearly mixed in with these
concepts of causality are the concepts of to what extent can the
current state be known, to what extent can the past be known, and to
what extent is the future determined uniquely by the past. The old
classical concept of causality necessarily implied that the future was
unique, even if we were not knowledgeable enough to calculate what it
was going to be. You can call that fatalism if you like, but it is a
clear implication of the classical concept of causality. Quantum
mechanics, or more properly the experiments that led to the
formulation of quantum mechanics, seems to argue strongly that the
situation is much fuzzier. We can neither know the present state
precisely, nor predict the future precisely. Still there is some
degree of causality in that the probabilities for future states can be
calculated based on what can be known about the present state.
Another interesting varient on the idea is the possibility that the
future can influence the past. (Don't worry, I'm not advocating this,
just presenting it as a philosophical possibility.) There are some
formulations of electromagnetism that do this, and they apparently
work just fine. As a conceptual example, consider an accelerating
charged particle that exchanges a photon with another charged
particle. We normally describe this as the first particle emitting a
photon (and suffering a recoil as result) followed by the future
particle recoiling as it absorbs the photon. Conceptually this could
also be described as the first particle colliding with a particle in
the future. Our reluctance to accept the latter point of view hints
at a key aspect of our mindset: that we think of the future as open
and not determined yet. We think that when the photon is emitted by
the first particle that it has the whole universe in front of it to
interact with, and that we cannot at that moment do more than estimate
probabilities of which charged particles it will end up reacting
with. It is possible, again from a philosophical point of view, that
the destination of the photon has been determined already when it was
emitted. After all, the relatavistic separation of the two events is
zero. If you explore the implications of this viewpoint, it is
possible to get back to a concept of causality that is very close to
determinism, except that it is impossible to precisely know the
current state. (Again, I'm not necessarily advocating this viewpoint,
only considering it as a possiblity.)
Again, the question everyone seems to skirt around is our concept of
the future. Is it "always in motion" as Yoda would say, or is it
fixed and immutable as classical mechanics would imply. I think that
is still an open question, although current interpretations of quantum
mechanics, and I think most scientists inclinations, are to favor the
former view.
I think also that both quantum mechanics and relativity must be
considered in all this. Relativity places severe constraints on the
relationship between the past, present and future. These
relationships are completely different from what quantum mechanics
imposes (not incompatible, I think, but different). Somehow we still
need to formulate a conceptual understanding of these issues around
time evolution that are consistent with both.
Rich L.
Peter
> On Aug 19, 4:10 pm, Peter <end...@dekasges.de> wrote:
> > > Oh No wrote:
> ...
> > > I think it is very important, particularly in the context with quantum
> > > theory (QT), to distinguish between causality and determinism.
> > Yes, indeed
> > > I'd define a theory or model as causal as follows: A theory is called
> > > causal if one knows the state of the system exactly at a certain time
> > > t0, if its state is known exactly for all earlier times t<t0.
> > This is a definition of determinism, not of causality; for this, it agrees
> > with your definition of determinism below
> ...
> > > Further, I call a theory deterministic, if the exact knowledge of the
> > > state of the system...
> > > ...implies to know the values for all possible
> > > observables of this system.
> > - Show quoted text -
> There are degrees of the causality concept. In classical mechanics it
> was assumed that, in principle at least, the current state can be
> determined exactly and that the future can be calculated exactly.
This holds true for Schrödinger's wave mechanics as well
Much confusion is caused by interchanging the different notions of state being
in use
> Whether you interpret this as fatalism or not is not a question for
> physics. I would call this "strong causality", although "determinism"
> would also be a good term. Quantum mechanics softened this concept
> such that the current state can only be known as a wavefunction with
> somewhat indeterminate position/momentum and with only predictions of
> the probabilities for various possible future development.
This points to the error to assign classical properties to quantum particles/
systems
> still causal, but is much weaker. Clearly mixed in with these
> concepts of causality are the concepts of to what extent can the
> current state be known, to what extent can the past be known, and to
> what extent is the future determined uniquely by the past. The old
> classical concept of causality necessarily implied that the future was
> unique, even if we were not knowledgeable enough to calculate what it
> was going to be. You can call that fatalism if you like, but it is a
> clear implication of the classical concept of causality. Quantum
> mechanics, or more properly the experiments that led to the
> formulation of quantum mechanics, seems to argue strongly that the
> situation is much fuzzier. We can neither know the present state
> precisely, nor predict the future precisely. Still there is some
> degree of causality in that the probabilities for future states can be
> calculated based on what can be known about the present state.
Schrödinger's and Dirac's equations are deterministic
Exactly that is the goal of Dirac's 1949 approach to bring special relativity
and Hamiltonian mechanics together
Thank you,
Peter
harry
"Hendrik van Hees" <Hendrik...@theo.physik.uni-giessen.de> wrote in
message news:g89ufv$r6a$1...@news2.open-news-network.org...
> Oh No wrote:
>
>> The meaning of causality was clear with a determinist universe, but
>> becomes difficult to state in quantum theory. For example
>> "microcausality" is simply the locality condition in relativistic qft.
>> If causality is not to mean determist causality, how can we put it
>> into words? If it just refers to probabilistic results, does that
>> reflect our usual understanding of causality? Do we need a different
>> word for quantum causality, and how are we to define it?
>
> I think it is very important, particularly in the context with quantum
> theory (QT), to distinguish between causality and determinism.
>
> I'd define a theory or model as causal as follows: A theory is called
> causal if one knows the state of the system exactly at a certain time
> t0, if its state is known exactly for all earlier times t<t0.
But then the state is exactly determined by its earlier state - thus I'd
call that deterministic. In other words, I don't find that definition
helpful to distinguish between the two.
To me a theory is causal if a change of state is due to an event.
> All our physical theories, including quantum mechanics, are causal in
> that sense. They are even a little bit more than that, namely they are
> local in time. It is sufficient to know the state of the system at one
> time t0<t to know it at time t.
>
> Further, I call a theory deterministic, if the exact knowledge of the
> state of the system implies to know the values for all possible
> observables of this system. Classical physics (including classical
> relatvistic theories of course) is causal and deterministic. QT is of
> course causal, but it's not deterministic since in principle if you
> know the state of a system (represented by a ray in Hilbert space)
> exactly at one instant of time, t0, you know it at any later time
> either.
Probably in the above sentence a "no" is lacking.
> It is not deterministic, since the state tells you which
> observables have a certain value, namely if any of the ray's
> representant Hilbert-space vectors is an eigenvector of the
> self-adjoint operator, representing that observable. All other
> observables are have no certain values, and thus even if one knows the
> exact state of the system one does not know the values of all
> observables but only of a certain set of observables.
>
> A very concise and clear elaboration of all this, you can find as the
> introductory chapter in
>
> J. Schwinger, Quantum Mechanics, Symbolism for atomistic measurements,
> Springer.
Regards,
Harald
Oh No
>
>Much confusion is caused by interchanging the different notions of state being
>in use
Much confusion is caused by the lack of an interpretation of quantum
mechanics. The issue of interpretation is essentially one of identifying
a correct definition of state.
>
>Schrödinger's and Dirac's equations are deterministic
This is again confusing. The wave equations are deterministic, but the
mechanics is not.
>
>
>Exactly that is the goal of Dirac's 1949 approach to bring special relativity
>and Hamiltonian mechanics together
>
I think this is done quite effectively in qed.
Oh No
>
>> What about the state of one of the particles in an EPR experiment?
>> Presumably, if we cannot separately define the state of an
>> individual particle in an entangled pair, this must force on us
>> non-locality?
>
>Within the usual concept of QT that's very easy. If you have an
>entangled particle pair, the single-particle state is (by definition)
>not known and to be described by a probability operator which is not
>of the form |psi><psi|. In the usual slang, one says it's not in a
>pure but a mixed state. In other words, the single particle is not in
>a definite quantum state.
>
>What do you mean by "non-locality"? Now we enter the issue of defining
>locality which has been put into another thread. In the usual
>business of, e.g., the entangled-twophoton experiments ("Aspect
>like") we have long-ranged correlations. You have to specify clearly,
>what you mean with "non-locality" here.
I shall try to put my thoughts on these topics together into a new
thread, with application to EPR.
>
>> I think this takes us forward, but I would grumble that to define
>> causality in terms of "state" only resolves the issue if we have a
>> clear interpretation of "state", so that it reduces the issue of
>> causality to the perennial issue of interpretation. Since I use an
>> interpretation according to which "state" is defined in terms of the
>> probabilities of measurement results, I would like to modify this,
>> along the lines of saying:
>>
>> There is a causal relation between two events (situations?) if the
>> outcome of a measurement at one event alters the probability of the
>> outcome of some measurement at the other.
>
>What I meant with "states" were "pure states" in the sense of usual
>quantum-mechanics slang.
>
Indeed, but unless one is prepared to define these mathematically (i.e.
according to the mathematical structure, and without reference to what
they mean physically) the usual qm slang suffers from lack of
interpretation.
Peter
www.killfile.org)
> >Thus spake Peter <end...@dekasges.de>
(I have added a '>' to bring the reference to the same level as the quoted
text)
> >Much confusion is caused by interchanging the different notions of state
> >being in use
> Much confusion is caused by the lack of an interpretation of quantum
> mechanics.
QM (Heisenberg, Schrödinger & equivalents) is the mechanics for oscillators
without turning points. Most confusion arises from trying to assign to these
oscillators properties they just not have got, such as x(t) and p(t).
> The issue of interpretation is essentially one of identifying
> a correct definition of state.
psi(x,t) is a Laplacian state function, while |psi(x,t)| and <psi(x,t)H|psi
(x,t)> are Newtonian ones. Both notions exhibit advantages and disadvantages,
but they are well defined both in CM and QM.
The quantum numbers are essentially stationary (Newtonian) state numbers (and
were originally called so.
> >Schrödinger's and Dirac's equations are deterministic
> This is again confusing.
Not at all
> The wave equations are deterministic, but the mechanics is not.
This is confusing - let's define 'mechanics'!
> >Exactly that is the goal of Dirac's 1949 approach to bring special
> >relativity and Hamiltonian mechanics together
> I think this is done quite effectively in qed.
Dirac, seemingly, did not think so. QED as presented usually (and on your
website) presupposes the Lorentz transformation. This is justified only by
the results. In contrast, Dirac's transformation formulae are justified by
their derivation.
Best wishes,
Peter
Hendrik van Hees
> Indeed, but unless one is prepared to define these mathematically
> (i.e. according to the mathematical structure, and without reference
> to what they mean physically) the usual qm slang suffers from lack
> of interpretation.
There is no other way to define what quantum theory means than
mathematics. I'm a proponent of the so-called "Minimal Statistical
Interpretation" (MSI). Especially I think that the Copenhagen
interpretation (CI) is more a confusion than an interpretation of
quantum theory, and EPR were very right to point this confusion out,
but CI is not necessary to formulate QT as a physical theory.
A very concise introduction to the MSI can be found in
L. Ballentine, Quantum Mechanics, Addison-Wesley
or (shorter) in the paper by the same author
Ballentine, Leslie E.: The Statistical Interpretation of Quantum
Mechanics, Reviews of Modern Physics 42(4), APS, 358?381, 1970
Oh No
>(I have added a '>' hoping the reference is not eaten, again, by
>www.killfile.org)
>> >Thus spake Peter <end...@dekasges.de>
>(I have added a '>' to bring the reference to the same level as the quoted
>text)
>
>> >Much confusion is caused by interchanging the different notions of state
>> >being in use
>
>> Much confusion is caused by the lack of an interpretation of quantum
>> mechanics.
>
>QM (Heisenberg, Schrödinger & equivalents) is the mechanics for oscillators
>without turning points.
that may describe wave mechanics, but it does not describe qm. qm
contains the notion of collapse.
>Most confusion arises from trying to assign to these
>oscillators properties they just not have got, such as x(t) and p(t).
>
>> The issue of interpretation is essentially one of identifying
>> a correct definition of state.
>
>psi(x,t) is a Laplacian state function, while |psi(x,t)| and <psi(x,t)H|psi
>(x,t)> are Newtonian ones. Both notions exhibit advantages and disadvantages,
>but they are well defined both in CM and QM.
We should require that fundamental properties of physics have physical
definition, not just mathematical definition. A metaphysical definition
is not sufficient if physical behaviour (collapse) is not modelled.
>
>The quantum numbers are essentially stationary (Newtonian) state numbers (and
>were originally called so.
>
>> >Schrödinger's and Dirac's equations are deterministic
>
>> This is again confusing.
>
>Not at all
oh yes it is. :-)
>
>> The wave equations are deterministic, but the mechanics is not.
>
>This is confusing
See, I told you it was confusing.
>- let's define 'mechanics'!
I think mechanics should describe the physical behaviour of matter.
>> >Exactly that is the goal of Dirac's 1949 approach to bring special
>> >relativity and Hamiltonian mechanics together
>
>> I think this is done quite effectively in qed.
>
>Dirac, seemingly, did not think so.
QED was not much developed in 1949.
>QED as presented usually (and on your
>website) presupposes the Lorentz transformation. This is justified only by
>the results.
Why do you ignore Einstein?
>In contrast, Dirac's transformation formulae are justified by
>their derivation.
The Lorentz transform is derived on my site, and in many other places.
Peter
> Oh No wrote
> >Thus spake Peter <end...@dekasges.de>
> >> >Much confusion is caused by interchanging the different notions of state
> >> >being in use
> >> Much confusion is caused by the lack of an interpretation of quantum
> >> mechanics.
> >QM (Heisenberg, Schrödinger & equivalents) is the mechanics for oscillators
> >without turning points.
> that may describe wave mechanics, but it does not describe qm. qm
> contains the notion of collapse.
I agree that Schrödinger's wave mechanics and Heisenberg's matrix mechanics
doesn't deal with the so-called collapse
> >Most confusion arises from trying to assign to these
> >oscillators properties they just not have got, such as x(t) and p(t).
> >> The issue of interpretation is essentially one of identifying
> >> a correct definition of state.
> >psi(x,t) is a Laplacian state function, while |psi(x,t)| and
> > <psi(x,t)H|psi(x,t)> are Newtonian ones. Both notions exhibit advantages
> >and disadvantages, but they are well defined both in CM and QM.
> We should require that fundamental properties of physics have physical
> definition, not just mathematical definition.
What exactly are "fundamental properties of physics"?
What is your definition of state?
> A metaphysical definition
why do you call it "metaphysical"?
> is not sufficient if physical behaviour (collapse) is not modelled.
psi as state function etc. refer to Schrödinger wave mechanics
> >The quantum numbers are essentially stationary (Newtonian) state numbers
> >(and were originally called so).
> >> >Schrödinger's and Dirac's equations are deterministic
> >> This is again confusing.
> >Not at all
> oh yes it is. :-)
not at all, because psi(x,t) is uniquely determined by psi(x,0)
> >> The wave equations are deterministic, but the mechanics is not.
> >This is confusing
>
> See, I told you it was confusing.
no, you did not - instead you mix the equations and their
interpretation/exploration
> >- let's define 'mechanics'!
> I think mechanics should describe the physical behaviour of matter.
not "the"; electromagnetism describes physical behaviour of matter, too, but
is not mechanics
> >> >Exactly that is the goal of Dirac's 1949 approach to bring special
> >> >relativity and Hamiltonian mechanics together
> >> I think this is done quite effectively in qed.
> >Dirac, seemingly, did not think so.
> QED was not much developed in 1949.
>
> >QED as presented usually (and on your
> >website) presupposes the Lorentz transformation. This is justified only by
> >the results.
> Why do you ignore Einstein?
I'm not ignoring him at all, but he has given a kinematical derivation only,
and this is incomplete, as Dirac's results demonstrate
> >In contrast, Dirac's transformation formulae are justified by
> >their derivation.
> The Lorentz transform is derived on my site, and in many other places.
But all these derivations provide an incomplete foundation only, because they
are only kinematical ones or for free bodies, no one applies to interacting
systems - we have discussed this several times in this group, I don't
understand, why you repeat your view without proof :-(
Best wishes,
Peter
Oh No
>
>
>> Indeed, but unless one is prepared to define these mathematically
>> (i.e. according to the mathematical structure, and without reference
>> to what they mean physically) the usual qm slang suffers from lack
>> of interpretation.
>
>There is no other way to define what quantum theory means than
>mathematics. I'm a proponent of the so-called "Minimal Statistical
>Interpretation" (MSI). Especially I think that the Copenhagen
>interpretation (CI) is more a confusion than an interpretation of
>quantum theory, and EPR were very right to point this confusion out,
>but CI is not necessary to formulate QT as a physical theory.
>
>A very concise introduction to the MSI can be found in
>
>L. Ballentine, Quantum Mechanics, Addison-Wesley
>
>or (shorter) in the paper by the same author
>
>Ballentine, Leslie E.: The Statistical Interpretation of Quantum
>Mechanics, Reviews of Modern Physics 42(4), APS, 358?381, 1970
>
statistically, or probabilistically, but I think we can do better than
just define it in terms of mathematical structure. We want to be able to
say why we have this particular mathematical structure, and we want to
be able to state in words how a quantum state relates to physical
structure. This is what I seek to do at http://www.teleconnection.info/r
qg/FoundationsOfQuantumTheory
Oh No
>
>> Oh No wrote
>> >Thus spake Peter <end...@dekasges.de>
>
>> >> >Much confusion is caused by interchanging the different notions of state
>> >> >being in use
>
>> >> Much confusion is caused by the lack of an interpretation of quantum
>> >> mechanics.
>
>> >QM (Heisenberg, Schrödinger & equivalents) is the mechanics for oscillators
>> >without turning points.
>
>> that may describe wave mechanics, but it does not describe qm. qm
>> contains the notion of collapse.
>
>I agree that Schrödinger's wave mechanics and Heisenberg's matrix mechanics
>doesn't deal with the so-called collapse
>
>
>> >Most confusion arises from trying to assign to these
>> >oscillators properties they just not have got, such as x(t) and p(t).
>
>> >> The issue of interpretation is essentially one of identifying
>> >> a correct definition of state.
>
>> >psi(x,t) is a Laplacian state function, while |psi(x,t)| and
>> > <psi(x,t)H|psi(x,t)> are Newtonian ones. Both notions exhibit advantages
>> >and disadvantages, but they are well defined both in CM and QM.
>
>> We should require that fundamental properties of physics have physical
>> definition, not just mathematical definition.
>
>What exactly are "fundamental properties of physics"?
>
>What is your definition of state?
I have answered this before, and in depth on
http://www.teleconnection.info/rqg/FoundationsOfQuantumTheory
I identify "state" with ket. A ket is a formal conditional clauses
describing the likelihood of particular measurement results.
>> A metaphysical definition
>
>why do you call it "metaphysical"?
The description of wave mechanics as "the mechanics of oscillators...."
is clearly metaphysical if one cannot exhibit physical oscillators. In
fact qm, taken as a whole, does not even allow such physical
oscillators.
>
>> is not sufficient if physical behaviour (collapse) is not modelled.
>
>psi as state function etc. refer to Schrödinger wave mechanics
Not really. It refers to the component <x|f> in a basis of position
states.
>> >> >Schrödinger's and Dirac's equations are deterministic
>
>> >> This is again confusing.
>
>> >Not at all
>
>> oh yes it is. :-)
>
>not at all, because psi(x,t) is uniquely determined by psi(x,0)
The word determinism applies to physical reality, not to the evolution
of probabilities, and it is confusing to misapply it. Moreover the
evolution of psi(x,t) is not uniquely determined in measurement.
>> >> The wave equations are deterministic, but the mechanics is not.
>
>> >This is confusing
>>
>> See, I told you it was confusing.
>
>no, you did not - instead you mix the equations and their
>interpretation/exploration
The thread is about causality. It is about the interpretation of
equations. You confuse things by drawing attention to something which is
not disputed and which has no ultimate bearing on the topic.
>> >- let's define 'mechanics'!
>
>> I think mechanics should describe the physical behaviour of matter.
>
>not "the"; electromagnetism describes physical behaviour of matter, too, but
>is not mechanics
I disagree. e.m. is the mechanics of interaction between electrons and
photons.
>> >> >Exactly that is the goal of Dirac's 1949 approach to bring special
>> >> >relativity and Hamiltonian mechanics together
>
>> >> I think this is done quite effectively in qed.
>
>> >Dirac, seemingly, did not think so.
>
>> QED was not much developed in 1949.
>>
>> >QED as presented usually (and on your
>> >website) presupposes the Lorentz transformation. This is justified only by
>> >the results.
>
>> Why do you ignore Einstein?
>
>I'm not ignoring him at all, but he has given a kinematical derivation only,
>and this is incomplete, as Dirac's results demonstrate
>
>> >In contrast, Dirac's transformation formulae are justified by
>> >their derivation.
>
>> The Lorentz transform is derived on my site, and in many other places.
>
>But all these derivations provide an incomplete foundation only, because they
>are only kinematical ones or for free bodies, no one applies to interacting
>systems - we have discussed this several times in this group, I don't
>understand, why you repeat your view without proof :-(
I do not understand your objection. Lorentz transformation applies to
vector quantities defined in coordinates. There is no distinction
between kinematic and dynamic in this context. I do not see why you say
there is, or why you say the result is incomplete. I do not recall any
such issue in the Dirac paper, but in any case I cannot find it. If you
would send it again I will see what I think he is talking about.
Cl.Massé
news:a8750b2c-1230-42d9...@34g2000hsh.googlegroups.com...
> Another interesting varient on the idea is the possibility that the
> future can influence the past. (Don't worry, I'm not advocating this,
> just presenting it as a philosophical possibility.) There are some
> formulations of electromagnetism that do this, and they apparently
> work just fine. As a conceptual example, consider an accelerating
> charged particle that exchanges a photon with another charged
> particle. We normally describe this as the first particle emitting a
> photon (and suffering a recoil as result) followed by the future
> particle recoiling as it absorbs the photon. Conceptually this could
> also be described as the first particle colliding with a particle in
> the future. Our reluctance to accept the latter point of view hints
> at a key aspect of our mindset: that we think of the future as open
> and not determined yet.
Yet that approach works if we consider that the past is what is known, and
the future is what is unknown. The idea that the future is not known is
very near to the one that it is not determined yet, but distinct. But now,
why isn't it know if it is already fully determined? So if determinism
holds, there is no way to distinguish past and future, a point against it
that escaped our predecessors, and which explains the existence of QM.
Also, if the future can influence the present, it can be known to some
extent and is then no longer the future, unless there is some subtle
"cloaking" mechanism that cancels all information.
--
~~~~ clmasse on free F-country
Liberty, Equality, Profitability.
Rich L.
...
> Also, if the future can influence the present, it can be known to some
> extent and is then no longer the future, unless there is some subtle
> "cloaking" mechanism that cancels all information.
>
> --
> ~~~~ clmasse on free F-country
> Liberty, Equality, Profitability.
Actually, I believe we might be seeing the influence of the future on
the "now". This is point of view is a bit off-the-wall, but I'm
wondering if it might have some advantages.
Consider an atom "that emits a photon". We can observe and measure
the recoil of the atom and estimate, within limits, in what direction
the photon left. The standard interpretation is that a particle, the
photon, is taking the energy emitted by the atom off into space to
find something that will absorb it. In QED this is part of an
interaction between two charged particles. This interaction is
described symetrically in QED such that whether the other particle is
in the past or future is immaterial. If the other particle was in the
past we would say the particle in the past influenced the particle in
the present. When a particle emits a photon into the future, why not
say that the present particle is being influence by the particle in
the future? If the photon's destination is already determined at the
moment of emission, then it would truely be the future interacting
with the past.
There are a couple of catches, of course. First, for this view to
make sense the interaction of the emitted photon must be set already.
This is contrary to our philosophical frame of mind, and the
experimental evidence is subtle, to put it mildly. If you think about
it carefully, the experiments to test Bell's Inequality support this
interpretation, although I don't feel they are conclusive. If the two
entangled photons, by the moment of "emission", have already
determined the charged particles in the two detectors that they will
each interact with, then there is no problem explaining the
polarization correlations. And there is no need to resort to
mysterious "instantaneous wave function collapse" that have no
understandable mechanism and cannot be described in a relativisticly
consistent way.
The second catch is that you don't see any "sinks" in the future,
similar to sources in the past. This probably isn't clear yet. When
we observe a light source, such as a street lamp, we are very aware of
a stream of photons from a particular position in space. We are
absorbing the photons emitted by the light. The analogous object in
the future would be a photon sink that would draw photons out from a
source in the present. We don't observe this: an atom relaxes and
emits a photon seemingly independent of matter around it that might
absorb any emitted photon. However an atom in a box that does not
support an EM mode at the photons frequency would not relax and emit
the photon. It would be stuck in it's excited state. Virtually all
experiments have been conducted in an environment where the photon has
a wide choice of destination particles available to choose from. In
these conditions it appears that the relaxation time constant can be
calculated without detailed knowledge of the configuration of the
matter in the future.
This is not a totally new idea, just an unpopular one. Bell
recognized this explaination of the experiments, but dismissed it
immediately. People do not WANT to be part of a universe whose future
is fixed already. We like to think that we can, by our choices,
influence the future. It certainly SEEMS like we are able to do
that. I'm not convinced this alternate view is either correct or
advances our understanding yet. What I am conviced of, however, is
that after almost a century of quantum mechanics and relativity
theories, we are still stuck reconcilling them to each other. I
suspect there is some fundamental conceptual roadblock that is keeping
us from seeing how they are compatible and how to proceed with further
developments. As such I think it would be useful for more people to
consider crazy ideas like this, in the hopes that someone will be
inspired to come up with the breakthrough idea that moves us forward.
Rich L.
harry
"Peter" <end...@dekasges.de> wrote in message
news:guest.20080815191810$13...@news.killfile.org...
>
> In order to avoid misunderstandings in several (sub)threads, I would like
> to
> propose to find a common understanding of certain basic notions, such as
> 'causality'. (For 'locality', see the foregoing new thread.)
>
> To begin with, my understanding of 'causality' is, that an event has got a
> cause.
>
> The principle of causality states that *all* events have got a cause
> ('nihil
> sine ratione', as Leibnitz said).
>
> obscured by unlucky historical developments. It was incorrectly connected
> with classical-mechanical motion. Quantum mechanics has revealed that such
> motion does not take place within atoms. Unfortunately, not only the
> classical-mechanical motion was rejected, but also causality in general.
> This
> confusion seems to endure to our days...
>
> Thus, I would like to agree with you, that causality is a universal
> principle, where the concrete form of realization depends on the specific
> area of application.
>
>
> Looking forward,
> Peter
Note the disagreement as indicated in two Wikipedia articles:
- http://en.wikipedia.org/wiki/Causality
(cause must happen before its effect)
- http://en.wikipedia.org/wiki/Causality_(physics)
(cause must happen before or at same time as its effect)
I propose to stick to the causality idea of the physics article, according
to which (for all practical purpose) a cause may happen simultaneously with
its effect.
Regards,
Harald
Oh No
>
>Note the disagreement as indicated in two Wikipedia articles:
>- http://en.wikipedia.org/wiki/Causality
>(cause must happen before its effect)
>- http://en.wikipedia.org/wiki/Causality_(physics)
>(cause must happen before or at same time as its effect)
>
>I propose to stick to the causality idea of the physics article, according
>to which (for all practical purpose) a cause may happen simultaneously with
>its effect.
The physics article is wrong here. If two events are simultaneous, they
can only be causally related if they are not separated in space. If we
have only one spacetime coordinate, we have only one event. I.e. this
must refer to two different descriptions of the same thing, thereby
confusing logical cause (meaning merely restatement in different words)
with physical cause. The example given, of Newton's second law, in which
it is (dubiously) stated that Force is cause and acceleration is effect
is incorrect in claiming that the two are simultaneous, since
acceleration simply means that velocity has changed *after* the event.
There is another problem with the notion of cause preceding effect. This
is a view which we impose on the world as a result of our experience of
the macroscopic behaviour of matter, resulting from the law of entropy.
The laws of physics, otoh, are time symmetrical. I do not consider that
we can justify this constraint at a fundamental level, and nor do I even
see how to describe it in a quantum context.
harry
"Rich L." <ralivi...@sbcglobal.net> wrote in message
news:83f1dade-bf8f-4ac2...@x35g2000hsb.googlegroups.com...
Yes that's an interesting approach. If 'm not mistaken, it's called
absorption theory.
As a slight distraction: sometimes it's attempts are made to make it
plausible by stating that for a photon zero time passes "in flight" between
emission and absorption and the distance is therefore like zero.
However, after doing some reflection on that argument, I consider it a fake
argument: for a photon, nothing happens in flight and the universe is a
point. IOW, such a description simply compresses the universe to zero. That
simply obscures a meaningful physical description.
> There are a couple of catches, of course. First, for this view to
> make sense the interaction of the emitted photon must be set already.
> This is contrary to our philosophical frame of mind, and the
> experimental evidence is subtle, to put it mildly. If you think about
> it carefully, the experiments to test Bell's Inequality support this
> interpretation, although I don't feel they are conclusive. If the two
> entangled photons, by the moment of "emission", have already
> determined the charged particles in the two detectors that they will
> each interact with, then there is no problem explaining the
> polarization correlations. And there is no need to resort to
> mysterious "instantaneous wave function collapse" that have no
> understandable mechanism and cannot be described in a relativisticly
> consistent way.
Indeed, that's very attractive. However, it also implies that the whole
universe is pre-determined: nothing you will do can hinder an emitted photon
from reaching its intended destination. Anyway, you also emphasize that
below.
Yes crazy ideas must be considered; and also seemingly disproved
common-sense ideas must be kept in mind, as proves are sometimes found to be
flawed.
Regards,
Harald
Ken S. Tucker
On Aug 25, 5:50 am, Oh No <N...@charlesfrancis.wanadoo.co.uk> wrote:
...
> There is another problem with the notion of cause preceding effect. This
> is a view which we impose on the world as a result of our experience of
> the macroscopic behaviour of matter, resulting from the law of entropy.
> The laws of physics, otoh, are time symmetrical. I do not consider that
> we can justify this constraint at a fundamental level, and nor do I even
> see how to describe it in a quantum context.
Perhaps we need to understand Action (h), Energy (E=h/t)
and Power (E/t=h/t^2). There seems to be a tendency to
focus on Energy and it's conservation.
Is Action conserved? Is Power conserved?
Let's consider Action. System S1 passes an action h
to system S2,
S1-h ---->h----> h+S2
using any of the 4 known conventional forces, we
term, gravity, EM, weak and strong. The Action h
is held to be a constant and invariant, apart from
the nature of the causal force.
The concept of Action also has angular momentum
(spin) and also the product of two charges "a" and "b",
each is fundamental.
Where the so-called "arrow of time" is concerned,
Power has a t^2 so it makes NO diff if "t" is +/- , to
give us a lead. A problem still exists (for me) that
the charge product can + or - that is h = +/- |h|,
but that renders the concepts of "repulsion" and
"attraction".
We theoreticians seem bound by rules that respect
the indivisability of "h", so an ordinary derivative
CANNOT be employed to "h", there is no such
thing as "dh", empirically, (that I ever read).
Let's use "D" as a finite increment and study an
implication, such as, E = h/t == Dh/Dt.
That Dh/Dt cannot be taken to the limit dh/dt, so
that means we need a new calculus.
Regards
Ken S. Tucker
Peter
> The physics article is wrong here. If two events are simultaneous, they
> can only be causally related if they are not separated in space. If we
> have only one spacetime coordinate, we have only one event. I.e. this
> must refer to two different descriptions of the same thing, thereby
> confusing logical cause (meaning merely restatement in different words)
> with physical cause. The example given, of Newton's second law, in which
> it is (dubiously) stated that Force is cause and acceleration is effect
> is incorrect in claiming that the two are simultaneous, since
> acceleration simply means that velocity has changed *after* the event.
Yes, poor Newton! He wrote
Lex II. The change in motion [momentum, see Definitions] is proportional to
the force impressed and proceeds in the direction of that force.
Here is no 2nd time-derivative.
His "hypotheses non fingo" is also misunderstood in the Wikipedia article
(where also Mach is quoted without warning that he has well described only
what he had actually understood, where he claimed to have all understood),
but this is another story.
Back to Lex II above. It states that the momentum change is determined by the
external force applied. Doesn't this qualify the latter as cause for the
former?
What means "after" in an equal-time formula like (p=mv, m=const)
m dv(t) = F(t) dt ?
Best wishes,
Peter
harry
"Oh No" <No...@charlesfrancis.wanadoo.co.uk> wrote in message
news:8mJu5fBY...@charlesfrancis.wanadoo.co.uk...
> Thus spake harry <harald.vanlin...@epfl.ch>
>>
>>Note the disagreement as indicated in two Wikipedia articles:
>>- http://en.wikipedia.org/wiki/Causality
>>(cause must happen before its effect)
>>- http://en.wikipedia.org/wiki/Causality_(physics)
>>(cause must happen before or at same time as its effect)
>>
>>I propose to stick to the causality idea of the physics article, according
>>to which (for all practical purpose) a cause may happen simultaneously
>>with
>>its effect.
>
> The physics article is wrong here. If two events are simultaneous, they
> can only be causally related if they are not separated in space.
That article doesn't suggest otherwise! The essential point is that a single
physical event (whereby any physical event has a certain duration) can
describe a causal relationship.
> If we
> have only one spacetime coordinate, we have only one event. I.e. this
> must refer to two different descriptions of the same thing, thereby
> confusing logical cause (meaning merely restatement in different words)
> with physical cause. The example given, of Newton's second law, in which
> it is (dubiously) stated that Force is cause and acceleration is effect
> is incorrect in claiming that the two are simultaneous, since
> acceleration simply means that velocity has changed *after* the event.
Within the boundaries of Newtonian theory, any force that is applied at a
body causes the simultaneous acceleration of that body - and such an
acceleration takes place over a certain period of time. It's a property of
acceleration that the velocity has changed after the acceleration period.
Interestingly, the amount of acceleration is caused by both the quantity of
force and the quantity of mass which resists the acceleration; moreover, the
description can be refined by considering the particles of which a body is
made up, together with its elasticity. Probably your theory differs from
that of Newton by postulating a time delay for any effect, even at a
fundamental level; but not all theories contain such an assumption and it
does not affect their causality.
> There is another problem with the notion of cause preceding effect. This
> is a view which we impose on the world as a result of our experience of
> the macroscopic behaviour of matter, resulting from the law of entropy.
I hoped to emphasize with the above references that the causality view is
not based on entropy considerations (what did Newton think about entropy?);
thus I think that the law of entropy is based on the idea of causality and
not the inverse.
> The laws of physics, otoh, are time symmetrical. I do not consider that
> we can justify this constraint at a fundamental level, and nor do I even
> see how to describe it in a quantum context.
Our memory is causal. Time symmetry of physical laws does not imply memories
coming from the future. Physical processes are described as function of
time, and as I proposed in the thread on time:
"Time (duration) is a standard of measure for the progress of physical
processes, based on a cyclical process. The old standard used the rotation
period of the earth relative to the sun. Nowadays the standard is based on
the natural resonance period of caesium atoms under standard conditions.
Interestingly, physical processes don't regress by definition; consequently,
physical time can only proceed and not regress either."
Murray called that "an operational definition we all know", but perhaps you
disagree. :-)
Cheers,
Harald
Oh No
>
>"Rich L." <ralivi...@sbcglobal.net> wrote in message
>news:83f1dade-bf8f-4ac2...@x35g2000hsb.googlegroups.com...
>>> Also, if the future can influence the present, it can be known to some
This put me in mind of the quantum zeno effect
http://en.wikipedia.org/wiki/Zeno_effect
How can watching for a radioactive decay affect whether that decay takes
place if the future cannot affect the past?
Oh No
>Hi Charles.
>On Aug 25, 5:50 am, Oh No <N...@charlesfrancis.wanadoo.co.uk> wrote:
> ...
>> There is another problem with the notion of cause preceding effect. This
>> is a view which we impose on the world as a result of our experience of
>> the macroscopic behaviour of matter, resulting from the law of entropy.
>> The laws of physics, otoh, are time symmetrical. I do not consider that
>> we can justify this constraint at a fundamental level, and nor do I even
>> see how to describe it in a quantum context.
>
>Perhaps we need to understand Action (h), Energy (E=h/t)
>and Power (E/t=h/t^2). There seems to be a tendency to
>focus on Energy and it's conservation.
>Is Action conserved? Is Power conserved?
>
>Let's consider Action. System S1 passes an action h
>to system S2,
>
>S1-h ---->h----> h+S2
>
>using any of the 4 known conventional forces, we
>term, gravity, EM, weak and strong. The Action h
>is held to be a constant and invariant, apart from
>the nature of the causal force.
>
>The concept of Action also has angular momentum
>(spin) and also the product of two charges "a" and "b",
>each is fundamental.
In my view, action is no more than a mathematical trick. It is not
fundamental.
>
>Where the so-called "arrow of time" is concerned,
>Power has a t^2 so it makes NO diff if "t" is +/- , to
>give us a lead. A problem still exists (for me) that
>the charge product can + or - that is h = +/- |h|,
>but that renders the concepts of "repulsion" and
>"attraction".
>
>We theoreticians seem bound by rules that respect
>the indivisability of "h", so an ordinary derivative
>CANNOT be employed to "h", there is no such
>thing as "dh", empirically, (that I ever read).
Actually there is such a concept. The principle of least action makes
dh=0. We only have empirical law when action is minimised.
Oh No
>Back to Lex II above. It states that the momentum change is determined
>by the external force applied. Doesn't this qualify the latter as cause
>for the former?
>
>What means "after" in an equal-time formula like (p=mv, m=const)
>
> m dv(t) = F(t) dt ?
>
to the difference in two times, t and t + dt, where t + dt is after t.
When we form the limit, we refer to this difference for all small times
delta_t>0, not for delta_t=0, and we may only talk about t + dt -> t, we
may not set t+dt = t. In the early days of calculus, many false were
found from this mistake.
Oh No
>
>"Oh No" <No...@charlesfrancis.wanadoo.co.uk> wrote in message
>news:8mJu5fBY...@charlesfrancis.wanadoo.co.uk...
>> Thus spake harry <harald.vanlin...@epfl.ch>
>>>
>>>Note the disagreement as indicated in two Wikipedia articles:
>>>- http://en.wikipedia.org/wiki/Causality
>>>(cause must happen before its effect)
>>>- http://en.wikipedia.org/wiki/Causality_(physics)
>>>(cause must happen before or at same time as its effect)
>>>
>>>I propose to stick to the causality idea of the physics article, according
>>>to which (for all practical purpose) a cause may happen simultaneously
>>>with
>>>its effect.
>>
>> The physics article is wrong here. If two events are simultaneous, they
>> can only be causally related if they are not separated in space.
>
>That article doesn't suggest otherwise!
Nor did I say it did. You have interrupted a logical argument mid
deduction.
>The essential point is that a single
>physical event (whereby any physical event has a certain duration) can
>describe a causal relationship.
An event is defined to be instantaneous. I.e. no duration.
>> If we
>> have only one spacetime coordinate, we have only one event. I.e. this
>> must refer to two different descriptions of the same thing, thereby
>> confusing logical cause (meaning merely restatement in different words)
>> with physical cause. The example given, of Newton's second law, in which
>> it is (dubiously) stated that Force is cause and acceleration is effect
>> is incorrect in claiming that the two are simultaneous, since
>> acceleration simply means that velocity has changed *after* the event.
>
>Within the boundaries of Newtonian theory, any force that is applied at a
>body causes the simultaneous acceleration of that body - and such an
>acceleration takes place over a certain period of time. It's a property of
>acceleration that the velocity has changed after the acceleration period.
>Interestingly, the amount of acceleration is caused by both the quantity of
>force and the quantity of mass which resists the acceleration; moreover, the
>description can be refined by considering the particles of which a body is
>made up, together with its elasticity. Probably your theory differs from
>that of Newton by postulating a time delay for any effect, even at a
>fundamental level; but not all theories contain such an assumption and it
>does not affect their causality.
This has nothing to do with my theory. You are misreading what I said
about Newton's law.
>> There is another problem with the notion of cause preceding effect. This
>> is a view which we impose on the world as a result of our experience of
>> the macroscopic behaviour of matter, resulting from the law of entropy.
>
>I hoped to emphasize with the above references that the causality view is
>not based on entropy considerations (what did Newton think about entropy?);
>thus I think that the law of entropy is based on the idea of causality and
>not the inverse.
Newton knew nothing of entropy, as the concept has only been understood
since Boltzman. Your claim that entropy is based on causality is a
mathematical fallacy. Do not have opinion on something for which there
is mathematical proof.
>
>> The laws of physics, otoh, are time symmetrical. I do not consider that
>> we can justify this constraint at a fundamental level, and nor do I even
>> see how to describe it in a quantum context.
>
>Our memory is causal. Time symmetry of physical laws does not imply memories
>coming from the future.
Indeed not. Memory of the past is seen as a consequence of entropy, and
the law of entropy is derived from time symmetrical laws (together with
an initial condition which is obviously not time symmetrical).
>Physical processes are described as function of
>time, and as I proposed in the thread on time:
>"Time (duration) is a standard of measure for the progress of physical
>processes, based on a cyclical process. The old standard used the rotation
>period of the earth relative to the sun. Nowadays the standard is based on
>the natural resonance period of caesium atoms under standard conditions.
>Interestingly, physical processes don't regress by definition; consequently,
>physical time can only proceed and not regress either."
>
>Murray called that "an operational definition we all know", but perhaps you
>disagree. :-)
It agrees perfectly with the definition I have given at
http://www.teleconnection.info/rqg/FoundationsOfSpecialRelativity
There is no other definition of the quantity time used in physics.
However, it is possible to count backwards. What we mean by time
reversal symmetry is that if you count backwards, a law remains
unchanged. The law of entropy is not time symmetric. Almost all other
laws, and all fundamental laws are time symmetric.
Cl.Massé
It is the transactional interpretation of QM. Though being promising, I
don't think that today it has been formulated in a working manner. Anyway,
the influence of the future can't be seen even in this interpretation,
because of some "cloaking" mechanism I spoke about. It would be very
enlightening to express special relativity from such a mechanism, where for
example causally forbidden influences exactly cancel, a little like in the
Huygens principle, where the backward wave cancels.
> And there is no need to resort to
> mysterious "instantaneous wave function collapse" that have no
> understandable mechanism and cannot be described in a relativisticly
> consistent way.
Although giving a meaning to past and future, QM seems to imply that the
future is already set. The disentanglement of that paradox is probably the
key of the mystery. In CM, the future is already set and influences the
present, since the equations can be exactly inverted. getting rid of all
of this influence would give a fully random Universe, since the future
shouldn't keep any information from the past. So the theories must preserve
a little of it in a very subtle and "keen" way.
> The second catch is that you don't see any "sinks" in the future,
> similar to sources in the past. This probably isn't clear yet. When
> we observe a light source, such as a street lamp, we are very aware of
> a stream of photons from a particular position in space. We are
> absorbing the photons emitted by the light. The analogous object in
> the future would be a photon sink that would draw photons out from a
> source in the present.
That is because of the second principle of thermodynamics.
> We don't observe this: an atom relaxes and
> emits a photon seemingly independent of matter around it that might
> absorb any emitted photon. However an atom in a box that does not
> support an EM mode at the photons frequency would not relax and emit
> the photon. It would be stuck in it's excited state.
There is an explanation of that, it isn't necessary to introduce the
influence of the future. Actually, there is a quantized electromagnetic
field that evolves according to the wave equations, until a measurement is
done. No experiment puts the atom in the box and measure the recoil in less
time than the one necessary for the light to cross the box. The field
reflected by the walls of the box cancels unallowed transitions through
destructive interferences. Described in a sequential manner, the photon is
emitted, then its reflection causes the inverse transition. This phenomenon
can't be observed, since the measurement disturbs the system.
> Virtually all
> experiments have been conducted in an environment where the photon has
> a wide choice of destination particles available to choose from. In
> these conditions it appears that the relaxation time constant can be
> calculated without detailed knowledge of the configuration of the
> matter in the future.
There is no way to perform an experiment to test that, since it would entail
the knowledge of the environment, which is then no longer in the future. It
is a purely interpretational issue.
> This is not a totally new idea, just an unpopular one. Bell
> recognized this explaination of the experiments, but dismissed it
> immediately. People do not WANT to be part of a universe whose future
> is fixed already. We like to think that we can, by our choices,
> influence the future. It certainly SEEMS like we are able to do
> that. I'm not convinced this alternate view is either correct or
> advances our understanding yet. What I am conviced of, however, is
> that after almost a century of quantum mechanics and relativity
> theories, we are still stuck reconcilling them to each other. I
> suspect there is some fundamental conceptual roadblock that is keeping
> us from seeing how they are compatible and how to proceed with further
> developments. As such I think it would be useful for more people to
> consider crazy ideas like this, in the hopes that someone will be
> inspired to come up with the breakthrough idea that moves us forward.
Even a fully deterministic theory doesn't necessarily fix the future, like
the chaos theory. With it, the future allows a more precise determination of
the present than a direct measurement. So, such an idea probably need to
give up linearity, then also the superposition principle, therefore in a
very "keen" way.
Rich L.
> the chaos theory. With it, the future allows a more precise determination of
> the present than a direct measurement. So, such an idea probably need to
> give up linearity, then also the superposition principle, therefore in a
> very "keen" way.
>
> --
> ~~~~ clmasse on free F-country
> Liberty, Equality, Profitability.
>
I have a suspicion that part of the answer of integrating QM with GR
will be dealing with nonlinear effects. The process of
renormalization in QED impresses me a bit like a sophisticated version
of epicycles. It is an aproximation technique that gets arbitrarily
close to the correct answer if you include enough terms. And if the
effect you are trying to aproximate is not too different from the
sorts of effects the basic math can simulate. The problem, it seems
to me, is GR is not naturally represented by solutions to
Schrodinger's Equation and requires more than an infinite number of
terms. If we could formulate our physics in a fundamentally nonlinear
way, and if we had the mathematics to work with such things, I suspect
it would be a big step towards getting QM and GR to work together.
But it is far from clear.
Rich L.
Ken S. Tucker
> Thus spake Ken S. Tucker <dynam...@vianet.on.ca>
I think, prior to conclusions, we might consider
Weinbergs "Grav&Cosmo", chp 12, "THE ACTION
PRINCIPLE", since he worked hard to write it.
Eq.(12.3.2), begins an interest followed by,
"*Thus the energy-momentum tensor... if and only
if the matter action is a scalar*".
Ok, I can produce improvements, on his approach,
but "mathematical trick" is disagreeable.
> >Where the so-called "arrow of time" is concerned,
> >Power has a t^2 so it makes NO diff if "t" is +/- , to
> >give us a lead. A problem still exists (for me) that
> >the charge product can + or - that is h = +/- |h|,
> >but that renders the concepts of "repulsion" and
> >"attraction".
>
> >We theoreticians seem bound by rules that respect
> >the indivisability of "h", so an ordinary derivative
> >CANNOT be employed to "h", there is no such
> >thing as "dh", empirically, (that I ever read).
>
> Actually there is such a concept. The principle of least action makes
> dh=0. We only have empirical law when action is minimised.
Thanks for the correction.
> Regards
> Charles Francis
> http://www.teleconnection.info/rqg/MainIndex
Cheers
Ken S. Tucker
Peter
> Thus spake Oh No <No...@charlesfrancis.wanadoo.co.uk>
> >Back to Lex II above. It states that the momentum change is determined
> >by the external force applied. Doesn't this qualify the latter as cause
> >for the former?
> >
> >What means "after" in an equal-time formula like (p=mv, m=const)
> >
> > m dv(t) = F(t) dt ?
> This is a differential formula, and therefore not equal time. It refers
> to the difference in two times, t and t + dt, where t + dt is after t.
> When we form the limit, we refer to this difference for all small times
> delta_t>0, not for delta_t=0, and we may only talk about t + dt -> t, we
> may not set t+dt = t. In the early days of calculus, many false were
> found from this mistake.
I agree that at the moment where the force starts its action, the velocity is
not yet changed. In the eq. above, at time=t, dt=0 => dv=0.
v(t+dt) is hidden in that eq., this has seduced me ;-)
Thank you,
Peter
harry
>> Oh No <No...@charlesfrancis.wanadoo.co.uk> writes:
>
>> The physics article is wrong here. If two events are simultaneous, they
>> can only be causally related if they are not separated in space. If we
>> have only one spacetime coordinate, we have only one event. I.e. this
>> must refer to two different descriptions of the same thing, thereby
>> confusing logical cause (meaning merely restatement in different words)
>> with physical cause. The example given, of Newton's second law, in which
>> it is (dubiously) stated that Force is cause and acceleration is effect
>> is incorrect in claiming that the two are simultaneous, since
>> acceleration simply means that velocity has changed *after* the event.
>
> Yes, poor Newton! He wrote
>
> Lex II. The change in motion [momentum, see Definitions] is proportional
> to
> the force impressed and proceeds in the direction of that force.
>
> Here is no 2nd time-derivative.
Indeed. Neither is any time delay assumed.
> His "hypotheses non fingo" is also misunderstood in the Wikipedia article
> (where also Mach is quoted without warning that he has well described only
> what he had actually understood, where he claimed to have all understood),
> but this is another story.
>
> Back to Lex II above. It states that the momentum change is determined by
> the
> external force applied. Doesn't this qualify the latter as cause for the
> former?
Obviously.
> What means "after" in an equal-time formula like (p=mv, m=const)
>
> m dv(t) = F(t) dt ?
I see no "after". It's about a physical process with a duration dt (which
can be made arbitrarily small) that yields - during dt - a change in
velocity dv. The change in velocity is not supposed to suddenly happen at
the end of the (arbitrary) time interval! I also don't know any commentators
who read an action-reaction time delay into that (see e.g. Poincare in
http://www.thesciencebookstore.com/etext/poincarephysics.html )
Regards,
Harald
Peter
> In my view, action is no more than a mathematical trick.
This is too few; the heuristic value should not be underestimated
> It is not fundamental.
This is correct, if "fundamental" are only those entities which have got a
fundamental unit of measurment
> ... The principle of least action makes
> dh=0. We only have empirical law when action is minimised.
I would rather say that the invariance of h makes the application of the
calculus of variations within quantum theory doubtful. The virtual
neighbouring integration paths would yield h+&h, if existing. Bound
stationary wave functions have got 'neighbouring' ones that diverge
unphysically (beyond L2).
Best wishes,
Peter
harry
> Thus spake harry <harald.vanlin...@epfl.ch>
>>
>>"Oh No" <No...@charlesfrancis.wanadoo.co.uk> wrote in message
>>news:8mJu5fBY...@charlesfrancis.wanadoo.co.uk...
>>> Thus spake harry <harald.vanlin...@epfl.ch>
>>>>
>>>>Note the disagreement as indicated in two Wikipedia articles:
>>>>- http://en.wikipedia.org/wiki/Causality
>>>>(cause must happen before its effect)
>>>>- http://en.wikipedia.org/wiki/Causality_(physics)
>>>>(cause must happen before or at same time as its effect)
>>>>
>>>>I propose to stick to the causality idea of the physics article,
>>>>according
>>>>to which (for all practical purpose) a cause may happen simultaneously
>>>>with
>>>>its effect.
>>>
>>> The physics article is wrong here. If two events are simultaneous, they
>>> can only be causally related if they are not separated in space.
>>
>>That article doesn't suggest otherwise!
>
> Nor did I say it did. You have interrupted a logical argument mid
> deduction.
>
>>The essential point is that a single
>>physical event (whereby any physical event has a certain duration) can
>>describe a causal relationship.
>
> An event is defined to be instantaneous. I.e. no duration.
Instantaneous does not imply no duration. All physical events have a
duration - zero duration has zero existence.
>>> If we
>>> have only one spacetime coordinate, we have only one event. I.e. this
>>> must refer to two different descriptions of the same thing, thereby
>>> confusing logical cause (meaning merely restatement in different words)
>>> with physical cause. The example given, of Newton's second law, in which
>>> it is (dubiously) stated that Force is cause and acceleration is effect
>>> is incorrect in claiming that the two are simultaneous, since
>>> acceleration simply means that velocity has changed *after* the event.
>>
>>Within the boundaries of Newtonian theory, any force that is applied at a
>>body causes the simultaneous acceleration of that body - and such an
>>acceleration takes place over a certain period of time. It's a property of
>>acceleration that the velocity has changed after the acceleration period.
>>Interestingly, the amount of acceleration is caused by both the quantity
>>of
>>force and the quantity of mass which resists the acceleration; moreover,
>>the
>>description can be refined by considering the particles of which a body is
>>made up, together with its elasticity. Probably your theory differs from
>>that of Newton by postulating a time delay for any effect, even at a
>>fundamental level; but not all theories contain such an assumption and it
>>does not affect their causality.
>
> This has nothing to do with my theory. You are misreading what I said
> about Newton's law.
Causality is used in different theories, and in Newtonian theory the basic
action-reaction causality is supposed to happen without delay. Delay times
even were seen as a complication for Newton's principle, see for example
Poincare ("but it will not be simultaneous"):
http://www.thesciencebookstore.com/etext/poincarephysics.html
>>> There is another problem with the notion of cause preceding effect. This
>>> is a view which we impose on the world as a result of our experience of
>>> the macroscopic behaviour of matter, resulting from the law of entropy.
>>
>>I hoped to emphasize with the above references that the causality view is
>>not based on entropy considerations (what did Newton think about
>>entropy?);
>>thus I think that the law of entropy is based on the idea of causality and
>>not the inverse.
>
> Newton knew nothing of entropy, as the concept has only been understood
> since Boltzman.
That's what I meant.
> Your claim that entropy is based on causality is a
> mathematical fallacy. Do not have opinion on something for which there
> is mathematical proof.
That's interesting for sure - note that I did not make a claim. What I
remarked is that your claim (about the notion of cause preceding effect
resulting from the law of entropy) does not seem to fit with history.
Anyway, it's immaterial for this thread.
>>> The laws of physics, otoh, are time symmetrical. I do not consider that
>>> we can justify this constraint at a fundamental level, and nor do I even
>>> see how to describe it in a quantum context.
>>
>>Our memory is causal. Time symmetry of physical laws does not imply
>>memories
>>coming from the future.
>
> Indeed not. Memory of the past is seen as a consequence of entropy, and
> the law of entropy is derived from time symmetrical laws (together with
> an initial condition which is obviously not time symmetrical).
That may be your view. I see memory of the past as a consequence of cause
and effect; and as I pointed out, this probably also was the general view at
the time that the causality principle was developed - simply because entropy
was not considered at that time.
>>Physical processes are described as function of
>>time, and as I proposed in the thread on time:
>>"Time (duration) is a standard of measure for the progress of physical
>>processes, based on a cyclical process. The old standard used the rotation
>>period of the earth relative to the sun. Nowadays the standard is based on
>>the natural resonance period of caesium atoms under standard conditions.
>>Interestingly, physical processes don't regress by definition;
>>consequently,
>>physical time can only proceed and not regress either."
>>
>>Murray called that "an operational definition we all know", but perhaps
>>you
>>disagree. :-)
>
> It agrees perfectly with the definition I have given at
>
> http://www.teleconnection.info/rqg/FoundationsOfSpecialRelativity
>
> There is no other definition of the quantity time used in physics.
That's good news, for it allows meaningful discussions.
> However, it is possible to count backwards. What we mean by time
> reversal symmetry is that if you count backwards, a law remains
> unchanged. The law of entropy is not time symmetric. Almost all other
> laws, and all fundamental laws are time symmetric.
Yes.
Regards,
Harald
Oh No
>
>"Oh No" <No...@charlesfrancis.wanadoo.co.uk> wrote in message
>news:FAUAgIEu...@charlesfrancis.wanadoo.co.uk...
>> Thus spake harry <harald.vanlin...@epfl.ch>
>>>The essential point is that a single
>>>physical event (whereby any physical event has a certain duration) can
>>>describe a causal relationship.
>>
>> An event is defined to be instantaneous. I.e. no duration.
>
>Instantaneous does not imply no duration. All physical events have a
>duration - zero duration has zero existence.
I should have said instantaneous and no duration are synonymous. We have
a semantic issue. An event in theoretical physics is defined to mean a
particular space-time coordinate. This is, of course, an idealisation.
You are talking of processes. I think you may reasonably say that
physics should be described in terms of processes, rather than events,
but we should try to ensure we are speaking the same language.
At any point in history certain laws must be taken as fundamental. Later
generations may find more fundamental properties, or (as in this case)
theorems, which show that what was once considered fundamental need no
longer be taken as such. I am less interested in the history than in the
enquiry into what is really fundamental.
In this instance I think there is a reasonably clear logical path (as
distinct from historical path of discovery).
Time symmetric fundamental laws + assymmetric boundary conditions
----> entropy
----> the classical notion of cause and effect as reflected in human
memory
There is a second classical idea bearing on causality, namely
determinism, but in this thread we have also been trying to understand
what causality might mean in a quantum context, when we do not have
determinism. Interestingly enough classical statistical mechanics, in
which probabilities are governed by unknowns in a deterministic theory,
extends very easily to incorporate statistics governed by probabilities
from quantum theory. The logical path described above continues to
apply. That is all very well for statistical results, but I have been
seeking a definition of causality which applies to individual process in
quantum mechanics.
harry
>> Thus spake Oh No <No...@charlesfrancis.wanadoo.co.uk>
>
>>> Back to Lex II above. It states that the momentum change is
>>> determined by the external force applied. Doesn't this qualify the
>>> latter as cause for the former?
>>>
>>> What means "after" in an equal-time formula like (p=mv, m=const)
>>>
>>> m dv(t) = F(t) dt ?
>
>> This is a differential formula, and therefore not equal time. It
>> refers to the difference in two times, t and t + dt, where t + dt is
>> after t. When we form the limit, we refer to this difference for all
>> small times delta_t>0, not for delta_t=0, and we may only talk about
>> t + dt -> t, we may not set t+dt = t. In the early days of calculus,
>> many false were found from this mistake.
>
> I agree that at the moment where the force starts its action, the
> velocity is not yet changed.
That's right of course. The interaction (which causes a change of state)
starts at time t, and it ends at time t+dt. The change of velocity happens
during the period dt, while the force is applied.
> In the eq. above, at time=t, dt=0 =>
> dv=0.
>
> v(t+dt) is hidden in that eq., this has seduced me ;-)
Wow! v(t) simply means v at any instant t, as function of t; and F(t) in the
same equation means F at the same instant t. Isn't that very basic?
Cheers,
Harald
======================================= MODERATOR'S COMMENT:
I recommend to consult Huygens' principle in mechanics ;-)
Peter
>
> "harry" <harald.vanlin...@epfl.ch> writes:
>> "Oh No" <No...@charlesfrancis.wanadoo.co.uk> wrote
...
> Causality is used in different theories, and in Newtonian theory the basic
> action-reaction causality is supposed to happen without delay. Delay times
> even were seen as a complication for Newton's principle, see for example
> Poincare ("but it will not be simultaneous"):
> http://www.thesciencebookstore.com/etext/poincarephysics.html
Please don't call "Newton's theory" what Newton himself has seen several
times expressis verbis to be a (FAPP, as Bell would say) working mathematical
formula, but not a physical theory
...
> > However, it is possible to count backwards. What we mean by time
> > reversal symmetry is that if you count backwards, a law remains
> > unchanged. The law of entropy is not time symmetric. Almost all other
> > laws, and all fundamental laws are time symmetric.
>
> Yes.
No
Best wishes,
Peter
Rich L.
vs unset future, etc. a crazy thought occured to me. I mention it
here in the spirit of a pholosophical discussion (I think this is the
right group for that), not because I necessarily advocate it.
It occured to me that a lot of the difficulty with causality has to do
with the assumed time sequence. We are always treating time as a
special thing compared to space. (This is natural given how our minds
perceive time differently) What if that is an error? If the future
is as fixed and definite as the past (as speculated earlier), then
perhaps the correct physics is to learn the constraints that apply
between events and particles and other forms of energy in the matrix
of events we call space-time. Our most fundamental equations are
symmetric in time. We generally use them to take a past configuration
and the present configuration and calculate the most likely future
configuration. In principle we could do the reverse using the future
and present (if we could know the future). The appeal of this idea
would be that concepts like "spooky action at a distance" and
"instantaneous wave function collapse" would go away. In their place
would be an understanding of how events in spacetime interconnect,
much like a truss work mechanical structure maintains its
configuration due to the constraints between the joints of the
trusses.
Of course something would have to be done about entropy. There would
have to be some principle that explains why "past" configurations of
the universe always have lower entropy than future configurations. I
don't think saying "because that is the way it is" would be enough (at
least not for me).
Just an idea...
Rich L.
======================================= MODERATOR'S COMMENT:
Have you ever told your employer that you would like to work 8 meters rather than 8 hours per day, and your architect that your terace should be 30 squared minutes? ;-)
Oh No
>While meditating on all this causality/determinism/time symmetry/set
>vs unset future, etc. a crazy thought occured to me. I mention it
>here in the spirit of a pholosophical discussion (I think this is the
>right group for that), not because I necessarily advocate it.
>
>It occured to me that a lot of the difficulty with causality has to do
>with the assumed time sequence. We are always treating time as a
>special thing compared to space. (This is natural given how our minds
>perceive time differently) What if that is an error? If the future
>is as fixed and definite as the past (as speculated earlier), then
>perhaps the correct physics is to learn the constraints that apply
>between events and particles and other forms of energy in the matrix
>of events we call space-time. Our most fundamental equations are
>symmetric in time. We generally use them to take a past configuration
>and the present configuration and calculate the most likely future
>configuration. In principle we could do the reverse using the future
>and present (if we could know the future). The appeal of this idea
>would be that concepts like "spooky action at a distance" and
>"instantaneous wave function collapse" would go away. In their place
>would be an understanding of how events in spacetime interconnect,
>much like a truss work mechanical structure maintains its
>configuration due to the constraints between the joints of the
>trusses.
I do not think this should be considered a crazy thought. Rather it may
be what quantum theory is telling us, and should be considered very
carefully.
>
>Of course something would have to be done about entropy. There would
>have to be some principle that explains why "past" configurations of
>the universe always have lower entropy than future configurations. I
>don't think saying "because that is the way it is" would be enough (at
>least not for me).
>
There is such a principle, namely the big bang, an initial condition in
a low entropy state.
harry
> Thus spake harry <harald.vanlin...@epfl.ch>
>>
>> "Oh No" <No...@charlesfrancis.wanadoo.co.uk> wrote in message
>> news:FAUAgIEu...@charlesfrancis.wanadoo.co.uk...
>>> Thus spake harry <harald.vanlin...@epfl.ch>
>
>>>> The essential point is that a single
>>>> physical event (whereby any physical event has a certain duration)
>>>> can describe a causal relationship.
>>>
>>> An event is defined to be instantaneous. I.e. no duration.
>>
>> Instantaneous does not imply no duration. All physical events have a
>> duration - zero duration has zero existence.
>
> I should have said instantaneous and no duration are synonymous.
> We have a semantic issue.
> An event in theoretical physics is defined to
> mean a particular space-time coordinate. This is, of course, an
> idealisation. You are talking of processes.
Idealisations are OK, as long as we don't forget what we talk about. Your
use of "event" is incompatible with the first use of that term in this
thread by Peter: "The principle of causality states that all events have got
a cause" - at a truly single point in time nothing happens and no cause for
nothing is required.
According to the dictionary, the use of "event" as during zero time is
specific for relativity theory; but even so, from the way I have seen it
used it may nevertheless imply a small amount of time (such as a bomb
blast), negligibly small for calculations. Anyway, causality and
specifically action-reaction refer to physical processes (or physical
events, although you reserve that term for a point in time).
< I think you may
> reasonably say that physics should be described in terms of
> processes, rather than events, but we should try to ensure we are
> speaking the same language.
See above. Of course, this thread is very much about talking the same
language.
>> Causality is used in different theories, and in Newtonian theory the
>> basic action-reaction causality is supposed to happen without delay.
Good - that was the only thing I tried to point out! :-)
It adds to the original formulation by Peter, although I'm not sure yet
about how to formulate it.
>> Delay times even were seen as a complication for Newton's principle,
>> see for example Poincare ("but it will not be simultaneous"):
>> http://www.thesciencebookstore.com/etext/poincarephysics.html
[...]
>>>> Our memory is causal. Time symmetry of physical laws does not imply
>>>> memories coming from the future.
>>>
>>> Indeed not. Memory of the past is seen as a consequence of entropy,
>>> and the law of entropy is derived from time symmetrical laws
>>> (together with an initial condition which is obviously not time
>>> symmetrical).
>>
>> That may be your view. I see memory of the past as a consequence of
>> cause and effect; and as I pointed out, this probably also was the
>> general view at the time that the causality principle was developed
>> - simply because entropy was not considered at that time.
>
> At any point in history certain laws must be taken as fundamental.
> Later generations may find more fundamental properties, or (as in
> this case) theorems, which show that what was once considered
> fundamental need no longer be taken as such. I am less interested in
> the history than in the enquiry into what is really fundamental.
>
> In this instance I think there is a reasonably clear logical path (as
> distinct from historical path of discovery).
>
> Time symmetric fundamental laws + assymmetric boundary conditions
> ----> entropy
> ----> the classical notion of cause and effect as reflected in human
> memory
I regard the biological notion of human memory to be an effect from physical
causes such as light and chemistry; thus I agree to disagree on that.
Probably the difference is that for me physical processes are primary, and
physical laws are merely human descriptions of emerging properties of
fundamental processes.
> There is a second classical idea bearing on causality, namely
> determinism, but in this thread we have also been trying to understand
> what causality might mean in a quantum context, when we do not have
> determinism. Interestingly enough classical statistical mechanics, in
> which probabilities are governed by unknowns in a deterministic
> theory, extends very easily to incorporate statistics governed by
> probabilities from quantum theory. The logical path described above
> continues to apply. That is all very well for statistical results,
> but I have been seeking a definition of causality which applies to
> individual process in quantum mechanics.
I wish you success!
Harald
harry
"harry" <harald.vanlin...@epfl.ch> wrote in message
news:1219752...@sicinfo3.epfl.ch...
For the moderator: The Huygens principle is rather graphical, but it can be
expressed in words as follows (from a Dutch physics book, I'll try to
translate it). "Every point [in a medium] that is reached by a wave, acts as
a secondary source of vibration which emits waves in all directions. If a
continuous line of such sources is present, the wave fronts of these
secondary sources combine into resulting wave fronts that are the tangent
lines of the original ones."
Not clear what the Huygens' principle has to do with the above relationship
between force on and acceleration of a single particle, except for using it.
Harald
harry
>
>>
>> "harry" <harald.vanlin...@epfl.ch> writes:
>
>>> "Oh No" <No...@charlesfrancis.wanadoo.co.uk> wrote
> ...
>
>> Causality is used in different theories, and in Newtonian theory the
>> basic
>> action-reaction causality is supposed to happen without delay. Delay
>> times
>> even were seen as a complication for Newton's principle, see for example
>> Poincare ("but it will not be simultaneous"):
>> http://www.thesciencebookstore.com/etext/poincarephysics.html
>
> Please don't call "Newton's theory" what Newton himself has seen several
> times expressis verbis to be a (FAPP, as Bell would say) working
> mathematical
> formula, but not a physical theory
It's unclear to me what the connection is between your comment here and what
I wrote here above... Do you perhaps claim that Newtonian theory does not
contain the action-reaction principle, so that Poincare was terribly
mistaken when he called it "the principle of Newton"?
> ...
>
>> > However, it is possible to count backwards. What we mean by time
>> > reversal symmetry is that if you count backwards, a law remains
>> > unchanged. The law of entropy is not time symmetric. Almost all other
>> > laws, and all fundamental laws are time symmetric.
>>
>> Yes.
>
> No
Please clarify why according to you Charles' above remark is wrong.
Thanks,
Harald
======================================= MODERATOR'S COMMENT:
I thought you will explain why it is correct ;-)
Oh No
>> Oh No <No...@charlesfrancis.wanadoo.co.uk> writes:
>
>> In my view, action is no more than a mathematical trick.
>
>This is too few; the heuristic value should not be underestimated
I do not wish to poor scorn on the value of tricks, but I do not think
the principle of least action gives physical insight.
>> It is not fundamental.
>
>This is correct, if "fundamental" are only those entities which have got a
>fundamental unit of measurment
>
>> ... The principle of least action makes
>> dh=0. We only have empirical law when action is minimised.
>
>I would rather say that the invariance of h makes the application of the
>calculus of variations within quantum theory doubtful. The virtual
>neighbouring integration paths would yield h+&h, if existing. Bound
>stationary wave functions have got 'neighbouring' ones that diverge
>unphysically (beyond L2).
>
I think Feynman's path integral casts some light on the principle. Only
at the minimum do with have constructive interference, otherwise the
contributions of paths cancel. Of course, this means we are not actually
applying the calculus of variations in this context. Moreover, since I
do not accept the physical reality of waves (another mathemagical trick,
imv), I again do not think this gives physical insight.
Oh No
>
>"Peter" <end...@dekasges.de> wrote in message
>news:guest.20080826124138$15...@news.killfile.org...
>>
>>>
>>> "harry" <harald.vanlin...@epfl.ch> writes:
>>
>>>> "Oh No" <No...@charlesfrancis.wanadoo.co.uk> wrote
>>> > reversal symmetry is that if you count backwards, a law remains
>>> > unchanged. The law of entropy is not time symmetric. Almost all other
>>> > laws, and all fundamental laws are time symmetric.
>>>
>>> Yes.
>>
>> No
>
>Please clarify why according to you Charles' above remark is wrong.
>
>Thanks,
>Harald
>
>
>======================================= MODERATOR'S COMMENT:
> I thought you will explain why it is correct ;-)
>
It hardly needs explanation. You might quibble that really I mean PCT
symmetry.
Peter
> Oh No <No...@charlesfrancis.wanadoo.co.uk> writes:
> >>> > However, it is possible to count backwards. What we mean by time
> >>> > reversal symmetry is that if you count backwards, a law remains
> >>> > unchanged. The law of entropy is not time symmetric. Almost all other
> >>> > laws, and all fundamental laws are time symmetric.
...
> It hardly needs explanation. You might quibble that really I mean PCT
> symmetry.
It may not, but a foot passenger like me needs ;-)
Thus, let me ask, is the statement
'the change of the momentum vector is proportional to the applied force
vector'
time symmetric?
I guess, with 'time symmetric' you mean not only t->(-t), but also to change
the sign of certain other quantities, correct?
Thank you,
Peter
harry
>
> "Peter" <end...@dekasges.de> wrote in message
> news:guest.20080826124138$15...@news.killfile.org...
>>
>>> > However, it is possible to count backwards. What we mean by time
>>> > reversal symmetry is that if you count backwards, a law remains
>>> > unchanged. The law of entropy is not time symmetric. Almost all other
>>> > laws, and all fundamental laws are time symmetric.
>>>
>>> Yes.
>>
>> No
>
> Please clarify why according to you Charles' above remark is wrong.
>
> Thanks,
> Harald
>
> ======================================= MODERATOR'S COMMENT: I thought you
> will explain why it is correct ;-)
I guessed that something may have been overlooked. It's generally easier to
show what has been overlooked, than to show that nothing has been
overlooked. :-)
Cheers,
Harald
Oh No
Yes. t -> - t implies dt -> -dt, and hence v = dx/dt -> - v = -dx/dt.
Thus the sign of mv is changed.
Peter
> > 'the change of the momentum vector is proportional to the applied force
> > vector'
> >
> >time symmetric?
> >
> >I guess, with 'time symmetric' you mean not only t->(-t), but also to
> >change the sign of certain other quantities, correct?
> Oh No <No...@charlesfrancis.wanadoo.co.uk> writes:
> Yes. t -> - t implies dt -> -dt, and hence v = dx/dt -> - v = -dx/dt.
> Thus the sign of mv is changed.
I see. Now, let me repeat my question (slightly changed): Is the statement
'the change of the momentum is proportional to the applied force
and along the same direction as that force'
time symmetric?
Thank you,
Peter
Peter
> >>> > However, it is possible to count backwards. What we mean by time
> >>> > reversal symmetry is that if you count backwards, a law remains
> >>> > unchanged. The law of entropy is not time symmetric. Almost all other
> >>> > laws, and all fundamental laws are time symmetric.
> >>> Yes.
> >> No
> > Please clarify why according to you Charles' above remark is wrong.
> >
> > Thanks,
> > Harald
> >
> > ======================================= MODERATOR'S COMMENT:
> > I thought you will explain why it is correct ;-)
> I guessed that something may have been overlooked. It's generally easier to
> show what has been overlooked, than to show that nothing has been
> overlooked. :-)
>
> Cheers,
> Harald
Yes - I should have put my comment in a regular posting rather than in the
Moderator's comment field, sorry!
The overlookings are discussed in the neighbouring subthread (it's not so
straightforward, as you may imagine ;-), please have a look there,
Thank you,
Peter
Oh No
course, the law refers to rate of change of momentum.
Peter
> >> Yes. t -> - t implies dt -> -dt, and hence v = dx/dt -> - v = -dx/dt.
> >> Thus the sign of mv is changed.
> >I see. Now, let me repeat my question (slightly changed): Is the statement
> >
> > 'the change of the momentum is proportional to the applied force
> > and along the same direction as that force'
> >
> >time symmetric?
> Is the answer not obvious, or are you hiding some sneaky trick?
No trick, but I have in mind a certain sequence of questions. Because the 2nd
question could influence the answer to the 1st one, however, I'm hiding it so
far.
> Of course, the law refers to rate of change of momentum.
Does this mean 'no' or 'yes'? (Sorry for my incomplete command of English; I
don't see to what "Of course" refers.)
Thank you,
Peter
Ken S. Tucker
On Aug 27, 6:30 am, Oh No <N...@charlesfrancis.wanadoo.co.uk> wrote:
> Thus spake Peter <end...@dekasges.de>
>
> >> > 'the change of the momentum vector is proportional to the applied force
> >> > vector'
>
> >> >time symmetric?
>
> >> >I guess, with 'time symmetric' you mean not only t->(-t), but also to
> >> >change the sign of certain other quantities, correct?
>
> >> Oh No <N...@charlesfrancis.wanadoo.co.uk> writes:
>
> >> Yes. t -> - t implies dt -> -dt, and hence v = dx/dt -> - v = -dx/dt.
> >> Thus the sign of mv is changed.
>
> >I see. Now, let me repeat my question (slightly changed): Is the statement
>
> > 'the change of the momentum is proportional to the applied force
> > and along the same direction as that force'
>
> >time symmetric?
>
> Is the answer not obvious, or are you hiding some sneaky trick? Of
> course, the law refers to rate of change of momentum.
Consider dv/dt = d^2(x)/(dt^2) = acceleration.
The (dt^2) is a +quantity in either direction of dt.
Sometimes people play a movie film backward
that's the same as dx(forward) = -dx(backward),
with "dx" being spatial, to depict a reversal of time,
but aren't we really reversing the film in spatial
displacement?
Thanks Peter, didn't realize that question until
you made the point of arguing.
> Regards
> Charles Francis
> http://www.teleconnection.info/rqg/MainIndex
I don't see "a trick".
Cheers
Ken S. Tucker
Oh No
>> Oh No <No...@charlesfrancis.wanadoo.co.uk> writes:
>
>> >> Yes. t -> - t implies dt -> -dt, and hence v = dx/dt -> - v = -dx/dt.
>> >> Thus the sign of mv is changed.
>
>
>> >I see. Now, let me repeat my question (slightly changed): Is the statement
>> >
>> > 'the change of the momentum is proportional to the applied force
>> > and along the same direction as that force'
>> >
>> >time symmetric?
>
>
>> Is the answer not obvious, or are you hiding some sneaky trick?
>
>
>No trick, but I have in mind a certain sequence of questions. Because the 2nd
>question could influence the answer to the 1st one, however, I'm hiding it so
>far.
That sounds like a trick.
>
>> Of course, the law refers to rate of change of momentum.
>
>Does this mean 'no' or 'yes'?
>(Sorry for my incomplete command of English; I
>don't see to what "Of course" refers.)
Of course refers to the fact that we should correctly state Newton's
second law, if indeed what you are doing is really physics not a trick.
N2 is time symmetric.
Oh No
>Hi Peter and Charles.
>
>On Aug 27, 6:30 am, Oh No <N...@charlesfrancis.wanadoo.co.uk> wrote:
>> >I see. Now, let me repeat my question (slightly changed): Is the statement
>>
>> > 'the change of the momentum is proportional to the applied force
>> > and along the same direction as that force'
>>
>> >time symmetric?
>>
>> Is the answer not obvious, or are you hiding some sneaky trick? Of
>> course, the law refers to rate of change of momentum.
>
>Consider dv/dt = d^2(x)/(dt^2) = acceleration.
>The (dt^2) is a +quantity in either direction of dt.
>Sometimes people play a movie film backward
>that's the same as dx(forward) = -dx(backward),
>with "dx" being spatial, to depict a reversal of time,
>but aren't we really reversing the film in spatial
>displacement?
No. Space reversal, or Parity Symmetry is also true of classical laws.
We can mirror image the film, and the laws don't change, but this is not
being applied here. The reversal of the sign of v is due to the reversal
of the sign of dt, not that of dx.
>> Regards
>> Charles Francis
>
>I don't see "a trick".
It wouldn't be a trick if one could see it. But don't believe Peter's
protestations of innocence. He will be misstating a question, so that it
asks something silly. Thus he is trying to induce a silly answer. Then
he will snip his original question and laugh, and pretend that any
silliness in what he wrote was due to his command of English.
In this case it is particularly suspicious that he has written change of
momentum, misstating Newton's law and asking about something which is
not, in general, true. If I answer by charitably replying as though he
had asked something sensible, he will pounce.
Cl.Massé
news:1219757...@sicinfo3.epfl.ch...
> I regard the biological notion of human memory to be an effect from
> physical causes such as light and chemistry;
Memory entails a feedback mechanism, for any type of memory including
electronic ones. Also, memory is contradictory with entropy increase, since
it detains information.
--
~~~~ clmasse on free F-country
Liberty, Equality, Profitability.
Peter
> >I see. Now, let me repeat my question (slightly changed): Is the statement
> >
> > 'the change of the momentum is proportional to the applied force
> > and along the same direction as that force'
> >
> >time symmetric?
> Of course refers to the fact that we should correctly state Newton's
> second law, if indeed what you are doing is really physics not a trick.
> N2 is time symmetric.
What means "correctly state Newton's second law"? As it stands, it's a
trivia, iow, I would like to see your formulation.
The formulation above is close to Newton's original formulation (which I do
consider to be fundamental for classical mechanics).
Looking forward,
Peter
Oh No
not made such a trivial mistake unless deliberately. The statement above
is incorrect and, meaningless. N2 refers to the rate of change of
momentum, not the change of momentum. You made have no mention of the
time period over which momentum changes, and one cannot say, in the
absence of knowledge of how force changes during that time period
whether your statement is time symmetric or not. Moreover, the claim I
made referred to the laws of physics, not to your statement which is not
a law.
Peter
> >Thus spake Ken S. Tucker <dyna...@vianet.on.ca>
> >> >I see. Now, let me repeat my question (slightly changed): Is the
> statement
> >>
> >> > 'the change of the momentum is proportional to the applied force
> >> > and along the same direction as that force'
> >>
> >> >time symmetric?
> >> Is the answer not obvious, or are you hiding some sneaky trick? Of
> >> course, the law refers to rate of change of momentum.
> >...I don't see "a trick".
> It wouldn't be a trick if one could see it. But don't believe Peter's
> protestations of innocence. He will be misstating a question, so that it
> asks something silly. Thus he is trying to induce a silly answer. Then
> he will snip his original question and laugh, and pretend that any
> silliness in what he wrote was due to his command of English.
Poor Charles, not knowing Newton's original formulation of Law II, he gets
headache from searching a trick in the sentence
'the change of the momentum is proportional to the applied force
and along the same direction as that force'
In your words, one has, of course, to take the correct website. One (of the
very few, admittedly) is
http://www.gfisher.org/appendix_to_2.htm
with:
** start of quotation **
Newton's Second Law. "The change of motion is proportional to the motive
force impressed and is made in the direction of the right line in which that
force is impressed." <![endif]>
The motion of a body is defined by Newton to be the product of a quantity
called the mass of the body, which measures its reluctance to change its
state, with the velocity of the body, which measures the rate at which its
distance from some reference point is changing, and also specifies a
direction in which the change is taking place. This is called momentum
today. The velocity and/or direction may change at each instant of time.
The change in motion is actually the rate of change of momentum [PE: Why
Newton didn't write 'rate of change of motion'?]. Except in a few simple
cases a quantitative statement that this rate of change of momentum is
proportional to impressed forces requires the techniques of the mathematical
discipline known as calculus. To say the rate of change of momentum is
proportional to the impressed forces is to say that it is some fixed number
multiplied by the quantity which measures the force at each point of space
and instant of time. The particular fixed number or constant to be used is
different for different units of measurement for time, distances and forces
(second or years, meters or feet or miles, pounds or dynes, etc.). Often
impressed forces are different at each point of space, but at any one given
point are the same for each instant of time.
** end of quotation **
Ok, if you think that knowing the 2nd question will make it easier to you,
here it is.
Is the equation
dp = F dt
time reversal invariant?
> In this case it is particularly suspicious that he has written change of
> momentum, misstating Newton's law and asking about something which is
> not, in general, true.
See quoted text above
> If I answer by charitably replying as though he
> had asked something sensible, he will pounce.
I don't understand this sentence, so I cannot comment it (let's hope that
it's nicer than your unfounded accusations above ;-)
Best wishes,
Peter
PS: A moderator should more than anybody else refrain from personal attacks,
shouldn't he?
harry
>>> I see. Now, let me repeat my question (slightly changed): Is the
>>> statement
>>>
>>> 'the change of the momentum is proportional to the applied force
>>> and along the same direction as that force'
>>>
>>> time symmetric?
>
>
>> Of course refers to the fact that we should correctly state Newton's
>> second law, if indeed what you are doing is really physics not a
>> trick. N2 is time symmetric.
>
>
> What means "correctly state Newton's second law"? As it stands, it's a
> trivia, iow, I would like to see your formulation.
Dictionary.com:
"matters or things that are very unimportant, inconsequential, or
nonessential; trifles; trivialities."
That law does seem to be merely a refined definition of "force", as long as
Hooke's Force law is not taken as primary (and perhaps he omitted it because
he did not want to include Hooke?). However, it would perhaps not do history
justice to swallow that; in particular, Newton does not seem to define
"force" in general. If we take bending ability as definition of force, the
second law for "applied force" does become a law since it quantifies a
change in (m*v) as function of such a bending ability. By chance I just
looked up some references to a few of Newton's definitions for another
discussion:
"The quantity of motion is the measure of the same, arising from the
velocity and quantity of matter conjunctly."
"Every body perseveres in its state of rest, or of uniform motion in a right
line, unless it is compelled to change that state by forces impressed
thereon."
- http://members.tripod.com/~gravitee/definitions.htm
- http://members.tripod.com/~gravitee/axioms.htm
Note that the causality of action-reaction is there nicely sketched a mutual
interaction between bodies:
"vis insita, may, by a most significant name, be called vis inertiae, or
force of inactivity. But a body exerts this force only, when another force,
impressed upon it, endeavours to change its condition; and the exercise of
this force may be considered both as resistance and impulse; it is
resistance, in so far as the body, for maintaining its present state,
withstands the force impressed; it is impulse, in so far as the body, by not
easily giving way to the impressed force of another, endeavours to change
the state of that other."
- Note the word "impress" (if correctly translated): Dictionary: c.1374, "to
apply with pressure, make a permanent image in," from L. impressus, pp. of
imprimere "press into or upon, stamp," from in- "into" + premere "to press".
That does appear to be an implicit recognition of a definition of force
according to Hooke's law.
- Note also the choice of the word "when" (if correctly translated) which
does not imply a delay.
> The formulation above is close to Newton's original formulation
> (which I do consider to be fundamental for classical mechanics).
Regards,
Harald
Peter
> Thus spake Peter <end...@dekasges.de>
> >> >I see. Now, let me repeat my question (slightly changed): Is the
> statement
> >> >
> >> > 'the change of the momentum is proportional to the applied force
> >> > and along the same direction as that force'
> >> >
> >> >time symmetric?
> >
> >
> >> Of course refers to the fact that we should correctly state Newton's
> >> second law, if indeed what you are doing is really physics not a trick.
> >> N2 is time symmetric.
> >What means "correctly state Newton's second law"? As it stands, it's a
> >trivia, iow, I would like to see your formulation.
> >
> >The formulation above is close to Newton's original formulation (which I do
> >consider to be fundamental for classical mechanics).
> This is why you are accused of trickery. I find it clear that you have
> not made such a trivial mistake unless deliberately. The statement above
> is incorrect and, meaningless.
You have just stated that Newton's original formulation "is incorrect and,
meaningless".
IOW, you claim that
- the change of momentum is *not* proportional to the applied force,
- the change of momentum is *not* along the direction of that force.
What, then, is dp in dp=Kdt?
> N2 refers to the rate of change of
> momentum, not the change of momentum.
Why Newton wrote "change" and not "rate of change"?
> You made have no mention of the
> time period over which momentum changes
Not me, Newton himself :-)
> , and one cannot say, in the
> absence of knowledge of how force changes during that time period
> whether your statement is time symmetric or not.
This comes close to my goal :-) One cannot make the same statements obout the
time reversal symmetry of
delta p ~ K (a)
and of
dp = K dt (b)
(for symplicity, be K=const). But both (a) and (b) are correct.
> Moreover, the claim I
> made referred to the laws of physics, not to your statement which is not
> a law.
It's N2 in the original formulation - please criticize Newton, not me for his
formulations ;-)
Anyway, I have discussed this in length, in order to learn about the
invariance of equations.
Consider d'Alembert's wave equation (speed=1).
&^2f/&t^2 = &^2f/dr^2 (*)
Only even powers of t enter, dr is invariant, f enters homogeneously => this
eq. is invariant, correct?
Now, consider the set
&f/&t = g
&g/&t = &^2f/dr^2
This set is invariant only, if we require (claim?!) additionally that f or g
switch the sign together with t. Is this not an essential difference to the
manifest invariance of (*)?
Best wishes,
Peter
Oh No
>
>> Thus spake Peter <end...@dekasges.de>
>
>> >> >I see. Now, let me repeat my question (slightly changed): Is the
>> statement
>> >> >
>> >> > 'the change of the momentum is proportional to the applied force
>> >> > and along the same direction as that force'
>> >> >
>> >> >time symmetric?
>> >
>> >
>> >> Of course refers to the fact that we should correctly state Newton's
>> >> second law, if indeed what you are doing is really physics not a trick.
>> >> N2 is time symmetric.
>
>
>> >What means "correctly state Newton's second law"? As it stands, it's a
>> >trivia, iow, I would like to see your formulation.
>> >
>> >The formulation above is close to Newton's original formulation (which I do
>> >consider to be fundamental for classical mechanics).
>
>
>> This is why you are accused of trickery. I find it clear that you have
>> not made such a trivial mistake unless deliberately. The statement above
>> is incorrect and, meaningless.
>
>You have just stated that Newton's original formulation "is incorrect and,
>meaningless".
I have said no such thing. I have never seen it translated as you have
given it.
>
>IOW, you claim that
>- the change of momentum is *not* proportional to the applied force,
>- the change of momentum is *not* along the direction of that force.
>
>What, then, is dp in dp=Kdt?
That is a small change in a small time period, which is as you should
have translated the law.
>
>> N2 refers to the rate of change of
>> momentum, not the change of momentum.
>
>Why Newton wrote "change" and not "rate of change"?
He wrote neither, since he wrote in latin.
>
>> You made have no mention of the
>> time period over which momentum changes
>
>Not me, Newton himself :-)
One would need to study his whole treatise to see what was intended by
the word Mutationem, but it is always translated as rate of change in
motion.
>
>> , and one cannot say, in the
>> absence of knowledge of how force changes during that time period
>> whether your statement is time symmetric or not.
>
>This comes close to my goal :-) One cannot make the same statements obout the
>time reversal symmetry of
>
> delta p ~ K (a)
>
>and of
>
> dp = K dt (b)
>
>(for symplicity, be K=const). But both (a) and (b) are correct.
Even taking a, it is time symmetrical because it says nothing about the
value of the constant of proportionality.
>> Moreover, the claim I
>> made referred to the laws of physics, not to your statement which is not
>> a law.
>
>It's N2 in the original formulation - please criticize Newton, not me for his
>formulations ;-)
>
>
>Anyway, I have discussed this in length, in order to learn about the
>invariance of equations.
>
>Consider d'Alembert's wave equation (speed=1).
>
> &^2f/&t^2 = &^2f/dr^2 (*)
>
>Only even powers of t enter, dr is invariant, f enters homogeneously => this
>eq. is invariant, correct?
>
>Now, consider the set
>
> &f/&t = g
>
> &g/&t = &^2f/dr^2
>
>This set is invariant only, if we require (claim?!) additionally that f or g
>switch the sign together with t. Is this not an essential difference to the
>manifest invariance of (*)?
>
I wouldn't say so.
harry
Compare with
- http://members.tripod.com/~gravitee/axioms.htm
- http://rack1.ul.cs.cmu.edu/is/newton/ p.83:
" LAW II.
The alteration of motion is ever proportional to the motive force impressed;
and is made in the direction of the right line in which that force is
impressed.
If any force generates a motion, a double force will generate double the
motion, a triple force triple the motion, whether that force be impressed
altogether and at once, or gradually and successively. And this motion
(being always directed the same way with the generating force), if the body
moved before, is added to or subtracted from the former motion, according as
they directly conspire with or are directly contrary to each other; or
obliquely joined, when they are oblique, so as to produce a new motion
compounded from the determination of both."
Apparently gfisher's text is based on the above translation by by Andrew
Motte (1729).
[...]
>
> Ok, if you think that knowing the 2nd question will make it easier to
> you, here it is.
>
> Is the equation
>
> dp = F dt
>
> time reversal invariant?
Hmm... I'm starting to wonder! But why not:
dp = F dt
dt'= -dt
F = F
=> dp' = -dp
However, I regard such time inversion as just a movie played backward; cause
and effect are not changed at all.
In contrast, the inverse physical process would be:
dp = F dt
dt'= dt
F = -F
=> dp' = -dp
>> In this case it is particularly suspicious that he has written
>> change of momentum, misstating Newton's law and asking about
>> something which is not, in general, true.
>
> See quoted text above
>
>
>> If I answer by charitably replying as though he
>> had asked something sensible, he will pounce.
>
> I don't understand this sentence, so I cannot comment it (let's hope
> that it's nicer than your unfounded accusations above ;-)
>
> Best wishes,
> Peter
>
>
> PS: A moderator should more than anybody else refrain from personal
> attacks, shouldn't he?
Yes indeed!
Harald
Peter
> > Peter wrote:
> > What means "correctly state Newton's second law"?
> "harry" <harald.vanlin...@epfl.ch> writes
> That law does seem to be merely a refined definition of "force",
Only superfacially - actually, Newton's axioms concentrate on state
conservation and state change. What could be more fundamental for a celestial
mechanics?
> as long as
> Hooke's Force law is not taken as primary
it is rather special, and it is useless for celestial mechanics
> (and perhaps he omitted it because
> he did not want to include Hooke?).
that the conflict with Hooke played its role is probable, yes
> However, it would perhaps not do history
> justice to swallow that; in particular, Newton does not seem to define
> "force" in general.
he classifies the types he needs for celestial mechanics in the Definitions
> If we take bending ability as definition of force, the
> second law for "applied force" does become a law since it quantifies a
> change in (m*v) as function of such a bending ability. By chance I just
> looked up some references to a few of Newton's definitions for another
> discussion:
>
> "The quantity of motion is the measure of the same, arising from the
> velocity and quantity of matter conjunctly."
> "Every body perseveres in its state of rest, or of uniform motion in a
> right
> line, unless it is compelled to change that state by forces impressed
> thereon."
> - http://members.tripod.com/~gravitee/definitions.htm
> - http://members.tripod.com/~gravitee/axioms.htm
great, so I hope you will study carefully not only 'The Laws', but also 'The
Definitions'; eg, you will find here the imagination of a field what is often
ascribed to Lichtenberg and Faraday
> Note that the causality of action-reaction is there nicely sketched a
> mutual interaction between bodies:
>
> "vis insita, may, by a most significant name, be called vis inertiae, or
> force of inactivity. But a body exerts this force only, when another force,
> impressed upon it, endeavours to change its condition;
as was formulated already in the basic laws of Descartes and Huygens
> and the exercise of
> this force may be considered both as resistance and impulse; it is
> resistance, in so far as the body, for maintaining its present state,
> withstands the force impressed; it is impulse, in so far as the body, by
> not
> easily giving way to the impressed force of another, endeavours to change
> the state of that other."
yes, an impressive example of Newton's thinking - note the reference to
"state"
> - Note the word "impress" (if correctly translated): Dictionary: c.1374,
> "to
> apply with pressure, make a permanent image in," from L. impressus, pp. of
> imprimere "press into or upon, stamp," from in- "into" + premere "to
> press".
for translation issues, I would most of all consult the new translation by
Cohen 1999 - from a physical rather than linguistic point of view, it is
probable that Newton had in mind the impact of bodies
> That does appear to be an implicit recognition of a definition of force
> according to Hooke's law.
this might have played a role, too - recall, however, that, by considering mv
to be a vector, N. could correct the faults in Descartes' laws of impact, who
considered only |mv| as 'quantity of motion'
> - Note also the choice of the word "when" (if correctly translated) which
> does not imply a delay.
there is no delay between dp and K, all agree about that :-)
Best wishes,
Peter
Peter
> Oh No <No...@charlesfrancis.wanadoo.co.uk> writes:
> > Thus spake Peter <end...@dekasges.de>
> >> >> > 'the change of the momentum is proportional to the applied force
> >> >> > and along the same direction as that force'
> >> This is why you are accused of trickery. I find it clear that you have
> >> not made such a trivial mistake unless deliberately. The statement above
> >> is incorrect and, meaningless.
The statement above seems to be compatible with recognized translations, see
Harald's posting
> >You have just stated that Newton's original formulation "is incorrect and,
> >meaningless".
> I have said no such thing. I have never seen it translated as you have
> given it.
"seeing is believing" ;-)
-> see Harald's posting and my reply about several (!) translations (which
are, moreover, compatible with that by Mach and other ones into German I know
- available at request)
perhaps, you know only *interpretations*, such as F=m.a (exploited, but never
published by Newton)
my formulation is compatible with all translations quoted so far in this
thread - in case you have got an essentially different one, please quote it,
and I'm the last who is willing to return to the original Latin text :-)
NB: in this case, it would, perhaps, be helpful to consult also 'De
Gravitatione'
> >IOW, you claim that
> >- the change of momentum is *not* proportional to the applied force,
> >- the change of momentum is *not* along the direction of that force.
> >
> >What, then, is dp in dp=Kdt?
> That is a small change in a small time period, which is as you should
> have translated the law.
yes, I'm doing so, indeed :-))
what is your answer to my question?
> >> N2 refers to the rate of change of momentum, not the change of momentum
> >Why Newton wrote "change" and not "rate of change"?
> He wrote neither, since he wrote in latin.
Latin is - according to its rules - considered to be the most exact language
Europeans ever have created/developed.
(BTW: This does *not* exclude poetic features, eg, 'with highest honor' is
not expressed as 'cum summa laude', but as 'cumma sum laude')
hence, your reply is void, unfortunately
> >> You made have no mention of the time period over which momentum changes
> >Not me, Newton himself :-)
> One would need to study his whole treatise to see what was intended by
> the word Mutationem, but it is always translated as rate of change in
> motion.
nobody, except you, have *translated* it this way (see above)
> >This comes close to my goal :-) One cannot make the same statements obout
> >the time reversal symmetry of
> >
> > delta p ~ K (a)
> >
> >and of
> >
> > dp = K dt (b)
> >
> >(for symplicity, be K=const). But both (a) and (b) are correct.
> Even taking a, it is time symmetrical because it says nothing about the
> value of the constant of proportionality.
please reread this statement, and you will see its flaws :-)
> >> Moreover, the claim I
> >> made referred to the laws of physics, not to your statement which is not
> >> a law.
> >It's N2 in the original formulation - please criticize Newton, not me for
> >his formulations ;-)
obviously, you cannot ;-)
> >Anyway, I have discussed this in length, in order to learn about the
> >invariance of equations.
> >
> >Consider d'Alembert's wave equation (speed=1).
> >
> > &^2f/&t^2 = &^2f/dr^2 (*)
> >
> >Only even powers of t enter, dr is invariant, f enters homogeneously =>
> >this eq. is invariant, correct?
yes or no?
> >Now, consider the set
> >
> > &f/&t = g
> >
> > &g/&t = &^2f/dr^2
> >
> >This set is invariant only, if we require (claim?!) additionally that f or
> >g switch the sign together with t. Is this not an essential difference to
> >the manifest invariance of (*)?
> I wouldn't say so.
ok, whatyou would say?
P.
Peter
> > 'the change of the momentum is proportional to the applied force
> > and along the same direction as that force'
> >
> >
> > http://www.gfisher.org/appendix_to_2.htm
> >
> > with:
> >
> > ** start of quotation **
> >
> > Newton's Second Law. "The change of motion is proportional to the
> > motive force impressed and is made in the direction of the right line
> > in which that force is impressed." <![endif]>
> "harry" <harald.vanlin...@epfl.ch> writes:
> Compare with
> - http://members.tripod.com/~gravitee/axioms.htm
> - http://rack1.ul.cs.cmu.edu/is/newton/ p.83:
> " LAW II.
> The alteration of motion is ever proportional to the motive force
> impressed;
> and is made in the direction of the right line in which that force is
> impressed.
according to Cohen 1999, here ends Law 2 and starts the additional,
explanatory text
> If any force generates a motion, a double force will generate double the
> motion, a triple force triple the motion, whether that force be impressed
> altogether and at once, or gradually and successively. And this motion
> (being always directed the same way with the generating force), if the body
> moved before, is added to or subtracted from the former motion, according
> as they directly conspire with or are directly contrary to each other; or
> obliquely joined, when they are oblique, so as to produce a new motion
> compounded from the determination of both."
>
> Apparently gfisher's text is based on the above translation by by Andrew
> Motte (1729).
I agree
Cohen's translation (p.416):
"Law 2 A change in motion is proportional to the motive force impressed and
takes place along the straight line in which that force is impressed."
(It is thus justified by all these versions that I have claimed my
formulation above to be "close to Newton's original formulation")
> > Ok, if you think that knowing the 2nd question will make it easier to
> > you, here it is.
> >
> > Is the equation
> >
> > dp = F dt
> >
> > time reversal invariant?
> Hmm... I'm starting to wonder! But why not:
> dp = F dt
> dt'= -dt
> F = F
> => dp' = -dp
This is the usual (and Charles') answer, and I don't contradict it - I just
wish to point out the difference to equations, where no such changes of other
quantities is necessary (see the neighbouring subthread, where I refer to
d'Alembert's wave equation)
> However, I regard such time inversion as just a movie played backward;
> cause and effect are not changed at all.
> In contrast, the inverse physical process would be:
> dp = F dt
> dt'= dt
> F = -F
> => dp' = -dp
yes - time reversal invariance does not invoke a different (the reverse)
physical process
Thank you for this point,
Peter
Oh No
>
>> Oh No <No...@charlesfrancis.wanadoo.co.uk> writes:
>
>> > Thus spake Peter <end...@dekasges.de>
>
>> >> >> > 'the change of the momentum is proportional to the applied force
>> >> >> > and along the same direction as that force'
>
>> >> This is why you are accused of trickery. I find it clear that you have
>> >> not made such a trivial mistake unless deliberately. The statement above
>> >> is incorrect and, meaningless.
>
>The statement above seems to be compatible with recognized translations, see
>Harald's posting
>
The statement is not compatible with the modern language of mathematics
as used in physics, and remains a mistranslation.
>
>perhaps, you know only *interpretations*, such as F=m.a (exploited, but never
>published by Newton)
>
I am sure any good linguist will tell you that in translation one should
express the intended meaning, rather than literally translate the words.
Usage is not the same in different ages or in different languages.
>my formulation is compatible with all translations quoted so far in this
>thread -
It is not compatible with correct modern usage. For Newton to express
himself as he did, at a time when he was also developing the calculus
and rigor had not been introduced was one thing. His meaning is clear in
the equations he solved. To give this formulation now is inexcusable.
One should teach schoolboys better than that.
>
>> >IOW, you claim that
>> >- the change of momentum is *not* proportional to the applied force,
>> >- the change of momentum is *not* along the direction of that force.
>> >
>> >What, then, is dp in dp=Kdt?
>
>> That is a small change in a small time period, which is as you should
>> have translated the law.
>
>yes, I'm doing so, indeed :-))
>
>what is your answer to my question?
You know my answer. Stop playing games.
>
>> >This comes close to my goal :-) One cannot make the same statements obout
>> >the time reversal symmetry of
>> >
>> > delta p ~ K (a)
>> >
>> >and of
>> >
>> > dp = K dt (b)
>> >
>> >(for symplicity, be K=const). But both (a) and (b) are correct.
>
>> Even taking a, it is time symmetrical because it says nothing about the
>> value of the constant of proportionality.
>
I have read the statement. You do not affect time reversal symmetry by
expressing the law as a proportion rather than as an equation. (a) is
still true if you substitute p -> -p.
>
>> >> Moreover, the claim I
>> >> made referred to the laws of physics, not to your statement which is not
>> >> a law.
>
>> >It's N2 in the original formulation - please criticize Newton, not me for
>> >his formulations ;-)
>
>obviously, you cannot ;-)
It is not appropriate. See above.
>
>> >Anyway, I have discussed this in length, in order to learn about the
>> >invariance of equations.
>> >
>> >Consider d'Alembert's wave equation (speed=1).
>> >
>> > &^2f/&t^2 = &^2f/dr^2 (*)
>> >
>> >Only even powers of t enter, dr is invariant, f enters homogeneously =>
>> >this eq. is invariant, correct?
>
>yes or no?
>
>> >Now, consider the set
>> >
>> > &f/&t = g
>> >
>> > &g/&t = &^2f/dr^2
>> >
>> >This set is invariant only, if we require (claim?!) additionally that f or
>> >g switch the sign together with t. Is this not an essential difference to
>> >the manifest invariance of (*)?
>
>> I wouldn't say so.
>
>ok, whatyou would say?
I would say you are playing games and not making much sense if you seek
to include errors in order to dispute time reversal symmetry.
Peter
> Oh No <No...@charlesfrancis.wanadoo.co.uk
> >'the change of the momentum is proportional to the applied force
> >and along the same direction as that force'
> >> The statement above is incorrect and, meaningless.
> >The statement above seems to be compatible with recognized translations,
> >see Harald's posting
> The statement is not compatible with the modern language of mathematics
> as used in physics, and remains a mistranslation.
There is a German translation, where the translator has written what he
thinks is the correct understanding and usage of the 'Principia' - it's just
terrible
> >perhaps, you know only *interpretations*, such as F=m.a (exploited, but
> >never published by Newton)
> I am sure any good linguist will tell you that in translation one should
> express the intended meaning, rather than literally translate the words.
> Usage is not the same in different ages or in different languages.
this holds true for any translation, including poetry - for this, I'm
recommending the new translation by Cohen 1999
> >my formulation is compatible with all translations quoted so far in this
> >thread -
> It is not compatible with correct modern usage. For Newton to express
> himself as he did, at a time when he was also developing the calculus
> and rigor had not been introduced was one thing. His meaning is clear in
> the equations he solved. To give this formulation now is inexcusable.
> One should teach schoolboys better than that.
Newton just didn't add
'and proportional to the time this force is applied'
or the like. I wish only to understand, why?
BTW, you can "teach schoolboys better" only, if you know and understand
better, but you know and understand differently, this is not yet "better"
> >IOW, you claim that
> >- the change of momentum is *not* proportional to the applied force,
> >- the change of momentum is *not* along the direction of that force.
> >
> >What, then, is dp in dp=Kdt?
> That is a small change in a small time period, which is as you should
> have translated the law.
translation is not interpretation - I would agree that a translator should
comment in a footnote like that, in his formula for the force of gravity
between 2 bodies, Newton himself has exploited this law with adding the
proportionality to the time interval
> >yes, I'm doing so, indeed :-))
> >what is your answer to my question?
> You know my answer. Stop playing games.
It's not all a game you don't understand or cannot follow ;-)
> >> >This comes close to my goal :-) One cannot make the same statements
> obout
> >> >the time reversal symmetry of
> >> >
> >> > delta p ~ K (a)
> >> >
> >> >and of
> >> >
> >> > dp = K dt (b)
> >> >
> >> >(for symplicity, be K=const). But both (a) and (b) are correct.
...
> I have read the statement. You do not affect time reversal symmetry by
> expressing the law as a proportion rather than as an equation. (a) is
> still true if you substitute p -> -p.
no, because, then, the direction of delta_p is no longer that of K, but the
opposite one - don't be confused! :-)
> >> >Anyway, I have discussed this in length, in order to learn about the
> >> >invariance of equations.
> >> >
> >> >Consider d'Alembert's wave equation (speed=1).
> >> >
> >> > &^2f/&t^2 = &^2f/dr^2 (*)
> >> >
> >> >Only even powers of t enter, dr is invariant, f enters homogeneously =>
> >> >this eq. is invariant, correct?
> >> >Now, consider the set
> >> >
> >> > &f/&t = g
> >> >
> >> > &g/&t = &^2f/dr^2
> >> >
> >> >This set is invariant only, if we require (claim?!) additionally that f
> or
> >> >g switch the sign together with t. Is this not an essential difference
> >> >to the manifest invariance of (*)?
> >> I wouldn't say so.
> >ok, what you would say?
> I would say you are playing games and not making much sense if you seek
> to include errors in order to dispute time reversal symmetry.
Again, it's not all a game you don't understand or cannot follow ;-)
I see a difference in that
&^2f/&t^2 = &^2f/dr^2 (*)
is time reversal invariant eo ipso, while
&f/&t = g
defines that the time reversal symmetry of g is opposite to that of f (if f
has got one). If this is wrong, I'm happy to learn the correct reading :-)
Best wishes,
Peter
Oh No
>> himself as he did, at a time when he was also developing the calculus
>> and rigor had not been introduced was one thing. His meaning is clear in
>> the equations he solved. To give this formulation now is inexcusable.
>> One should teach schoolboys better than that.
>
>Newton just didn't add
>
> 'and proportional to the time this force is applied'
>
>or the like. I wish only to understand, why?
I think the answer is simple. When we work on the boundaries of human
knowledge, as Newton did, things are not clear. Greater clarification
becomes possible over the course of time once the laws are better known.
>
>BTW, you can "teach schoolboys better" only, if you know and understand
>better, but you know and understand differently, this is not yet "better"
It is better to express laws, and mathematics, more correctly. Remember
a great deal of confusion in the calculus was cleared up when
Weierstrass created the precise language of analysis. That was after
Newton.
>> >IOW, you claim that
>> >- the change of momentum is *not* proportional to the applied force,
>> >- the change of momentum is *not* along the direction of that force.
>> >
>> >What, then, is dp in dp=Kdt?
>
>> That is a small change in a small time period, which is as you should
>> have translated the law.
>
>translation is not interpretation - I would agree that a translator should
>comment in a footnote like that, in his formula for the force of gravity
>between 2 bodies, Newton himself has exploited this law with adding the
>proportionality to the time interval
Indeed he did. I think we have to accept Newton's intention in the law.
The way in which he applied the law shows that Mutationem meant to him
what we now call rate of change of momentum. If we give an over literal
translation we capture neither his intent, nor the physical meaning of
the law.
>> >yes, I'm doing so, indeed :-))
>
>> >what is your answer to my question?
>
>> You know my answer. Stop playing games.
>
>It's not all a game you don't understand or cannot follow ;-)
It is trivial. I do not find it interesting to create objections where
there are none.
>> >> >This comes close to my goal :-) One cannot make the same statements
>> obout
>> >> >the time reversal symmetry of
>> >> >
>> >> > delta p ~ K (a)
>> >> >
>> >> >and of
>> >> >
>> >> > dp = K dt (b)
>> >> >
>> >> >(for symplicity, be K=const). But both (a) and (b) are correct.
> ...
>> I have read the statement. You do not affect time reversal symmetry by
>> expressing the law as a proportion rather than as an equation. (a) is
>> still true if you substitute p -> -p.
>
>no, because, then, the direction of delta_p is no longer that of K, but the
>opposite one - don't be confused! :-)
I am not confused. The statement of proportionality
delta p ~ K
contains no direction information. It is satisfied whether the constant
of proportionality is positive or negative.
>> >> >Anyway, I have discussed this in length, in order to learn about the
>> >> >invariance of equations.
>> >> >
>> >> >Consider d'Alembert's wave equation (speed=1).
>> >> >
>> >> > &^2f/&t^2 = &^2f/dr^2 (*)
>> >> >
>> >> >Only even powers of t enter, dr is invariant, f enters homogeneously =>
>> >> >this eq. is invariant, correct?
>> >> >Now, consider the set
>> >> >
>> >> > &f/&t = g
>> >> >
>> >> > &g/&t = &^2f/dr^2
>> >> >
>> >> >This set is invariant only, if we require (claim?!) additionally that f
>> or
>> >> >g switch the sign together with t. Is this not an essential difference
>> >> >to the manifest invariance of (*)?
>
>> >> I wouldn't say so.
>
>> >ok, what you would say?
>
>> I would say you are playing games and not making much sense if you seek
>> to include errors in order to dispute time reversal symmetry.
>
>Again, it's not all a game you don't understand or cannot follow ;-)
>
>I see a difference in that
>
> &^2f/&t^2 = &^2f/dr^2 (*)
>
>is time reversal invariant eo ipso, while
>
> &f/&t = g
>
>defines that the time reversal symmetry of g is opposite to that of f (if f
>has got one). If this is wrong, I'm happy to learn the correct reading :-)
>
This has already been answered. Under time reversal symmetry dx/dt
changes sign, because dt changes sign. The same applies to &f/&t = g.
Nothing new is introduced here.
Peter
> >Newton just didn't add
> >
> > 'and proportional to the time this force is applied'
> >
> >or the like. I wish only to understand, why?
> Oh No <No...@charlesfrancis.wanadoo.co.uk> writes
>
> I think the answer is simple. When we work on the boundaries of human
> knowledge, as Newton did, things are not clear. Greater clarification
> becomes possible over the course of time once the laws are better known.
yes, of course
> >BTW, you can "teach schoolboys better" only, if you know and understand
> >better, but you know and understand differently, this is not yet "better"
>
> It is better to express laws, and mathematics, more correctly. Remember
> a great deal of confusion in the calculus was cleared up when
> Weierstrass created the precise language of analysis. That was after
> Newton.
yes
> >translation is not interpretation - I would agree that a translator should
> >comment in a footnote like that, in his formula for the force of gravity
> >between 2 bodies, Newton himself has exploited this law with adding the
> >proportionality to the time interval
>
> Indeed he did. I think we have to accept Newton's intention in the law.
> The way in which he applied the law shows that Mutationem meant to him
> what we now call rate of change of momentum. If we give an over literal
> translation we capture neither his intent, nor the physical meaning of
> the law.
I agree that here could be one of the omissions Truesdell has spoken of
> It is trivial. I do not find it interesting to create objections where
> there are none.
None objections were intended, just questions whether there is a deeper,
forgotten meaning behind N's original formulation
> >no, because, then, the direction of delta_p is no longer that of K, but
> >the opposite one - don't be confused! :-)
> I am not confused. The statement of proportionality
>
> delta p ~ K
>
> contains no direction information. It is satisfied whether the constant
> of proportionality is positive or negative.
I should have written
delta \vec{p} ~ \vec{K}
sorry!
> >I see a difference in that
> >
> > &^2f/&t^2 = &^2f/dr^2 (*)
> >
> >is time reversal invariant eo ipso, while
> >
> > &f/&t = g
> >
> >defines that the time reversal symmetry of g is opposite to that of f (if f
> >has got one). If this is wrong, I'm happy to learn the correct reading :-)
> This has already been answered. Under time reversal symmetry dx/dt
> changes sign, because dt changes sign. The same applies to &f/&t = g.
> Nothing new is introduced here.
Yes, I'm not claiming to introduce anything new here. Would you agree that
one cannot say that an definition like
g := df/dt
is time reversal symmetric?
Thank you,
Peter
Oh No
>> This has already been answered. Under time reversal symmetry dx/dt
>> changes sign, because dt changes sign. The same applies to &f/&t = g.
>> Nothing new is introduced here.
>
>Yes, I'm not claiming to introduce anything new here. Would you agree
>that one cannot say that an definition like
>
> g := df/dt
>
>is time reversal symmetric?
>
behaviour of nature, not to all quantities which can be given
mathematical definition.
harry
> Thus spake Peter <end...@dekasges.de>
>>
>>> Oh No <No...@charlesfrancis.wanadoo.co.uk
>>
>>
>>> express himself as he did, at a time when he was also developing
>>> the calculus and rigor had not been introduced was one thing. His
>>> meaning is clear in the equations he solved. To give this
>>> formulation now is inexcusable. One should teach schoolboys better
>>> than that.
>>
>> Newton just didn't add
>>
>> 'and proportional to the time this force is applied'
>>
>> or the like. I wish only to understand, why?
>
> knowledge, as Newton did, things are not clear. Greater clarification
> becomes possible over the course of time once the laws are better
> known.
>>
>> understand better, but you know and understand differently, this is
>> not yet "better"
>
- That's fine as long as the improvement is admitted.
- It's a scientific crime if the correction is not admitted and instead lies
are propagated.
[...]
>>>> What, then, is dp in dp=Kdt?
>>
>>> That is a small change in a small time period, which is as you
>>> should have translated the law.
>>
>> translation is not interpretation - I would agree that a translator
>> should comment in a footnote like that, in his formula for the force
>> of gravity between 2 bodies, Newton himself has exploited this law
>> with adding the proportionality to the time interval
>
> Indeed he did. I think we have to accept Newton's intention in the
> law. The way in which he applied the law shows that Mutationem meant
> to him what we now call rate of change of momentum. If we give an
> over literal translation we capture neither his intent, nor the
> physical meaning of the law.
I suspect that his intend was instead the proportionality that Peter
suggested here below (but why write K instead of F?).
[...]
>>>>>> This comes close to my goal :-) One cannot make the same
>>>>>> statements obout the time reversal symmetry of
>>>>>>
>>>>>> delta p ~ K (a)
>>>>>>
>>>>>> and of
>>>>>>
>>>>>> dp = K dt (b)
>>>>>>
>>>>>> (for symplicity, be K=const). But both (a) and (b) are correct.
>> ...
>>> I have read the statement. You do not affect time reversal symmetry
>>> by expressing the law as a proportion rather than as an equation.
>>> (a) is still true if you substitute p -> -p.
>>
>> no, because, then, the direction of delta_p is no longer that of K,
>> but the opposite one - don't be confused! :-)
>
> I am not confused. The statement of proportionality
>
> delta p ~ K
>
> contains no direction information. It is satisfied whether the
> constant of proportionality is positive or negative.
This is probably just a matter of notation.
delta p ~ F whereby p and F are vectors. F and delta p have the same
direction:
"the change of the momentum is proportional to the applied force and along
the same direction as that force"
[...]
Harald
======================================= MODERATOR'S COMMENT:
'K' is the Capital in 'Kraft' (German word for 'force')
Peter
> "harry" <harald.vanlin...@epfl.ch> writes
> > Oh No wrote:
> > It is better to express laws, and mathematics, more correctly.
>
> - That's fine as long as the improvement is admitted.
> - It's a scientific crime if the correction is not admitted and instead
> lies are propagated.
What exactly do you have in mind?
...
> > I am not confused. The statement of proportionality
> >
> > delta p ~ K
> >
> > contains no direction information. It is satisfied whether the
> > constant of proportionality is positive or negative.
> This is probably just a matter of notation.
> delta p ~ F whereby p and F are vectors. F and delta p have the same
> direction:
> "the change of the momentum is proportional to the applied force and along
> the same direction as that force"
yes
(overlap with my response - we have very slow transfer from the submission to
the publishing part of this server)
Thank you,
Peter
> ======================================= MODERATOR'S COMMENT:
> 'K' is the Capital in 'Kraft' (German word for 'force')
Yes, in German, there are 'Kraft' (force) and 'Feld' (field), thus, it's
seducing to take 'K' for force and 'F' for field
Cl.Massé
> Compare with
> - http://members.tripod.com/~gravitee/axioms.htm
> - http://rack1.ul.cs.cmu.edu/is/newton/ p.83:
> " LAW II.
> The alteration of motion is ever proportional to the motive force
> impressed; and is made in the direction of the right line in which that
> force is impressed.
Translated in French (or German,) then translated back, I wouldn't trust it.
*Straight* line. *Amount of* motion = momentum.
maxwell
> Thus spake Peter <end...@dekasges.de>
Gentlemen: this is sounding a lot like a theological debate. Newton
was not Mohammed or St Paul, passing on 'the word of God'. He was a
natural philosopher trying to create a mathematical representation of
the motion of large-scale objects so that he could predict planetary
motion. As such he was dealing with a problem where any time delays
were completely irrelevant, hence all his formulations were inplicitly
'at the same time'. This set the tone for 'instantaneous physics'
that defined the subsequent mathematical development known as
'classical mechanics'. It was only the recognition that
electromagnetism (the dominant interaction at the human scale)
involved time delays that broke this Platonic stranglehold on
physics. The rest is history ...
PS Peter & Charles, your extended conversation here illustrates
another problem with Platonism: who guards the guardians? I'm sure if
others had pursued this intense game of 'head-tennis' in a moderated
forum then the moderators would have closed it down much sooner.
Peter
> Gentlemen: this is sounding a lot like a theological debate.
:-o
> Newton was not Mohammed or St Paul, passing on 'the word of God'.
I agree, because I agree with Truesdell's characterization of the 'Principia'
> He was a
> natural philosopher trying to create a mathematical representation of
> the motion of large-scale objects so that he could predict planetary
> motion.
I'm happy that you refer to the - largely forgotten - fact, that the
'Principia' were intended to be a celestial mechanics
> As such he was dealing with a problem where any time delays
> were completely irrelevant, hence all his formulations were inplicitly
> 'at the same time'. This set the tone for 'instantaneous physics'
> that defined the subsequent mathematical development known as
> 'classical mechanics'. It was only the recognition that
> electromagnetism (the dominant interaction at the human scale)
> involved time delays that broke this Platonic stranglehold on
> physics. The rest is history ...
Your reasoning ignores the fact that Newton himself was keen to avoid the
impression that he thinks the instantanousness of his force law to be
physically relevant.
The promotion of his physics on the continent by non-physicists - notably, by
Voltaire - has hidden this reservation and led to the wrong standard
statement in virtually all textbooks.
Your formulation "the rest ist history" ignores that many insights by Newton
(and others) has still not been exploited
> PS Peter & Charles, your extended conversation here illustrates
> another problem with Platonism: who guards the guardians?
I agree that this is the killing question to any 'dictatorship of the
goodones'
> I'm sure
:-o
> if others had pursued this intense game of 'head-tennis' in a moderated
> forum then the moderators would have closed it down much sooner.
Not in this very liberal group
(I have not rejected this posting - although it contains unjustified
accusations -, because I'm personally involved, and thus have left the
judging to the other moderators)
Moreover, it's the matter of a step-by-step arguing. I'm not claiming to have
pursuing my final goal (to show that the fundamental laws are *not* time-
versal symmetric) in an optimum way. - BTW, I'm not aware of having read
comments from you which would help me to avoid flaws.
Best wishes,
Peter
harry
"Cl.Massé" <a...@c.com> wrote in message
news:48b93e24$0$9628$426a...@news.free.fr...
> "harry" <harald.vanlin...@epfl.ch> a écrit dans le message de
> news:121993...@sicinfo3.epfl.ch...
>
>> Compare with
>> - http://members.tripod.com/~gravitee/axioms.htm
>> - http://rack1.ul.cs.cmu.edu/is/newton/ p.83:
>> " LAW II.
>> The alteration of motion is ever proportional to the motive force
>> impressed; and is made in the direction of the right line in which that
>> force is impressed.
>
> Translated in French (or German,) then translated back, I wouldn't trust
> it.
> *Straight* line. *Amount of* motion = momentum.
No. It's stated that it was "translated into English by Andrew Motte". Of
course, it's old English...
www.archive.org/details/100878576 .
Harald
harry
"maxwell" <sp...@shaw.ca> wrote in message
news:05148b14-fc3e-4b3d...@z11g2000prl.googlegroups.com...
> On Aug 28, 5:15 pm, Oh No <N...@charlesfrancis.wanadoo.co.uk> wrote:
>> Thus spake Peter <end...@dekasges.de>
>>...
> Gentlemen: this is sounding a lot like a theological debate. Newton
> was not Mohammed or St Paul, passing on 'the word of God'. He was a
> natural philosopher trying to create a mathematical representation of
> the motion of large-scale objects so that he could predict planetary
> motion. As such he was dealing with a problem where any time delays
> were completely irrelevant, hence all his formulations were inplicitly
> 'at the same time'.
That was indeed the point that I brought up, in order to correct some
suggested causality definitions that came up in this thread. It has nothing
to do with putting Newton on a pedestal, but about not overlooking common
definitions.
> This set the tone for 'instantaneous physics'
> that defined the subsequent mathematical development known as
> 'classical mechanics'. It was only the recognition that
> electromagnetism (the dominant interaction at the human scale)
> involved time delays that broke this Platonic stranglehold on
> physics. The rest is history ...
That is not quite right. For example in classical mechanics, light also
takes time to go from one place to another.However, with the coming of
electromagnetism the concept of time delays was refined. It is still assumed
that cause and effect at a single point are instantaneous (example:
Einstein's first E=mc^2 paper).
Regards,
Harald
Peter
> ... For example in classical mechanics, light also
> takes time to go from one place to another.
Let me remark, that the exploration of light is not part of classical
mechanics.
I'm stressing this because of its consequences for the theory of measurement
which are well known, though seldom discussed from this point of view (what
is the reason for many confusions and interchangings).
Best wishes,
Peter
harry
"Peter" <end...@dekasges.de> wrote in message
news:guest.20080901114746$73...@news.killfile.org...
>> "harry" <harald.vanlin...@epfl.ch> writes:
>
>> ... For example in classical mechanics, light also
>> takes time to go from one place to another.
>
> Let me remark, that the exploration of light is not part of classical
> mechanics.
Probably you distinguish between classical mechanics and classical optics,
while Newton included light with mechanics; and maxwell to whom I replied
seems to have meant that. - See also
http://en.wikipedia.org/wiki/Classical_mechanics#History
Alternatively, if we exclude light from classical mechanics, simply replace
"mechanics" in the above exchange by "optics".
> I'm stressing this because of its consequences for the theory of
> measurement
> which are well known, though seldom discussed from this point of view
> (what
> is the reason for many confusions and interchangings).
Please elaborate for I have no idea what you mean (is it related to
causality?).
Regards,
Harald
Salviati
news:1219757...@sicinfo3.epfl.ch...
> Idealisations are OK, as long as we don't forget what we talk about. Your
> use of "event" is incompatible with the first use of that term in this
> thread by Peter: "The principle of causality states that all events have
> got a cause" - at a truly single point in time nothing happens and no
> cause for nothing is required.
> According to the dictionary, the use of "event" as during zero time is
> specific for relativity theory; but even so, from the way I have seen it
> used it may nevertheless imply a small amount of time (such as a bomb
> blast), negligibly small for calculations. Anyway, causality and
> specifically action-reaction refer to physical processes (or physical
> events, although you reserve that term for a point in time).
I see you quite right in that a truly single point in time is irrelevant in
practice.
Ask people what they are calling events. Perhaps every so called
event is strictly speaking extended over a certain timespan.
Nonetheless we can mostly decide which event preceded an other one.
This decision is an alternative one like between positive and negative.
Ideally there is nothing in between.
In mathematics, there is the very number zero between positive and
negative numbers.
Zero is not a natural number when we start countingwith one.
What about zero as a real number? As long as we interprete numbers
like nested intervals, rational as well as irrational numbers are always
included between a larger and a smaller one. Example:
0.99...9<1<1.00...1 regardless how many repetitions are expressed
by the three points. Mathematics pragmatically substitutes the fictitious
ideal "real" numbers with would have an illusory infinite acuity
by "real"-called rationals of unspecified sufficient size.
These considerations are seemingly pointless.
However, ther are some consequences for mathematics as well as for physics.
Really (ir)real numbers could not at all be excluded from continuum,
and if one would be able to do so, then this would not do cause any change.
Buridan's donkey cannot suffer starvation because it is impossible to
exactly meet the point zero in IR.
In http://home.arcor.de/eckard.blumschein/M283.html
I gave an example for misleading intermediate values in integral tables.
When I suggested to use cosine transform (CT) in IR+ instead of complex
Fourier transform in IR, there was a counter argument: CT does not yield
anything for a sine function.
That's true but of no practical relevance because elapsed time grows,
and therefore the dead point has no duration.
Nonetheless, there is no reason to deny a distinct end of positive elapsed
time.
The real-valued CT is equivalent to a complex FT that additionally
encodes an arbitrarily chosen point of reference.
Aren't antisymmetrical functions and matrices also avoidable if we restrict
to positive quantities?
In this case we need an absolute exclusion of negative values. Use of FT
would not really change that because its use rests on step function,
even and odd component.
Regards,
Salviati:
... in ultima conclusione, gli attributi di eguale
maggiore e minore non aver luogo ne gl'infiniti,
ma solo nelle quantità terminate.
IR>|>IR+>|>IR
>
> < I think you may
>> reasonably say that physics should be described in terms of
>> processes, rather than events, but we should try to ensure we are
>> speaking the same language.
>
> See above. Of course, this thread is very much about talking the same
> language.
>
>>> Causality is used in different theories, and in Newtonian theory the
>>> basic action-reaction causality is supposed to happen without delay.
>
> Good - that was the only thing I tried to point out! :-)
> It adds to the original formulation by Peter, although I'm not sure yet
> about how to formulate it.
>
>>> Delay times even were seen as a complication for Newton's principle,
>>> see for example Poincare ("but it will not be simultaneous"):
>>> http://www.thesciencebookstore.com/etext/poincarephysics.html
>
> [...]
>
>>>>> Our memory is causal. Time symmetry of physical laws does not imply
>>>>> memories coming from the future.
>>>>
>>>> Indeed not. Memory of the past is seen as a consequence of entropy,
>>>> and the law of entropy is derived from time symmetrical laws
>>>> (together with an initial condition which is obviously not time
>>>> symmetrical).
>>>
>>> That may be your view. I see memory of the past as a consequence of
>>> cause and effect; and as I pointed out, this probably also was the
>>> general view at the time that the causality principle was developed
>>> - simply because entropy was not considered at that time.
>>
>> At any point in history certain laws must be taken as fundamental.
>> Later generations may find more fundamental properties, or (as in
>> this case) theorems, which show that what was once considered
>> fundamental need no longer be taken as such. I am less interested in
>> the history than in the enquiry into what is really fundamental.
>>
>> In this instance I think there is a reasonably clear logical path (as
>> distinct from historical path of discovery).
>>
>> Time symmetric fundamental laws + assymmetric boundary conditions
>> ----> entropy
>> ----> the classical notion of cause and effect as reflected in human
>> memory
>
> I regard the biological notion of human memory to be an effect from
> physical causes such as light and chemistry; thus I agree to disagree on
> that. Probably the difference is that for me physical processes are
> primary, and physical laws are merely human descriptions of emerging
> properties of fundamental processes.
>
>> There is a second classical idea bearing on causality, namely
>> determinism, but in this thread we have also been trying to understand
>> what causality might mean in a quantum context, when we do not have
>> determinism. Interestingly enough classical statistical mechanics, in
>> which probabilities are governed by unknowns in a deterministic
>> theory, extends very easily to incorporate statistics governed by
>> probabilities from quantum theory. The logical path described above
>> continues to apply. That is all very well for statistical results,
>> but I have been seeking a definition of causality which applies to
>> individual process in quantum mechanics.
>
> I wish you success!
>
> Harald
Peter
> >> ... For example in classical mechanics, light also
> >> takes time to go from one place to another.
> > Let me remark, that the exploration of light is not part of classical
> > mechanics.
> Probably you distinguish between classical mechanics and classical optics,
> while Newton included light with mechanics; and maxwell to whom I replied
> seems to have meant that.
Neither 'The Definitions', nor 'The Axioms' of the 'Principia' mention light,
and 'Opticks' is quite a separate work -> let me ask how did you come to this
statement?
> - See also
> http://en.wikipedia.org/wiki/Classical_mechanics#History
The classification of sciences is blurred by historical developments, and it
may well be that a unique one is even not possible ;-)
Anyway, according to this Wikipedia article, mechanics is a science exploring
bodies, and light is not a body.
> Alternatively, if we exclude light from classical mechanics, simply replace
> "mechanics" in the above exchange by "optics".
in classical optics, light has a finite speed of propagation, indeed, at
least since Olaf Römer
> > I'm stressing this because of its consequences for the theory of
> > measurement
> > which are well known, though seldom discussed from this point of view
> > (what is the reason for many confusions and interchangings).
> Please elaborate for I have no idea what you mean (is it related to
> causality?).
Indirectly, as those confusions and interchangings have contributed to the
(erroneous) conclusions about the vanishing of causality and determinism in
quantum physics.
What I meant: Measurement using light is measurement by means of quantum
particles. Thus,
- measurement of classical motion using light is
quantum measurement of classical events,
while
- measurement of quantum motion using light is
quantum measurement of quantum events.
I believe that this difference is crucial for understanding the differences
between measurements of classical (eg, path of bodies) and of quantum events
(eg, motion of electrons in atoms).
Please don't hesitate to ask further questions in case that this outline is
too short :-)
Thank you,
Peter
maxwell
> > maxwell <s...@shaw.ca> writes:
> ..
> Your reasoning ignores the fact that Newton himself was keen to avoid the
> impression that he thinks the instantanousness of his force law to be
> physically relevant.
'Principia' that his
theory needed the (at that time) zero time-difference limit for his
calculations to work; however he chose to hide his 'secret weapon' -
the calculus, so he only used geometric expositions for his
sophisticated readership. As such, his exposition is based on finite
time differences, not on a model of the continuum. He thought in
terms of finite impulses @I (were I have used @ for the Greek delta) -
these caused a finite change in momentum @p, Newton chose his units so
that these were not only proportional but 'equal'; IOW N2 was @I -> @p
or impulse CAUSES a change in momentum. Later mathematicians, using
the calculus restated this (in the limit) as @I(t) = p(t+@t) - p(t).
The switch to the equality symbol hides Newton's intention to model
causality (remember Newton did not include any algebra in 'Principia',
his technique was purely geometric with references to - at most -
length & area ratios). Furthermore, since mathematically, all
velocity changes were assumed to be 'smooth' (or analytic) then:
@I(t) = p(t) - p(t - @t). This misled mathematicians to propose that
time runs in either direction. I'm sure Newton (and all his
contemporary philosophers would have vigorously rejected this
interpretation).
> The promotion of his physics on the continent by non-physicists - notably, by
> Voltaire - has hidden this reservation and led to the wrong standard
> statement in virtually all textbooks.
only a concept defined in the continuum limit i.e. F(t) @t = @I(t).
> Your formulation "the rest is history" ignores the many insights by Newton
> (and others) that have still not been exploited
I agree 100% once again, but this was soon after natural philosophy
got hijacked by the Platonic mathematicians who preferred symbolic
manipulation to understanding.
>BTW, I'm not aware of having read comments from you which would help me to >avoid flaws.
Charles to your dialogue; I have only responded now because of your
implicit invitation.
Best wishes, 'Maxwell'
harry
>
>> >> ... For example in classical mechanics, light also
>> >> takes time to go from one place to another.
>
>> > Let me remark, that the exploration of light is not part of classical
>> > mechanics.
>
>> Probably you distinguish between classical mechanics and classical
>> optics,
>> while Newton included light with mechanics; and maxwell to whom I replied
>> seems to have meant that.
>
> Neither 'The Definitions', nor 'The Axioms' of the 'Principia' mention
> light,
> and 'Opticks' is quite a separate work -> let me ask how did you come to
> this
> statement?
Newton regarded light as particles, and generally particles are regarded as
part of mechanics. The
>> - See also
>> http://en.wikipedia.org/wiki/Classical_mechanics#History
>
> The classification of sciences is blurred by historical developments, and
> it
> may well be that a unique one is even not possible ;-)
>
> Anyway, according to this Wikipedia article, mechanics is a science
> exploring
> bodies, and light is not a body.
I referred not to the whole article but to a section, in particular the
sentences that refer to light:
"Newton, and most of his contemporaries, with the notable exception of
Huygens, worked on the assumption that classical mechanics would be able to
explain all phenomena, including light, in the form of geometric optics.
Even when discovering the so-called Newton's rings (a wave interference
phenomenon) his explanation remained with his own corpuscular theory of
light."
>> Alternatively, if we exclude light from classical mechanics, simply
>> replace
>> "mechanics" in the above exchange by "optics".
>
> in classical optics, light has a finite speed of propagation, indeed, at
> least since Olaf Römer
I would say at least since Newton and Huygens.
>> > I'm stressing this because of its consequences for the theory of
>> > measurement
>> > which are well known, though seldom discussed from this point of view
>> > (what is the reason for many confusions and interchangings).
>
>> Please elaborate for I have no idea what you mean (is it related to
>> causality?).
>
> Indirectly, as those confusions and interchangings have contributed to the
> (erroneous) conclusions about the vanishing of causality and determinism
> in
> quantum physics.
>
> What I meant: Measurement using light is measurement by means of quantum
> particles.
That statement is debatable. But for sure measurement using light is
measurement by means of quanta - the nature of those quanta is unknown, but
we know that they show wave and particle behaviour.
> Thus,
>
> - measurement of classical motion using light is
>
> quantum measurement of classical events,
>
> while
>
> - measurement of quantum motion using light is
>
> quantum measurement of quantum events.
>
> I believe that this difference is crucial for understanding the
> differences
> between measurements of classical (eg, path of bodies) and of quantum
> events
> (eg, motion of electrons in atoms).
Hmm... any motion "is" both "classical" and "quantum", for those are simply
descriptions (together with interpretation) of what we observe. However, you
probably mean that for big objects quantum behaviour is usually negligible.
And I guess that you are mean that the interaction of photons with the
quantum events that they measure is important, while this is negligible with
classical events.
> Please don't hesitate to ask further questions in case that this outline
> is
> too short :-)
I think that I got the picture; probably you meant that the exploration of
light is nowadays not part of classical mechanics.
Regards,
Harald
Cl.Massé
>> it. *Straight* line. *Amount of* motion = momentum.
"harry" <harald.vanlin...@epfl.ch> a écrit dans le message de
news:1220253...@sicinfo3.epfl.ch...
>
> No. It's stated that it was "translated into English by Andrew Motte". Of
> course, it's old English...
> www.archive.org/details/100878576 .
>From Latin? That doesn't make the translation good. Especially if Latin
was as badly used as English today.
Peter
> >> >> ... For example in classical mechanics, light also
> >> >> takes time to go from one place to another.
> >> > Let me remark, that the exploration of light is not part of classical
> >> > mechanics.
> >> Probably you distinguish between classical mechanics and classical
> >> optics,
> >> while Newton included light with mechanics; and maxwell to whom I replied
> >> seems to have meant that.
> > Neither 'The Definitions', nor 'The Axioms' of the 'Principia' mention
> > light,
> > and 'Opticks' is quite a separate work -> let me ask how did you come to
> > this statement?
> Newton regarded light as particles, and generally particles are regarded as
> part of mechanics.
To be exact, one should, within classical mechanics (CM), speak
about 'bodies'. For the 'genuine' part of it (that is the part which doesn't
know an interaction constant, ie, the mechanics of impacts), the essential
property of bodies is their impenetrability. Because the 'Principia' were
intended to describe celestial mechanics, Newton wrote more about the
properties of bodies which are important for a theory of gravity than for
impacts; though he mentions impenetrability and hardness (see also his
text 'De gravitatione...'). Light particles are, obviously, not impenetrable
and, consequently, not object of CM.
> >> - See also
> >> http://en.wikipedia.org/wiki/Classical_mechanics#History
> > The classification of sciences is blurred by historical developments, and
> > it may well be that a unique one is even not possible ;-)
> >
> > Anyway, according to this Wikipedia article, mechanics is a science
> > exploring bodies, and light is not a body.
See above for 'body'
> I referred not to the whole article but to a section, in particular the
> sentences that refer to light:
> "Newton, and most of his contemporaries, with the notable exception of
> Huygens, worked on the assumption that classical mechanics would be able to
> explain all phenomena, including light, in the form of geometric optics.
I would not see geometric optics to be a part of mechanics; where are the
forces? - However, I should first consult the 'Opticks'.
> Even when discovering the so-called Newton's rings (a wave interference
> phenomenon) his explanation remained with his own corpuscular theory of
> light."
which yields the opposite result to Snell's formula, ie, is, obviously, not
correct
> >> Alternatively, if we exclude light from classical mechanics, simply
> >> replace "mechanics" in the above exchange by "optics".
> > in classical optics, light has a finite speed of propagation, indeed, at
> > least since Olaf Römer
> I would say at least since Newton and Huygens.
It's about the same time ;-)
> >> > I'm stressing this because of its consequences for the theory of
> >> > measurement
> >> > which are well known, though seldom discussed from this point of view
> >> > (what is the reason for many confusions and interchangings).
> >> Please elaborate for I have no idea what you mean (is it related to
> >> causality?).
> > Indirectly, as those confusions and interchangings have contributed to the
> > (erroneous) conclusions about the vanishing of causality and determinism
> > in quantum physics.
> >
> > What I meant: Measurement using light is measurement by means of quantum
> > particles.
> That statement is debatable. But for sure measurement using light is
> measurement by means of quanta - the nature of those quanta is unknown, but
> we know that they show wave and particle behaviour.
"light quantum" is more appropriate, thank you
> > Thus,
> >
> > - measurement of classical motion using light is
> >
> > quantum measurement of classical events,
> >
> > while
> >
> > - measurement of quantum motion using light is
> >
> > quantum measurement of quantum events.
> >
> > I believe that this difference is crucial for understanding the
> > differences
> > between measurements of classical (eg, path of bodies) and of quantum
> > events (eg, motion of electrons in atoms).
> Hmm... any motion "is" both "classical" and "quantum", for those are simply
> descriptions (together with interpretation) of what we observe.
Classical motion means motion that can be described by means of CM
> However, you
> probably mean that for big objects quantum behaviour is usually negligible.
yes, this also
> And I guess that you are mean that the interaction of photons with the
> quantum events that they measure is important, while this is negligible
> with classical events.
I'm not sure whether the interaction is the essential point. There are
demolishionless quantum measurement, iirc. Measurement with classical means
is, eg, using impacts, there is also interaction.
One of the main differences between quantum particles and classical bodies is
their manner of space occupation. Unfortunately, most discussions discard
this and speak about the measurement of position and velocity of quantum
particles as if they would move along orbits. And they are wondering that
the 'measurement' of a non-existent quantity yields another result than in
CM, where this quantity exists :-o
> > Please don't hesitate to ask further questions in case that this outline
> > is too short :-)
> I think that I got the picture; probably you meant that the exploration of
> light is nowadays not part of classical mechanics.
I would not say "nowadays". Newton's theory of light is untenable; Euler
excluded light from CM in view of the fact that light rays penetrate each
another without visible distortion; in Helmholtz's 'Lectures on Mechanics',
light is not explored...
Best wishes,
Peter
harry
If light is modelled as particles which are small bodies, obviously the
contrary is true.
>> >> - See also
>> >> http://en.wikipedia.org/wiki/Classical_mechanics#History
>
>> > The classification of sciences is blurred by historical developments,
>> > and
>> > it may well be that a unique one is even not possible ;-)
>> >
>> > Anyway, according to this Wikipedia article, mechanics is a science
>> > exploring bodies, and light is not a body.
>
> See above for 'body'
>
>> I referred not to the whole article but to a section, in particular the
>> sentences that refer to light:
>> "Newton, and most of his contemporaries, with the notable exception of
>> Huygens, worked on the assumption that classical mechanics would be able
>> to
>> explain all phenomena, including light, in the form of geometric optics.
>
> I would not see geometric optics to be a part of mechanics; where are the
> forces? - However, I should first consult the 'Opticks'.
How could particles not be subject to forces?
>> Even when discovering the so-called Newton's rings (a wave interference
>> phenomenon) his explanation remained with his own corpuscular theory of
>> light."
>
> which yields the opposite result to Snell's formula, ie, is, obviously,
> not
> correct
Indeed - which wasn't the point. It was thought that all phenomena could be
explained with mechanics. By the way, even elastic waves consist of
particles.
>> >> Alternatively, if we exclude light from classical mechanics, simply
>> >> replace "mechanics" in the above exchange by "optics".
>
>> > in classical optics, light has a finite speed of propagation, indeed,
>> > at
>> > least since Olaf Römer
>
>> I would say at least since Newton and Huygens.
>
> It's about the same time ;-)
Oh my, indeed - I mistakenly placed him one century later!
Yes that's right, such as with laser tweezers.
> One of the main differences between quantum particles and classical bodies
> is
> their manner of space occupation. Unfortunately, most discussions discard
> this and speak about the measurement of position and velocity of quantum
> particles as if they would move along orbits.
That's OK if they model them as classical particles. But indeed, often such
descriptions are blurred.
> And they are wondering that
> the 'measurement' of a non-existent quantity yields another result than in
> CM, where this quantity exists :-o
>
>
>> > Please don't hesitate to ask further questions in case that this
>> > outline
>> > is too short :-)
>
>> I think that I got the picture; probably you meant that the exploration
>> of
>> light is nowadays not part of classical mechanics.
>
> I would not say "nowadays". Newton's theory of light is untenable;
Newton's theory of the celestial mechanics is untenable too but that doesn't
mean that it wasn't part of classical mechanics; and Huygen's theory of
light theory certainly was part of classical mechanics since it made full
use of wave mechanics.
> Euler
> excluded light from CM in view of the fact that light rays penetrate each
> another without visible distortion; in Helmholtz's 'Lectures on
> Mechanics',
> light is not explored...
That means that Euler's light theory (which I don't know) doesn't treat
light as part of mechanics. Of course, "classical mechanics" can mean
whatever mechanical theory existed in the past. :-)
Regards,
Harald