Re: [Phys-l] F causes a

["Bob LaMontagne" <rlamont@postoffice.providence.edu>]

I often use Word-Web (http://wordweb.info/ free download) for definitions
because I find that it follows the way a word is actually used in common
practice. It defines the word cause in two ways:

Noun: cause
   1) Events that provide the generative force that is the origin of
something.

Verb: cause
   1) Give rise to; cause to happen or occur. Not always intentionally

In physics, one of the conditions for causality is that there is not only an
order to the events, but that there is no frame of reference from which that
order can be reversed. In common usage, however, causality does not seem to
strongly require a temporal separation of the cause and resulting effect.

A ball bounces off a wall. What is the "cause" of that bouncing motion? If
the wall was absent, the ball would continue its motion. But the force on
the ball and the acceleration of the ball occur simultaneously. This is
where I have a hard time with the concept of "cause". I must admit that I
have not resolved this comfortably in my own mind. I would appreciate any
comments from the group on this particular example.

Bob at PC

-----Original Message-----
From: phys-l-bounces@carnot.physics.buffalo.edu
[mailto:phys-l-bounces@carnot.physics.buffalo.edu] On Behalf Of Hugh Haskell
Sent: Thursday, May 11, 2006 1:17 AM
To: Forum for Physics Educators
Subject: Re: [Phys-l] F causes a

At 23:31  -0400 5/10/06, John Denker wrote:
> It is a truism that we deal in models.
> It is a truism that most (perhaps all_ of these models are imperfect.
>
> The problem with truisms is that even though they are true, they are
> not very informative.  And to phrase the argument in Manichaen terms
> (right versus wrong, black versus white) is to throw judgement to the
> winds.
>
> The crucial point is that some models are much, much better than others.
> It is not our job to choose models versus no models;  our job is to choose
> better models over worse models.
>
> So the issue is whether
>       "F causes ma (and not vice versa)"         [a]
> is better/same/worse compared to
>       "F equals ma"                              [b]
> As far as I can see, every limitation and every correction that must
> be applied to [b] must also be applied to [a].  I see absolutely no
> logical or physical reason why anyone should prefer [a] over [b].
>
> In contrast, there is every reason to believe that the physical
> relationship between F and ma is symmetric, and is therefore better
> modelled by a symmetric statement than an unsymmetric statement.
> Equality is symmetric;  causality is not.
Are we missing a subtle anthropomorphism here? Objects don't know 
from beans about forces, accelerations or anyone named Newton. They 
just do what they do, and, as has been pointed out, we construct 
models to help us understand how to predict what it is that they will 
do under various conditions. No causality is implied in the motion 
itself. It is only our anthropomorphic prejudices that infer some 
form of causality. But that impulse to infer causality is very strong 
in humans, for better or for worse. So perhaps, if we have this 
overwhelming urge to apply causal laws to our physics, maybe we need 
to follow John's prescription and not write NSL as an equality, but 
as an implication: F => ma, and not the converse, although we could 
use the converse to figure out what the imputed force might have been 
that "caused" the observed ma. John has stated that a causal 
relationship cannot be inferred if the statement of the law is 
mathematically symmetric, and that is certainly correct. But it also 
seems to me that one might, in principle, conduct an experiment that 
could distinguish between the symmetry of a mathematical equality and 
the causal implication, and that is to look for the inevitable time 
delay that must exist between a cause and its effect. Admittedly, for 
close encounters detecting this time delay may be beyond our current 
technological capability, but certainly if one thinks about it for a 
while it may be possible to come up with a suitable experiment. If 
so, and if it does demonstrate a causal relationship, then that would 
imply that we have been writing NSL wrong all these many centuries.

I haven't come up with such an experiment yet (several possibilities 
have popped into my head but all have had fatal flaws that make them 
not workable), but hopefully someone else could. Or maybe such an 
experiment is truly impossible. At this point I don't know. Since we 
know that all interactions take place without the object being in 
actual "contact" (if fact, it seems to me that we in the final 
analysis, we cannot even define "contact" since we don't really know 
where the boundary of an object is), such an experiment would have to 
involve action at a distance great enough to enable the time delay to 
be observed, and then to decide if that time delay between the 
proximate event and the resulting effects does in fact constitute the 
desired verification of the "effect," or if we are just looking at 
action at a distance involving propagating fields, and have just 
moved the problem to a different level without actually solving it.

I don't know the answer to this question, but I am willing to 
entertain the possibility of changing the notation for NSL to make a 
causal connection explicit. I don't know if that would solve the 
issue or not, but it would satisfy our very human desire to see 
causal connections in nature. If it happens that there is no 
experiment that could establish the truth of a causal connection one 
way or the other, then I see no harm in making an arbitrary choice, 
as long as we are explicit about the fact.

Hugh
-- 

Hugh Haskell
<mailto:haskell@ncssm.edu>
<mailto:hhaskell@mindspring.com>

(919) 467-7610

Never ask someone what computer they use. If they use a Mac, they 
will tell you. If not, why embarrass them?
			--Douglas Adams
******************************************************
_______________________________________________
Forum for Physics Educators
Phys-l@carnot.physics.buffalo.edu
https://carnot.physics.buffalo.edu/mailman/listinfo/phys-l