- #1
Fra
- 4,104
- 606
The discussion about the status of lorentz invariance, and as I wanted to twist it, in general, the physical and scientific basis for symmetries started to diverge in my posts so I started a new thread for this to comment more without contaminating the original thread.
Thanks for the links atyy. I like parts and suggestions of this paper but it's arguments has limitations and doesn't address what I have in mind at a sufficiently deep level.
They write “The past exists only insofar as it is recorded in the present”
I like this.
I'd also like to ADD - for coherence of reasoning! - that, laws exist only insofar as they are represented in the present.
Here the effect of law are constraints on the observers action. This actions remains until environmental feedback suggest otherwise - in which case the current state is subject to destructuve/deformning forces.
This means that a system acts "as if the law didn't exists" until evidence that it does, arrives. (This is similar to the idea that a system follows a geodesic given no other forces, and even when other forces are present you could say that it still follows a geodesic, it's just that the geodesic changes)
This suggests a kind of relativity of law, and if this process is understood I think a lot of fine tuning problems can be explained without highly improbable intitial conditions.
The general basic idea put forward is that there would be no physical observer around to observer/record a entropy decreasing in an isolated system, which loosely reduces to that "improbable observations" (which is what violating second law is) would be impossible.
I would rather put this in an evolutionary perspective and suggest that improbable observations are possible, but observers that consistently do them are not stable and are thus not likely to be observed to populate this world.
So I miss the extension of the argument to include not only entropies and probabilities, but to include also actions and transition probabilites. Entropy is just one a priori information measure of possibilities, action is an a priori measure of possible changes due to the instability of the present.
I think the argument here can be sharpened and developed to suggest predictions on the population of observers and their interaction properties. Ie. matters microstructure and it's implict actions. Thus what is usually considered as mathematical consistency used to constrain by deduction interactions and system properties, can instead be used as an evolutionary selection principle and instead explain.
This would suggest that the present state of an observer, encodes the physical law, necessary to evolve the systme into the future. However this law only determines the infinitesimal evolution, since the law is constantly evolving. An external large observer, might be able to predict the deformation of law, and thus the action of the total system of observer + local environment by understanding this process. Then one might say that oh well, then the external view represents the real timeless law, but this is a mistake since the external observer is not unique. There are several of them, each having a different perspective. So there are only a collection of hierarchies of interacting observers.
This would be much more radical than that paper.
---------
My view of physical law, and symmetry constraining nature is that their purpose is to help predict the future, just as the goal of science is. IMO, science isn't about finding something as naive as eternal truths, it's about trying to infer from experince with nature, the patterns that helps us predict the evolution of our own environment, simply as an act of self preservation.
Thus I think this conjecture this brings two "constraints" to help us look for a better unified understand of the laws.
(1) Physical law must be inferrable by means of a physical process. In particular does this reject the methodology of having as a guide simplicitly as in "mathematically beauty". In particular would this also suggest that a simple low complexity observer - think elementary particles and subatomic systems - have a more constrained possibilities of inferrable laws. At some point down the *complexity of the inferenece system* there might not be a lot of choices.
So I think the most important scaling to be considered here is the scaling of the observers complexity, which is the same as the complexity of the "inference machinery".
So unification could be understood as how during this inference process, more complex LAWS become physically distinguishable as the observers complexity grows.
(2) A parallell problem is to understand the growth of complexity (probably related to the origin of mass). This leads to the second constraint that hte law must be physically encodable by the observer. Ie. law itself implies information and thus memory space.
All these things are IMO questions that are prior to notions such as 4D spacetime. The dimensionality of spacetime is one particular thing that must be infered during this process.
All all this is very similary to biological evolutionary processes.
Thus the phrasing that science is the search for absolute, timeless truth is IMHO a simplistic view of science and learning. Instead it's what we are doing with the "truths" that is the difference, it's the basis for our actions, and as such, our actions are based on our inferred constraints, wether eternally tru or not. The notion of eternal truths and universally "right" is simply a redundant notion that only confuses IMO, beucase there is no universal measure of right or wrong anyway.
Or put differently, yes you can say that laws are "timeless" but only to the extent implict in “The past exists only insofar as it is recorded in the present” in the case of limited memory records :)
This is also why I object to some peoples very narrow reasoning in the original thread. Clearly this boils down to the view of science and the scientific methd and it's connection to physics and physical interactions. That last time I looked I certainly consider a human to be as much as physical system as anything else. So some level of coherence of reasoning is what I request.
/Fredrik
atyy said:
Thanks for the links atyy. I like parts and suggestions of this paper but it's arguments has limitations and doesn't address what I have in mind at a sufficiently deep level.
They write “The past exists only insofar as it is recorded in the present”
I like this.
I'd also like to ADD - for coherence of reasoning! - that, laws exist only insofar as they are represented in the present.
Here the effect of law are constraints on the observers action. This actions remains until environmental feedback suggest otherwise - in which case the current state is subject to destructuve/deformning forces.
This means that a system acts "as if the law didn't exists" until evidence that it does, arrives. (This is similar to the idea that a system follows a geodesic given no other forces, and even when other forces are present you could say that it still follows a geodesic, it's just that the geodesic changes)
This suggests a kind of relativity of law, and if this process is understood I think a lot of fine tuning problems can be explained without highly improbable intitial conditions.
The general basic idea put forward is that there would be no physical observer around to observer/record a entropy decreasing in an isolated system, which loosely reduces to that "improbable observations" (which is what violating second law is) would be impossible.
I would rather put this in an evolutionary perspective and suggest that improbable observations are possible, but observers that consistently do them are not stable and are thus not likely to be observed to populate this world.
So I miss the extension of the argument to include not only entropies and probabilities, but to include also actions and transition probabilites. Entropy is just one a priori information measure of possibilities, action is an a priori measure of possible changes due to the instability of the present.
I think the argument here can be sharpened and developed to suggest predictions on the population of observers and their interaction properties. Ie. matters microstructure and it's implict actions. Thus what is usually considered as mathematical consistency used to constrain by deduction interactions and system properties, can instead be used as an evolutionary selection principle and instead explain.
This would suggest that the present state of an observer, encodes the physical law, necessary to evolve the systme into the future. However this law only determines the infinitesimal evolution, since the law is constantly evolving. An external large observer, might be able to predict the deformation of law, and thus the action of the total system of observer + local environment by understanding this process. Then one might say that oh well, then the external view represents the real timeless law, but this is a mistake since the external observer is not unique. There are several of them, each having a different perspective. So there are only a collection of hierarchies of interacting observers.
This would be much more radical than that paper.
---------
My view of physical law, and symmetry constraining nature is that their purpose is to help predict the future, just as the goal of science is. IMO, science isn't about finding something as naive as eternal truths, it's about trying to infer from experince with nature, the patterns that helps us predict the evolution of our own environment, simply as an act of self preservation.
Thus I think this conjecture this brings two "constraints" to help us look for a better unified understand of the laws.
(1) Physical law must be inferrable by means of a physical process. In particular does this reject the methodology of having as a guide simplicitly as in "mathematically beauty". In particular would this also suggest that a simple low complexity observer - think elementary particles and subatomic systems - have a more constrained possibilities of inferrable laws. At some point down the *complexity of the inferenece system* there might not be a lot of choices.
So I think the most important scaling to be considered here is the scaling of the observers complexity, which is the same as the complexity of the "inference machinery".
So unification could be understood as how during this inference process, more complex LAWS become physically distinguishable as the observers complexity grows.
(2) A parallell problem is to understand the growth of complexity (probably related to the origin of mass). This leads to the second constraint that hte law must be physically encodable by the observer. Ie. law itself implies information and thus memory space.
All these things are IMO questions that are prior to notions such as 4D spacetime. The dimensionality of spacetime is one particular thing that must be infered during this process.
All all this is very similary to biological evolutionary processes.
Thus the phrasing that science is the search for absolute, timeless truth is IMHO a simplistic view of science and learning. Instead it's what we are doing with the "truths" that is the difference, it's the basis for our actions, and as such, our actions are based on our inferred constraints, wether eternally tru or not. The notion of eternal truths and universally "right" is simply a redundant notion that only confuses IMO, beucase there is no universal measure of right or wrong anyway.
Or put differently, yes you can say that laws are "timeless" but only to the extent implict in “The past exists only insofar as it is recorded in the present” in the case of limited memory records :)
This is also why I object to some peoples very narrow reasoning in the original thread. Clearly this boils down to the view of science and the scientific methd and it's connection to physics and physical interactions. That last time I looked I certainly consider a human to be as much as physical system as anything else. So some level of coherence of reasoning is what I request.
/Fredrik