[Fis] It From Bit video

Wed May 27 11:16:36 CEST 2015

Dear John,

That makes it clearer, thanks.  

The notion of symmetry is at the basis of the definition of probabilities (exchangeabilty (de Finetti), which operationalises symmetry, for variables that lie in the orbit of some group action, but whose transformed values do not have observable consequences such as changes of energy).  Any state defined by the values taken by variables, such that they deviate from the equiprobable distribution required for exchangeablity, necessarily has a different value taken by any measure of distances between distributions.  In that sense, one can correlate information and lack of symmetry, when the symmetric state is taken as reference.

Howver, there is more to symmetry than merely providing a reference state.  The way the world is described by physics is via symmetry, indeed via local symmetry.  The freedom to allow group transformations to variables locally must be coupled with compensatory transformations elsewhere.  And this is how interactions get generated, and we have light and other bosonic force mediators. Further, what is facilitated by appeals to notions of symmetry as a primitive principle, are not only ieas that rely on invariance, when the observables under scrutiny are unaffected by the symmetry transformations, but also covariance, where observables get transformed in a particular manner that respects its algebraic/geometric status.  

Cheers,
Srinandan

> On 26 May 2015, at 22:19, John Collier <collierj at ukzn.ac.za> wrote:
> 
> Dear Srinandan,
>  
> He relation of geometry to information theory (and also of particle theory in the Standard Theory) is by way of group theory. Groups describe symmetries, which are reversible. What is left over are the asymmetries, which are the differences that can be identified as information. This is worked out in some detail by my former student, Scott Muller, in Asymmetry: The Foundation of Information. Springer: Berlin. 2007. Seth Lloyd relates the information concept to quantum mechanics via group theory and other means in his Programming the Universe: A Quantum Computer Scientist Takes on the Cosmos. More direct connections can be made via the entropy concept where the information is the difference between the entropy of a system and its entropy with all internal constraints relaxed, but it comes to the same thing in the end. There are several convergent ways to relate information to form, then, in contemporary physics. But basically it is in the asymmetries.
>  
> As far as the relation between the asymmetries and symmetries go, I think this is still a bit open, since the symmetries represent the laws. Some physicists like Paul Davies talk as if the symmetries add nothing once you have all the asymmetries, so the laws are a result of information as well. I don’t see through this adequately myself as yet, though.
>  
> John
>  
>   <>
> From: Srinandan Dasmahapatra [mailto:sd at ecs.soton.ac.uk <mailto:sd at ecs.soton.ac.uk>] 
> Sent: May 26, 2015 10:20 PM
> To: ulan at umces.edu <mailto:ulan at umces.edu>; John Collier
> Cc: fis
> Subject: Re: [Fis] It From Bit video
>  
> Re: boundary conditions, etc.
>  
> I struggle to understand many/most of the posts on this list, and the references to boundary conditions, geometry and information leave me quite befuddled as well. Is it being claimed that geometry the same as information? That the requirement of predictions makes the focus on physical laws irrelevant unless the boundary conditions are specified? Or even that the continuum is at odds with the speed of light, considering classical electromagnetism is a well-defined continuum field theory. As for galactic distances, the only scientific basis upon which we conceive of the large scale structure of the universe is via the field equations of gravity, which brings a coherent package of causal thinking built into it. I did understand the bit on Noether, as energy conservation is indeed a consequence of time translation invariance, but that comes embedded in a continuum description, typically.
>  
> In biological systems, energy input makes the picture specific to the system one cordons off for study, and often it is hard to adequately describe phenomena by scalar potentials alone due to the currents in the system. And Noether cannot deliver reversibility. 
>  
> To me the message of Sean Carroll in the YouTube video that an equivalent redescription of physics (or biology) in terms of information is not enough, strikes me as sane.  
>  
> Cheers, 
> Srinandan
> 
> 
> -------- Original message --------
> From: "Robert E. Ulanowicz" 
> Date:26/05/2015 16:16 (GMT+00:00) 
> To: John Collier 
> Cc: fis 
> Subject: Re: [Fis] It From Bit video 
> 
> I would like to strongly reinforce John's comments about boundary
> conditions. We tend to obsess over the laws and ignore the boundary
> statements. (Sort of a shell game, IMHO.) If boundary conditions cannot be
> stated in closed form, the physical problem remains indeterminate! (The
> aphorism from computer science, "Garbage in, garbage out!" is appropriate
> to reversible laws as well.)
> 
> Then there is the issue of the continuum assumption, which was the work of
> Euler and Leibniz, not Newton. Newton argued vociferously against it,
> because it equated cause with effect. The assumption works quite well,
> however, whenever cause and effect are almost simultaneous, as with a
> force impacting an object, where the force is transmitted over small
> distances at the speed of light. It doesn't work as well when large
> velocities are at play (relativity) or very small distances and times
> (quantum phenomena) -- whence the need arose to develop the "exceptional"
> sciences, thermodynamics, relativity and quantum physics.
> 
> I would suggest it doesn't work well at very large distances, either.
> Consider galaxies, which are on the order of 100,000 or more light years
> in diameter. (I was surprised to learn recently that we really don't have
> decent models for the dynamics of galaxies.) Gravitational effects are
> relatively slow to traverse those distances, so that cause and effect are
> not immediate. (Sorry, I don't think quantum entanglement is going to
> solve this conundrum.) If cause and effect are widely separated, then the
> continuum assumption becomes questionable and by implication,
> reversibility as well. Now Noether demonstrated that reversibility and
> conservation are two sides of the same coin. So I see it as no great
> mystery that we encounter problems with conservation of matter and energy
> at galactic scales or higher -- witness "dark" matter and "dark" energy.
> 
> Of course, I am neither a particle physicist nor an astrophysicist, but
> merely someone writing from my armchair. So I invite anyone on FIS to put
> me straight as regards my speculations on these issues.
> 
> Cheers,
> Bob U.
> 
> 
> > Interesting question, Ken. I was not overly impressed with the video
> > because it didnâ€™t explain one of the most crucial points about the use
> > of information in dealing with quantum gravity, for which we as yet have
> > no good theory. The issue with both black holes and the origin of the
> > universe process is that the boundary conditions are dynamical. You can
> > have as many laws as you could want and still not have a physics if the
> > boundary conditions are ignored. Usually they are added in as an initial
> > state, or sometimes ad hoc but when they are changing, especially if they
> > are mathematically inseparable from the laws, there is a problem with
> > relying on the laws alone to explain. With black holes there is a question
> > of whether or not information disappears at their event horizon. There is
> > a similar issue for the observable portion of the universe at any given
> > time. It is hard to see how the questions can even be posed without
> > referring to information. Any boundary in basic physics can be conceived
> > the same way, and if all masses and energies come from geometry (in a
> > Unified Theory) then information is all there is in basic physics.
> >
> > I have argued for some time now that biological systems are much more
> > defined by their boundary conditions, which are typically dynamical and
> > changing, than by their energy flows, so information flows dominate,
> > though energy flows place limits, so I have talked of the information and
> > energy budgets being partially decoupled in biological systems. So
> > information is important to biology because understanding its flow can
> > answer questions about dynamical boundaries, just like in basic physics.
> > The energy (and matter) flows I will leave to the biophysicists, but the
> > paragraph above suggests that these are information flows as well. I like
> > the potential for unification here.
> >
> > Cheers,
> > John
> >
> > From: Fis [mailto:fis-bounces at listas.unizar.es <mailto:fis-bounces at listas.unizar.es>] On Behalf Of Ken Herold
> > Sent: May 26, 2015 12:30 AM
> > To: fis
> > Subject: [Fis] It From Bit video
> >
> > Released recently--what about the biological?
> >
> > https://www.youtube.com/watch?v=-ATWa2AEvIY <https://www.youtube.com/watch?v=-ATWa2AEvIY>
> >
> > --
> > Ken
> > _______________________________________________
> > Fis mailing list
> > Fis at listas.unizar.es <mailto:Fis at listas.unizar.es>
> > http://listas.unizar.es/cgi-bin/mailman/listinfo/fis <http://listas.unizar.es/cgi-bin/mailman/listinfo/fis>
> >
> 
> 
> _______________________________________________
> Fis mailing list
> Fis at listas.unizar.es <mailto:Fis at listas.unizar.es>
> http://listas.unizar.es/cgi-bin/mailman/listinfo/fis <http://listas.unizar.es/cgi-bin/mailman/listinfo/fis>
Srinandan Dasmahapatra
Bldg 1/ Rm 2009
Electronics and Computer Science
University of Southampton
Highfield, Southampton SO17 1BJ, UK
Tel: +44 (0)2380594503

-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://listas.unizar.es/pipermail/fis/attachments/20150527/69ce4ec3/attachment.html>