Re: [Fis] Is information physical? A logical analysis

2018-06-05 Thread Bruno Marchal
Hi Joseph,

> On 31 May 2018, at 17:28, Joseph Brenner  wrote:
> Stan,
> Good, but things can also run in the opposite direction. How about variety 
> (plus more energy) generating more variety, more possibilities and allowing 
> new ‘information’ to emerge? Standard logical analysis is inadequate because 
> it cannot handle this picture.

Is not a simple counting algorithm, in base greater than 2, generating all 
information (in Shannon sense)?

0, 1, 10, 11, 100, 101, 110, 111, 1000, 1001, 1010, 1011, etc.

Of course, this does not generate 01, of 001, but they appear as easy 
subsequence like in 101 or 1001. You can also directly generate all finite 

Then a universal dovetailer, which is a program generating all programs (in 
some fixed programming language), and execute them on all input do much more, 
as it generates all information, and all ways all machine reacted to such 
information, and, assuming mechanism, that generates (in some sense) also their 
thoughts and experiences in that process (i.e, the information in the cognitive 
first person private sense). This is already less trivial than a counting 
algorithm which can be shown to be NOT Turing universal.

But this will never generate any physical realities. On the contrary, as I 
explain in my papers, the physical will have to be retrieved through a sum on 
all computations (which exists in arithmetic) below my substitution level, 
going through my states, and this can be used to show that physics has to be, 
fundamentally, a statistics, even a quantum-like statistics on all computations.

So, the arithmetical reality contains the seed of your varieties, I think, 
which generates all information(s). Now, to get physics from there, we need to 
structure that information space by using the canonical machine theory of 
self-reference, by Gödel, Löb, Solovay. That would be too long to describe here 
and now, but ask any question, as I do think this is relevant in this list, and 
especially in this thread on the nature of information, and, I add, on the 
nature of matter.


PS Please, take your time to answer, as this was my second post of this week. I 
will comment your possible remarks next week.

> Joseph
> From: Fis [ 
> ] On Behalf Of Stanley N Salthe
> Sent: jeudi, 31 mai 2018 16:21
> To: Burgin, Mark; fis
> Subject: Re: [Fis] Is information physical? A logical analysis
> Mark -- What Shannon referred to as 'entropy' was 'variety'. 'Information' 
> per se was achieved by way of a reduction or winnowing of this variety of 
> possibilities, leaving 'information' to survive. 
> On Wed, May 30, 2018 at 10:24 PM, Burgin, Mark  > wrote:
> Dear Loet,
> Only one remark. There is no Shannon-type information but there is Shannon's 
> measure of information, which is called entropy. 
> Sincerely,
> Mark
> On 5/23/2018 10:44 PM, Loet Leydesdorff wrote:
>> Dear Mark, Soren, and colleagues,
>> The easiest distinction is perhaps Descartes' one between res cogitans and 
>> res extensa as two different realities. Our knowledge in each case that 
>> things could have been different is not out there in the world as something 
>> seizable such as piece of wood.
>> Similarly, uncertainty in the case of a distribution is not seizable, but it 
>> can be expressed in bits of information (as one measure among others). The 
>> grandiose step of Shannon was, in my opinion, to enable us to operationalize 
>> Descartes' cogitans and make it amenable to the measurement as information. 
>> Shannon-type information is dimensionless. It is provided with meaning by a 
>> system of reference (e.g., an observer or a discourse). Some of us prefer to 
>> call only thus-meaningful information real information because it is 
>> embedded. One can also distinguish it from Shannon-type information as 
>> Bateson-type information. The latter can be debated as physical.
>> In the ideal case of an elastic collision of "billard balls", the physical 
>> entropy (S= kB * H) goes to zero. However, if two particles have a 
>> distribution of momenta of 3:7 before a head-on collision, this distribution 
>> will change in the ideal case into 7:3. Consequently, the probabilistic 
>> entropy is .7 log2 (.7/.3) + .3 log2 (.3/.7) =  .86 – .37 = .49 bits of 
>> information. One thus can prove that this information is not physical.
>> Best,
>> Loet
>> Loet Leydesdorff
>> Professor emeritus, University of Amsterdam
>> Amsterdam School of Communication Research (ASCoR)
>>  ; 
>> Associate Faculty, SPRU,  University of 
>> Sussex; 
>> Guest Professor Zhejiang Univ. , Hangzhou; 
>> Visiting Professor, ISTIC,  

Re: [Fis] Is information physical? A logical analysis

2018-06-05 Thread Bruno Marchal
Dear Arturo,

> On 21 May 2018, at 12:49, wrote:
> Dear Bruno, 
> You state that:
> "When poll are done at congress in cosmology or quantum computing, about half 
> of the physicists endorse the non collapse theory, as it is covariant, and 
> has no “measurement problem”.
> This means that the main tenet of your account, your "First Principle", is 
> not accepted by HALF of the scholars.
> How can you build your huge building on a so much controversial claim?

That was just a probably useless reply to the remark according to which few 
people endorse the non-collapse theory. In science, there is no poll, no vote, 
no mainstream, nor any similar form of argument per authority.

My claim for the non)collapse theory is that the collapse theory violate the 
Schroedinger equation, or the unitary evolution, and so is mainly a statement 
that quantum mechanics does not apply to … and the rest has never been made 
clear. There are almost as much collapse theory than there are physicists 
rejecting the SWE. Even Bohr, after expelling us that the collapse results from 
the perturbation of the act of measuring, admits to Einstein & AL., in his 
reply to the EPR paper (which shows that if there is a collapse due to 
measurement, it has to be Faster Than Light (FTL), that the collapse is not a 
physical phenomenon. Yet, he will later refuse to talk to Everett, and ciricize 
his idea of abandoning the collapse.

Now, you can say that I am biased, as I predicted the “many-worlds” much before 
I knew about quantum physics, in the form of the many computations which have 
been proved to be executed in virtue of very elementary truth. That has been 
seen by Gödel partially, and made clear by Church, Kleene and the works which 
followed. Today we know that even a simple Diophantine equation does already 
the job, and emulate or simulate exactly all computational processes. 

Unfortunately, when I will study quantum mechanics, I will take, like many, the 
collapse for granted, but I was still sure about the many computations, and I 
was still predicting the digital equivalent of a notion of “many-worlds”. It 
took me to read the EPR paper to begin to change my mind, but it is only after 
the careful study of Everett argument, that I will realise that quantum 
mechanics confirms the “many-worlds”, and thus the simplest and obvious, albeit 
shocking perhaps, consequence of Mechanism in the philosophy of mind (aka 
cognitive science). I predicted also the non-cloning theorem, indeterminacy, 
and non-locality, all as a natural phenomenology “lived” by any introspective 
universal machine.

> Furthermore, your links with Tegmark and Benacerraf confirm my thesis: your 
> account is a philosophical one,
Not at all. I have proven that if Mechanism isn correct, then physics has to be 
retrieved in a very precise way, as a mode of knowledge imposed by 
incompleteness in the machine or number self-reference theory, and I would have 
claim to have refute that form of Mechanism if the facts would have refuted 
this. But the facts confirm this, up to now.

So, Mechanism explains, until now, both consciousness and matter appearances, 
where physicalism and/or materialism failed, up to now, and indeed it would 
require a non-mechanist theory of mind, which usually are nonsense.

It explains consciousness by showing that any universal machine which 
introspect itself, and remains sound in that process, is confronted to 

- True,
- Non doubtable,
- Immediate,
- Non provable,
- Non definable.

Which is a good candidate for consciousness.

The theory also provides a role for consciousness, as it shows that such truth 
can be used by the machine to transform itself into a more speedy machine 
relative to the machine(s) supporting it, or even to simply other machines. The 
machine needs to be very cautious, because if it deduces that this is true, it 
becomes inconsistent and unsound, so the machine needs some act of faith, or to 
remember the question mark when invoking Mechanism. It is a subtle but 
important point.

> based on a logical principle, that, although fashinating and intriguing, is 
> highly controversial. 

Mechanism in cognitive science is an old and venerable theory, appearing in 
Indian and Greek antic text, renewed by Descartes and Diderot (who called it 

Note that mechanism in cognitive science refutes mechanism in physics. This is 
due to the fact that no universal machine can ever know for sure which machine 
she is, nor which computations (in arithmetic) supports us, and that any exact 
observable prediction has to be retrieved from a non computable statistics on 
all computations. But the propositional physical logic has been shown 
decidable, and you can consult my long texts to look at a theorem prover for it 
in LISP.

Is the theory of evolution controversial? It used implicitly mechanism, because 
non-mechanism would require infinite amount