*** Forwarded message, originally written by Brent Meeker on 07-Feb-01
Thanks for your thoughtful (and straightforward) reply. I think there
is a confusion of levels in discussing these TOE's. Here is a diagram
illustrates my view of what is being attempted:
(1) | direct perceptions & thoughts |
(2) | the physical world & me |
| descriptions of the physical |
(3) | and psychological world in |
| terms of mathematical laws |
| the information content of the | -------------------
(4) | mathematical description of |<====| other postulates | (5)
| the world | -------------------
The downward arrows indicate abstraction (or perhaps I should say
inductive inference). The abstraction from direct
perceptions to a physical world and from the (naive) physical world to
scienctific description of the world are deliberately constructed so as
allow predictive inference in the upward direction, opposite the arrows.
However, it seems that (3) contains purely mathematical abstractions and
we ignore the interpretation of these we are left with (4) the
content of our world view. We can't infer backwards along this
Why we can't is what puzzled John Wheeler when he asked, of physics
equations, "What makes it fly?" In other words, what picks out just
THIS information content to be the world? TOE's on this list propose to
(5) which is something like:
(5) All logically possible (i.e. non-contradictory) mathematical
(5) All finitely computable descriptions
These proposals avoid Wheeler's "Why?" by saying EVERY such (4)+(5) is
a world. But if they ONLY did that they wouldn't be very interesting.
They are interesting if they allow inference backwards, from (4)+(5) to
Most of the arguments I read here (and I find them instructive) have to
do with what goes in (5) and whether that choice of (5) allows one to
no-white-rabbits or linear evolution of the wave function or some other
feature of (2) or (3).
What seems to me *not* to be interesting, is which level is "really
In one sense (1) is most "real" because we have it without inference.
On the other hand we know it contains inconsistent elements like dreams,
optical illusions, and hallucinations (white-rabbits?); so in the sense
of being shared (objective) (2) is more "real". We don't generally
think of (3) as being "real" because scientific descriptions don't
i.e. accidents of history or boundary conditions. However, (4)+(5) may
include all historical particularities as well as law-like descriptions
in that sense we might be willing to call them really "real". I realize
the goal is to explain (1), but that is not a reason to call it
nor is the fundamental status of (1) a reason to call (2) non-existent.
I haven't read "Similacron III", but I have read Egan's "Permutation
City". I enjoyed it very much (as I have enjoyed Egan's other stories)
would not criticze it as a novel. However, at the end Egan cheats
philosophically. He proposes that whichever mathematical structure
(simulation) is more detailed is the more real one. But "more detailed"
is difficult to define - and besides, what does it mean to be "more"
or even to be "real".
I don't think what I have said above is in conflict with your ideas
(although I don't want to put words in your mouth). I only want to make
things clear to myself and to damp what seem to me to be purely semantic
arguments about what's "real" that follow after assertions like "only
this moment exists" and "space-time is an illusion."
On 07-Feb-01, Marchal wrote:
> Brent Meeker wrote:
>> Bruno, I can't tell whether you're being playfully obtuse or merely
>> poetic, but you're not being straightforward.
> I'm really sorry. I send that mail too quickly. I regret and I
> admit not having been straightforward. Note that James's question
> was very difficult, but that is not an excuse.
> I was neither clear nor very funny. OK.
>> Of course you know what
>> is meant by the 'physical world'. It is that complex that we infer
>> from our perceptions to provide a comprehensible account of the
>> relations among those perceptions.
> In *that* very large sense I believe in the 'physical world', sure.
> But why the adjective "physical"? Why on earth should the world
> necessarily be physical, and what does that mean?
> I agree that for most people, words like "physical", "natural" and
> even "empirical" have some easy meaning.
> My belief is that such an eazy meaning is illusory.
> My whole work try to explain that. And then I try to explain where
> the illusion comes from.
> That 'complex we infer from our perception' is arithmetical truth or
> mathematical truth (I am not sanguine about the arithmetical
> restriction). Stories like what we are living, including our
> perception and the laws we infer are partial aspects on arithmetical
> truth seen from inside. and "seen from inside" can be defined in terme
> of relative computations (at least with comp).
>> If you wish to take a radically
>> fundamentalist position, avoiding all inference, then you can only
>> speak of "softness" not "softness of velvet" and you cannot presume
>> that you "feel only the presence of persons...". You can only feel -
>> to assign a source to that feeling is an inference just as is the
>> physical world.
> This shows how much I was unclear, because I have no problem at all
> with inference. I have a problem only with the inference of
> material stuff or substance.
>> I think you have the direction of abstraction
> Indeed. I agree, and that's what is really difficult to accept
> in my work.
>> Perceptions and feelings are the raw stuff of being.
> Perhaps, perhaps not. With comp it is more easy to postulate that
> there is no raw stuff at all, only arithmetical truth like
> "the number x apply to the number y relatively to the number z
> gives the number t. Such collection of truth contains all
> truth on the behavior, belief, even knowledge (definissable
> axiomatically) of all possible machines embedded in all possible
> neighborhood. I don't believe there is stuff, that is why
> I feel obliged to explain why machines believes in stuff, why it
> is locally useful with respect to their most probable computational
> history, etc.
> I really think that if we are digital machines (= if we can survive
> with a digital brain) then "the law of physics" can be derived
> from "the law of psychology", which are themselves derivable from
> computer science, through the discourse of the sound universal
> machine. This is what I called the reversal between physics
> and machine's psychology.
>> Persons, velvet, and the physical world are abstracted from them.
>> Information is, as you say, another level of abstraction away from
>> the physical world - but in the direction away from raw experience.
> Mmmh ...
> For easyness I repeat James Higgo question:
>>> In conversations with friends, I am often asked why the minimal
>>> Kolmogorov complexity of Tegmark's schema has any relevance to the
>>> physical world. Why should information theory tell us anything about
>>> the 'real' world? What grounds do we have to believe that the stuff
>>> of the 'physical' world is information?
> If the word 'information' (which is a confusing word which has a
> lot of meaning) is taken in the lay psychological sense (like in the
> sentence "I know because I get the information", then my shortest
> suggestion to James would be based on the idea of virtual reality.
> BEGIN SHORT ANSWER 1
> Reality is dream sharing, which is information sharing. That is why
> when we look outside or inside close enough, information theory
> shows its nose.
> END SHORT ANSWER 1
> James can then suggest the reading of Daniel Galouye's
> SIMULACRON III, or the more recent PERMUTATION CITY by Greg Egan.
> These novels illustrate very well the idea.
> But I should without doubt make precise that people like David
> Deutsch, Ekert, Landauer (and a lot more) would answer James
> Higgo in another, quite opposite, way : they would answer that
> information theory play a role in our study of nature because
> we realise, through the discovery of quantum information science,
> that information is physical.
> I, personaly, don't believe that at all, I believe, on the
> contrary, that physics is informatical, right at the start.
> (Actually I argue that such reversal follows logically from
> comp, that's the purpose of my thesis, cf my URL).
> I will think for better 'short' answer.
*** End of forwarded message ***