Subverting the Laws of Physics

This is a reply to:

   http://lists.gnu.org/archive/html/guile-devel/2014-09/msg00124.html

> How can we know that the enemy isn't using some laws of physics that
> we weren't taught at school (and that he deliberately keeps that
> knowledge out of schools)?

This was probably meant as a facetious comment, but with what is for
good reason called "the devil's luck," you've hit the proverbial nail
right on the head. The law of physics that the enemy is using, the
knowledge of which he keeps out of schools (and universities), is the
law that says every measurement necessarily must have a physical
meaning, and that the meaning of any empirical measurement is always
in the context of the meaning of other empirical measurements: no
experimental measurement has an absolute meaning. And you will see
this as soon as you consider, on the one hand, what is the meaning of
a reading of a P.D. of "1 V" on a meter, and on the other hand, what
is the meaning of "1 MeV" as the energy of a particle in an
accelerator ring.  Eventually, the semantics of any physical
measurement boil down to one or more of the defined reference units in
some metric system such as the Système international d'unités or
S.I., where the fundamental units are seconds, metres, kilograms,
amperes etc.

The science of metrology is all of the knowledge people have as to
how to go about actually making measurements in terms of these defined
units: how to calibrate devices, which methods are the most accurate,
etc. And the metrologists who work in the various national institutes
of weights and measures around the world use various precisely
specified experimental methods to construct the reference data, called
"realizations of the units," by which people in that place calibrate
measurements.  So the actual measurements that any laboratory makes
are always determined by a complex body of physical theory.

For example, the second may be defined as a certain number of
oscillations of an ion at a certain temperature in a caesium fountain
clock. Now, in order to know what this means, you need to know exactly how and why a caesium fountain clock actually works. So the
accurate measurement of a one second interval embodies (and therefore
depends upon) a great deal of sophisticated theory about how caesium
atoms behave at very low temperatures. And, as I'm sure Oleg would be
able to explain to you, if you know this, then you know that a
measurement of "one second" or "one meter" is by no means an immediate
observation: it is an elaborate series of inferences from other
measurements of other units all defined in the same system.

But this detail is not taught in schools, and it doesn't seem to
figure much in pop-science "literature" either. Perhaps this is
because it is a technical subject, and it is thought that one can't
teach people what a second really is until one has taught them how a
caesium fountain clock works. Or it may be because, in a sense, these
details are arbitrary: the idea of physical relativity in its most
general form is that of a "gauge theory" where the results of local
physical measurements are assumed to be globally consistent, under a
mapping which takes the results of local measurements from one region
to a neighbouring one. This mapping is called the local gauge
transformation. Then if the idea of a gauge theory is consistent
(which, a priori, is by no means necessary) it won't matter what the
local gauge at any point actually is, in the sense that the
"objective" physics is supposed to be an invariant of the
transformations from one local gauge to the next.  So it may be that
because the particular set of experimental methods (the
realizations) used to establish the reference data in any particular
gauge are arbitrary, people don't see any need to teach any of them.

Or it could be for the disingenuous reason that teaching this would
seem to undermine the "authority" and the "objectivity" of scientific
knowledge, and thereby leave room for the idea that God is in fact the
origin of all actual Human knowledge.

Whatever the reasons, the units of measurement are more often than not
presented to physics students as being in some sense "absolute" or
"god given ('because the professor said so!')". But they are defined in
the S.I., or some moral equivalent of it, so for better or for worse,
the S.I. and the c.g.s. systems, taken together, are effectively "the
denotational semantics of physical measurement." It's a pity though
that so many people who, because they write books about it, seem
to think they know physics, don't seem to actually know how any
physical measurements are made in practice.

The relevance of all this is that what I am proposing is an analogous
method of defining the semantics of programming languages: instead of
taking them as 'absolute' in terms of "whatever the compiler binary
and the CPU actually does with my source code," (the analogue of "what
The Professor says they are") we define them in terms of some formally
specified, and therefore well-known, "standard interpretations."
Precisely what these are doesn't actually matter: as long as they are
complete, machine readable and well understood, then, provided we know
what we're talking about, we will be able to make the necessary
"gauge transformations" between any two different interpretations, by
defining translators of the semantics in one form to the other, and
vice versa. And the mutual coherence of these different
interpretations will be the actual knowledge we have that the
semantics we think we know are what we actually think they are.
This state of knowledge is what I call a semantic fixed-point.

This is very different to a syntactic fixed-point, where we simply
compare the forms of the representations that appear in a terminal
window (or in documentation on a web page, for pity's sake!)  and
assume we know the semantics, without actually having any evidence
whether or not that assumption is correct. It is analogous to the
difference between, on the one hand, doing a physical experiment
twice, using two substantially different metric systems and comparing
the results 'metrologically', by converting between the different
realizations of units; and on the other hand, simply doing the
experiment once, using whatever is the conventional metric system, and
'blindly' reading the results directly off the calibrated scales on
the instruments.

> Then our enemy will always be in control! 

Only whilst the only place that people actually learn anything is in
school, and I hope that nobody still reading this actually believes
that. Formal education is only a "first-stage bootstrap": at
school/university we need just to learn enough from our teachers to be
able to continue learning independently.

The root of the problem, at least in so far as the education system is
concerned, is that this first-stage bootstrap has an "initialization
trap door" in it. The single most important thing one needs to teach
students, if they are to continue learning under their own steam, is
this: "we only ever actually know something when we know how and
why it is that we know that thing." And it is the almost total
failure to effectively teach this which gives the game away. The
evidence I have that this is not taught (i.e. the reasons I think that
I actually know this to be the case) are the countless instances I
see where people mindlessly assert "facts" which they "know," but
cannot explain how it is that they actually know these things. For
example, when someone asserts: "Then our enemy will always be in
control!" as if this were an immediate consequence of the fact that
the enemy controls what is taught in schools.

> This reasoning, although paranoid, seems completely valid [...]

Many teachers and lecturers don't themselves know this fundamental
epistemological principle, so the conspiracy is to a large extent
self-perpetuating, and need not even be a conscious one. But because
the subject is fundamentally based on epistemology and logic, I think
it more likely than not that it is a conscious conspiracy in computer
science, as it is taught in "prestigious" universities in Britain, the
USA and other countries where English is widely spoken.

To decide whether or not this is "paranoid reasoning" we can consider
"means, motive and opportunity." Such institutions as these are
typically considered to be "leading research in the field," and so
they have a great deal of autonomy in the choice of what they
teach. They also rely on foreign students for a significant part of
their income, and they exploit this source, ravenously and more or
less indiscriminately. So to prevent the dissemination of "valuable
knowledge" too far and wide (which, besides putting it in the hands of
"the enemy", would also erode their advantage, and thereby undermine
their ability to exploit foreign students), they intentionally
dumb-down the exoteric computer science syllabus.

The systemic nature of this conspiracy theory is not very different
from the dumbing-down mechanism which is evident in the mass media,
and which Chomsky calls "Manufacturing Consent."  The argument I've
just made proves nothing of necessity, but it demonstrates
possibility, and since paranoia is an unfounded fear of persecution,
it needs to be shown to be invalid before the assertion, that there
has been a conscious effort, by some people, to dumb-down the study of
logic and computer science in certain universities, could justifiably
be claimed to be "paranoid reasoning".

> [...] but it does abuse the notion of "enemy", by putting it in an
> extremely asymmetical situation.

Have you not seen either episode IV of "Star Wars", or "Matrix:
Revolutions", or "The Fifth Element", or "Kill Bill", or "The Bourne
Ultimatum", or "Salt", or "Avatar", or "Transcendence" for that
matter? These all "abuse the notion of 'enemy'" in exactly that way.

Finally, you seem to need to be told that nobody who had studied
computer science at any sort of university, anywhere in the world,
could possibly have come up with anything quite so stupifyingly
insightful as you did when you wrote "... we could detect that by
generating all possible binary sequences and checking whether the
generated ones are the same as the viewed ones. Or could this process
also be sabotaged?"

That, my friend, is the sort of thinking that could one day earn you
a PhD in Economics, or Theoretical Physics, or maybe even a job
interview at a nondescript office-block just outside of McLean, VA, a
prelude to a long and distinguished career as a government-sponsored
under-cover WikipediA editor. So you have a great future ahead of you;
provided that is, you never set foot in the theater.

Happy WikipediA hacking.

Ian