RoR: Forum

If a thing is a part of reality, and interacts with the rest of reality, is it possible that one can interact with it and discover that it exists?
(Edited by Dean Michael Gores
on 5/13, 12:33am)

Post 1

Saturday, May 13, 2006 - 9:36am

Maybe you're really aiming for something more limited or something else, but I interpretted your question as including things like mathematical/computational decidability so had to say no.

Post 2

Sunday, May 14, 2006 - 8:11am

If a thing is a part of reality, and interacts with the rest of reality,

What is there to discover?

Post 3

Sunday, May 14, 2006 - 2:23pm

DMG: If a thing is a part of reality, and interacts with the rest of reality, is it possible that one can interact with it and discover that it exists?
Aa: Maybe you're really aiming for something more limited or something else, but I interpretted your question as including things like mathematical/computational decidability so had to say no.

If A = B
and if B=C
then A = C.

If computational decidability is a part of reality, and computational decidability interacts with the rest of reality, is it possible that one can interact with computational decidability and discover that computational decidability exists? To which, Aaron answered, "No."

And you think that we have problems because of a radio disc jockey?

Post 4

Sunday, May 14, 2006 - 4:43pm

Computational decidability refers to problems where a solution may exist but it is not possible in finite time to determine it. It's such problem solutions I referred to, not the concept 'computational decidability' - and that's not implying violating some fundamental law of logic here. See Turing, Godel or any text on theory of computation.

Post 5

Sunday, May 14, 2006 - 6:36pm

( http://www2.physics.umd.edu/~redish/Money/)

Well, Aaron, can you be more specific? It is something of a dodge, to say, "read these books." I can think of computationally intractable problems. You mention

G�del. A G�delized encryption goes like this:

seed prime postion exp'n

N 01 14 1^14 = 1

U 02 21 2^21 = 2097152

C 03 03 3^3 = 27

L 05 12 5^12 = 1953125

E 07 05 7^5 = 16807

A 11 01 11^ 1 = 11

R 13 18 13 ^ 18 = (etc)

Then you multiply together as one product all of the exponentiations. Because the factors are powers of primes, the resulting number is theoretically factorable. You can broadcast a long text message as a single (large) number. Getting the plaintext out of the cipher is computationally intractable. (from The Code Book: all about unbreakable codes and how to use them (3rd ed.), Marotta. Loompanics. 1987)

Another example is the n-body problem. Croatia honors native son Ruger Boscovic on their money.

An 18th century polymath, he solved one of the restricted three-body problems. (You can read books on this.) As far as I know, these restricted three body problems have no synthetic solutions -- and no solutions at all exist for systems more complicated. When sending a satellite into farther space, scientists track it and make course corrections, closer in their methods to (empirical) Chaldean astrologers than to (rational) Enlightment mathematicians. So, there, again, the solutions might be indicated, but they are computationally intractable.

On the other hand, the Federal government still uses its DES/DEA (Data Encryption Standard/Data Encryption Algorithm), for federal reserve bank money transfers because it has never been cracked because it is computationally intractable -- but remains unused in military application because it might not be. Similarly, the RSA algorithm preferred by private enterprise has been seeded with longer and longer primes to keep ahead of factoring theory and factoring machinery. The longest known prime number -- several in recent years, as always -- was discovered by a homebrew internet-driven network of desktop personal computers. (Published December 04, 2003 MSU student's prime number largest one yet http://www.lsj.com/news/local/031204_numbers_1a-10a.html)

There was a time when just getting a system of wheels to point to the second derivative was a challenge worthy of a worldclass mathematician and the equally brilliant daughter of a brilliant poet: Babbage and Ada. Would that they were alive now to see what we can do!

So, what is computationally intractable changes over time -- and is not metaphysically determined. (But I could be wrong.)

Post 6

Sunday, May 14, 2006 - 8:38pm

Intractability is different from undecidability - takes a 'bloody long time' vs. could take infinite time.

The classic example is to tell if a given computer program will terminate if run. It is a fact of reality that it either will or won't. And you can look at its source code, step through and examine it while it runs - interact with it. However, some programs which do not terminate would be impossible to know that fact for certain by anything less complex than running them - which if they don't terminate takes infinite time, hence you never really 'know' that fact.

There are other types of areas that would also make the answer 'no'. Essentially, I'd say any of the following cases:

1) Where knowing would take infinite time - eg. undecidability
2) Where interaction required to know would destroy the observer, hence they could never really 'know' - eg. perhaps finding the traits of a singularity
3) Where interaction to gain some knowledge necessarily disturbs the system or otherwise prevents knowledge of other facts - eg. knowing both the position and velocity of a particle - basic Heisenberg uncertainty principle

As worded and as I read Dean's question, those are areas of unknowability, so I said no. His question was really not very limited or specific though, so if you wanted to refine certain points or verbage (eg. is the velocity of a particle something that 'exists'? how about the fact that a program won't terminate?) then maybe you could formulate the question to avoid such scenarios.

Post 7

Tuesday, May 16, 2006 - 3:40pm

Poll question:
Is it true that if a thing or a relationship is real then its existence is knowable?
Aaron, I am willing to entertain the notion that you have an insight that I can benefit from. This is an old problem and one that Objectivists consider simply solved, and not debateable: the unknowable does not exist.

Yet, there was a time when people thought that Earth is flat like a drum or a coin. Aristotle showed why it is most likely a sphere, but pointed out that even if it were flat, it would be so large, that you could not "fall off the edge." Also, Aristarchos suggested that Earth and the other planets orbit the Sun. That prompted Archimedes to perform an experiment. Archimedes attempted to measure the parallax in the location of a star and when he could not, he said that this proves that either the Sun goes around the Earth, or the universe is much larger than we can imagine.

So, maybe you, Aaron, perceive something that we do not. Maybe some things are unknowable. I just do not find the unknowable in your examples.

You suggest that "some programs which do not terminate would be impossible to know that fact for certain by anything less complex than running them ..." Can you point to one? Can you write one? Has someone else written one which is running? Is something like this in Knuth's Algorithms? Every computer program is a finite state machine. This program does not terminate:

start: print time$()
goto start

It will run as long as the computer exists and given realtime upgrades in platform, it will run as long as the universe exists. We know that. The unclosed loop is to computing what dividing by zero is to algebra: you can gimmick up a proof around this shell of an error and make the reader think that you have done the impossible. At root any "non terminating program" must have an unclosed loop of one form or another.

start: Input "Is Wesley Mouch the hero of Atlas Shrugged?" A$
if A$ <>"Yes" then goto start else end

Theoreticaly terminable, that little bugger is never going to end if run on my machine! But it is still a known gimmick.

Aaron: 1) Where knowing would take infinite time - eg. undecidability

Well, yes, that is what we are trying to find out. You cannot assert the hypothesis.

2) Where interaction required to know would destroy the observer, hence they could never really 'know' - eg. perhaps finding the traits of a singularity

You don't need a "singularity." The same could be true of a common bonfire or pulling the trigger of a gun, etc., etc. Searching for an afterlife by committing suicide is a generalization of that. And perhaps this is solution. The existence of an afterlife not knowable.

3) Where interaction to gain some knowledge necessarily disturbs the system or ...

This is pretty common and we live with it. One reason that customers do not like goods that are not priced in advance is that they fear that the price changes upward when they ask the price. Myself, I accept that the price could go down for me, because I am a cool guy. But what are we asking? It is a fallacy that there is "one true price." The price is different for everyone all the time. No two quotes need be the same.

Ethnographers working with gorillas, for instance -- to say nothing of business investors being wined by the colonels of a small country -- work hard to get themselves out of their observations, and it may be impossible. But what is impossible is not the doing. In other words, asking if you can join a troop of apes and not change their behavior is illogical, wanting to eat your cake and have it, too.

"... basic Heisenberg uncertainty principle..."

This is why we do not put a medicine ball on the billiards table or pepper the volleyball court with birdshot. It really spoils the game for everyone. All Heisenberg means is that at some level, the least stable event is the physical primary -- and we are justified in capitalizing that: Physical Primary -- by definition. People make too much of Heisenberg and the Zen Masters of Postmodern Physics.

Again, I am willing to accept that there is a puzzle to be solved, Aaron, but so far, you have not articulated an example that I can identify as meeting that criterion and I freely admit that the limitation could be mine.

Post 8

Tuesday, May 16, 2006 - 4:27pm

I get the impression you are more zealous that the idea that everything must be knowable than I am zealous to bother arguing about it.

See halting problem for surprisingly good wikipedia coverage of it.

Singularity, etc. - not just referring to 'you can't know what it feels like to be dead', but 'you can't know the specific traits of something that will always destroy you or your instruments.'

HUP - sure some people make too big a deal of HUP at the macro level. But (assuming that a particle's position and momentum 'exist') even applied properly it serves as a counterexample against a claim that everything which exists must be knowable.

As I said, there may just a semantic difference to how we look at his poll question. I see 'yes' as being an 'always' statement which falls with a single counterexample, while 'no' means 'not always' rather than 'never', and 'exists' can refer even to traits of particles, algorithms, etc.

Post 9

Wednesday, May 17, 2006 - 1:24am

If something's really real, then it interacts (somehow) with other existing things -- and it is thus, potentially, knowable to the mind of man (if only indirectly).

Ed

Post 10

Wednesday, May 17, 2006 - 1:29am

Talking point ...

It is not possible for something to simultaneously exist and be (even potentially) unknowable to the mind of man. There is no inherent limit to reason and discovery. And that's the beauty of the human mind.

Ed

Post 11

Saturday, May 20, 2006 - 9:56am

Dean asks: "Is it true that if a thing or a relationship is real then its existence is knowable?"

I'm not sure I can ever know the answer to that question. (Just kidding!)

Seriously, you know that this question was debated exhaustively awhile back, and I'm not sure anyone changed his or her mind on it.

Of course, my answer was a resounding, NO! The reason is that something could exist that we're logistically incapable of ever discovering, like a chemical reaction on the far side of the universe that we can't get to in time to observe or whose effects we'll never be able to infer. So, what I'm saying is that while everything that exists COULD be knowable, it doesn't follow that it MUST be knowable simply because it exists. Unfortunately, I seem to be in the minority on this question, which isn't too surprising on an Objectivist list. :)

- Bill

Post 12

Saturday, May 20, 2006 - 1:38pm

Bill,

============
what I'm saying is that while everything that exists COULD be knowable, it doesn't follow that it MUST be knowable simply because it exists
============

I agree with the sentiment present by the words above.

Ed

Post 13

Saturday, May 20, 2006 - 1:39pm

I was going for even more obscure examples (since I'd consider the reaction on the far side of the universe outside the bounds of interaction), but it is nice to see another minority opinion here.

Post 14

Saturday, May 20, 2006 - 3:32pm

Dean asks: "Is it true that...

I didn't ask that.

Post 15

Wednesday, May 24, 2006 - 10:17pm