I think you’re misunderstanding the original post(s).

The scenario presented by Lawrence Krauss isn’t one where a civilization in the very distant future has somehow preserved scientific literature from billions of years in the past and then chooses to reject that documentation, but rather where a future civilization similar to ours today uses evidence available through observation to develop theories about the nature of the universe. His point was that what can be observed in the very distant future - 100 billion years or so from now - may be very different from what can be observed today.

Scientific theories must ultimately be corroborated with repeatable observations. A scientific theory cannot be based on something that cannot be observed (we call those “theories” either pseudoscience, or religion, or hallucinations, or …).

Some have responded to this thread that if something can’t be observed, then it really isn’t important to science. That “If there is a truth so remote that it has absolutely no influence on observable reality then knowing this truth gives you no benefit.” This is correct in one sense, but science isn’t just about materially benefiting from knowledge. Scientists seek the truth – they want to know how things actually are, and they often assume that if observations suggest how things actually are, then that’s how they actually are. That kind of assumption is not always correct. My guess is that it's rarely if ever correct.

The necessary reliance on observation is both a strength and weakness of science. Theories supported by observation can be wrong – and not wrong in just some fine details, but flat out wrong. Other theories that are not supported by observation may occasionally be right.

Imagine in the Lawrence Krauss scenario a future scientist who imbibes some future psychedelic substance and “sees” a universe not composed of a single galaxy, but one composed of innumerable galaxies, all moving away from each other at an accelerating rate. Imagine that he develops a theory to take these ideas into account. Further imagine that his theory is consistent with the prevailing “single galaxy in an infinite void” theory.

He shares his new revelations with his fellow scientists. And what is their response? They may laugh at him. They may find his ideas interesting. Or not. Regardless, they’ll correctly state that there is no evidence to support such an absurd theory. So in this future society, accepted theories about the nature of the universe are wrong, and if someone somehow “guessed” what is actually correct, his ideas would be rejected due to lack of supporting evidence.

I understand that the main point is that there is a general predicament when it comes to basing the acquirement of knowledge solely on empiricism, but I don't think all scientists act entirely on empirical standards as you've stated. My question is why would that future society not support a theory of the plurality of galaxies? This would entail that future scientists would strictly act on empirical methods and nothing more. If the history of science tells us anything, it's that scientists are evolving and advancing beyond the limitations of pure observation. I just can't help but think of distant future scientists transcending and reaching greater horizons of knowledge in which basing things solely off observation would be a relic of the past.

We could ask the same question about today’s scientists. Most mainstream scientists don’t support the “many worlds” interpretation of quantum mechanics, even though it more elegantly describes the nature of things than does the Copenhagen interpretation.

We could also ask why would future scientists even conceive of the idea of a plurality of galaxies? Prior to sufficiently powerful telescopes, our scientists thought that our galaxy was the only galaxy. It was only after observing other galaxies that the idea of multiple galaxies was even entertained. What would lead scientists of the distant future to consider the possibility of a multitude of galaxies? And even if they considered it a possibility, without any evidence it would never become a theory.

It’s true that brilliant scientific minds leap ahead of what can be empirically determined, but for their theories to be accepted, empirical evidence must be eventually forthcoming. In the Krauss scenario, there would not be any empirical evidence obtainable, even in theory, to support a “many galaxies accelerating away from each other” theory.

Well, that's why I stress the importance of the role of history and our documentation. But given the scenario where these are not of grave significance in comparison to contemporary observation, then you'd be correct in the assertion that scientists of the distant future would have no grounds to consider the plurality of galaxies.

I think real science always works with three instead of two worlds: not the worlds of false and true statements, but rather false statements, true statements and statements of wich their falsehood or truth cannot be proved.
Then it is the question if maybe the truth could still be known if it could never be proven.
Maybe there could in some cases be a sort of 'negative proof' like :we cannot directly proof X, but every plausible alternative for X is disproven so X is the most plausible thing we're left with.

Maybe in some cases you could also us a method of 'reversal evidence' like: we speculate X could be the case, than based on solely this speculation without altering or adding postulations, we come with a wider theory that would predict wich statements refering to the entire world-view of the theory could be proven and wich not. In that case the fact that we can proof Y but not X, could maybe be considered as reason for assuming it's truth. This is a slippery method because you could mold the theory to fit what you cannot proof.

The scenario didn’t suggest that our distant future Earth descendants will be the ones who observe. After all, the Earth will be long gone in 100 billion years. The scenario simply imagined a future planet whose evolution is similar to ours, but which gets started 100 billion years later.

In other words, for this imagined civilization, there is no documentation of what the universe was like 100 billion years ago. It’s not a question of how much emphasis these hypothetical future scientists choose to put on ancient documentation. There simply isn’t any.

This may be how science works in practice, but mathematics certainly doesn’t work that way. A mathematical theorem must be proven true to be accepted. Even if countless counter-examples, even if an infinitely large number of counter-examples, can be proven false, this has no bearing whatsoever on the truth or falsity of an unproven theorem.

(Read about Fermat’s Last Theorem as an interesting example.)

But then what have you to say about Gödel's incompleteness theorems?

One cannot say this with absolute certainty. Down the line of human existence, we may construct ways of documenting crucial realizations of the nature of our universe which may be made available long after the demise of our planet and our very existence.

I didn’t say that it’s not possible to somehow preserve historical information for 100 billion years (although that seems exceedingly unlikely), but rather that for the purposes of this imagined scenario it is assumed that no such historical information is available.

I think science essentially gave up on the problem of 'how things actually are' over 150 years ago. There are two independent but interacting worlds. 'How things actually are' and 'How things appear to the human intellectual apparatus.' Krauss' scenario is a beautiful illustration.

In science we use the human intellectual apparatus to construct models that satisfy two criteria. A nice model should be predictive and explanatory. The model must predict future occurrences that previously appeared unpredictable and it must explain past phenomena that were previously inexplicable.

When we construct such a model everyone gets excited and the newsmen say 'a new truth is found and the old truth was wrong.' This is a mistake that springs from blurring the scientific model with the underlying 'how things actually are.'

Sure scientists seek the truth, but we all know we can't really get there. The best we can do is approximate the truth faithfully within the human intellectual apparatus.

That said, it's working so far.

I see. Points of reference fading away. There could have been points of reference (not only thinking about planets) that are already lost in our timely dimension, leaving us an open end. Discovery would be more like recollection.



The limitations of science, again:



-René Magritte

If its "knowledge for the sake of knowledge" even of phenomena so remote that have no impact in our perceived reality, then could this thirst be quenched even by "faux knowledge"? That is also a question and something that many people might fall for.


If though something or its results are observable directly or indirectly, science can shed some light. Of course it has limitations, but thats hardly a reason to dismiss it or to fly off to "i can make up anything i want for something that cannot be observed by science and still be valid" mode.

What a coincidence; it saw this very same talk by Krauss last weekend, on it's highest dose of cactus yet. It was a fine occasion indeed.



It feels compelled to take issue with your phrasing; this is a limitation of not only science, but human knowledge and perception, whichever avenue it is gleaned through.

If something cannot be perceived, we cannot know it. Nothing "beyond" science magically changes this. Science is simply the best framework we have for assembling knowledge, and there is no other framework that will grant us knowledge hidden to science.

Even extreme claims claims which might seem to be outside the domain of science can be tested and, if necessary, assimilated by it. If Bob claims to see hyperspace dwelling aliens when he smokes DMT and wishes to prove that they are real, scientific experiments can be devised and conducted which might show the truth or falsehood of any hypotheses put forward. If Bob has feelings, convictions or deeply held beliefs about these aliens which he chooses to believe despite an ongoing lack of scientific evidence, they are not knowledge at all, but the kind of unsubstantiated convictions with which psychiatry is concerned.

Yes, of course this is a limitation that extends beyond science. But that doesn’t make it any less of a limit on science. The point is that there are things that actually exist which can’t ever be perceived. As a result, our understanding of reality will remain incomplete.

Regarding your claim that scientific experimentation is the only way to substantiate claims, I must disagree. I know without a doubt that I am conscious, yet there is no scientific test to prove this. Does the lack of scientific evidence for consciousness make my claim that I am conscious an “unsubstantiated conviction with which psychiatry is concerned”?

yup.
then again, psychiatry is not technically a science.
psychiatrists may employ some scientific methods, but the DSM IV is basically a pill catalog for overlapping symptoms. psychiatry has interpretations as ambiguous as many legal statutes; how convenient.

I’m not sure if you’re following the logic.

“Cactophage” implies that unsubstantiated convictions are a sign of mental illness. Convictions that cannot be scientifically tested are considered unsubstantiated. I hold the unsubstantiated conviction that I am conscious. Does it follow that I am mentally ill because I’m certain that I’m conscious?

If yes, then it follows that all people who claim to be conscious are mentally ill.
If no, then there exist certain unsubstantiated convictions that are not indicative of mental illness.

yes.
(R^2 = 0.9999)

there is a correlation between perceiving consciousness and mental illness.

Well, isn't claiming to be conscious is really so circular as to be a truism?

Granted, things can get quite murky when very smart people get very pedantic about the fundamental axioms upon which our knowledge is predicated (Gödel, et al). However, I think it's possible to get surprisingly far using scientific methods, even when dealing with something as slippery as consciousness. In fact, science seems to be making some really incredible inroads into the nature and foundation of consciousness itself in only the last few years.

Where hard, material science cannot be done, and measurements made in a laboratory, there remains enormous scope for progress by applying whatever tools of science remain available. Philosophy is largely the application of whatever can be scavenged from "hard" sciences such as mathematics (formal logic, etc) to matters such as consciousness, identity, morality - matters which hard science usually can't find direct purchase on.

Is it conclusive? Very rarely, if at all. Can it advance our understanding, enrich us? Indubitably, and greatly. Is it more successful than the alternatives? Well, I can't recall the last time someone crashed a passenger jet into a building because they felt strongly about utilitarianism, or felt Critique of Pure Reason to be widely misunderstood.