Suppose there is a claim that outside the observable universe (assuming that there is an outside, other than that we don't know anything about outside) there exists a substance A. Can we conclude that substance A has a zero or close to zero percent probability of actually existing, or do we have to conclude that the probability of substance A existing is inconclusive and unkowable?

  • Can you give an example of a claim outside the observable universe? You mean like the existence of God? Or a physical event before the Big Bang? Can you provide context for your question? What is Subtance A? Is it like Substance D in the great Philip K. Dick novel. A Scanner Darkly? – user4894 Mar 5 '15 at 1:44
  • For example, Substance A could be an object A with certain characteristics. I want the object to be unrelated to our universe so we good say that the substance A is a powerful being for example. I want to be able to say if the specific being is unlikely to be true, but I am unsure if i can. – CognisMantis Mar 5 '15 at 1:56
  • See Logical Positivism (philosophybasics.com/branch_logical_positivism.html) – R. Barzell Mar 5 '15 at 22:26

You either need to be more careful with your definition of "probability of actually existing," or at least use Bayesian statistics.

In Bayesian statistics, there is a factor which can be roughly translated as "knowedness." Thus it is valid to have some distribution (probability of true/false) with a 0 value for how known the value it. This leads to all sorts of interesting odd things like Jeffery's priors.

The other approach is being more careful with the wording. There is no actual scientific way to say "substance A has a % chance of existing" because that phrasing is ontological. It is a statement about the true universe, not what we know about it. An alternate phrasing would be "our empirical models suggest that the probability of A existing is very low."

This modeling step is critical. We have 0 information about whether this substance happens to exist on the dark side of the moon of a planet in another solar system. However, if our model of the universe suggests that most materials like substance A tend to be roughly evenly distributed across the universe, then we can sample the areas around us and do analysis.

In "pure" science, the way you'd do this is to create a null hypothesis that "substance A exists, but statistical anomalies account for why we haven't seen it yet." The alternative hypothesis is "substance A doesn't exist, or only occurs at some very low rate." You would make guesses as to the distribution of substance A (roughly uniform, in our case). You would then prove that it takes a great stroke of luck/bad luck to get a world where the distribution of substance A was just right to prevent us from detecting it.

Once you arrive at something like "there is a 0.01% chance that a random distribution of substance A would go undetected," you can say "we reject the null hypothesis" because we find it does a poor job of modeling our world. Then we get to claim the alternate hypothesis "substance A doesn't exist, or only occurs at some very low rate."

Note: we did not prove that hypothesis, we merely disproved the "prevailing" opinion that it exists, but random chance stopped us from observing it. This is a key difference in wording. Many wars have broken out between science and religion from individuals misinterpreting this path as "proving" substance A does not exist.

  • I was more asking if you did not know anything about the claim other than it is unverifiable, could you dismiss the claim and say that it is extremely unlikely or is it better to say that the claim is unconclusive. – CognisMantis Mar 5 '15 at 3:02
  • Ahh, in that case, the first part on Bayesian statistics will be the direction you are probably looking for. In particular, take a look at the Jeffreys' Prior, which is the direct approach to your question using Bayesian inference. – Cort Ammon Mar 5 '15 at 3:07
  • @CortAmmon Do Bayesian statistics apply "outside the observable universe?" Is the question even meaningful? Am I taking the OP too literally, and perhaps he just means "unknown" as opposed to having some type of truth value literally outside of the observable universe? Your answer makes sense if OP means "unknown" rather than outside the observable universe, which has a very vague meaning if any. – user4894 Mar 5 '15 at 4:51
  • @user4894 I was working off of his use of the word "probability." Unobserved/unmeasured and unobservable/unmeasurable function the same in bayseian statistics. The only difference is that one has the potential to be empirically studied in the future. However, as written, the question is about the present, not the future. – Cort Ammon Mar 5 '15 at 5:11
  • See Logical Positivism (philosophybasics.com/branch_logical_positivism.html) – R. Barzell Mar 5 '15 at 16:39

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.