6

How convinced are you of there being aliens in the universe? Sure, you might not be able to quantify this on any scale. But the level of conviction you feel is certainly there in your consciousness. You may feel more convinced by X than Y or vice versa. Sometimes, those convictions may change.

The question then is: how do you know if your convictions are rational and in some level map to the truth of the world? If Popper’s falsifiability criterion is used, it becomes clear that convictions seem unfalsifiable. Attaching a probability seems to do no good as well, not just because it is difficult to find out the true probability of a claim, but also because the very notion of this concept seems subjective and open to interpretation. After all, in the real world, claims are either true or false.

Perhaps a more modest defense of this can be falsifiable. Perhaps one can establish a reasonable standard of what one’s conviction should be in the face of certain evidence even if one can’t know the truth. But what standard should we use? And is this standard itself testable?

Despite the seeming subjectivity behind this, the feelings we feel are real and the convictions we feel are real. But if they can’t be falsified, how do we know if what we feel about a claim is accurate in any sense? Should they simply be discarded?

thinkingman
  • 6,354
  • 19
  • 53
  • 1
    Isn't that the same question as you asked yesterday? – Frank Mar 11 '23 at 17:33
  • No adult, even the most erudite of them, can answer this question. – Agent Smith Mar 11 '23 at 18:17
  • 1
    @AgentSmith Not quite - I answered it. – Frank Mar 11 '23 at 18:58
  • Despite the provocative title, I guess I don't know what you are asking. How we decide on a degree of belief? What has that got to do with falsificationism? –  Mar 11 '23 at 23:54
  • "How can we know *anything* is false, based on subjective guesswork?" Your question seems to assume that every question in science has the same level of ambiguity as "do space aliens exist?" – RonJohn Mar 12 '23 at 18:19

7 Answers7

5

Yes, they should be discarded, or at least suspended, because feelings alone are not a reasonable justification of anything.

Feelings can be altered depending on our emotional state, for example. They change over time. They can be very wrong: I can have a very strong feeling I can fly, but acting on that feeling may be disastrous. And feelings are personal: other individuals may have different or contradicting feelings with equal validity to yours.

You may have a strong personal feeling about the existence of aliens, but the next person may have an equally strong personal intuition that goes in the opposite direction to yours.

Now what? There is no mechanism for agreement in sight. Either you can seek agreement with others, by discarding your feelings and agreeing with others on objective observations, or you can stick with your feelings, but whatever claims you make based on those may not have any merit for others, or with respect to what actually is the case.

There is a rational need to look beyond opinions and feelings if you want to establish knowledge that is agreed upon by others. Feelings could validate only a type of "knowledge" that may not extend beyond yourself, and could be erroneous to boot.

Mark Andrews
  • 5,894
  • 5
  • 21
  • 38
Frank
  • 2,347
  • 1
  • 14
  • 17
3

I realized early in my philosophic journey, that how we know what we know, was a VERY critical question. Disagreements over how one should do epistemology, put too many philosophic investigations or dialogs off track from the outset.

The question can best be addressed as a pragmatic one -- how can we be most effective in resolving philosophic disputes or questions?

I found that of the methods I have seen used, those of methodological naturalism seem to be the most useful. Threat philosophic questions similarly to how one would treat science questions, and try to spell out a problem, a hypothesis about that problem, figure out ways to test the (or many) hypothesis(es), modify or reject hypotheses that fail tests, and iteratively approach better and better understanding, and plausibly consensus. This is the Popperian approach.

Popper offers a significant improvement over his predecessors, Hume and the Logical Positivists. Hume's rationalist perfectionism threw doubt on this whole methodology (solved by adopting pragmatism instead), and the LP focus on confirmations was VERY susceptible to confirmation bias.

As with all of philosophy, other philosophers have critiqued Popper, and found that falsifiability is not possible in absolute terms (Quine), that scientists don't really abide by it (Kuhn), and that there are exceptions in the way science is actually done (Feyerabend).

In response to these critiques, Popper reformulated his thinking in terms of a network of claims, with central and ancillary statements in the hypothesis family. Popper's reformulation works pretty well, but Lakatos's concept of Research Programmes does an even better job of describing how science both can and should operate: http://www.mantleplumes.org/Lakatos.html

In a Research Programme approach, we adopt it if it is explanatorily useful, and has few and declining problem areas, and there are not better Research Programmes available to us. We then abandon it when the problems are long term, and accumulating, and there are better options.

Problems, and their resolution, can be approximated by falsifications, refutations, statistical confidence, etc. But trying to formulate these concepts explicitly fails. Lakatos' efforts to define progressivity and regressivity in explicit terms, were shown to have logical faults. Same with Popper's efforts to show that science theories incrementally approach truth (increasing verisimilitude) -- this too was shown to have logic faults.

We therefore are forced to operate off judgement, and the intersubjective consensus of people OF good judgement. Applying judgement to Research programs, solutions vs problems, progressivity vs. regressivity, is not that hard for easy cases, and for hard/marginal cases-- we probably SHOULD have multiple competing research programmes running concurrently!

This is a pragmatic, judgement based approach to epistemology. But it is one that follows a set of principles that are FAR more useful than unstructured convictions.

Popper's falsifiability is a good first approximation approach to doing epistemology, and the masters level class is Lakatos' Research Programmes.

Dcleve
  • 9,612
  • 1
  • 11
  • 44
  • what has this got to do with degrees of belief? –  Mar 12 '23 at 00:00
  • 1
    One thing that should be said, is that scientists are not waiting on Popper, Lakatos or anybody. They have experiments to conduct, theories to formulate, try to reconcile those as best as they can and move on. Quine, Kuhn, Feyerabend come after the fact and try to put the pieces together, rather than dictating what scientists do or how they work. – Frank Mar 12 '23 at 00:05
  • @zero. In general, we are unable to assign precise degrees of belief. The attempts to do so create a false presumption of accuracy. And I know this is the case with 98.2% confidence! For a very few cases that we have successful history of analyzing, we may be able to come up with probability and confidence levels for events. I offer an example. The Space Shuttle was probably the best analyzes designed product in history. The reliability evaluation looked at 10,000 different failure modes, and estimated mean flights between failure at 40-140. The actual rate was 2 for 144, or 72. – Dcleve Mar 12 '23 at 04:03
  • 1
    However, NASA admin didn’t like that number, and reassigned a new team, who estimated 10,000 mean flights between failure, which was the published expectation. We are talking about a nearly 100:1 ratio between the competent estimate and a less competent estimate, both with detailed engineering justifications. And even the competent one had a 3.5:1 span between high and low estimates. With these degrees of errors for very thorough estimates, we are mostly just fooling ourselves when we claim to quantify degrees of beliefs. This answer offers a different and more useful paradigm. – Dcleve Mar 12 '23 at 04:15
  • @Dcleve If a degree of belief cannot be justified, why not just toss the concept altogether? Person A having a high degree of belief or a low degree of belief in something doesn’t imply anything about the object of belief – thinkingman Mar 12 '23 at 04:57
  • @thinkingman In some circumstances even a very imprecise metric can be very useful. – Dcleve Mar 12 '23 at 05:35
  • @thinkingman "If a degree of belief cannot be justified, why not just toss the concept altogether?" This is why so many people have tossed belief in the supernatural. – RonJohn Mar 12 '23 at 18:22
  • @RonJohn -- I challenge any dismissers of God or ghosts or souls as "supernatural" to defend that appelation. I consider it has been thoroughly falsified, and lacks any justifications. – Dcleve Mar 13 '23 at 14:55
  • @Dcleve what is a god (not the one-eyed god known as television, but the kind that people have worshipped for millennia, to get the gods to *do* or *not do* something), if not "above nature"? If they were *in nature* and affecting us on a regular basis, something in the Standard Model would point to "magic" particles. Thus, gods are by definition supernatural. – RonJohn Mar 13 '23 at 15:48
  • @RonJohn You just asserted that falsification testing refutes all Gods. That makes Gods natural/testable claims and therefore explicitly not supernatural. You also argued the false claim that everything that exists reduces to the SM of QM. This second assertion is also a naturalistic claim, and easily shown to be false as SM particles will never explain romantic love, the British Constitution, the relative value of different approaches to epistemology, emergence, or the predator/prey population relationship in an ecosystem. See section 5 of SEP on Scientific Reductionism. – Dcleve Mar 14 '23 at 13:17
  • @Dcleve your first sentence is utter nonsense: we can't refute a supernatural being; the best we can do is say "there's no evidence for supernatural beings 'stirring the atoms' to make miracles happen." – RonJohn Mar 14 '23 at 15:57
  • @Dcleve and the part about love and the British Constitution is bordering on Wrong, since we don't *yet* know enough about how the brain works. – RonJohn Mar 14 '23 at 15:58
  • @RonJohn -- repeating a falsehood does not make it any less false. You claimed a God requires SM of QM particles. This is false, but the principle behind the claim is a claim about how much we know about physics, and the necessity of everything to be physics, thererfore refuting a non SM God. DENIAL that you made a naturalist argument about God is -- blatantly untrue. Repeating it when called out -- is just blind dogmatism. Your following up with the BS about "supernatural beings", when you can't show Gods are supernatural is to triple down on dogmatism. – Dcleve Mar 15 '23 at 04:31
  • @Dcleve I think you're trolling me. – RonJohn Mar 15 '23 at 06:33
  • @RonJohn Applying falsification tests to ideological dogma is the opposite of trolling. – Dcleve Mar 15 '23 at 07:27
1

Frequency distributions are measurements.

Probabilities are mathematical facts contingent on measurements about frequency distributions.

Guesses about probabilities are guesses about mathematical facts contingent on measurements, therefore they are guesses about measurements, therefore they are in-principle falsifiable.

I have a guess that flipping a fair coin has a 93% chance of landing on heads. Conceive of a set of measurements which will falsify my guess.

Guesses about the probability of a whole universe having or not having a particular thing or class of things are falsifiable if and only if you can: 1 - measure a frequency distribution whose elements are whole universes; or 2 - establish a model, supported by a statistically significant frequency distribution of measurements that either agree with or disagree with the model.

1: Preferably a large number of universes, so that we have a good P value and can get published. Spock and Evil Mirror Spock are both aliens, but we really need a few hundred universes to be sure, and we really do need to measure the whole universe each time, or at least keep measuring until we find aliens.)

2: Such a model, if we had one, would be usable to correlate guesses about the probabilities at one end of the model (e.g. the probability of universes like this one containing Spock) to frequencies of non-Spock measurements here on Earth - how well do chemistry and biology experiments comport with the model, and how big are the error bars on our data. How exactly that would work I have no idea, or I'd be too busy getting my Nobel Prize in biology to type this.

For 2 - I have a guess that flipping a fair coin a thousand years from now will have a 93% probability of landing on heads. Conceive of a set of measurements that support a model which correlates them to my prediction which will falsify my guess while we're still alive.

g s
  • 2,868
  • 1
  • 4
  • 16
  • You don't always need that much. To falsify that the Earth is flat, you just need one observation. – Frank Mar 12 '23 at 00:48
  • @Frank likewise, to falsify that the universe contains *no* aliens, you only need one Spock. The reason is that the set of all planet Earths only has one member. To falsify the guess that 83% of planets are flat, you need a considerably larger sample set, either of planets (Earth being round could be a fluke!) or of general observations of the laws of physics, which could be derived without ever leaving a cave, and used to convincingly prove that of any given selection of planets that follow the laws of physics, 0% are flat. – g s Mar 12 '23 at 01:17
  • Note that in either case w/r/t 83% of planets, you need a statistically significant, in-principle-measurable sample set, either of physics experiments in caves, or of planets measurable with telescopes, to which my guess about planet flatness probability can be correlated either 1:1 (we look at planets and see if they are round) or with a correlation determined by the model we have for planets, physics, and flatness. – g s Mar 12 '23 at 01:22
  • I should, however, add a note to my response addressing the use of indirect measurements for predicting the likelihood of Spock on the basis of non-Spock measurements. – g s Mar 12 '23 at 01:27
1

Are all probabilistic beliefs unfalsifiable?

I would say no. In fact, I would go as far as saying that all probabilistic beliefs are indeed falsifiable.

I refer to two types of statements as probabilistic: Some P are not Q and Some P are Q. That is, the true state of affairs is somewhere between the twin certainties of No P are Q and All P are Q, and, or between 0% and 100%. Whatever this particular belief might be, it is not certain; it can only be held to a probability.

So, the belief that some swans are white is falsified by proof that no swans are white. The belief that some swans are not black is falsified by proof that all swans are black. These examples, of course, ignore the practical problems of proof; the point here is to show the falsifiability of the initial statement.

Mark Andrews
  • 5,894
  • 5
  • 21
  • 38
0

If the beliver is not an author of his belife(blind faith), it will be difficult to find his experience of the believe, but not impossible, because usually believe based on unconscious. So Freud method of psychoanalysis is usually works or something.

Incidentally your believe that you can prove that belife based(or not based) on a "logic"(formal one) is unconscious belife too. This is not a logic, this is self-understanding-logic, self logic limits.

άνθρωπος
  • 527
  • 2
  • 9
0

Degrees of belief formally represent the strength with which we believe the truth of various propositions.

If we establish something is false, then anything but zero degree of belief is falsified; even if we can only do that to near certainty, every other degree of belief is inconclusively falsified.

Some philosophers think that, despite Popper's claims to have solved the problem of induction, accepting some basic statement (there is a black swan), and so the entire scientific process, is conventionalist: it is just a matter of convention whether or not a theory has been falsified (not much different to Hume's original treatment of the riddle of induction)

Popper’s answer is that the acceptance or rejection of basic statements depends upon a convention-based decision on the part of the scientific community.

From a logical point of view, the testing of a theory depends upon basic statements whose acceptance or rejection, in its turn, depends upon our decisions. Thus it is decisions which settle the fate of theories. (2002: 91)... accepted by an act, by a free decision.

If you are wondering how we can be certain about anything, especially science and what has been falsified, then Popper's answer evolved. Early in his career, he thought falsification was certain, but this claim was abandoned, just as he introduced a proxy to confirmation (corroboration).

I see nothing particular to statistics and psychology here: introducing a percentage does not add any further paradoxes or issues to a lack of certainty.

  • Yes, it is really not conclusive whether the earth is flat or not (for example), as you point out, nothing in science is conclusive. – Frank Mar 12 '23 at 00:07
  • I see what you mean @Frank but it does seem that Popper thought we have not 100% falsified any scientific theory. that's philosophy for you (the 0.000% chance no-one exists but you) –  Mar 12 '23 at 00:12
  • For sure, but it doesn't quite matter in practice. – Frank Mar 12 '23 at 00:14
  • I absolutely agree @Frank –  Mar 12 '23 at 00:15
  • Like I said, scientists are not waiting on Popper to tell them what to do. Popper comes after the fact and tries to make sense of what scientists have done. – Frank Mar 12 '23 at 00:16
  • I don't think philosophers have much say in what gets funding in physics, though I have been told that some of the soft sciences (sociology e.g.) are very heavily influenced by philosophy @Frank –  Mar 12 '23 at 00:17
  • 1
    Maybe that's a good demarcation criterion between hard sciences and soft "sciences" - whether philosophers or scientists rule the roost :-) – Frank Mar 12 '23 at 00:24
0

Your question is in a sense self-contradictory. We have feelings about whether something might be true or false precisely because the possibility is not immediately falsifiable- in the absence of a cast-iron rationale, we must fall back on judgements. Remember that the human brain is the product of evolution, so how we make judgements about the world around us is rarely a matter of applied logic.

Every day, humans collectively make billions of judgements- some turn out to be justified and some not. The types of judgements we make vary enormously. To take a tiny number of examples at random, consider the following questions:

Will it be warm enough to eat outside next Tuesday? Should I invest in crypto-currency? Should my film company remake Titanic? What's the best way to secure peace in the Middle East? Will this marketing message be effective? Will my wife like this present? How should I price this tender? Was the damage to this sapling caused by a badger? Do I still have time to get to the station in time for the train?

Questions like those are not decided by conscious logic- we make judgements, often snap judgements. Sometimes they are proved right and sometimes they are proved wrong and sometimes we never get to find out whether they were right or wrong. Different people make conflicting judgements on the same evidence. So you can't know whether a judgement is accurate until time proves it one way or the other- if it were possible to know, people wouldn't make millions of wrong judgements every day. That said, you would be wrong to discard all your beliefs- the trick is to become better at recognising which ones matter and which ones are more trustworthy. Are there aliens in space? For most people, the question is utterly irrelevant, and most people, if asked, would make a snap judgement- they wouldn't try to reach a reasoned position based on a careful analysis of all the relevant factors.

Marco Ocram
  • 8,686
  • 1
  • 8
  • 28