Philosophy of Science 4 - Goodman's Problem of Induction

[@keremdogan122]

Only video that is meaningful to ask "who is watching this in 2021?" Great explanation as always!

[@Leo-io4bq]

2023

[@Trynottoblink]

You ever consider monetizing these videos with a brief ad in the middle or at the end? You deserve to make at least some money off this work.

[@ianhruday9584]

The predicate looks odd, not because of the time dependence but it refers back to observation. So, what would emerald be which is never observed? To say that "all emeralds are greed," is not to say that "all emeralds will be observed," or even "all emeralds when observed will be seen to be green," since we can always appeal to deviant viewing conditions. One thing which looks odd about the grue predicate is that it implicitly makes reference to the conditions under which we look for inductive evidence. Moreover, it refers to two different sets of viewing conditions. "Green" is not like this. We have a sense of what "green" refers to, but we don't know what grue or bleen look like. One possible constraint on projectability might be to say that the observations which confirm the inductive inference should remain relatively stable across viewing conditions, and where they do not - e.g. viewing an emerald under red light, we should have or be able to find good explanation for their instability. I think this is pretty intuitive. If we say "watching the sun rise today confirms the induction it will rise every day," and it equally confirms the induction "it will rise every day except tomorrow." We are making a mistake, because we expect different observations to confirm the same inference. I can already see an objection to this, but I'm not sure if it can be answered. Suppose I want to confirm the hypothesis that the buss only runs every second day. I do think I can confirm this inductively. So, maybe what's odd about these deviant predicates is that they hide the alternate observations, or the alternate observations occur irregularly. The buss schedule still follows a simple regular pattern whereas the grue predicate doesn't. I don't know does this work? Can it be modified to work?

[@chuckbartelt4920]

I've been hoping you would do this. Thank you.

[@Sunfried1]

To any logician put there: Does not the Grue problem incorporate a contradiction (Green, then not green)? Thus would a deduction for Gruemess not be vulnerable to the Principle of Explosion where one can infer any conclusion from the green-not green contradiction? I'm no logician, just asking.

[@batkinson130]

I know I’m commenting on a 5-year-old video, but I’m not really convinced by the argument that you can make green and blue time dependent by defining them in terms of grue and bleen. For instance, it’s 2022 now and I’m wearing a green shirt. Suppose we lived in grue world, and suppose someone were to go to a thrift store tomorrow and find this shirt. How would they know if it was grue or bleen? They would need to know when the shirt was made, and they don’t have any way of knowing that. (Sure, maybe in grue world they’d print the manufacture date on the tag, but let’s say the tag was ripped off.) However, they could still know for sure that the shirt is green, since just by looking at it they would know that if it was made before 2020 then it’s grue, and if it was made after 2020 then it’s bleen, so in either case it’s green by the definition you give. Therefore, even in grue world, green and blue are time independent and make for more suitable projectable predicates. They only seem to be time dependent in grue world because they need to “cancel out” the time dependence baked into the definitions of grue and bleen. This isn't to say that the argument that deviance can depend on language is wrong, just that the example you give doesn't seem to illustrate your point.

[@StatelessLiberty]

I wonder if this is related to Wittgenstein’s rule following paradox. We want to assume in science that like causes produce like results, nature will “go on in the same way.” But there are infinitely many ways you could “go on in the same way” depending on how you interpreted the past incidents.

[@StatelessLiberty]

Oh you already said it. Nvm

[@thefinnishbolshevik2404]

This isnt exactly the simplicity argument, but wouldnt we simply want to use the most precise common attribute for the emeralds? We've only found green ones so its reasonable to assume they're all probably green. Why assume they might be blue-green? Why assume they might be red-green? Lets just stick to green. Maybe im misunderstanding but grue seems to be a broader category which includes green and blue, but how could we predict the emeralds would be blue, since we didnt observe any? The term grue seems to be assuming the future blue emeralds which we could not have observed yet. Or am I missing something?

[@jagsittermedsimonochjobbar]

Why haven’t I come across this channel before? Your videos are great :-)

[@pmyou2]

The concept of 'grue', were it a math concept, would not be continuously differentiable. It has a discontinuity. To use 'grue' correctly, the speaker / listener would need to understand this. That is the reason for the failure of grueness predictions. It fails to note the intrinsic discontinuity in the concept.

[@Oskar1000]

Hmm, I like this idea. But aren't there continuous differentiable functions that would sort of look like this function. It would be low for a long time and then jump up quickly (though not instantly).

[@pmyou2]

@@Oskar1000 It is that it has a singularity, a point at which it is not differentiable that I am underscoring because that is what grue appears to be for me. If you have an example of something 'gruelike' where it is continuously differentiable, that would be interesting. Do you have one in mind?

[@Oskar1000]

@@pmyou2 Well sine-waves are continously differentiable right. And sums of any number of sine-waves of different magnitudes would also be continously differentiable. And it's a known fact that you can combine sine-waves to produce all sorts of lines. This technique is used in compression of music (and images). So we could form something that would look like a line that's at 0 for a long time and then (almost) instantaneously jumps up to 1. It would take an insane amount of sine-waves but it should work.

[@ArbiterElegantiaee]

so lucid and succinct. thank you!

[@danielwa4819]

Goodman himself "solved" the problem - its not the problem of language, nor simplicity, nor natural kinds, but a problem of projectible predicates based NOT on successful usage of language, but that we have used much more projectable hypotheses using "green" than "grue" in the past and they have worked out in favor. Goodman was ultimately not skeptic of epistemology but a pragmatist (and I read Hume in the same way too - it's not a problem to be solved but to be defused in a Wittgensteinian sense ).

[@antonioesposito2242]

what was that lol @11:00

[@KaneB]

No idea. Sounds like "Heeeeere's Johnny". Somebody must have been watching The Shining.

[@johannmeerbaum7798]

@@KaneB I thought the exact same thing!

[@SingedAndZoeGaming]

Sounds like “yeah buddy” to me

[@ayeshazainali5708]

Very very effective video💚💚

[@azimuthal1392]

Nice exposition! I think emeralds are kinda bluish-green anyway :p

[@horsymandias-ur]

Any day now

[@ibukontraktor-taqwaproperty]

These guys who reject induction need to learn more about things other than just sitting at their den and think that they are smarter than other people. My God! This is such a 7-year old way of thinking.

[@Ansatz66]

It's no wonder that induction is so seductive when you can prove anything you like by using it. It feeds perfectly into people's confirmation bias. Just pick a hypothesis that sounds plausible and induction is guaranteed to agree with you, making induction the perfect ally in any situation.

[@DarrenMcStravick]

Problem is, same thing applies to deduction. Just look at Curry's paradox.

[@Ansatz66]

@@DarrenMcStravick : Very few people take self-referential sentences seriously. Any sentence that begins "If this sentence is true..." is obvious silliness, so if that is what you have to do to twist deduction into "proving" what you want, then deduction is on pretty solid footing.

[@DarrenMcStravick]

@Ansatz66 Curry's paradox isn't the only problem, you also have the problem of deduction, the logo-centric predicament, and the uninformativeness/triviality of logical truths. Also, how exactly does one not "take self-referential sentences seriously"? Do you have a reason for not permitting them into logical arguments?

[@Ansatz66]

@@DarrenMcStravick : The problem of deduction is just a case of what can happen when people fail to communicate due to having different interpretations for certain words. That problem is not really about deduction so much as it is a fundamental problem for all human communication, regardless of what people are trying to talk about. The logo-centric predicament is just a misunderstanding of the purpose of axioms. Axioms are not supposed to be claims to be justified; they are rules of discourse that are accepted to help facilitate communication, not because they represent real truths about the world. If X is an axiom, that means we treat X as being true in our discussion regardless of whether it is really true, just so that we have a common foundation upon which we can build. You do not need to believe a thing in order to assume it for the sake of discussion, and you do not need to justify a belief that you do not hold. Deductive truths may seem trivial and uninformative because you use them so pervasively and effortlessly. People mostly use deduction without even giving it a single conscious thought, but if we were somehow forbidden from using deduction then all communication would quickly fall apart. It seems trivial when we take it for granted, but we would miss it terribly if we did not have it. "How exactly does one not 'take self-referential sentences seriously'? Do you have a reason for not permitting them into logical arguments?" Not all sequences of words are meaningful, even if each word in the sequence is a well-accepted English word with clear meaning in isolation. For example, "Apples kite koala rustic amber." We might ask why this sequence of words has no meaning, but there is no good reason. It just so happens that the English speaking world has not collectively devised a meaning for it. Maybe someday in some slang dialect that sequence will become a meaningful sentence, but not today. That is why we exclude self-referential sentences from logical arguments, because they are usually meaningless.