ฝัง
- เผยแพร่เมื่อ 15 ก.ค. 2024
- This videos outlines counterfactual skepticism, which proposes that most ordinary counterfactuals are false, defended most famously by Alan Hájek.
I offer private tutoring in philosophy. For details please email me: kanebaker91@gmail.com
Support me on Patreon: / kanebaker91
Donate to my PayPal: paypal.me/kanebaker91
My Discord: / discord
0:00 - Introduction
1:38 - Chance undermines "would"
12:03 - "Might" undermines "would"
15:07 - Underspecification of antecedents
19:58 - Disjunctive poisoning
24:10 - Quasi-miracles
29:57 - Contextualism
36:31 - Objections to contextualism
![[TH] VCT Pacific Stage 2 - Playoffs Day 2 // PRX vs GEN | DRX vs TLN](http://i.ytimg.com/vi/gmrSk-_ahlA/mqdefault.jpg)
Not sure why the TH-cam algorithm recommended a 47-minute Powerpoint explaining a niche logical and linguistic position, but I watched all of it regardless.
Great video, very well-explained.
Thanks!
By the way, true to form, analytic philosophers deeply consider the possibility that a glass quantum-tunnels through the Earth, but ignores the possibility that the glass falls on a rug and survives...
I kept thinking it could bounce, but when I think of a glass I'm biased because I don't drink alcohol and so I don't expect something like a wine glass - something more like what you'd drink lemonade or milk from, where others could be more likely to assume a glass that more believably can't bounce.
I don't think analytic philosophy and ordinary language can be friends. I wouldn't invite them both to the same party, because I don't trust them not to fight, or even to politely ignore each other.
Counterfactuals are basically natural language's shortcut way of doing two things: 1) visualizing what would happen if the world was the same except "one thing" (which is actually a really complicated thing to try to parse out in a formal theory) and 2) assessing a high bayesian probability as to what would happen if the world were set up in this way (also really complicated to precisely delineate for any philosophical system). The brain is really smart so it came up with a shortcut/cheat for expressing these two things at once, but humans trying to philosophize are really stupid so our formal systems break when counterfactuals (basically a really useful cognitive tool) are put into formal proposition systems.
The formal system ends up classifying things as "false" which really aren't actually formal truth claims, but the response will always be "K but counterfactuals are great and your system sux because it doesn't have them". The fact that prohibiting "false" things leads to "no counterfactuals" is so intellectually expensive that prohibiting "false" things becomes really (obviously) stupid.
@@tyrjilvincef9507 And you know what is also a really stupid analytic philosopher? Computers! Trying to figure out how in the hell our brains manage to come up with the idea of counterfactuals and how to simulate that in computers is probably one of the most difficult steps in getting a 'true' artificial intelligence.
This clarity delivery experience and teaching that you explain carefully and pondered is so refreshing and soothing, also so tempered explained. Revigorating!Thank you alot! ^^
I have never thought about it that way. thanks kane b!
All the air molecules moving to the corner of the room isn't like quantum tunnellling through the earth, in that it not only won't happen because it's so vanishingly improbable, but actually there isn't the same inescapable chanciness, with stat mech, as you've said, it assumes the underlying is deterministic, and without the unspecified initial conditions containing a shockwave on the horizon or something equivalently artificial/low-entropy, it probably simply can't happen.
The objections to counterfactuals seem to fail if we think of counterfactual conditionals as epistemic rather than ontic. A true or acceptable counterfactual is typically one that is the consequence of the most plausible theory or model we have. There will always be rival models, so we can tolerate a contrasting 'might be'. But in most cases we are able to single out the most plausible model.
I mean you mentioned dialtheism. But what is with fuzzy truth-values? The truth-value of „if I would have not flip the coin then it would not have landed tails“ might be the probability of the scenario that the coin would not have landed tails if I did not flip the coin.
Also, what do you think about the proposition „If I would not have drop the glass, then it likely would not have been broken“?
On chance (first argument), the statement "If I dropped the glass, it would *probably* have shattered" is a counterfactual, but remains true. All we need to do is (implicitly) assign probabilities to counterfactuals, which would be a good idea anyway, and problem solved.
PS: This actually works against all of the arguments listed in the video. They all assume counterfactuals are intended to be understood as absolute, necessary, with a probability of exactly 1. In reality, people making counterfactual statement practically always intend the probability to be high, but less than 1.
@@whycantiremainanonymous8091 I don't think that weaking the consequent in that way is going to help with all of these problems:
(I*) If we were in Rome, or Venice, or Florence, or Naples, or Baghdad, then we would probably be in Italy
This seems straightforwardly false, no matter how many Italian provinces you add to the antecedent. If we were in Rome, we just would be in Italy. If we were in Baghdad, we would not. Probability doesn't seem to help here. I think that what (I*) has in common with
(G) If I dropped the glass, it would have shattered
is that we judge both to be approximately true. In the case of (G), we can transform an approximate truth into a pure truth by probabilizing the consequent: "If I had dropped the glass, it probably would have shattered." (Or rather, we can transform an approximate truth into a more-precise-but-still-approximate truth by doing this. In my view there are idealizations involved even in the latter counterfactual, such as fixing a referent for "the glass". But that's beside the point with respect to Hajek's arguments.) For (I*), we transform an approximate truth into a truth by deleting just one of the disjuncts. (I*) and (G) in different ways both approximate some other counterfactual that is true, or closer to the truth.
@@KaneB I don't think your objection works. "If we're in either Rome, Florence, Naples, Venice, or Baghdad, then we're probably in Italy" would be a true statement, so long as we're also given the information that our being in Baghdad is not very likely. It would be false only if we claim total ignorance about the probability distribution (or if we know Baghdad is at least as likely to be our location as the other cities combined), but Hajeck acknowledges the poisoning disjuncts are all low-probability events.
Hajeck's argument works against unqualified counterfactuals. But then the same argument can be made against unqualified factual statements too (because there are always possible, even if extremely unlikely, scenarios in which our information about the facts is incorrect). Unqualified statements (meant literally, which they never are in real life) are generally false. But that shouldn't be surprising.
@@whycantiremainanonymous8091 (I*) doesn't say anything about the probability of us being in Baghdad, and that probability assessment just seems like a separate claim, irrelevant to the assessment of the truth value of (I*). "If we're in Rome or Baghdad, then we're in Italy" is straightforwardly false, no matter how low we take the probably of us being in Baghdad to be. After all, we're stipulating that it's a counterfactual, so we're assuming that we're *not* in Baghdad. We can set the probability of us being in either Rome or Baghdad at 0. Even so, "If we're in Rome or Baghdad, then we're in Italy or Iraq" is true, and "if we're in Rome or Baghdad, then we're in Italy" is false... or so it seems to me.
@@whycantiremainanonymous8091Compare: "If we're in Rome or the Andromeda Galaxy, then we're in Italy." Given our background knowledge, the prior probability of us being in the Andromeda Galaxy will be very low. No human being has ever been in the Andromeda Galaxy, nor do we currently have the technological capacity to send anybody to the Andromeda Galaxy. Even so, this counterfactual is false, and we can't fix it by probabilizing the consequent: "If we're in Rome or the Andromeda Galaxy, then we're probably in Italy" is also false. I think that if we add a bunch of Italian provinces to the antecedent -- "If we're in Rome or Naples or Florence or Venice or ... or the Andromeda Galaxy, then we're in Italy" -- then it's reasonable to say that the counterfactual is an approximately true. But I don't think it's probability that's doing the work here.
curious if you chose the names Rubin and Cherise as an homage to the Jerry Garcia/robert hunter song?
No it was just a coincidence that I had the names "Rubin", "Cherise", a parade, with Cherise watching Rubin dance, and that I also happen to be a big Deadhead.
as far as I know, the only indeterminacy in the universe comes from the collapse of the quantum wave function; so unless we can directly tie a counterfactual to a quantum collapse, to make it true either the laws of physics would have to be changed or the starting conditions of the universe would have to be changed. And it seems to me there would be unimaginably far-reaching side effects of either choice that makes the whole point of the counterfactual moot in the first place
It seems like the only real thread is that ordinary language isn't philosophically rigorous, which we're all well aware of. If you're having an ordinary conversation and you say "If you dropped that glass it would have shattered" the response of "Well actually it might have quantum tunneled through the Earth" marks you as someone who can't interact with other people.
Ordinary language almost universally supposes a "probably", but it's unwieldy to insert that into every statement we make. Though we are also very likely to promote those almost statements from probable to determinate, which causes all the various problems you get when you use inductive evidence for deductive arguments.
The case of a magical coin that can only land heads or tails is at least a little more interesting, and there are some interesting things to say about probabilistic events and how we understand them but it doesn't seem like Hajek is hitting any of that, but is instead focused on finding complicated ways to demonstrate how ordinary language isn't logically rigorous. That doesn't even rise to the level of higher order truths about Chmess.
Regardless, love your channel, and I'm happy to see your continued growth and success. Being a public philosopher is noble work, and skepticism is very under-appreciated in the field.
What about a counterfactual formulation of a thermodynamic law, say.
'If energy were observed in a closed system then it would remain there' is true.
The counterfactual expresses a (nonaccidental) exceptionless regularity. It amounts to a phenomenological law rather than a theoretical one, so there's no need to rule out quasi-miracles ... The counterfactual is true in all those worlds where the accepted laws of nature obtain.
That's a statement that is by definition true.
A closed system is defined as energy not entering or leaving.
Most counterfactuals will not be like that.
But perhaps it is only appropriate to say "WOULD" when it is by definition true and use "might" "Probably" or "almost certainly" in all other cases. I think it's pretty clear that in normal language, we take "will" to mean something else than absolute, gun-to-the-head certainty
@@loganvanderwier8866 I can see why you say that. Under conceptual analysis the statement is true by definition. I guess it has more to do with my crow-barring many presuppositions into a single you tube comment...
The thing is the term 'closed system' has legitimate meaning here in virtue of certain scientific background conditions. A scientifically respectable notion of 'closed system' involves intuitive (explanatory) elements. Our nontrivial understanding of what a closed system is will go beyond the observation reports. We can say that the laws of nature only apply within the (closed) experimental set-ups we observe, but doesn't science want to say some something more about the physical world. If that's right, then scientific statements of the above sort are contingently true. For instance, we might come across some anomalous data which forces us to revise our understanding of those concepts in current use.
Any chance of a vid on wittgensteins later philosophy? personally not sure how philosophy (of this manner) continues after the lack of any strong argument against his metaphilosophy
I don't find Wittgenstein very interesting, so probably not any time soon.
@@KaneB Really? Do you not think that the private language argument, if sound, affects pretty much most areas of philosophy? Especially things, from what I can tell, you have an interest in (truth, modality [and metaphysics more generally], knowledge/scepticism, etc)
@@yyzzyysszznn It might be relevant to them, but lots of arguments are relevant to those topics. I would also distinguish Wittgenstein's rather poor (in my view) work from the very interesting work done by some people who were elaborating on ideas they found in Wittgenstein. For instance, Kripke's presentation of meaning skepticism is a fascinating skeptical problem. Also, I don't think Wittgenstein's broader meta-philosophy actually poses any challenge to more traditional approaches to philosophy, even on its own terms. Even if I were to accept everything Wittgenstein said, I wouldn't feel under any obligation to alter my approach to philosophy; I would simply conceive of it as a special kind of language game. I think the most damaging criticism that Wittgenstein can make of traditional philosophy is that traditional philosophers have a mistaken conception of what it is that they are doing... but that doesn't strike me as a particularly serious problem. (After all, given how much disagreement there is about meta-philosophy, I would expect that a large proportion of philosophers have a mistaken view of their own activities.)
By analogy, consider the arts. Here is a story that might be something of a myth, but that hopefully is illustrative. In the past, artists aimed to create realistic representations of nature. They assumed that there was a distinction between realistic and non-realistic representation, and that the proper aim of art was realistic representation. Over the 19th and early 20th centuries, both of these assumptions were challenged. Art need not aim for realism, and perhaps "realism", in the sense of accurate representation of nature, is not even a coherent idea. This opens up the space for artists to pursue different goals and different styles. But there is no argument here that artists ought to stop creating art in the style that was traditionally labelled "realist". Many artists continue to refine that style; it's just that today, we understand the style in a slightly different way. I can deny that there is such a thing as accurate representation of nature, or deny that this is the aim of art, without having any objection to the artworks of Chuck Close and Richard Estes. I think the same goes for traditional philosophy, given the sorts of criticisms raised by Wittgenstein.
@@KaneB I think the comparison with art suggests a misunderstanding, if I may say so. There are many philosophical problems which Wittgenstein's work elucidates as being plain nonsense, or non-problems resulting from confusions about concepts.
It follows from the private language argument (PLA) that behaviour is a criterion for concepts we have traditionally deemed 'internal', such as thinking, or pain. The PLA shows that pain is public if it is to make sense. Now lets look at something like Chalmers' P-Zombies, who in our case, mutatis mutandis, does not have the internal feeling of pain. Wittgenstein shows this is no problem-if a wheel turns nothing, it is not part of the mechanism.
You seem to be saying, "but hey, lets talk about it anyway, we still can, because its just a change like an artistic change." It isnt-it has been shown to be nonsense. The referentialism behind a problem such as the sorites paradox acts like "we dont know whats a heap and whats not" is a deep philosophical problem. It isnt-we do know a heap from a grain, and in between, there just isnt a straight referential link. You say, what-we should still try to find the metaphysical nature of the heap?
I remember you said in a video about Lewis that you hated the style of philosophy wherein "possible worlds" becomes reified and an object of metaphysical analysis. Wittgenstein shows why such endeavours are nonsense.
@@yyzzyysszznn I was addressing Wittgenstein's broader metaphilosophy, which is what you initially asked me about, not the private language argument specifically. As I said, the private language argument -- or at least some versions of it as it has been developed by other people -- might be interesting, but it's just one among many interesting arguments that bear on these topics.
>> You say, what-we should still try to find the metaphysical nature of the heap?
I say you can either tell me what error is made in the Sorites argument, at which point you're playing the philosophy game, or you can just ignore the discussion. There's nothing wrong with ignoring the discussion, but I also don't see what's wrong with engaging in the discussion by Wittgenstein's own standards, even if the discussion comes out as nonsensical according his favoured conception of meaning. (Additionally, I don't think Wittgenstein is successful in showing that any of these things are nonsense. But that's beside the point. Just to be clear about my view, though: Lewis's talk of "possible worlds" isn't nonsense, it's simply mistaken, and I arrive at the conclusion that it's mistaken by engaging in rather traditional philosophical analysis.)
Wouldn't this apply to chancey conditionals in general?
"If I drop the glass, it will break" -- Let's say I do in fact drop the glass, and it breaks. Then the conditional is straightforwardly true, since it has a true antecedent and true consequent. I do drop the glass, and it doesn't break: then the conditional is straightforwardly false, since it has a true antecedent and false consequent. What if I don't drop the glass? Then the conditional is vacuously true, at least on the standard semantics for indicative conditionals.
With that said, there are folks who have argued that Hajek's arguments entail the falsehood of most conditionals in general, which leads to a much more radical skepticism than he intends to defend. Check out David Boylan's article "Counterfactual skepticism is (just) skepticism"
@@KaneB thank you, the first 10 pages of the article articulates essentially what I was thinking. I guess what I had in mind by "chancy conditionals in general" was future indicatives in particular, but Boylan takes it much further than I was willing to go 😂
Is Newtonian mechanics one giant counterfactual?
what every kid was thinking when their mom would pull out this argument 😑 lol
I have met someone who told me they wouldn't change their beliefs, even if they were brought to understand they are false
Also, I know spiders are not scary. But I don't believe it. And there are many others who would not act upon knowledge, but rather on beliefs
In my opinion the problem is that Hayek is demanding that natural language be precise but, while no natural language is as precise as Hayek would like, English is a language that revels in ambiguity. This ambiguity is sometimes useful. For example the ceasefire agreement at the end of the six day war included the line 'occupied territories will be returned to the Palestinians'. The UN drafters were careful not to apply quantifiers or articles to 'occupied territories' in order the Israeli negotiators could think 'Some Occupied Territories' and the Palestinian negotiators could think 'All Occupied Territories'. Of course, the ambiguity fell apart in 1973.
If most conterfactuals were false, this statement would also have been false.
The statement "most counterfactuals are false" is not a counterfactual.
My favorite line from the movie 'Clerks' is when Randall says "This job wouldn't be so bad if it weren't for all the customers."
The line is funny because it is both true and false. While the customers are a source of annoyances, the job itself would not exist if there were no customers.
I do not see that this is anything more than declaring that (most?) fictions are fictional. I thought the point of counterfactuals is that they are not true, the clue is in the name. I am intrigued however by what counterfactual could be true. On counterfactuals generally we can legitimately estimate probabilities, even though we can still dispute the specific probabilities.
The antecedents are not true, but as I understand it, the standard view is that counterfactuals can be true. When I say, "If P had been, Q would have been", what's counter to fact is the supposition that P had been, not the connection the counterfactual asserts between P and Q.
@@KaneB So given that Q is part of the defining qualities of P, we can say: if P a priori > Q, even though P is false. Hence most, rather than all counterfactuals are false.
Generally though I think the point of counterfactuals is that they are not true, but that they express expectations, whose likelihood can be disputed. The interest in such counterfactuals would be when they provide a causal account of events that did not happen but might have happened; if they don't do this, other than perhaps being a fun diversion, there can be little interest. Religious people would not agree though because if God > Q is a staple of their faith.
is adequately but not overly informative (quantity maxim)
the speaker does not believe to be false and for which adequate evidence is had (quality maxim)
is relevant (maxim of relation or relevance), and
is clear, unambiguous, brief, and orderly (maxim of manner).
Many counterfactual will be false in a manner wholly irrelevant to information being communicated.
In context the sentence "If you had dropped the glass it would shatter." Could have meant that glass is worth 1,000 euros and if we break it we would most likely have to pay for it. The truth of the counterfactual and knowledge of its truthyness are irrelevant what is relevant is what has gone unsaid.
"The stolen coin would have landed heads"
"That's not true"
"So you think it would have landed tails"
I think this is the usual understanding of counterfactuals. A better example might not involve claiming psychic powers:
"You stopped me from walking into traffic. I would have died!"
"That's not true"
"I'm a lot smaller than that semi!"
It seems the truth value of counterfactuals doesn't pin on the 'would' part of the sentence at all. That part is assumed or not truth apt, or else you're not talking about the same thing.
I think the possible worlds contextual stuff is closer, but I really doubt anyone means anything like that when they speak. It would be very surprising if even trained philosophers ever gave that a serious thought irl.
If I met someone who was genuinely confused about these (like someone who without studying philosophy somehow thought these were mostly false as a rule) I'd say in the coin case the speaker is telling you in stopping the coin from landing the thief stopped the coin from landing and they believed the result the thief stopped was heads. Say the thief caught the coin and ran - the counterfactual is confirming you both agree he stopped the coin landing, and it's a confusion to think the truth value landed on would in the regular case.
Here's the common irregular case:
"You stopped me walking into traffic! I would have died!"
"Don't be so dramatic the road is empty."
Here, it doesn't seem they're saying the counterfactual is false, but a plain misunderstanding. They didn't just misread the result, but the shared understanding regular people have with eachother about their surroundings when talking.
(Edit: possible worlds doesnt seem anywhere near as plausible as saying our regular mostly shared understanding of the present projects into the future a bit. If you make "projection" something like a model we assume is shared, we can also say sensible stuff that contradicts physics like "if coins landed on their side a third of the time, then heads would also land a third". Our brains are riddled with shortcuts - it's strange to think without ever mentioning anything of the sort that we refer to whole other worlds)