Infinitism

According to infinitism in epistemology, justification requires an infinite series of non-repeating reasons. This video outlines the motivation for infinitism and discusses some of the objections.

0:00 - Introduction
1:22 - The regress of reasons
8:57 - No final questions
16:58 - The finite mind problem
25:47 - The origin problem
30:28 - Alternative chains
34:55 - Specifying the basic beliefs

[silverharloe]

Infinitism: when the three year old asking "why" on loop gets *soooo* frustrated with their parents stopping at "because I said so" that they are forced to write a philosophical retort calling their parents arbitrary and thus unjustified.

[facundocesa]

I have 2 parents.
I have 4 grandparents.
I have 8 great grandfathers.
And so on.

But I don't have infinite ancestors. They start repeating themselves in my lineage.

Reasons and justifications also have parents and grandparents.
The more parents and grandparents they have, the more trustworthy they seem to me, in principle.

If some of the parents of one of your reasons turns out to be false, well, it has others, so it's fine.

And reasons, for me, work like ancestors, but with time travel. So some lineages CAN contain loops, but there's nothing wrong with that because the elements of the loops have other connections.

Interestingly, we have inherited reasons from our ancestors, not only in the form of culture but also in the form of the senses we use to perceive things.

[Lojak-exe]

LETS GOO!!! Honestly there couldn’t be a better timing for this video, at least for me! Thank you

[zenoncortez]

I actually first encountered this in a math textbook! They described sets as primitive objects and said they couldn’t rigorously define them because of the issues posed by the regress of reasons.

[InventiveHarvest]

Myself, I use circularity as we can differentiate between progressive circles and degenerate circles. Progressive circles lead to new knowledge, creating a positive feedback loop. Degenerate circles go nowhere.

However, this video has shown me that I am also sympathetic to partial justification. The stopping point can be defined pragmatically at the point where marginal benefit equals marginal cost. With this in mind, one can determine a number of times to go round the circular reasoning and consider themselves justified.

[loryugan]

Wake up babe. New KaneB video just dropped.

[bankiey]

On acid, you can take any little thing that is said in earnest and instantly see the next regression, and the next, and the next, in fading clarity, and it strongly reminds me of the cascade of images when two mirrors face each other, except if every other image was rotated 90 degrees.

[TenaciousWombat]

The idea that comes to mind for me is to consider a set of propositions with a "reason for" relation. Then you can define an infinite justification recursively:

J_0: p
J_n+1: x such that x is a reason for J_n that has not already appeared at some stage m<=n

Then you don't need the entire infinite object, you just need "J_n exists for all n". If the chain ends then it must do so at a specific finite stage, so instead of asserting an infinite proposition you are denying all propositions of the form "there is no J_n" for specific n.

But really there would be a lot of different possible chains, and maybe some of them end and some don't. You could define the tree of potential justifications as:

J_0: {p}
J_n+1: {x | x is a reason for an element of J_n}

And then you want that J_n is nonempty for all n, which only requires one chain to continue. If there are infinitely many propositions then it seems pretty plausible to me that there would be infinite chains. Like, the claim that they all terminate or are all eventually circular seems more bold.

[exandil]

I think infinitism's the way to go. In my opinion, people and philosophers struggle with it because they believe they have definite knowledge. People want to think they are absolutely right and that there are no reasons others may want to refute that. Another issue, I think, lies with the word knowledge. People equivocate different senses of knowledge: sensual experience with rational critique, intellectual intuition with description of sensual reality. I think there is place for foundationalist and coherentist motifs in pragmatically "justifying" sensual beliefs and justifyably understanding abstract belief systems, but all and all the two principles of infinitism seem to perssuade me. We don't have absolute knowledge, only relative knowledge, all beliefs need to be justified to be knowledge and only beliefs may justify beliefs. Since anybody can always seek to refute a belief, justification must in principle go to infinity. I think anybody who argued against an opponent about politics or ethics can vouch for this. Sorry for the rambling, I was making it up as I went.

[brandonsaffell]

Hey, another great one. I was discussing metaphysical infinitism the other day, and it seems like metaphysical infinitism would necessitate epistemological infinitism.

Regardless, I think epistemological infinitism is the way to go, and I find it very surprising the field is so opposed to it. Especially as compared to alternatives like seeming / brute facts / mere facts / circular webs of belief. If only we had found those golden first principles we could insert into our formal logic machine, life would be so much simpler.

[Opposite]

-What worries me about infinitism are the rules of inference, I mean any rule for epistemic/logical reasoning and not just deductive rules.
-I may justify a rule A with the set of Rules S. Now S has to exclude A to avoid circularity.
-Now I want to justify another rule B from S with a smaller set of rules S*. But S* is basically S just without B.
-It seems like I have now less and less rules until only one or a irreducible small set of rules is left over.
-It seems like that a infinite chain of reasoning is not possible for rules of inference, infinitism therefore fails at least when it comes to rules.

-When it comes to foundationalism, then I would probably use not believes about appearances but instead the appearances themselves as the foundation.
-I would tend to think that I don’t need representations like believes for something I have direct access to.
-But I think what worries me the most about foundationalism are the rules of inference.
-If there is no solid foundation for the truth tracking tendency of those rules then it seems like I can’t do much with a rock solid foundation in the form of appearances.

[jordanh]

Kane thank you so much for this video! I have long loved infinitism. Cheers

[MrGabrucho]

The position mistakes questions for justification which question on clarifying the justification. When you ask 'does f give you reason to believe B?' you're not throwing the discussion back to the chain of justifications(reasons), you are opening a new chain of questions, aimed at understanding the position.

[linus._.]

Well, Kripke's argument for the impossibility of analytical reduction of mental acts could be used here as an attempt, in ny view, to restore the possibility of actual synthetic a priori judgement. If we take, for instance, as Kripke puts, that what makes pain hurt is not some hidden inner nature that you may be unaware of, but rather the way you simply feel it (and you can't hallucinate about that, because even if there's no material respald for you pain but your mind makes it like it's real, it's real in a specific mental domain D), we can say: the sentence "I'm in pain" will be true, by necessity, once I'm in pain. Or, we could say: I'm in pain ---> I'm in pain (A ---> A). In this case, we can actually assert A in a specific time t and say that, the fact that "I'm in pain" will be true by necessity for a specific mental domain D, which is only known privately for myself; yet, although it'd be true by necessity (I can say it's necessarily true since I'm able to feel it and then assume that, once I feel it, I can't deny its existence anymore since the way I feel it is enough to tell it hurts, necessarily), it wouldn't be a proper analytic judgment, since, analytically speaking, it "could be false" ("I'm in pain" is clearly not a tautology, as there are possible words where it is clearly false). Which makes me believe this is a way to somehow restore the possibility for a synthetic a priori judgement. And such synthetic a priori judgement puts, as it seems, a final point to our inquiry (regarding the scope of sentences that could be derived from this primitive fact known synthetically, whereas still a priori) where it cannot get any more primitive than it is. Which can, then, be used as an argument against infinitism.

[Gurogun]

your channel is really interesting, glad I found it

[batkinson]

Regarding the very last point, about how we’re more confident in 1+1=2 than in cosmic inflation because of 1+1=2’s self-evidence rather than its shorter chain of reasoning, I’m not sure I agree.

Surely there are plenty of self-evident propositions that we can reasonably cite within the chain of reasoning for cosmic inflation, they’re just much further back in the chain. These propositions, together with the empirical claims cited in support inflation, should give us much more confidence in inflation than in 1+1=2. Why don’t they?

It’s because of the length of the chain. The self-evident propositions supporting inflation are so far removed from the central assertion about inflation itself that they’ve essentially lost any justifying power. This would seem to be the case with most chains of reasoning; at some point the reasons we’re citing no longer have anything to do with our main assertion, and so lose their strength as reasons. And if reasons lose their strength the further they are down the chain, how could we claim that a longer chain provides more doxastic justification than a shorter one?

[madra]

Excellent job. Would what Matt dillahaunty usually presents fall with coherence justification?
It seems also at some point the reasons given at a certain level of employment ( via scientific method) the 'guess' so to say Would be verified no?

[FreeWill_is_unintelligible]

Hi Kane! Are you planning to make a video on epistemic and rule circularity, by any chance?

[RemotHuman]

I think option #2 is probably how it actually works in our minds/brains. Or maybe option #1.
#2 is how we rationalize our beliefs logically, #1 is how we actually form the beliefs (through life experiences, which we just learn from without justification, probably because that helps with natural selection. So the bottom of the chain is natural selection maybe.

Or am I supposed to be questioning what I just said, no because I'm not trying to justify my belief I'm just positing it

[drey]

you had me until 18 min mark when you said infinitists aren't trying to be skeptical. The entire time I was assuming this is what this was about. Now I'm stumped as to why anyone would define themselves as an infinist if their goal is not pan-skepticism