As I said in my reply to Ying, radical doubt should be used rather than solved because as a tool it serves to remind ourselves to check and recheck our beliefs and as a problem it's unsolvable. — TheMadFool

Yes, doubt pushes us forward to discover more. We are all explorers.

Do any of you have any idea what that/those truth(s) is/are? — TheMadFool

Well the cogito is a pretty good candidate for indubitable truth, but the trouble is it doesn't necessarily connect to any other truths, it leaves you in a position where you have to preface every claim with "it seems to me that ..." - but of course we want to know whether is, not just seems.

As others have said, the process of doubt is usually limited, it depends on prior acceptance of some truth(s). To doubt, you need some truths as a lever. If you extend doubt infinitely, then you automatically limit yourself to the cogito, in fact to strict solipsism, and you stay there so long as you're infinitely extending your doubt.

So that said, the "indubitable" truths that we work with on a day-to-day basis are mostly simple, perceptual level truths, truths that are based on the animal level of perception that your ancestors used to survive and pass on their genes with. That's not a guarantee of absolute certainty, and they're always subject to correction, but they're the closest things we have to absolute certainty in the empirical realm.

I think a lot of it (particularly with Descartes methodology) is a confusion between the processes of ordinary knowledge-gathering and the processes of mathematical/logical thinking.

In ordinary knowledge-gathering, you're actually making discoveries, you're learning something new about the world, new information. But this has to mean that empirical knowledge is always provisional.

In maths and logic, on the other hand, with a complex or difficult problem, you feel subjectively like you're making a discovery, like you would in the everyday process of knowledge-gathering, but what you're discovering was already implicit in the axioms, etc., that you start out with, so you're not in fact discovering new information, just unfolding what was implicit. (It may be new information to you, but that novelty is strictly subjective.)

What this means is that the standard of certainty in maths and logic cannot be applied to empirical knowledge, except in the sense that in empirical knowledge-gathering, you are creating an internally-consistent projection or model of how the world is, that you then test against eventuating reality. There's mathematical certainty, deductive certainty, within the model and the implications for testing that you can draw from it. But you can never be certain that the model you're using is the right model for the occasion.

And that's where empirical doubt is always possible, and it's basically the putting side-by-side of two internally consistent models. IOW, an anomaly crops up in experience, which means that there must be something wrong with the model you've been assuming to be true up till now; so then you figure out some other possible model for the world, and match your two models against each other, and filter the right one out on the basis of homely, perceptual level truths (measurements, meter readings, etc.) that you are less doubtful about.

Of course there are truths. — Michael Ossipoff

Can you name one truth that you're 100% sure of? — TheMadFool

I'll name two:

1. If all dogs are mammals, and all mammals are animals, then all dogs are animals.

2, A posting attributed to Mad Fool asked me, " Can you name one truth that you're 100% sure of?"

Michael Ossipoff

Well, radical doubt will question the certainty with which you assert truths. Are you sure it's not a demon manipulating your mind?

Sorry I flagged your comment by mistake. Some changes since I last visited the site. Sorry again.

Anyway, thanks for your reply. I learned a lot from it.

Well, radical doubt will question the certainty with which you assert truths. Are you sure it's not a demon manipulating your mind? — TheMadFool

Then you believe in demons. I don't share that belief of yours.

Aside from that, are you referring to the doubt, or the asserted certainty?

Asserted certainty:

I do claim that there are uncontroversial metaphysical statements.

...but I don't make as many assertions as you might think, because I make no claim that the complex system of inter-referring timeless abstract if-then facts about hypotheticals, that i refer to, is "real" or "existent". (...whatever those words would mean.)

Doubt:

How sure are you that words can accurately and completely describe all of Reality?

If you can't say for sure that that's true, then any claim about all of Reality is questionable.

In any case, I'm not interested in debating or convincing anyone about the limits of discussion or description. I'd rather just discuss metaphysics, which is agreed, by most, to be discussable.

Michael Ossipoff

If you can't say for sure that that's true, then any claim about all of Reality is questionable. — Michael Ossipoff

That's what I mean. Radical doubt is, must be, painful to philosophers because it undermines everything, from their axioms to their logic.

I should say my piece in favour of Descartes' method.

I think it's right to say that Descartes was looking for indubitable truths to serve as foundations from which he could infer anything else he was going to believe. He begins the search by considering his cognitions in general and, in particular, his sensory experiences. He tries to locate cognitions that are self-certifying. An explication of "self-certifying" is best found by illustration. Consider the practice of gazing into a crystal ball in order to contact the dead. The gazer comes to form various beliefs about what the dead are saying. It is possible to ask why we should take what the gazer says at all seriously. Why should we think that what he says is true? He will likely retreat back to the claim that his crystal ball allows him to contact the dead, but we will immediately wonder why we should take this at all seriously. Why should we think that his crystal ball is reliable about these things? If we cannot find any answer to that question, there is a sense in which we will fail to assure ourselves that what the gazer says is true. That is, there will be a "why" question which he simply cannot answer.

All of these same issues arise for any and every method of belief formation. For example, I might answer a why question about my belief that there is a laptop in front of me by saying that I can see one, but this will lead to the question of why I should take sense perception to be reliable. I read Descartes as noticing this fact, that someone who is genuinely curious about the truth of their beliefs in general will raise why questions about any and every such belief. He seeks a way out of it - to satisfy his curiosity about the truth of his beliefs. To that end, he looks for beliefs about which a further why question makes no sense. He then tries to build the rest of his philosophy on top of those beliefs. His criterion for foundations ends up being clarity and distinctness. He thinks that when something is, in some sense, clear and distinct, there is no further question it makes sense to ask about why it should be accepted. He also thought that propositions that were clear and distinct were absolutely certain, but we should note that whether or not clear and distinct propositions are such that no further why question makes sense about them, and whether or not they are absolutely certain are logically two completely different matters.

Many people think something like the following about Descartes. Descartes is lead into solipsism because he requires absolute certainty for his beliefs. If you don't require absolute certainty, then you can escape and believe everything that you normally do. Doubt is thus just a silly game of demanding certainty when it isn't needed. It is true that Descartes pursues absolute certainty in the Meditations. His foundations are absolutely certain, and he insists only on deductive inference which would preserve certainty. But it isn't true that just by abandoning that standard you can avoid the issues Descartes gets himself into. Suppose that we relax the standard of certainty to probability. We could then, in principle, allow merely probable foundations and probable inferences. We are still limited, however, to propositions such that no further why question makes sense. Philosophers who have relaxed the standard seem to me to have failed far worse than Descartes ever did. They either fail to recognize that intelligible "why" questions can still be raised about their foundations or they simply don't see the regress of why questions; they only see certainty, and dismiss Descartes for being obsessed with certainty.

From this perspective, consider Gurugeorge's Model Building ideas:

you are creating an internally-consistent projection or model of how the world is, that you then test against eventuating reality. There's mathematical certainty, deductive certainty, within the model and the implications for testing that you can draw from it. But you can never be certain that the model you're using is the right model for the occasion. — gurugeorge

How does a model get "tested against reality"?

an anomaly crops up in experience, which means that there must be something wrong with the model you've been assuming to be true up till now; so then you figure out some other possible model for the world, and match your two models against each other, and filter the right one out on the basis of homely, perceptual level truths (measurements, meter readings, etc.) that you are less doubtful about. — gurugeorge

These ideas obviously assume that sense perception is reliable, and it is painfully easy to ask, why think that sense perception is reliable? And Gurugeorge seemingly has no answer to this question. The model building idea strikes me as very similar to Falsificationism, and I am attracted to that doctrine, but it has to be propped up by the claim that sensory experiences yield beliefs about which no further why question can be raised, and (since deduction is needed to deduce that a model is false given certain experiences) the claim that the rules of deductive inference have that status too. There may be ways to make these claims stick, but to do so one would have to employ either Descartes' idea of clarity and distinctness, or something which does the same job of answering that final why question. But those who say that Descartes' was just obsessed with certainty (perhaps guru doesn't hold this), never give such accounts. This is why I think they fail even more than Descartes did.

Now, it might be that part of the model we are building is the assumption that sense perception and deduction are reliable, but if it is, what sense does it make to say that the model itself is tested by appeal to sense experience? To test a model which contains the theory that sense perception and deduction are reliable by appeal to sense perception and deduction is clearly question begging. Thus, these ideas about models always leave one with an unanswered "why" question about the key assumptions of the model. It isn't just that the assumptions aren't absolutely certain. Its worse than that. Its that no reason at all has been given to think the assumptions are true, not even a fallible or probable one.

Hence, construed as an attempt to solve the problem Descartes tried to solve, the model building ideas are a failure. If, however, the model building ideas aren't an attempted solution to that problem, I admit that I have no real idea how to understand them.

I should apologize to for picking on his post as an illustration. I wanted some way to make the Cartesian perspective clear by contrasting it with some other ideas. His post was the most well thought out and lucid of those that seemed appropriate for this purpose.

Best
PA

and it is painfully easy to ask, why think that sense perception is reliable? — PossibleAaran

Great post! All I'll say here is that this is the nub of it: the only reason one gets so much as the idea that there's an anomaly (that casts doubt on one's model) in the first place, is because one is already relying on sense perception as valid enough to tell you that there's an anomaly; therefore, unless one is given reason to suppose otherwise, the subsequent model-testing sense-perceptions can have at least the same level of reliability. (And the "reasons to suppose otherwise" usually involve some break in the conditions for normal perception, with the possibility of correcting perception being already priorly accepted.)

IOW, if you think there's an anomaly at all, that information comes to you from sense perception, therefore (unless there's good reason to suppose otherwise - that's where the burden of proof lies) whatever perceptions you use to test your revised conception of reality must be on exactly the same level of reliability as the perceptions that gave you the very idea that there was an anomaly in the first place. The more certain you are that there was an anomaly, the more certain you must be that sense perception is at least in principle reliable enough to resolve the anomaly.

Going back to the arguments from illusion (and the gradual progression of Cartesian doubts in the course of the Meditations): the only reason why you think you were subject to an illusion (or why Descartes noticed he'd been in error about various things) is because of some corrective perception that reveals that your previous perception was an illusion. Therefore you're already implicitly allowing the validity of at least some sense perceptions: the corrective sense perceptions at least must be valid, for the illusion to be genuinely an illusion. Therefore you can't use the argument from illusion to globally doubt the validity of sense perception on the basis that sometimes you're subject to illusion.

the only reason why you think you were subject to an illusion (or why Descartes noticed he'd been in error about various things) is because of some corrective perception that reveals that your previous perception was an illusion. Therefore you're already implicitly allowing the validity of at least some sense perceptions: the corrective sense perceptions at least must be valid, for the illusion to be genuinely an illusion. Therefore you can't use the argument from illusion to globally doubt the validity of sense perception on the basis that sometimes you're subject to illusion. — gurugeorge

Thanks for the reply guru. It seems we are slowly isolating the key disagreements between us. Here is one of them. You think that what Descartes does is something like this. He first notices that he has sometimes been mistaken. From this fact he extrapolates that he might be mistaken in any given instance, and that is why he suspends judgement prior to finding his proof of God's existence and such. If this were how he was thinking, then I'd say you would be right that to get this started, he has to take fore-granted that at least some of his sensory experiences are veridical. It seems right to me to say that he can't really do that unless he is already assuming the reliability of sense perception, and so that assumption would be a presupposition of the possibility of his doubting it. A remarkable transcendental argument against scepticism!

My reply would be that even if Descartes did proceed in the way you think, he need not have done so. Descartes need not assume that sometimes he is mistaken, in order to ask "why believe that sense perception is reliable?". He can simply ask it, can he not? Of course, he cannot sensibly ask it if he is also assuming that sense perception is never mistaken or if he is assuming that sense perception is reliable. But he need not assume that. He can simply raise, relentlessly, with respect to every opinion he has, "why believe that?", until he reaches some beliefs for which this question cannot meaningfully be posed.

But, given your initial remarks about "getting the idea that there is an anomaly" in one's model, perhaps you think that for some reason Descartes cannot sensibly raise this question about the reliability of sense perception. It would be great if that were so, but how could it be?

Best
PA

I'd said:

If you can't say for sure that that's true, then any claim about all of Reality is questionable.

You replied:

That's what I mean. Radical doubt is, must be, painful to philosophers because it undermines everything, from their axioms to their logic. — TheMadFool

But I was just referring to doubt about the controversial claim that all of Reality is discussable and describable.

I wouldn't say that logic is in doubt.

I claim that uncontroversial things can be said about metaphysics, and that much about what is, is unontroversially discussable, describable. That discussable, describable domain of what is (other than the physical sciences), I refer to as metaphysics.

But sure, all assumptions should be subject to question, and radical doubt, skepticism, is the right approach to philosophy.

For example, I suggest that people be skeptical about Materialism's big brute-fact.

Michael Ossipoff

But, given your initial remarks about "getting the idea that there is an anomaly" in one's model, perhaps you think that for some reason Descartes cannot sensibly raise this question about the reliability of sense perception. It would be great if that were so, but how could it be? — PossibleAaran

No, of course he can ask it. But the short answer to the question "Why believe that sense perception is reliable?" is because it is in fact reliable.

And you can show that sense perception is in fact reliable because you can distinguish reliable perceptions from unreliable ones, and show that we have more reliable ones than unreliable ones: which means that sense perception is reliable. It's not absolutely guaranteed to give a nugget of truth every time, but it sometimes does and sometimes doesn't, and the fact that it mostly does is the very meaning of "reliable."

Again, it's the two concepts together that apply to reality, not just the one. If perception always gave guaranteed nuggets of truth, we wouldn't have the concept reliable, because we wouldn't have the concept unreliable, there would be no contrast, so we wouldn't notice, so the question wouldn't even come up.

The longer answer would involve evolutionary biology, anatomy, neurobiology, etc. And it would also involve looking at perceptions in the context of desires and expectations (whether they're fulfilled or not) - perception (for animals and us) is a phase of action in service of desire, and under expectations set by models as aforesaid. This would be connected with pragmatism, but I think pragmatism goes too far in tying truth connections simply to fulfillments. We are talking about truth, which is to say we are ultimately talking about the models/projections of how reality is, and using perceptions to filter the model. But the fact that using our perceptions gets us what we want is definitely tied in with what we mean by the reliability of perception: if we were baulked at every turn by following our perceptions, and ended up starving in a ditch, we wouldn't think much of perception.

(Now in all the above, I can sense you champing at the bit: you no doubt want to say, "But aren't sense perceptions being used in the very process of checking out whether sense perceptions are reliable?" This seems to be homing in on our disagreement even more: somehow, you think this is circular. But why?)

Btw, I can't resist it: if Descartes relentlessly asks "why?" then ultimately he's in the position of the child relentlessly asking "why?" in the comedian Louis CK's skit, and his intelocutor is entitled to lose patience with him at some point: "WHY? Aw fuck you, eat your french fries you little shit, goddamit." :D

https://www.bing.com/videos/search?q=louis+ck+why&view=detail&mid=9D4912E6BB08967131D39D4912E6BB08967131D3&FORM=VIRE

Skepticism is a problem for philosophy because there is no absolute certainty in it. How does one overcome it? Do we fall back on pragmatism or do we just ignore it? — TheMadFool

We can start by clearly differentiating belief from truth. Doubt pairs with belief, not truth.

One way of viewing scepticism is as a failure of nerve. The sceptic does not have the courage to commit.

We can start by clearly differentiating belief from truth. Doubt pairs with belief, not truth. — Banno

Thank you for that. Sometimes this site is worth the price of admission.

But sure, all assumptions should be subject to question, and radical doubt, skepticism, is the right approach to philosophy. — Michael Ossipoff

Why?

But sure, all assumptions should be subject to question, and radical doubt, skepticism, is the right approach to philosophy. — Michael Ossipoff

Why? — Banno

Because there's no assurance that an assumption is right.

But it was just a comment, not an assertion, and it isn't something that I'd argue or debate.

I don't mean or want to tell others what to believe or what their attitude should be. So, if you don't think that your assumptions are subject to question, then I have no argument with, or criticism of, that.

Michael Ossipoff

Because there's no assurance that an assumption is right. — Michael Ossipoff

Nor an assurance that it is wrong.

So why is it rational to doubt without reason, yet not to believe without reason?

Because there's no assurance that an assumption is righ — Michael Ossipoff

Nor an assurance that it is wrong.

So why is it rational to doubt without reason, yet not to believe without reason? — Banno

Because when you admit to yourself that you don't have reason to believe something, then, by definition, you have doubt about it.

Michael Ossipoff

Because when you admit to yourself that you don't have reason to believe something, then, by definition, you have doubt about it. — Michael Ossipoff

Then when you admit to yourself that you don't have reason to doubt something, then, by definition, you have belief in it.

Then when you admit to yourself that you don't have reason to doubt something, then, by definition, you have belief in it. — Banno

Yes.

If you don't have reason to doubt it, that's because there's reason to believe it's true, and no reason to believe that it isn't true..

Or maybe you don't care about the proposition (whatever it might be), or have enough interest in it to know or care whether you doubt it or believe it.

But then you don't believe it, and, by definition you doubt it, even if you completely-disinterestedly doubt it.

Michael Ossipoff

Now in all the above, I can sense you champing at the bit: you no doubt want to say, "But aren't sense perceptions being used in the very process of checking out whether sense perceptions are reliable?" This seems to be homing in on our disagreement even more: somehow, you think this is circular. But why?) — gurugeorge

Excellent stuff. I was tempted by this thought, and I did write a response explaining why I thought the argument was circular, but I don't think that's right after all. Your argument never uses the premise "sense perception is reliable" and so it isn't logically circular. I'm not sure that the argument is precise enough for me to disagree at this stage.

And you can show that sense perception is in fact reliable because you can distinguish reliable perceptions from unreliable ones, and show that we have more reliable ones than unreliable ones: — gurugeorge

How, exactly, can you distinguish veridical sense perceptions from non-veridical ones? And how can you show that we have more of the former than the latter?

This is the crux of it. If you can do that, then you have an answer to those pesky "why" questions.

Btw, I can't resist it: if Descartes relentlessly asks "why?" then ultimately he's in the position of the child relentlessly asking "why?" in the comedian Louis CK's skit, and his intelocutor is entitled to lose patience with him at some point: "WHY? Aw fuck you, eat your french fries you little shit, goddamit." — gurugeorge

My old supervisor criticized my conception of skepticism for being "childish". I agree that there is a parallel between the child's constant questioning and the sceptical one. But I don't see why that makes the sceptical questioning objectionable. It isn't as though if children do P, then necessarily P isn't sensible.

So why is it rational to doubt without reason, yet not to believe without reason? — Banno

The word "rational" is vague. I think a philosophical tradition that starts in Socrates and is carried through by various people to Descartes and - in some places - to Russell, defines rational belief as belief for which one has good reasons to hold. I don't think that is the "ordinary" meaning if there even is such a thing as that - I doubt it. But it is a good thing to have reasons for the things you believe and so good to strive to have them wherever possible.

If you don't have reason to doubt it, that's because there's reason to believe it's true, and no reason to believe that it isn't true.. — Michael Ossipoff

The underlying assumption here is that belief requires justification. Why should it?

How, exactly, can you distinguish veridical sense perceptions from non-veridical ones? And how can you show that we have more of the former than the latter? This is the crux of it. If you can do that, then you have an answer to those pesky "why" questions. — PossibleAaran

Count them. Seriously, just count them. Think of all the times when you've proceeded as if your sense perceptions have been correct, and your desires and expectations have been fulfilled by proceeding on the assumption that they're correct, versus the times you misperceived. You get up in the morning, you see what looks like a toothbrush, you pick it up and find you can brush your teeth with it. You reach for what looks like a door handle and find you can use it to open the door and get out of the house. You go to the train station, you step into what looks like a train and you find it's taken you to what looks like your place of work, which look like it has your workstation, where indeed the work is as you remember leaving it, etc., etc., etc. Maybe on the way home you encounter a situation like this:-

"I thought I saw a banker's clerk descending from a bus,
I looked again and saw it was a hippopotamus."

So there you have a whole slew of desires and expectations fulfilled by taking sense perceptions as veridical, and you have one misperception, one expectation baulked. The ratio I'd say is par for the course for the average day.

What else are you to conclude other than that sense perceptions are reliable? What else would reliability consist in, other than ... this sort of thing?

Is it logically possible that any segment of that sequence, or the whole sequence, might be systematically mistaken? Sure, but give a reason for it - until there's a reason to take seriously the idea that there's been systematic, thoroughgoing erroneous perception, then the hypothesis of systematic error is (as I've insisted elsewhere) mere idle imagination.

But even then, you'll always be juxtaposing correct perceptions against erroneous perceptions, even then correct perceptions still have to be possible in order to demonstrate that the whole sequence of perceptions above was erroneous. And that's because perception does that job: that's the burden of empiricism. Perceptions are truth-makers for propositions, that's the place they have in the economy of thought, and we have no other thing to take their place, certainly not schmerception (which is the truncated, presuppositionless way of looking at perception).

My old supervisor criticized my conception of skepticism for being "childish". I agree that there is a parallel between the child's constant questioning and the sceptical one. But I don't see why that makes the sceptical questioning objectionable. It isn't as though if children do P, then necessarily P isn't sensible. — PossibleAaran

It's really more that the sceptic or the endless why-questioner isn't quite getting the game. "Why" questions have a limited ambit, always, they're delimited in a given universe of discourse, against a background in which some things are accepted as true. The extrapolation and extension is basically just continually moving the goalposts.

But we have to be careful here, because sometimes (e.g. a careful detective or journalist, or indeed a scientist or philosopher) pursuing questions a layer or two deeper than the original will discover something useful or interesting. (In this connection, see this wonderful Richard Feynman clip.) But that - knowing when to pursue a "why" question and when to drop it - is what makes inquiry partly an art and a game as well as a science - and partly a matter of judgement arising from long experience with particular fields.

Count them. Seriously, just count them. Think of all the times when you've proceeded as if your sense perceptions have been correct, and your desires and expectations have been fulfilled by proceeding on the assumption that they're correct, versus the times you misperceived. You get up in the morning, you see what looks like a toothbrush, you pick it up and find you can brush your teeth with it. You reach for what looks like a door handle and find you can use it to open the door and get out of the house. You go to the train station, you step into what looks like a train and you find it's taken you to what looks like your place of work, which look like it has your workstation, where indeed the work is as you remember leaving it, etc., etc., etc. Maybe on the way home you encounter a situation like this:-

"I thought I saw a banker's clerk descending from a bus,
I looked again and saw it was a hippopotamus."

So there you have a whole slew of desires and expectations fulfilled by taking sense perceptions as veridical, and you have one misperception, one expectation baulked. The ratio I'd say is par for the course for the average day. — gurugeorge

This wasn't quite what I wanted. I understand that you think that one can use a track-record argument for the claim that sense perception is reliable. Sense perception got things right on occasions X, Y, Z, N, N+1... therefore sense perception is reliable. My question is, why believe, in any particular case, that sense perception got it right? I look into my bathroom and form the belief that there is a toothbrush on the sink. Why should I believe that there is? Remember, at this point we haven't established that sense perception is reliable, so we cannot appeal to that. Why, then, should I take it that sense perception is getting things right in this particular instance if I can't take it to be reliable yet? If the track record argument works, there must be some reason to believe its premises.

It's really more that the sceptic or the endless why-questioner isn't quite getting the game. "Why" questions have a limited ambit, always, they're delimited in a given universe of discourse, against a background in which some things are accepted as true. The extrapolation and extension is basically just continually moving the goalposts. — gurugeorge

I think that most sceptics knew perfectly well that they weren't playing the usual 'game' that is played in ordinary life. I don't think that Descartes was foolish enough to think that in ordinary life we pursue why questions all the way through. He even points this out himself in the Meditations. He saw that ordinary discourse involves taking many things for-granted. But his philosophy wasn't a description of discourse in London. It was an attempt to answer all of those why questions that aren't ordinarily answered. In doing this he recognizes that he's pursuing matters much further than they are usually pursued, but he has goals which he thinks are best achieved by doing this.

This wasn't quite what I wanted. — PossibleAaran

Yeah, I know it's kind of a boring answer. :)

I understand that you think that one can use a track-record argument for the claim that sense perception is reliable. Sense perception got things right on occasions X, Y, Z, N, N+1... therefore sense perception is reliable. My question is, why believe, in any particular case, that sense perception got it right? — PossibleAaran

There's no reason to believe it in any particular case, not by "reading off" from the perception (or even schmerception) in isolation. The perception's validity isn't given alongside the givenness of the perception.

But that's not a problem, because the reason to believe (trust in) the validity of any random given instance of perception comes from trust in the general series, which includes also the possiblity of occasional error. It seems kind of paradoxical, but it really isn't. We aren't guaranteed the validity of any given perception, taken in isolation, but we have a reasonable degree of proven confidence in the series as a whole, which means that the given perception is likely to be valid, but may occasionally not be.

It would be a problem if we were in the position of having to "read off" any given perception's validity from the perception itself in order to have knowledge. But we aren't, so it isn't.

I look into my bathroom and form the belief that there is a toothbrush on the sink. Why should I believe that there is? Remember, at this point we haven't established that sense perception is reliable, so we cannot appeal to that. — PossibleAaran

You should believe it's a toothbrush because you're having a toothbrush-like perception, that's good enough reason to TRUST that you'll be able to brush your teeth with the damn thing.

Now that trust may turn out to have been misplaced (maybe it's an alien spaceship), but we know it's highly unlikely to have been misplaced (the chances of it being an alien spaceship, or any of an infinity of other logically possible things, including your hallucination as a brain in a vat, or whatever, is vanishingly tiny, and would require strong proof to counterbalance the chain of expectations-fulfilled that we've gotten by going along with perception as generally valid).

Why, then, should I take it that sense perception is getting things right in this particular instance if I can't take it to be reliable yet? If the track record argument works, there must be some reason to believe its premises. — PossibleAaran

You can't take any given perception in isolation to be reliable, its reliability is something that can only be validated subsequently as part of a linked network of expectations fulfilled (so long as you're using perceptions as a guide to reality, in order to fulfil those expectations).

But the expectations' fulfillments are themselves things perceived, after all, so what else do you have to go on?

IOW, it's not a bug, but a feature. ("It" here being the lack of one-to-one guaranteed intrinsic relationship between a perception and its validity, that's somehow given along with givenness of the perception itself.)

In doing this [Descartes] recognizes that he's pursuing matters much further than they are usually pursued, but he has goals which he thinks are best achieved by doing this. — PossibleAaran

Yes, and that's been a valuable exercise, but the fact that it hasn't led anywhere is what's instructive.