You are viewing the historical archive of The Philosophy Forum.
For current discussions, visit the live forum.
Go to live forum

Some science will just never be correct

Georgios Bakalis March 26, 2021 at 19:21 7900 views 29 comments
I’ll admit that this title is a bit ‘click-bait’ if you like, because it is just a question rather than a claim, however I will make a claim, but just to see the explanation of others. Note: I am not claiming to have come up with this.

1. A lot of (if not all) Science is based on drawing conclusions from patterns
2. To be certain that a pattern is always reliable (that there are no anomalies) you have to test something infinitely
3. We do not test things infinitely (in fact we cannot)
4. Therefore, any science based on drawing a conclusion from a pattern is not reliable


Furthermore, when scientists test something, they often reproduce the test at most a couple hundred times, which is, in the larger scheme of things, not a huge amount.

However, one part of me is unsure. Most of the science based on spotting patterns is often in chemistry and biology where we have seen large medical advances.
My question is for someone to spot the mistake in the above.

Comments (29)

Deleted User March 26, 2021 at 19:25 #515034
This user has been deleted and all their posts removed.
Georgios Bakalis March 26, 2021 at 19:27 #515037
Reply to tim wood By that, I assume you are suggesting I explain what I mean by ‘reliable’. If so, I will try. We cannot draw a conclusion from a flawed system, so therefore it is un reliable.
javi2541997 March 26, 2021 at 19:30 #515038
Quoting Georgios Bakalis
1. A lot of (if not all) Science is based on drawing conclusions from patterns
2. To be certain that a pattern is always reliable (that there are no anomalies) you have to test something infinitely
3. We do not test things infinitely (in fact we cannot)
4. Therefore, any science based on drawing a conclusion from a pattern is not reliable


Flawed criteria... Why we have to taste everything infinitely? I guess if science and their theories are relatable and important is because they literally work despite they are not being proved indefinitely
We all agree that if we throw an object through window it falls down the floor due to gravity theory.
Should we prove this infinitely?
Georgios Bakalis March 26, 2021 at 19:39 #515043
Reply to javi2541997 I partially agree. I will try to explain why I think we need to test things infinitely to be indefinitely flawed.
Let’s say you wanted to prove that you cannot compress a solid. I suppose you might go around with different solids and compress them (at the same force) and record if you can compress them. The first 200 times, (I doubt) any solids are going to compress, but maybe on the 201st time you will. You just couldn’t know, because it is a pattern. Of course, the more and more you get the same result, the more likely it is that the next result will be the same, however you cannot prove this until you do it.
Essentially, to prove that you will always get the same result, you have to actually carry out the test.

I agree that we do not need to prove things indefinitely but still.
Deleted User March 26, 2021 at 19:39 #515044
This user has been deleted and all their posts removed.
Georgios Bakalis March 26, 2021 at 19:41 #515046
Reply to tim wood Fair enough, I do assume I should not have used the word ‘reliable’ twice in different contexts.
Huh March 26, 2021 at 19:50 #515050
What do you mean?

javi2541997 March 26, 2021 at 19:56 #515055
Quoting Georgios Bakalis
Let’s say you wanted to prove that you cannot compress a solid. I suppose you might go around with different solids and compress them (at the same force) and record if you can compress them. The first 200 times, (I doubt) any solids are going to compress, but maybe on the 201st


I understand your point here. It is interesting what are you saying but it looks like you are defending we are forced to search absolute perfection. This could be dangerous. Just because one is not answering to the premises doesn’t make the rest invalid. This exactly happens, for example, in covid vaccine.
So, just because 2 or 3 % of the patients are suffering issues do we have stop providing others until we reach a perfect 0 % in issues? I guess this is impossible.
Tim3003 March 26, 2021 at 21:02 #515106
If you believe we need to test a pattern an infinite number of times to prove it is reliable you don't believe in science at all! Scientific advance is not the result of an infinite coincidence of the same results from an action, it is the acceptance that the results of a specifically designed test prove a previously stated theory.

As an example. If I toss an unbiassed coin enough times, the longer I go on, the closer the results will come to exactly 50% heads and 50% tails. By your criteria, since I probably won't ever reach an exact 50/50 result - and even if I do, the next toss will break it; then I cannot say the chances of heads vs tails are 50/50.
The scientist presupposes that the chances are exactly 50/50, and over time the test result tends ever more closely to that. Statisiticans use confidence limits to rate how sure they are that an achieved appearance of a result is trustworthy. You can never be 100% confident - as you say, but to progress you have to say "99.99% is close enough", or something like that. There must come a stage where you decide you have ruled out any possible error in your test.. That's why at the LHC they have 2 totally separate teams analysing the results of every experiement.
counterpunch March 26, 2021 at 21:11 #515117
Reply to Georgios Bakalis

Nice computer Georgios! Science is true because it works. My question would be how do you not recognise science as truth? And why do you want so desperately that science is not true? I consider science "true"- even accepting endless epistemic philosophical complications, it's an increasingly valid and coherent understanding of reality that's really come into focus in the computer age; knowledge it behoves us to pay attention to - and in my opinion, is our best bet for any kind of sustainable future.
Banno March 26, 2021 at 21:14 #515120
Reply to Georgios Bakalis You've spotted the problem of induction.

There's a lot of background here.
DingoJones March 26, 2021 at 21:29 #515134
Reply to Georgios Bakalis

Science isn’t about patterns it’s about repeatability, at least in the context you’re using. That’s an important distinction to your argument. Once you swap out “patterns” for “repeatability” your question doesn’t need to be asked anymore.
It would preclude the part of your argument that depends on 100% certainty as well, as science is concerned only with repeatability not certainty.
Banno March 26, 2021 at 21:34 #515138
Reply to DingoJones What are patterns if not repetitions?
DingoJones March 26, 2021 at 21:48 #515145
Reply to Banno

A complex series of variable repetitions. The way the thread is not the shirt.
SophistiCat March 26, 2021 at 21:50 #515146
Quoting Georgios Bakalis
My question is for someone to spot the mistake in the above.


There is no mistake in what you said (other than some questionable choice of words). The mistake is in your misplaced expectation of certainty. Science and empirical investigation in general provides plausible beliefs, not certain knowledge.
Joshs March 26, 2021 at 22:30 #515160
Reply to Georgios Bakalis “Some science will just never be correct.”

Ah, the ascetic ideal rears its ugly head.
Banno March 26, 2021 at 22:50 #515171
Reply to DingoJones Is the pattern the thread or the shirt?
Streetlight March 26, 2021 at 23:13 #515183
Yes. This is why anything scientific is falsifiable. If it cannot, in principle, be proven wrong by further observation, that's not science, that's religion. The revisibility of science is a feature, not a bug.
simeonz March 26, 2021 at 23:17 #515185
Reply to Georgios Bakalis
As @SophistCat mentioned, you could replace conclusions with beliefs, but that may not have changed much. Knowledge, I consider, is just a word for true belief, and we cannot verify which belief is true. That is, in practice, all strong beliefs are considered knowledge by the bearer until proven otherwise. It is meta-scientific distinction.

You may allude to the idea that we test things indefinitely. That is, at least methodologically, science stays open to novel experience, continuously tests and rectifies itself when faced by contradiction. And as methodology, I agree that this is the best approach there could be. But there are two problems. First, those new amendments make the laws accurate retroactively, but does not prove that they are now predictive. And we are not just seeking retrospectively accurate models. Also, since most of the science is probabilistic, even if not in theory like thermodynamics and QM, it would be as a measurement practice, it cannot actually distinguish confirmation from refutation. Therefore, discovery of confirmation or contradiction is substituted for discovery of likely and unlikely occurrences. It seems to come down to belief, or instinct, or naturally coerced resolution, if you will. Science is a state of being scientific.
DingoJones March 26, 2021 at 23:19 #515186
Reply to Banno

I would say the shirt.
Banno March 26, 2021 at 23:22 #515187
Reply to DingoJones Nup. A 'complex series of variable repetitions' is a pattern.
Joshs March 26, 2021 at 23:23 #515188
Reply to StreetlightXquote="StreetlightX;515183"]If it cannot, in principle, be proven wrong by further observation, that's not science, that's religion.[/quote]

Nietzsche would beg to differ. Even Popper’s neo-Kantian inspired ‘falsifiability’ , to the extent that it maintains a role for truth , would be considered by him to be an expression of the religiosity of the ascetic ideal.
Streetlight March 26, 2021 at 23:26 #515192
Reply to Joshs Yes but this is why I don't get my science from Nietzsche.
Joshs March 26, 2021 at 23:30 #515194
Reply to StreetlightX Wouldn’t Deleuze side with Nietzsche on science? I thought you agreed with his philosophy.
Pop March 26, 2021 at 23:32 #515196
Quoting Georgios Bakalis
Therefore, any science based on drawing a conclusion from a pattern is not reliable


If this was true then nothing would be reliable, as everything reduces down to a drawing of conclusions from patterns (of information).

It is the logical flaw in your assertion.

** consciousness is a drawing of conclusions from patterns of information ( information integration ). If science is flawed on the basis of doing this then so is consciousness. By extension so is your assertion, as it is a drawing of conclusions from patterns of information.
Streetlight March 26, 2021 at 23:33 #515198
Reply to Joshs I agree or disagree with claims, not people. And gosh I love Nietzsche but he can be full of shit like anyone else.
Joshs March 27, 2021 at 00:03 #515207
Reply to StreetlightX I understand that, but I’m asking because I was impressed with Deleuze’s book on Nietzsche and am curious as to whether you think Deleuze might have departed from Nietzsche’s view of science.
Streetlight March 27, 2021 at 00:10 #515213
Reply to Joshs It's not really relevant to the OP, and the question in any case is too broad. If you show your work, I might do the same, but not on this thread.
Joshs March 27, 2021 at 03:05 #515270
Reply to StreetlightX I don’t have any work to show, Just wanted your quick take on the subject, which doesn’t have to be done on this thread.