My new YouTube video about atheism

[FlatAssembler]

But can't you see this fundamental limitation?

I am not sure what you mean.

Even in maths?

Well, at least, in mathematics, you don't have to deal with unknown intial variables. You can prove he four-color theorem by brute-forcing, something like that is probably impossible in any other science.

Even in logic?

Again, I am not sure what you mean.

Axioms?

Again, I am not sure what you mean.

That doesn't make the research less scientific.

But it does. Scientific methods that are a normal thing in natural sciences are rarely applied in social sciences. In physics, any experimental result has a p-value, usually on the order of magnitude of one over a million. In linguistics, p-value is rarely calculated. And when it is, it's usually on the order of one over hundred or so.

There are still controlled double-blind trials involved in some of these fields, along with objective measures constructed to measure variables of interest (and even the means to establish their reliability and validity).

Sure, but such are very rare.

And some of these trials, believe it or not, are true experiments.

I am not saying it's never possible to make systematic observations or controlled experiments in social sciences. I am saying it's usually not possible. And when it is possible, it's quite often not done that way. In linguistics, for example, it was common knowledge for almost a century that vowel quality was determined primarily by the shape of the tongue, when there was never experimental evidence to support that, and experimental results suggesting otherwise began to appear as early as the 1920s.

they don't realize a lot of effort is often put to ensure they work as intended

Often it is. Usually, it isn't. Polls predict the results of the elections no better than guessing. And, when there are more rigorous forms of science, the results of surveys mostly prove to be wrong. Surveys found a huge correlation between brain tumors and cellphone usage, rigorous studies find no correlation whatsoever. Surveys found a significant negative correlation between vitamin E intake and lung cancer, rigorous studies found that vitamin E can even cause lung cancer. Surveys generally show calcium decreases the risk of heart diseases, rigorous studies tend to show it either makes no difference or actually increases the risk of heart disease.

Plus, given enough time, technology, and further research, there will hopefully be more certainty regarding various aspects of human nature, both individually and collectively, presently and historically.

Social sciences generally rely on dumb luck for the hypotheses to be proven or disproven. A big difference between the De Saussure's prediction of the Indo-European laryngeals and the prediction of background cosmic radiation is that astronomy was able to guarantee that, given the right kind of technology, that hypothesis can be proven or disproven. Linguistics couldn't guarantee that, in some possible world very close to ours no Anatolian language was attested, or none of their attestations survived.