Hacker News new | past | comments | ask | show | jobs | submit login

> "That's not an understanding problem. That's lack of experimental evidence."

I find that phrase very interesting. I don't wish to single you out, as the underlying sentiment is something recognisable, but I would suggest you can't have scientific understanding without experimental evidence.

The scientific method relies on both definition of a hypothesis and exploration of that hypothesis through experimentation. If you only have a series of untested hypotheses then at best they're a bunch of educated guesses. Verification through experimentation is at the core of what elevates science beyond idle conjecture.




You have summarized logical positivism. It holds for the physical sciences where, say, dropping a stone from a certain height has repeatability. Social sciences involve people, and sometimes even the same person responds differently to similar stimuli.


It holds for both, the difference is in what is classed as a successful result. With the physical sciences you expect the outcome of your experiment to be something repeatable, as it implies you've understood the variables well enough to do so. On the other hand, with social science we accept that our knowledge of the variables is too limited to expect repeatability, so our experiments rely on statistical correlation in order to show whether a hypothesis was worthy of further exploration.

In both cases, you need experimentation to back up your claims. Social science without experimental results is just collective dogma.


"With the physical sciences you expect the outcome of your experiment to be something repeatable." Is it? Look at statistical mechanics. Repeatable, kind of, but on statistical basis.


You're splitting hairs. A single replication is not enough, I never suggested otherwise.


An experiment is just collecting data and analyzing it. Sure, you try to control for variables, but in some cases that's best done through statistical methods. The concept of an experiment as something you have to do, ie "pour two substances together into a beaker and see what happens," is just a way to collect data. Space probes have experiments for example, even though they're doing nothing but collecting data. The important part is whether your hypothesis explains the data from your experiment.


there was science before empiricism. aristotle wasn't very much in touch in reality, but i think he believed he did science


Aristotle was notably precisely because he was an empiricist. He was the first philosopher to actually observe things (mostly plants and animals). This was a departure from Platonic rationalism, which told you that knowledge of reality was knowledge of forms.




Guidelines | FAQ | Lists | API | Security | Legal | Apply to YC | Contact

Search: