3 ms·

Repeatable? Sorry in advance for quibbling, but -- suppose that a star goes nova, and you weren't expecting it, but you happen to have telescopes gathering inte

by pjscott 10y ago

Repeatable? Sorry in advance for quibbling, but -- suppose that a star goes nova, and you weren't expecting it, but you happen to have telescopes gathering interesting data from this once-in-a-stellar-lifetime event. If you learn interesting things from it, is that science? I'd say yes, but this isn't repeatable. It's also not really an experiment. It's wonderful evidence, though, really priceless stuff that you couldn't get any other way. What is the essence of science? I'm not convinced that it involves repeatability or experimentation in any really fundamental way.

vacri 10y ago
It's science, but not scientific method. The two are related but different. It's an observation that is rigorously recorded, but not an experiment. Science is full of this stuff - look at the field of taxonomy, for example. The idea that only those concepts which have been experimentally tested against the null hypothesis can be labelled 'science' goes against both reason and the spirit of the term.
Avshalom 10y ago
Yeah I've opined about this a couple times here. But yeah I don't like "experiments" either. I like data, I like observation; experiments are just a way to control the rate we gather those observations, gather that data, instead of (and I did astronomy as an undergrad so I am particularly sympathetic to this field of examples) hoping a star explodes while you're looking at it. Gamma Ray Bursts are a great example of things we spent decades pretty much just hoping we'd be looking at when they happened. And there's weird one offs that we're basically just guessing on. But we build better instruments and make better observations and gather more data.
snowwrestler 10y ago
Yes, repeatable. The specific event of that nova will not be repeated, of course, but the interesting observations, when compared with existing theory, will suggest repeatable experiments. And that's where the science really gets done. Not in the initial observation, but in the controlled experiments that follow, in which individual hypotheses can be tested. Experiments do not need to be run in a lab, BTW. As long as the variables are accounted for, natural observations are just as valid as those that are synthesized in a lab. One of the most famous "experiments" in physics history was the observation of gravity-deflected star images during an eclipse, which helped support Einstein's general theory of relativity.

by pjscott 10y ago

vacri 10y ago

It's science, but not scientific method. The two are related but different. It's an observation that is rigorously recorded, but not an experiment. Science is full of this stuff - look at the field of taxonomy, for example. The idea that only those concepts which have been experimentally tested against the null hypothesis can be labelled 'science' goes against both reason and the spirit of the term.

Avshalom 10y ago

Yeah I've opined about this a couple times here. But yeah I don't like "experiments" either. I like data, I like observation; experiments are just a way to control the rate we gather those observations, gather that data, instead of (and I did astronomy as an undergrad so I am particularly sympathetic to this field of examples) hoping a star explodes while you're looking at it. Gamma Ray Bursts are a great example of things we spent decades pretty much just hoping we'd be looking at when they happened. And there's weird one offs that we're basically just guessing on. But we build better instruments and make better observations and gather more data.

snowwrestler 10y ago

Yes, repeatable. The specific event of that nova will not be repeated, of course, but the interesting observations, when compared with existing theory, will suggest repeatable experiments. And that's where the science really gets done. Not in the initial observation, but in the controlled experiments that follow, in which individual hypotheses can be tested. Experiments do not need to be run in a lab, BTW. As long as the variables are accounted for, natural observations are just as valid as those that are synthesized in a lab. One of the most famous "experiments" in physics history was the observation of gravity-deflected star images during an eclipse, which helped support Einstein's general theory of relativity.