3 ms·

I would then ask: "what is science?", "who defines it?", "who is science?" Readiing the last article you linked, it seems like a legitimate debate about the sc

I would then ask: "what is science?", "who defines it?", "who is science?" Readiing the last article you linked, it seems like a legitimate debate about the scientific method and what constitutes scientific theories. Every scientists (but especially "science fans") need to acknowledge the fact that all science rests on certain axioms (partially laid out in the scientific method) that can't be proven by science. So going with the newest developments in the philosophy of science, the people developing science itself, seems only reasonable. What you call "getting rid of evidence" is only the exploration of other types of evidence aside from classical empirical evidence

whatshisface 9y ago
Up until this weird new thing where people want to dream up models and not worry about empiricism, "science" has always been characterized by the fact that you can strip out all the dependence on philosophy and reduce it to statements about procedures and algorithms. Every metaphysics has to reckon with our ability to act in the world, but they don't all have to include particular statements about the ontology of (say) numbers. As a result, statements about similarities between processes (for example the way counting in your head, a neurochemical process, is similar to putting stones in a bucket, a mechanical process) can be "independent" of the underlying ontological model. After all, whatever "similarity," means on the inside, those processes are "similar." In a sense, it's similar to how mathematicians write proofs in natural-ish language and call them true under the idea that they could convert them to purely formal symbolic manipulations if they wanted to put in the effort. Scientists talk about things as if they had all kinds of ontological statuses that they can't be shown to have - (ever heard a chemist describe electrons as "wanting" to occupy the lowest energy state?) - but that's nothing but language. The only reason any of it works is because you could cut out all of the weird stuff if you really wanted to.
- roenxi 9y ago
  I can't quite tell if I agree with you; so I shall attach this opinion. Science is characterised by being able to strip everything out of it down to a procedure to gather data and some guesses at how future data will look. Taking gravity as the classic example, in theory any person with any set of biases can put together a table with column A being mass and column B being distance to the ground and column C being time to fall the distance. From that, they could reconstruct Newton's Laws of Gravity if they measure everything carefully enough and are honest. Maybe the theory they come up with is F = g m / r^2 and the mechanism is tiny goblins. Cool. They can call that science as long as the goblins will always follow the rules. But the rules of evidence are very important, we can't claim the goblins are going to do something differently sometimes if we haven't observed it in the data. The data is the science. The formula and mechanisms are useful, but they are inevitably wrong in detail. Gravity isn't ever going to be the only effect measured in a data set, so no data set can ever perfectly confirm a theory correct, and certainly cannot confirm a mechanism. We don't need to rely on the scientists getting the mechanisms right. But we do need science to have a very high standard of measurement. It would help if they hold themselves back from coming up with crazy theories, but I don't know what a crazy theory looks like so I'm not going to try and police that today.
  - whatshisface 9y ago
    The models are also the science, because they come packaged along with statements about how well they fit the data. It's the "mechanism for the sake of having a mechanism" stuff that usually ends up as egg-on-face. Strings sit in the same group as "flux-as-actual-fluid," Boltzmann's atoms, and caloric fluid: intellectually interesting guesses that remain guesses until the empirical side gets advanced enough. The fluids lost, atoms won, and we're waiting on strings.

whatshisface 9y ago

Up until this weird new thing where people want to dream up models and not worry about empiricism, "science" has always been characterized by the fact that you can strip out all the dependence on philosophy and reduce it to statements about procedures and algorithms. Every metaphysics has to reckon with our ability to act in the world, but they don't all have to include particular statements about the ontology of (say) numbers. As a result, statements about similarities between processes (for example the way counting in your head, a neurochemical process, is similar to putting stones in a bucket, a mechanical process) can be "independent" of the underlying ontological model. After all, whatever "similarity," means on the inside, those processes are "similar." In a sense, it's similar to how mathematicians write proofs in natural-ish language and call them true under the idea that they could convert them to purely formal symbolic manipulations if they wanted to put in the effort. Scientists talk about things as if they had all kinds of ontological statuses that they can't be shown to have - (ever heard a chemist describe electrons as "wanting" to occupy the lowest energy state?) - but that's nothing but language. The only reason any of it works is because you could cut out all of the weird stuff if you really wanted to.

roenxi 9y ago
I can't quite tell if I agree with you; so I shall attach this opinion. Science is characterised by being able to strip everything out of it down to a procedure to gather data and some guesses at how future data will look. Taking gravity as the classic example, in theory any person with any set of biases can put together a table with column A being mass and column B being distance to the ground and column C being time to fall the distance. From that, they could reconstruct Newton's Laws of Gravity if they measure everything carefully enough and are honest. Maybe the theory they come up with is F = g m / r^2 and the mechanism is tiny goblins. Cool. They can call that science as long as the goblins will always follow the rules. But the rules of evidence are very important, we can't claim the goblins are going to do something differently sometimes if we haven't observed it in the data. The data is the science. The formula and mechanisms are useful, but they are inevitably wrong in detail. Gravity isn't ever going to be the only effect measured in a data set, so no data set can ever perfectly confirm a theory correct, and certainly cannot confirm a mechanism. We don't need to rely on the scientists getting the mechanisms right. But we do need science to have a very high standard of measurement. It would help if they hold themselves back from coming up with crazy theories, but I don't know what a crazy theory looks like so I'm not going to try and police that today.
- whatshisface 9y ago
  The models are also the science, because they come packaged along with statements about how well they fit the data. It's the "mechanism for the sake of having a mechanism" stuff that usually ends up as egg-on-face. Strings sit in the same group as "flux-as-actual-fluid," Boltzmann's atoms, and caloric fluid: intellectually interesting guesses that remain guesses until the empirical side gets advanced enough. The fluids lost, atoms won, and we're waiting on strings.