Epistemology
Most recent answer: 10/22/2007
Q:
Eh-hem...w-wat tis de scientifico definition for de words, "fact" and "truth"? H-how do theis words compares? Also...w-what ill zet takes te make it a "fact" or "truth"...h-ow many or what kindos of evidencos?
- Anonymous
- Anonymous
A:
Why are you writing in such an odd dialect? Actual speakers of another
language may stutter a bit in speech, but not in writing. So I assume
you’re the same character who has recently written several questions in
this form. Please use normal language in the future.
As for the questions, scientists mean the same thing by ’truth’ as anybody else means. What’s to define?
As for ’facts’, we also don’t mean anything fundamentally different from what other people mean. However, because science has built up a coherent system of very general, highly confirmed rules and of reliable methods, we are often able to see facts by somewhat less direct methods than are available to individual people, wthout the advantages of cooperative study, advanced instruments, mathematical reasoning, etc.
For example, we consider it a fact that the universe is about 13.7 billion years old, although the collection of observations and calculations that go into that conclusion are not immediately accessible to any individual looking around on his own.
You probably have something more definite in mind, so why not just ask in plain language?
Mike W.
On the subject of "truth", I can talk about a narrow application of how the word can be used in statistical analysis of data, just to separate that idea from what our knowledge is.
Physical parameters generally have "true" values which experimentalists try to measure. Electrons have mass, and the true value of the amount of mass a single electron has is a number. We don’t know exactly what this number is, but we can try to measure it. A lot of measurements can be made, usually with random variations in the outcomes, as well as systematic biases. We try to think of all the ways our measurements can be biased and either account for them or try to design a better way of measuring something that doesn’t have so many problems. We can average a bunch of similar measurements together to reduce the uncertainty due to random fluctuations. In the end, when we run out of patience or money or both, we quit, and quote a measured value for the electron mass. Is it the "true" value? Not very believable. We assign an error estimate to our measurement and claim that the true value is somewhere close to the measured value, and it should be within the error bars (or maybe two, possibly three, assuming Gaussian statistics).
Our knowledge then is the measured value, which is our approximation of the truth until a better measurement comes out. We could have gotten lucky and actually reported the exact true value, but we’d never really know that for sure if we have any sources of uncertainty. Just about all kinds of knowledge have (or should have) detailed lists of the assumptions that go into the statements and uncertainties reported.
Sometimes measurements conflict with each other either because some assumptions are wrong, or a mistake has been made, or a statistical fluctuation has happened. Then we’re stuck (until we do some work to try to reconcile the answers). This is a good reason for reporting uncertainties along with measurements so that they can be compared to see if they are consistent or if they disagree. An old proverb says that a man with one watch always knows what time it is, but a man with two watches is never quite sure.
Tom
As for the questions, scientists mean the same thing by ’truth’ as anybody else means. What’s to define?
As for ’facts’, we also don’t mean anything fundamentally different from what other people mean. However, because science has built up a coherent system of very general, highly confirmed rules and of reliable methods, we are often able to see facts by somewhat less direct methods than are available to individual people, wthout the advantages of cooperative study, advanced instruments, mathematical reasoning, etc.
For example, we consider it a fact that the universe is about 13.7 billion years old, although the collection of observations and calculations that go into that conclusion are not immediately accessible to any individual looking around on his own.
You probably have something more definite in mind, so why not just ask in plain language?
Mike W.
On the subject of "truth", I can talk about a narrow application of how the word can be used in statistical analysis of data, just to separate that idea from what our knowledge is.
Physical parameters generally have "true" values which experimentalists try to measure. Electrons have mass, and the true value of the amount of mass a single electron has is a number. We don’t know exactly what this number is, but we can try to measure it. A lot of measurements can be made, usually with random variations in the outcomes, as well as systematic biases. We try to think of all the ways our measurements can be biased and either account for them or try to design a better way of measuring something that doesn’t have so many problems. We can average a bunch of similar measurements together to reduce the uncertainty due to random fluctuations. In the end, when we run out of patience or money or both, we quit, and quote a measured value for the electron mass. Is it the "true" value? Not very believable. We assign an error estimate to our measurement and claim that the true value is somewhere close to the measured value, and it should be within the error bars (or maybe two, possibly three, assuming Gaussian statistics).
Our knowledge then is the measured value, which is our approximation of the truth until a better measurement comes out. We could have gotten lucky and actually reported the exact true value, but we’d never really know that for sure if we have any sources of uncertainty. Just about all kinds of knowledge have (or should have) detailed lists of the assumptions that go into the statements and uncertainties reported.
Sometimes measurements conflict with each other either because some assumptions are wrong, or a mistake has been made, or a statistical fluctuation has happened. Then we’re stuck (until we do some work to try to reconcile the answers). This is a good reason for reporting uncertainties along with measurements so that they can be compared to see if they are consistent or if they disagree. An old proverb says that a man with one watch always knows what time it is, but a man with two watches is never quite sure.
Tom
(published on 10/22/2007)