devilsadvocate
New Member
Reading the discussion about what is the difference between a law and a fact in the "Evolution by natural selection" -thread got me thinking about the problem laws of nature poses for empiricism again. First, here's some definitions that we hopefully agree on:
Laws of nature are,
- Not analytic, ie. Not true by virtue of definition, but derived a posteriori from observation.
- Contingent, meaning they could be different from what they are
- (at least usually) presented in the form of universal conditional, "All A's are B's", "All copper conducts electricity". They are general, not particular like facts are.
- Need to be true to be called "laws". In empiricist terms this means that all observations we make do not contradict the law.
So far this has been very empiricist-friendly definition, but I'm fairly sure we need to add at least one more definition in order to differentiate laws of nature from accidental generalizations. The proposition "All the cars I own, or will ever own, will be blue" seems, quite reasonably, not to be a law of nature, although it can fulfill all the definitions above. The definition is therefore not sufficient. We need to add at least one more condition,
- Laws of nature necessitate behaviour of the stuff the law governs.
This added definition demarcates generalizations from actual laws of nature nicely. It is no longer true that "if I own a car, it will be blue" is a law of nature, because there is nothing that necessitates this to be the case. It might turn out to be true, but only by accident, unlike observations we make about stuff that is governed by laws of nature.
However, the definition poses serious problems for empiricism. To claim that there are fundamentally unobservable - Remember, empiricists only claim to look at strings of particular facts. Nobody ever claims to have direct experience of a law itself - governing modes of reality, that not only happen to be true, but necessitate how objects must behave, seems like a huge dip in metaphysics empiricism tries so hard to avoid.
Any solutions to the problem?
Laws of nature are,
- Not analytic, ie. Not true by virtue of definition, but derived a posteriori from observation.
- Contingent, meaning they could be different from what they are
- (at least usually) presented in the form of universal conditional, "All A's are B's", "All copper conducts electricity". They are general, not particular like facts are.
- Need to be true to be called "laws". In empiricist terms this means that all observations we make do not contradict the law.
So far this has been very empiricist-friendly definition, but I'm fairly sure we need to add at least one more definition in order to differentiate laws of nature from accidental generalizations. The proposition "All the cars I own, or will ever own, will be blue" seems, quite reasonably, not to be a law of nature, although it can fulfill all the definitions above. The definition is therefore not sufficient. We need to add at least one more condition,
- Laws of nature necessitate behaviour of the stuff the law governs.
This added definition demarcates generalizations from actual laws of nature nicely. It is no longer true that "if I own a car, it will be blue" is a law of nature, because there is nothing that necessitates this to be the case. It might turn out to be true, but only by accident, unlike observations we make about stuff that is governed by laws of nature.
However, the definition poses serious problems for empiricism. To claim that there are fundamentally unobservable - Remember, empiricists only claim to look at strings of particular facts. Nobody ever claims to have direct experience of a law itself - governing modes of reality, that not only happen to be true, but necessitate how objects must behave, seems like a huge dip in metaphysics empiricism tries so hard to avoid.
Any solutions to the problem?