James R
12-16-02, 09:56 PM
The question sometimes arises on sciforums as to whether scientific laws actually describe reality or whether they are just inventions of humans. How do we know that there aren't different laws of nature which would work equally as well as the ones we have now?
Also, what makes something a "law of nature" in the first place?
The following is adapted from a short piece I wrote a couple of years ago which deals with some of these issues. It is based largely on notes for a course called "The Big Questions" given at the Australian National University.
------------
<b>1. What is a "law of nature"?</b>
The first thing is to establish what we're talking about.It seems that out of our sample of biologists, physicists, chemists and others, there are a few different views of what constitutes a natural law. Obviously some type of observed regularity is necessary. Laws tend to take the form "All As are B" or "When A happens, B happens". However, is mere regularity enough to constitute a law, or do we need some additional factor? For example, "All my siblings have blond hair" is a true statement, but would not be regarded by anyone as a law of nature.
Proponents of the so-called <i>anti-Humean</i> view argue that a causal relationship is necessary. In other words, it is not enough that whenever A happens, B happens - B happening must be explainable by A. The fact that my brother is blond is not caused by his being my brother. On the other hand, if we consider the statement "All metals expand when heated", the fact that expansion is observed when a particular metal is heated is in some way a direct result of its being a metal. The fact that it expands can be explained by referring to the law. "Why did it expand? It's a metal." (Compare: "Why is he blond? He's my brother." Not a very good explanation.)
Another feature of laws is that they constrain what is possible. If something violates a law, it is impossible, whereas it is possible to violate a mere regularity. We can say for certain that no metal will be discovered which contracts on heating, but it is possible that I could have a non-blond brother (although in fact I don't).
The anti-Humean philosophy assumes that laws refer to physically necessary relations between things. Other philosophers attack this view on the grounds that the laws don't really explain things at all. It is not at all obvious why the fact that something is a metal requires that it expands on heating. An alternative view, the <i>Mill-Ramsey-Lewis (or MRL)</i> theory, built on the Humean assumption that there are no "necessary relations", holds that laws of nature are in fact just a subset of all the mere regularities, albeit a special one. This subset is in some sense the minimal one required to explain the universe. Another way of saying this is that laws are the "most interesting" and general of the regularities. A law of nature must describe a very wide array of instances in a relatively simple way.
Do laws control the universe, or describe it? Perhaps the fundamental difference between the anti-Humean view and the MRL view is that anti-Humeans have a gut feeling that the laws somehow make things happen the way they do, whereas the MRL people think that the laws are merely an interesting way of describing how things happen.
<b>2. Are laws of nature merely human constructs, or do they have an independent existence?</b>
The relevant philosophical categories here are called <i>constructivism</i> and <i>Platonism</i>. Followers of Plato hold that the laws of nature are "out there", waiting for us to discover them. Constructivists say that we invent laws to explain our observations.
Although the categories are not synonymous, Platonists tend to be anti-Humean, and constructivists tend to adhere to the MRL explanation of laws. (I may be an exception, in that I'd describe myself as a Platonist, but I lean towards the MRL view.)
Platonism (and hence anti-Humeism) is now, and has been in the past, quite popular with physicists. Einstein talked about knowing the "thoughts of the old one". Paul Davies talks about the "mind of God". The assumption seems to be that there is some kind of "master plan" to the universe, which we merely have to figure out. The alternative, which I don't think is given enough attention, is that there is no such plan that the universe is obliged to follow, and the laws of nature in no way actually cause the universe to be as it is.
<b>3. Do the laws of nature describe an actual underlying reality?</b>
This is yet another can of worms. Our laws of nature refer to many entities which we can't directly perceive, such as quarks, curved spacetime and genes. Are these things real, or just convenient human constructs? Again, there are multiple views, forming different philosophical schools:
<i>a. Realism</i>
This is the view that the things our theories talk about have real physical existence.
<i>b. Constructive empiricism</i>
This is the view that unobservable entities don't necessarily exist. The best we can say is that the theories built on the postulated entities give the correct answers for the observable entities which they supposedly affect indirectly by means of the unobservable entities.
<i>c. Instrumentalism</i>
This view holds that theories are only "really" about the observable world. They are true if what they say about the observable facts is true. But the abstract things they refer to are only introduced to make the theory more concise or elegant. Theories don't "actually" talk about the things they purport to talk about.
For example, suppose we use the curved spacetime of the theory of General Relativity to explain the bending of light around a star. The theory says that light is bent because it is following the closest thing to a "straight line" through the curved space.
Now, a realist will take the view that the theory is telling us that there is curved space around the star and that we have every reason to believe that this curvature causes the light to bend. A constructive empiricist will agree that the theory says there is curved space around the star, but will say that there are no grounds for believing that the space is really curved or that the curvature causes the bending. What matters to the realist is that the theory tells us the truth about nature. The constructive empiricist is not interested in absolute truth, only in the observed facts being adequately described.
An instrumentalist, on the other hand, disagrees that the theory is talking about curved spacetime at all. Curved spacetime is just a convenient way of explaining the observations. What the theory is really saying is merely shorthand for "if you look at the light near a star it will be seen to bend by this amount".
I won't go into the arguments for and against each of the views. As with all philosophy, it's definitely worthwhile to consider the question yourself, before referring to others' views.
The distinctions that philosophers make are often subtle, and can require several readings to get straight. In the end, though, there are no "right" answers. You're free to pick the view you like best, or dismiss the whole issue as unimportant, though hopefully not a waste of time.
----------
So, what's your view?
Also, what makes something a "law of nature" in the first place?
The following is adapted from a short piece I wrote a couple of years ago which deals with some of these issues. It is based largely on notes for a course called "The Big Questions" given at the Australian National University.
------------
<b>1. What is a "law of nature"?</b>
The first thing is to establish what we're talking about.It seems that out of our sample of biologists, physicists, chemists and others, there are a few different views of what constitutes a natural law. Obviously some type of observed regularity is necessary. Laws tend to take the form "All As are B" or "When A happens, B happens". However, is mere regularity enough to constitute a law, or do we need some additional factor? For example, "All my siblings have blond hair" is a true statement, but would not be regarded by anyone as a law of nature.
Proponents of the so-called <i>anti-Humean</i> view argue that a causal relationship is necessary. In other words, it is not enough that whenever A happens, B happens - B happening must be explainable by A. The fact that my brother is blond is not caused by his being my brother. On the other hand, if we consider the statement "All metals expand when heated", the fact that expansion is observed when a particular metal is heated is in some way a direct result of its being a metal. The fact that it expands can be explained by referring to the law. "Why did it expand? It's a metal." (Compare: "Why is he blond? He's my brother." Not a very good explanation.)
Another feature of laws is that they constrain what is possible. If something violates a law, it is impossible, whereas it is possible to violate a mere regularity. We can say for certain that no metal will be discovered which contracts on heating, but it is possible that I could have a non-blond brother (although in fact I don't).
The anti-Humean philosophy assumes that laws refer to physically necessary relations between things. Other philosophers attack this view on the grounds that the laws don't really explain things at all. It is not at all obvious why the fact that something is a metal requires that it expands on heating. An alternative view, the <i>Mill-Ramsey-Lewis (or MRL)</i> theory, built on the Humean assumption that there are no "necessary relations", holds that laws of nature are in fact just a subset of all the mere regularities, albeit a special one. This subset is in some sense the minimal one required to explain the universe. Another way of saying this is that laws are the "most interesting" and general of the regularities. A law of nature must describe a very wide array of instances in a relatively simple way.
Do laws control the universe, or describe it? Perhaps the fundamental difference between the anti-Humean view and the MRL view is that anti-Humeans have a gut feeling that the laws somehow make things happen the way they do, whereas the MRL people think that the laws are merely an interesting way of describing how things happen.
<b>2. Are laws of nature merely human constructs, or do they have an independent existence?</b>
The relevant philosophical categories here are called <i>constructivism</i> and <i>Platonism</i>. Followers of Plato hold that the laws of nature are "out there", waiting for us to discover them. Constructivists say that we invent laws to explain our observations.
Although the categories are not synonymous, Platonists tend to be anti-Humean, and constructivists tend to adhere to the MRL explanation of laws. (I may be an exception, in that I'd describe myself as a Platonist, but I lean towards the MRL view.)
Platonism (and hence anti-Humeism) is now, and has been in the past, quite popular with physicists. Einstein talked about knowing the "thoughts of the old one". Paul Davies talks about the "mind of God". The assumption seems to be that there is some kind of "master plan" to the universe, which we merely have to figure out. The alternative, which I don't think is given enough attention, is that there is no such plan that the universe is obliged to follow, and the laws of nature in no way actually cause the universe to be as it is.
<b>3. Do the laws of nature describe an actual underlying reality?</b>
This is yet another can of worms. Our laws of nature refer to many entities which we can't directly perceive, such as quarks, curved spacetime and genes. Are these things real, or just convenient human constructs? Again, there are multiple views, forming different philosophical schools:
<i>a. Realism</i>
This is the view that the things our theories talk about have real physical existence.
<i>b. Constructive empiricism</i>
This is the view that unobservable entities don't necessarily exist. The best we can say is that the theories built on the postulated entities give the correct answers for the observable entities which they supposedly affect indirectly by means of the unobservable entities.
<i>c. Instrumentalism</i>
This view holds that theories are only "really" about the observable world. They are true if what they say about the observable facts is true. But the abstract things they refer to are only introduced to make the theory more concise or elegant. Theories don't "actually" talk about the things they purport to talk about.
For example, suppose we use the curved spacetime of the theory of General Relativity to explain the bending of light around a star. The theory says that light is bent because it is following the closest thing to a "straight line" through the curved space.
Now, a realist will take the view that the theory is telling us that there is curved space around the star and that we have every reason to believe that this curvature causes the light to bend. A constructive empiricist will agree that the theory says there is curved space around the star, but will say that there are no grounds for believing that the space is really curved or that the curvature causes the bending. What matters to the realist is that the theory tells us the truth about nature. The constructive empiricist is not interested in absolute truth, only in the observed facts being adequately described.
An instrumentalist, on the other hand, disagrees that the theory is talking about curved spacetime at all. Curved spacetime is just a convenient way of explaining the observations. What the theory is really saying is merely shorthand for "if you look at the light near a star it will be seen to bend by this amount".
I won't go into the arguments for and against each of the views. As with all philosophy, it's definitely worthwhile to consider the question yourself, before referring to others' views.
The distinctions that philosophers make are often subtle, and can require several readings to get straight. In the end, though, there are no "right" answers. You're free to pick the view you like best, or dismiss the whole issue as unimportant, though hopefully not a waste of time.
----------
So, what's your view?