Thursday, May 29, 2008

The authority of science and the A-theory of time

The A-theory of time holds that there is an objective distinction between past, present and future (or past and future, or past and non-past, or non-future and future, or present and past, vel caetera). A standard argument against the A-theory of time is that scientific theories do not make this distinction, and hence the distinction is bogus. This is a bad argument (for a true conclusion!). Here are some reasons why.

First, as has been noted before, it is simply false that scientific theories do in fact fail to make this distinction. It seems like a perfectly fine scientific theory to say that birds "arose nearly 200 million years ago". But of course this claim presupposes a distinction between past, present and future, since otherwise the "ago" makes no sense.

Maybe this objection is uncharitable. Perhaps the correct claim is that no fundamental scientific theory presupposes a distinction between past, present and future. A worry I have about this version of the argument is the question of the criteria for fundamentality—there may be a circularity in that we reject the fundamentality of a theory that does make a distinction between past, present and future.

Perhaps, then, we should say:

  1. The laws of nature, as we know them, make no reference to the objective distinction between past, present and future, but treat all times on par.
But it is now rather hard to fill in the details in the argument. Is there a suppressed premise like the following?
  1. If the laws of nature make no reference to some distinction, then that distinction is probably not objective
But surely here is a counterexample to (2): the laws of nature make no reference to the distinction between right and wrong, but the distinction is nonetheless objective. This counterexample won't convince the hard-nosed irrealist about ethics, and perhaps one might modify the suppressed premise by saying:
  1. If the laws of nature make no reference to some distinction, and the distinction concerns a topic amenable to scientific examination, then the distinction is likely not objective.
Here we need the additional premise that the nature of time is amenable to scientific examination. I think (3) is dubious, in part due to the vagueness of the concept of amenability. But I now want to pursue a different criticism of the argument against the A-theory.

If we can know a priori that science must come to a particular conclusion, then our argument for the conclusion is no longer a scientific one. For the peculiar epistemic authority of science is based on scientific methods' responsiveness to empirical data, and in such a case we have no responsiveness. Given an a priori argument that science must come to a particular conclusion, we might accept the conclusion, but if we do so, we do so on the basis of a meta-scientific argument, not a scientific one.

Now, I think we can know a priori that the laws of nature that science arrives at will not have an objective distinction between past, present and future. For the laws of nature that science formulates are timeless. Consider what a law that makes such a distinction would be like. It might, for instance, say that some property is had by an entity now but not in the future. But then the sentence will shortly no longer be true. And that is not the sort of sentence that formulates anything we recognize as a scientific law.

But if we know a priori that the laws of nature that science arrives at will not make a distinction between past, present and future, then we cannot use the lack of such a distinction as a scientific argument against the A-theory.

Of course there may be other and maybe even scientific arguments against the A-theory.

18 comments:

Unknown said...

I find your comments here interesting. Of course, you assume that the nature of empirical knowledge is at best approximate (to wink at Kierkegaard a bit). This assumption forces me to reflect upon the nature of a priori knowledge, and whether or not one can have knowledge of a proposition p (or an argument a), and the content of both p and a be about the empirical/external world, and yet one’s knowledge of p or a, not be at best probable (or subject to possible defeat, a la fallibilism).

(1) Science must come to a particular conclusion.

If your assumption about the nature of a priori knowledge holds, and if science really is principally responsive reflection upon empirical data then I can’t help but see the truth of your first paragraph, viz., that (1) is shown true by way of a meta-scientific argument not a scientific one. If well ordered science was understood in such a way that it incorporated the mathematical foundations of science (e.g. field equations for the special and general theories of relativity for example), then the debatable point about the Minkowskian interpretation of those field equations, in which space and time become geometricized, would support a b-theory of time but such a support relation could be properly dubbed “scientific.”
It does seem to me however, that we come to know about nomic laws via responsive reflection upon empirical data.
Lastly, I’m not sure how the formulation of a nomic law which does not admit temporal becoming, or a hard and fast objective distinction between the past, present, and future, counts as evidence against the a-theory. Why couldn’t the a-theoriest or the presentist simply reformulate, or restate such laws in a way that was thoroughly consistent with the a-theory? For example, she could simply make such formulations tenseless, and therefore they would be (if true) tenselessly true?

Writing off of the cuff…

Anonymous said...

You resolve much of this difficulty if you stop thinking about the "laws of nature" as some equations written down by God at Creation, waiting for us to discover them.

Nature is not regulated by the equations written down in physics textbooks. Nature works as it does and we have physics textbooks to help us predict its behavior.

Martin Cooke said...

It might, for instance, say that some property is had by an entity now but not in the future. But then the sentence will shortly no longer be true. And that is not the sort of sentence that formulates anything we recognize as a scientific law.

Might it deal with the general nature of the collapse of wavefunctions, which may have to occur at the present moment?

Tim Lacy said...

The question of whether the A-Theory holds would seem to depend upon the topic to which science is being applied. For instance, are there not scientific theories that apply to aging---a process that inherently has a past present, and future? Are there not theories, and perhaps laws, that apply to certain eras in the life of living things? - TL

Clark Goble said...

Isn't the main argument against A-theory that of special relativity where an universal present makes no sense? The argument against this is that QM requires a background dependent system (one reason why it is incompatible with relativity). Thus one can justifiably treat relativity as epistemically but not realistically correct. (i.e. it explains measurements but not the way things are) The counter-argument against this is that any successful grand unified theory must be background independent. (Lee Smolin most recently making this argument)

Anonymous said...

" it explains measurements but not the way things are"

Scientific theories only explain measurements, they never explain "the way things are."

Clark Goble said...

That's only if you buy into instrumentalism.

Alexander R Pruss said...

Tankadin:

Certainly, the B-theoretic laws can be reformulated in A-theoretic terms. But then they become more complex, with no explanatory gain.

Enigman:

Surely it won't make sense to claim that collapse happens only now. Obviously, if collapse happens now, it also happened in the past, and will happen in the future.

That said, you may be right that if scientific claims are allowed to revise logic, say by denying excluded middle, then my claim that scientific laws are bound to be B-theoretic in wording will be falsified, because one might have a theory on which collapse actually changes the future.

TL:

The process of aging is the same process whether the person aging is already now dead (so that the process is all in the past), or is not yet conceived (so that the process is all in the future), or is now alive (so that the process is partly past and partly future).

CG:

Yes, that is the main science-based argument against the A-theory. I wasn't refuting all science-based arguments against the A-theory, just one of them. :-)

Anonymous:

The theory of evolution (or, more generally, the theory of common ancestry) explains why organisms of different species all have genetic commonalities.

Besides, a measurement, when correct, expresses the way things are.

Alexander R Pruss said...

Let me put the argument somewhat differently. "If the difference between past, present and future is so fundamental to the reality of time, it is a wonder that fundamental physics does quite well in treating of space-time without this distinction." I think the same criticisms I made of the original argument apply to this one, but this one seems more rhetorically powerful.

Anonymous said...

"Besides, a measurement, when correct, expresses the way things are."

Kind of. Many measurements are theory-dependent.

The theory of evolution doesn't explain much, it just gives us a framework to categorize and generalize about what happened historically as some animals produced offspring and others didn't.

It may be a wonder that the physics equations have that symmetry, but physics actually always deals with time just the way we all do. It couldn't do anything otherwise, because we have no other way of understanding anything.

Anonymous said...

Let me state it another way. The theory of evolution doesn't explain, it helps us describe.

Clark Goble said...

"If the difference between past, present and future is so fundamental to the reality of time, it is a wonder that fundamental physics does quite well in treating of space-time without this distinction."

When you say fundamental physics does well treating spacetime without past, present, and future exactly what do you mean? Put an other way, what do you mean by fundamental physics? Some would say that fundamental physics hasn't done well at treating space time in general. The closest it GR but it works only by excluding a significant portion of what is fundamental. And QM only by excluding gravity.

Martin Cooke said...

Go Goble! And while any successful grand unified theory must be background independent there is a question of just what the background is.

"Obviously, if collapse happens now, it also happened in the past, and will happen in the future." But there are special dates, e.g. year zero. They are and then later they were, but they're not special when present except insofar as all things are. The thing about the collapses is that they are arguably fundamental physics (especially if we've souls) and that they give the present time being present a fundamental importance, obviously.

Incidentally, my apologies for my rant on Prosblog about modal ontological arguments, earlier in the year. I mention it because (a) I just came across the same thing in your PSR book, and it was the only bit (yet) that annoyed me (brilliant book, of course), and (b) I feel the same way about your reference to LEM, I think. That our predicates may be imperfectly defined in subtle ways should not surprise us, or make us question our logical laws.

Alexander R Pruss said...

I don't see why the presentness of the collapses matters, except in a logic which denies excluded middle. If our final theory includes collapse, surely it will equally describe past collapses, present collapses and future collapses. Now, maybe, it was important that the past collapses were present when they happened. But I don't see what's so special about collapse there. Every event is present when it happens. That's a pretty trivial claim about events.

Alexander R Pruss said...

Theory dependent measurements, nonetheless, tell us something about reality. If anything, they tell us that a certain measurement has been made. That instrument X shows number Y on gauge Z is, after all, a fact about reality.

Alexander R Pruss said...

If saying that all organisms have a common origin doesn't explain why they have genetic commonalities, then I don't know what you mean by explanation. This is a perfectly ordinary sort of explanation, used outside of biology as well. "Why are all the witnesses making the same mistake? Because there is a common origin to their statements--they have all been coached by the same lawyer." "Why do manuscripts B, C and F all read 'tou' rather than 'to' in the other manuscripts? Because manuscripts B, C and F all derive from a now no longer extant manuscript produced in Alexandria around AD 350."

Anonymous said...

"If saying that all organisms have a common origin doesn't explain why they have genetic commonalities, then I don't know what you mean by explanation."

Yes, that is a fact about the way things are that we conclude from evidence. Like saying "the sun is a big giant ball of gas".

But that is not what I am walking about. I am talking about the laws of gravitation (and nuclear physics and whatever) that we use to describe how a ball of gas turns into a star. Or the "law" of "survival of the fittest" or whatever. These are things that help us describe what is going on, but they are not actual active principles at work. Evolution does not "follow" survival of the fittest (or whatever other evolutionary principle you want to mention), and the Sun does not "follow" the rules of physics.

People do, though, follow rules. That is the metaphor we apply to physical systems, and then we reason based on that metaphor, and we get all confused.

Anonymous said...

Time is something that will take me years to think about properly, so you may well be right, for all I know; but I was thinking that our best theory of collapse may involve a unique reality (rather than many worlds) and genuine indeterminism (rather than hidden variables). Apologies if the following is gibberish...

My typing of "Time" was a real event, but before I typed it there were only uncollapsed possibilities. Then, there was no actual collapse into that event, waiting to happen in the future. Either time (past, present and future) is changing over time, which is absurd, or else the correct description of collapse applies only to present collapses:

The true description of a past collapse (e.g. my typing of "Time") must include the actual event, while the true description of a future collapse may not even be able to specify what the wavefunction will be like, or even refer to the collapse (but there is no collapse to refer to if there is only the present, so I don't see how LEM enters into it).