Reply to Michael Martin on the Cosmological Argument: Part One

Michael Martin is an atheistic philosopher and an expert on the problem of induction.  The reason I've chosen to respond to him on the cosmological argument is because of a) his in-depth analysis; and b) his ability to write on a popular level.  More complicated objections to cosmological arguments, such as those of Graham Oppy, deserve a response, but Martin's arguments likely have a wider influence.  By successfully answering Martin's objections to a number of cosmological arguments, my hope is that we can instill a greater confidence in theists who defend the existence of an Unmoved Mover/necessary entity/first cause.

I'll begin with some preliminary remarks.  Martin prefaces his treatment of various cosmological arguments with the following comment about cosmological arguments in general (his words will be in blue):

In its simplest form the cosmological argument is this: Everything we know has a cause.  But there cannot be an infinite regress of causes, so there must be a first cause.  This first cause is God. (Atheism: A Philosophical Justification, p. 96)

As anyone familiar with cosmological arguments will be quick to point out, the argument is almost never "everything has a cause."  Rather, it's usually something like, "every contingent thing has a cause," or "whatever begins to exist has a cause."  In the case of the argument from motion, the causal premise is: everything in motion has a sustaining cause.  To his credit, however, Martin recognizes that there are more sophisticated versions of the cosmological argument which incorporate these qualified causal premises.

It is well to state the problems with this simple version of the argument, since, as we shall see, they are found in some of the more sophisticated versions as well.  Perhaps the major problem with this version of the argument is that even if it is successful in demonstrating a first cause, this first cause is not necessarily God.  (p. 97)

What the reader has to note here is that the first part of any cosmological argument only attempts to show that there is a first cause of some sort.  Additional argumentation is needed to show that the first cause possesses attributes most consonant with theism.  Thomas Aquinas, for example, spends very little time demonstrating the existence of a first cause, and then a great deal of time showing that the first cause must be omnipotent, etc.  Martin himself alludes to this in one of his footnotes:

Because of this, an argument for a first cause must be supplemented with some other argument that attempts to show that the first cause is God.  Indeed, sometimes the cosmological argument is considered to have two parts.  In the first part a first cause is established, and in the second part the first cause is identified with God.  (p. 492)

I'm glad Martin takes the time to point this out, since it's not uncommon to find folks who knock down caricatures of the cosmological argument.  We often hear things like: "There may be a first cause, but that doesn't mean Christianity is true."  Well, of course the objector is technically correct.  What he (not necessarily Martin) overlooks, however, is that the cosmological argument, and in particular the first part of the cosmological argument, is only meant to be one aspect of a greater cumulative case for the truth of Christian theism, or of theism in general.

A first cause need not have the properties usually associated with God.  For example, a first cause need not have great, let alone infinite, knowledge or goodness.  A first cause could be an evil being or the universe itself.  In itself this problem makes the argument quite useless as support for the view that God exists.  (p. 97)

Here is where I disagree with Martin.  In order to legitimately deal with the cosmological argument as an argument for God's existence, one needs to acknowledge and respond to the arguments that philosophers, such as Thomas Aquinas, offer in support of the conclusion that the first cause is God.  We will come back to this point at a later time.  For now, I only want to deal with Martin's objections to the existence of an Unmoved Mover/first cause.

However, it has at least one other equally serious problem.  The argument assumes that there cannot be an infinite sequence of causes, but it is unclear why this should be so.  (p. 97)

I'll stop here for a moment only to point out that the proponent of the cosmological argument does not merely "assume" there cannot be an infinite regress of causes.  Instead, he argues that such a regress cannot be infinite.

Experience does not reveal causal sequences that have a first cause, a cause that is not caused.  So the idea that there can be no infinite sequences and that there must be a first cause, a cause without a cause, finds no support in experience.  (p. 97)

This may be true, but it's irrelevant for at least two reasons.  First, experience is not the sole criterion of demonstration.  One may have a priori reasons to reject the possibility of an infinite regress.  Secondly, there is no experience of an infinite regress, either.  If one is going to hold up experience as our only guide of knowledge, then one can neither claim that there is an infinite regress nor that there isn't one.  Martin makes a passing acknowledgement of this latter fact:

This is not to say that experience indicates an infinite sequence of causes.  Rather, the presumption of the existence of a first cause seems to be a nonempirical assumption that some people see as obvious or self-evident. (p. 97)

The problem here is that he once again refers to the argument as an "assumption."  Moreover, not all nonempirical conclusions are based on self-evidence.  There's no reason we cannot make inferences that are themselves not self-evident that are based on a priori truths.

From a historical point of view, however, any appeal to obviousness or self-evidence must be regarded with suspicion, for many things that have been claimed to be self-evidently true - for example, the divine right of kings and the earth as the center of the universe - have turned out not to be true at all. (p. 97)

This is a bit of a red herring.  Of course there are things that were once considered self-evident that are no longer considered true.  However, would Martin include "2+2=4" among such a list?  The fact is, there are conclusions that were once made on observational experience that are no longer considered true.  That the earth was thought of as the center of the universe was actually based on experience.  Because of stellar parallax, Aristotle concluded that the heavens revolved around the earth.  This was not based on "self-evidence," but rather on the experience that Martin holds up.  In this case, further experience showed that a limited experience bore false conclusions.  However, it has yet to be seen whether the arguments against an infinite regress of causes fall under Martin's list of false conclusions, or whether they are actually true.

Further, we have no experience of infinite causal sequences, but we do know that there are infinite series, such as natural numbers.  One wonders why, if there can be infinite sequences in mathematics, there could not be one in causality.  No doubt there are crucial differences between causal and mathematical series; but without further arguments showing precisely what these are, there is no reason to think that there could not be an infinite regression of causes.  (p. 97)

Just one difference between an infinite regress of causes and an infinite sequence of natural numbers is that the former would constitute something concrete, whereas the latter (if they exist at all) constitute something abstract.  Moreover, not all mathematicians agree that the sequence of natural numbers can be actually infinite.  Intuitionists, for example, reject such a notion.  However, since they're a minority of mathematicians, we can at least grant the possibility of there being infinitely-many natural numbers.  But, it's a category mistake to apply infinite set-theory (which is abstract) to an infinite regress of causes (which is concrete).

Some recent defenders of the cosmological argument have offered just such arguments, and I examine these arguments later.  But even if they are successful, in themselves they do not show that the first cause is God.  (p. 97)

I'm not aware of any defender of the cosmological argument who would say otherwise.  In my next post, I will respond to Martin's objections to the first major contention of the cosmological argument.