To grasp why Aquinas’s arguments throughout his corpus of work (De Ente, Summa, etc) lead ineluctably to God’s existence, we must begin by appreciating the specific type of causal reasoning Aquinas is engaged in. This is critical because one very common point of confusion — in fact, the perennial confusion — when it comes to Aquinas’s argumentation for God pertains to not differentiating between a per se and per accidens causal series. Once this distinction is made, however, it becomes obvious why Aquinas demands a primary member in one type of causal series (hence why the objection, “Why can’t there be an infinite backward string of causes?” has no force) but not in the other. Let’s review.
OK, so. Step one in causal reasoning is to identify what the causal property is that we’re seeking an explanation (or cause) of and then discern what that explanation (or cause) is. Imagine my bedroom is being illumined in the evening, to which we can then ask if the moon — bright and round — is the cause of the illumination. The answer to which is yes (in one sense) but no (in another). Simply put, the moon is the proximate but not ultimate cause of the bedroom being illuminated. Why? Because we know the moon does not possess the power of illumination inherently, it cannot stand as the ultimate cause of the causal property (illumination) under consideration. Instead, the moon possesses the power to reflect light, not produce it. For the production of light, we require the sun. Put differently, the moon simply “borrows” or “derives” the power of illumination from the sun and thereby acts as an intermediary (or instrumental, subordinate) cause in the process of illuminating the bedroom.
That is what makes a causal series (x) -> (y) -> (z) per se: (x) must borrow a power from (y) to produce (z). This is clearly the case with the moon deriving light from the sun to illumine the bedroom. Whereas in a causal series (x) -> (y) -> (z) which is per accidens, (x) does not need to borrow a power from (y) to produce (z). Think Harry begetting Bob and then Bob begetting John. Once Bob is begotten, Harry becomes irrelevant and can drop out without interruption to the causal series, because Bob has the power inherently to beget John. Bob doesn’t need to borrow the power of begetting from his father Harry in the process of begetting (that would be weird); he (Bob) has it in virtue of being a biologically well-functioning male. The point? Because members in a per accidens causal series have the inherent power to produce the effect under consideration, Aquinas grants that we need not trace to anything more primary, hence he maintains that such a causal series could regress infinitely.
But not so with a per se ordered causal series. Notice – tracing back to our original example — it doesn’t matter how many moons we add to the per se causal series. Posting even an infinite number of intermediary or instrumental causes which do not possess the causal property inherently – that is, do not, by principles internal to their nature, have the ability to produce the causal property under consideration – would be insufficient to introduce the causal property into the series in the first place. Just as we cannot stack white tiles to produce a purple floor (even if we had an infinite number of white tiles) we face a similar construction problem when attempting to stack causes which are mere “borrowers” of a causal property to explain the generation of the causal property. Borrowers can never add up into generators, and it is a category mistake to assume otherwise. Thus, even if there were an infinite number of intermediary causes – an infinite number of moons, in our example – there would still need to be a primary member (in our case, a sun) either within or BEYOND that series to explain how the causal property of illumination got there in the first place.
So Aquinas’s arguments rely upon per se (versus per accidents) causal reasoning, and it is by coming to understand the nature of a per se causal series that we understand why a primary member is necessary to explain the causality of the series. In short, when it comes to a per se causal series, some cause is the cause OF THE CAUSALITY of the series. No primary member, no causality of the series. No sun, no light reflecting from the moon.
Moving on.
Another mistake — one common among internet skeptics — is to think a per se causal series is defined by notions of hierarchy or simultaneity rather than dependency. Indeed, while a per se causal series may be often arranged hierarchically and involve cause-effect relations which are simultaneous (though not necessarily instantaneous) that is NOT what defines a per se causal series. Again, what defines a per se causal series is when the causal property in question is not possessed inherently by subsequent members of the series. When (x) must borrow a power from (y) to produce (z). THAT is what makes a causal series per se, regardless of if organized hierarchically, linearly, etc. This we must always keep in mind.
Final thought: Another objection is the idea that there can be a cause that is primary in one causal series but not primary in another, even if it is one and same substance. The sun, for example, may be the primary cause of illumination in the example we gave above, but not the primary cause of the causality of some other series, whatever that may be. This is occasionally brought up as a rebuttal to Aquinas inasmuch as it suggests that to find *a* primary member is not the find *the* or *only* primary member. This statement is true but trivial and irrelevant, because Aquinas’s way to God is concerned with the most fundamental property of all, which is existence. So, whatever is the cause of the existence of things is the cause of ALL CAUSALITY of ANY SERIES (since every other causal series depends upon it having existence) such that whatever is the cause of the existence of things is primary in the most unique and fundamental sense. It would, in fact, be *the* primary cause, not just *a* primary cause.
The issue with people rushing into Aquinas (hostiles and friendlies alike) is they assume Aquinas’s arguments for God are some haphazard collection of attempts demonstrate theism meant to put the respective atheist or agnostic in their place. But that is not what Aquinas is after. Rather, Aquinas is a metaphysician seeking first and foremost to understand the structure of reality – the nature of causation, the distinction between essence/existence, and so on — a structure which when properly understood happens to inevitably entail the existence of God. He is not doing an apologetic project; he is doing a metaphysical project, and there is a difference. Thus, to engage seriously with Aquinas’s arguments for God demands engaging seriously FIRST with his metaphysics generally. Failure to do the latter almost always results in straw-manned or question begging critiques of the former.
Related
How to Think About God – $1 eBook on Amazon
Philosophy Friday: Arguing Essence and Existence with Dr. Gaven Kerr