Duns Scotus has especially interesting and important things to say about the distinction between causal series ordered accidentally and those ordered essentially -- a distinction that plays a key role in Scholastic arguments for God’s existence. I discuss the distinction and Scotus’s defense of it in Scholastic Metaphysics, at pp. 148-54. Richard Cross, in his excellent book, Duns Scotus, puts forward some criticisms of Scotus’s position. I think Cross’s objections fail. Let’s take a look at them.
First, a brief review of the distinction. As longtime readers are aware, the key difference between the two kinds of causal series has to do with whether their members have their causal power in a derived or underived way. Consider the causal sequence: x → y → z. The series is essentially ordered if, in the very act of causing z, y borrows from x the power to do so. Hence, in the stock example, the stick pushes the stone only insofar as it derives from the hand the power to push it. A series is accidentally ordered if, in the act of causing z, y does not borrow from x the power to do so. Hence, in the stock example, a son can beget a son of his own whether or not his own father is still alive. (I’ve discussed this distinction at greater length in earlier posts, such as this one, and in various books and articles. Again, see Scholastic Metaphysicsfor detailed discussion.)
Cross labels an essentially ordered series of causes an “E-series,” and an accidentally ordered series of causes an “A-series.” His main criticism of Scotus’s use of the notion of an E-series is contained in the following passage:
In the late Reportatio (closely paralleled in the Ordinatio) Scotus argues from the following premise: “In essentially ordered causes… each second cause, in so far as it is causing, depends upon a first.” Put in this way, it follows straightforwardly that there must be a first member of an E-series. But the premise is question-begging, and I can see no reason for wanting to accept it. It requires that a first cause is necessary as well as sufficient for any effectin an E-series. But this is not so… [I]n any causal series there is asense in which the existence of earlier causes is necessary for the existence of later causes. But we cannot inferfrom this that a first cause is necessary for some effect. There are sometimes many different ways in which the same effect can be produced.
Taking account of this objection, we could loosely reformulate the premiseas follows: “In essentially ordered causes, any later cause, in so far as it is causing,depends upon an earlier cause.” Put thus, the premise looks wholly plausible. But there would be no problem with an infinite E-series thus construed. Howsoever many prior causes there were, any one of them would be logically sufficient for any later effect. (p. 19)
There are several things to say about this. To begin with, note that there are two senses in which something might be characterized as “first.” We might mean that it comes at the head of some sequence. This is what we have in mind when we say that Fred was first in line for the movie, or that Ethel was the first to arrive at the party. We mean first as opposed to second, third, fourth, fifth, etc. Let’s use “firsts” when what is intended is this sequential sense of the word. But we might mean instead, when we characterize something as “first,” that it is in some way more fundamental or essential relative to other things, or that in some respect it has a higher status. This is what we have in mind when we characterize something as being “of the first rank,” when we describe someone as “first among equals,” or when we give the title “First Lady” to the wife of the President of the United States. We mean first in the sense of principal or primary as opposed to secondary. Let’s use “firstf” when what is intended is this sense of the word, which involves some kind of fundamentality or eminence.
Now, something can be firsts without being firstf, and something can be firstf without being firsts. The U.S. Army Chief of Staff would be the firstf soldier in the Army even if he were not the firststo join the Army, indeed even if he were the last to join. And in theory a certain Army private could be the firsts soldier insofar as he joined before any other living soldier did, even though he has never gotten any further in rank and thus is far from being firstf. The FirstfLady of the United States is obviously not the firsts lady ever to have lived in the United States. Descartes’ Meditations on First Philosophy is not concerned with the firsts philosophy ever devised by a philosopher (Thales, say) but rather with firstfphilosophy, i.e. that branch of philosophy which deals with the most fundamental philosophical issues. When First Comics was founded in the 1980s, the company was not claiming to be the firsts comic book company, but rather aspiring to be the firstf comic book company. And so forth.
Now, suppose that when Scotus or some other Scholastic says that “in essentially ordered causes… each second cause, in so far as it is causing, depends upon a first,” what is meant is firsts. Then it is easy to see why Cross would raise the objections he does. For why should a second (or third, or fourth) cause require a firsts? If Scotus were just stipulating that you couldn’t have a second, third, fourth, etc. cause without a firstscause, then he would be begging the question (as Cross accuses him of doing), since Scotus’s critic doesn’t see why a firsts cause is needed and hasn’t been given a reason to change his mind. And if Scotus reformulates his position as the claim that “in essentially ordered causes, any later cause, in so far as it is causing,depends upon an earlier cause,” then (as Cross indicates) even if this is true, it will not entail that there is a firsts cause.
The problem, though, is that this is simply notwhat Scotus and other Scholastics mean. In the proposition that “in essentially ordered causes… each second cause, in so far as it is causing, depends upon a first,” what is mean is firstf, not firsts. In particular, the claim is that in essentially ordered causal series, causes which have their causal power in a merely secondary or derivative way require a cause which has its causal power in a primary or underivative way. And there is nothing question-begging about that, even if the point needs greater spelling out than Scotus gives it in that one quoted sentence considered in isolation.
When I point out that a stick cannot move a stone by itself but requires something else to impart to it the power to move stones and other things, I am not begging any questions but rather saying something that no one would deny, not only because we all know from experience that sticks don’t move stones by themselves but also because it is evident from the nature of sticks that the reason they don’t in fact move other things by themselves is that they can’t do so. For they simply don’t have the built-in power to do so. Neither am I begging any questions when I point out that the same thing is true of the arm which movies the stick. Like sticks, arms all by themselves not only never do move other things but couldn’t do so given their nature.
Nor am I begging any questions when I go on to conclude that such a series of causes requires something which imparts the power to move things without deriving it from anything else -- for example, a human being, who can use his arm to move the stick to move the stone, without the need for someone else to pick him up and move him while he does so. Here too I am saying something which is not only obvious from experience, but also evident from reflection on the natures of the causes involved. For one thing, human beings have by nature a built-in power of movement that sticks, stones, and arms do not. For another thing, in general what is derivative presupposes that from which it is derived. Even Scotus’s critic would have to admit that the stick’s movement of the stone cannot be accounted for unless we appeal to something from which the stick derives its causal power, such as the arm. And the critic would have to admit that accounting for the arm’s movement requires a similar appeal, for the same reason. But any further member we posit which, like the stick and the arm, lacks built-in power, will just raise the same problem all over again. So, we cannot account for the motion we started out with -- that of the stick as it moves the stone -- until we get to something which does have built-in or underivative causal power.
Positing an infinite regress of derivative causes is no alternative. Suppose I owe you money, you demand that I pay up immediately, and I offer you an IOU instead. Suppose you refuse to accept it on the grounds that you doubt I’ll ever be able to back it up with real money. Suppose that, in order to ease your doubts, I offer you a second IOU to back up the first. Naturally, you refuse that IOU too, and on the same grounds. Now it would be absurd to suppose that if I go on (Dumb and Dumber style) to offer you an infinite series of IOUs, each backing up the previous one, then you will suddenly have a reason to abandon your doubts and accept my IOUs. Similarly, it is absurd to suppose that positing an infinite regress of causes having merely derivative causal power somehow solves the problem that positing one, two, three, etc. derivative causes was unable to solve.
In any event, even if someone were for some reason to try to resist this line of argument, there is nothing question-begging about it, and neither does it fail to offer a reason for thinking that there must be a cause with built-in or underived causal power. So, Cross’s charge that Scotus either begs the question or fails to give any reason for supposing that an E-series requires a first member cannot be maintained, at least if what Scotus has in mind (as he surely does) is a firstfcause and not merely a firsts cause.
What about Cross’s point that “we cannot infer… that a first cause is necessary for some effect [since] there are sometimes many different ways in which the same effect can be produced”? The idea here seems to be that even if in the case of the stick moving the stone (say), the stick does so only because a person moves the stick with his arm, there are nevertheless other ways in which the stick might be moved. For example, it could be tied to some machine which moves it about, and by which it is able to move a stone. But the problem with this objection is that it shows only that, in the E-series in question, this or that particularfirstf cause is not necessary. It does not show that some firstfcause or other is not necessary in any E-series.
Cross raises a couple of further objections in an endnote. First, he suggests that:
We might be inclined to argue that, if there were no first cause to an E-series, we could not find the real cause of any effect… Richard Swinburne notes that this argumentfalls victim to what he labels the ‘compIetist fallacy’: if y causes z, then it really does explain the existence of z, even if y itself requires explanation. (pp. 161-62; Cross is referring to remarks made by Swinburne in the second edition of his book The Existence of God)
The trouble with this objection is that to say that something is not a “complete” cause is simply not the same thing as to say that it is not a “real” cause, and to say that something is not a “complete” explanation is simply not the same thing as saying that it is not a “real” explanation. Does the stick in our example reallymove the stone? Of course. Does its motion really explain the motion of the stone? Yes indeed. But is the stick the completecause of the motion of the stone? Of course not. And neither does its motion completely explain that of the stone, precisely because it would have no power to move the stone at all if it did not derive it from the person who uses it to move the stone.
Finally, Cross says:
Scotus's argument is made more complicated by his claim that even if per impossibile there were an infinite series of causes, each one would have to depend on some first cause that was outside the series… But this just blurs the distinction between an E-series and an A-series. On Scotus's initial definitions, an E-series will be self-sufficient; it will not depend on any cause outside itself. (p. 162)
The reason Cross thinks this blurs the distinction between an E-series and an A-series, it seems, is that Scotus and other Scholastics hold that an A-series need not have a first member, whereas an E-series must have one. But Scotus’s allowing for the sake of argument that an E-series might regress infinitely will seem to blur the distinction between an E-series and an A-series only if we fail to keep in mind the distinction between a firsts cause and a firstfcause. When Scotus allows for the sake of argument that an E-series might regress infinitely, he is not saying, even for the sake of argument, that an E-series might lack a firstfcause. Rather, he is allowing for the sake of argument that it might lack a firsts cause, and saying that even if it lacked one, it would still require a firstf cause.
For example, suppose the stone was being pushed by a stick, which was being pushed by another stick, which was being pushed by yet another stick, and so on ad infinitum. Such a series would not have a firstsmember. But there would still have to be a firstf member outside the series to impart motion to it, because of themselves a mere series of sticks, however long, would have no power to move at all.
So, Cross’s objections all fail. But someone might still wonder how all this supports an argument for God’s existence. For of course, Scotus, like Aquinas and other Scholastics, intends to argue for a single and divine first cause. Yet a person who moves a stone with a stick is only one firstf cause alongside many others, and a non-divine one at that.
But pointing out that an E-series must have a firstf member, and illustrating the idea with the stick example, is by no means the whole of a First Cause argument for God’s existence. It is only part of a much larger line of argument. For one thing, while a person who moves a stone with a stick is a firstf cause relative to that particular series, it does not follow that he is a firstf cause absolutely, full stop. Indeed, relative to other E-series, he will himself be an effect. For example, his existence at any moment depends upon the existence and proper configuration of his micro-level material parts. And in a metaphysically more fundamental way, it depends on his substantial form being conjoined with prime matter, and his essence being conjoined with an act of existence. The regress this entails will be vicious unless it terminates in a cause which is purely actual and thus need not be actualized by anything else. And a purely actual cause turns out on analysis to have the divine attributes.
But that’s a whole other story (for which see chapter 3 of Aquinasand several of the essays on natural theology in Neo-Scholastic Essays).