Dr Martin proposes "Feature X"  as a symbol for a common pattern he observes in St Thomas' arguments, which can be studied in full in the section "There is no God?". The following will provide a key to the discussion:

"Though the Five Ways can be seen to follow something of a common pattern, they differ essentially in detail. They differ essentially, in that though in each case what is being argued for is the need that the world as a whole has for an explanation which is distinct from it, in each of the Five Ways the world as a whole is being considered under a different aspect, under a different description, as a systematic whole, representing various different manners of systematisation. This is clear from the first sentence of each of the Five Ways. St Thomas begins by identifying a particular feature — call it feature X — a different feature for each of the Five Ways. This feature is displayed by this or that bit of the world: in each case St Thomas takes this claim as being obviously true, and does not argue for it. One of the distinguishing marks of feature X is that anything which displays feature X requires an explanation in terms of its relation to something else: the fact of displaying feature X means that whatever displays it requires an explanation, and does not explain itself. This is clearly a vulnerable point of each in the Five Ways: it is possible for the critic to argue that the feature which St Thomas identifies does not have this distinguishing mark. At various points we will see St Thomas defending his identification of a feature X against obvious criticisms of this kind: a good example is the defence he offers for the principle "omne movens ab alio movetur", everything which is in process of change has that change initiated in it by something else, in the First Way"

The "Feature X" identified in the First Way appears to me to be "being in process of change", in Latin "moveri". There is some ambiguity in this Latin word: it can be used to mean "moving", in the intransitive sense of the English word, i.e., precisely that of being in a certain process of change. Though the English word is usually limited to the notion of change of place, there seems no essential use of this specific sense (which the Latin word can also have) in the First Way. Moreover, the parallel passage in the Summa contra gentes shows us St Thomas explicitly discussing a number of different kinds of change, so it would appear that the notion of change of place is not foremost in his mind. But the more serious ambiguity consists in this: the same word "moveri" is also the passive form of "movere", to move in the transitive sense, to initiate change in something. It has been alleged by Kenny that Aquinas is thinking principally of the regular local movements of the Aristotelian heavens, on the one hand, and, on the other, is hopelessly confused by the ambiguity of "moveri" as between "being in process of change" and "having a change initiated in oneself". Here we can perhaps see the first advantage of observing the structure we have outlined. Such an ambiguity, if it existed, would invalidate the identification of the relevant feature X as both being a feature which (at least) parts of the world uncontentiously display, and also being a feature in virtue of which that which displays that feature requires an explanation in terms of a relationship to something else. In the sense "being in process of change", "moving" in the intransitive sense, most of the world displays the feature of "moveri"; and in the sense of "having one’s change initiated by something else", "being moved by something", (using "moveri" as the strict passive of "moving" in the transitive sense), it is obvious that everything that displays the feature of "moveri" requires an explanation in terms of a relationship to something else. The fallacy would consist in holding, clearly mistakenly, that there is a single feature adequately expressed by "moveri" in its two different sense.

It should become clear, as we examine the First Way, that St Thomas does not commit any such fallacy of equivocation. He spends a good deal of the First Way arguing that everything that is in process of change has that change initiated in it by something else: that is, he argues that whatever "is moved" in one sense also "is moved" in the other. It may be that his arguments fail to prove his point, and indeed they are slightly obscure: but it is clear that he has not been confused into a fallacy by the mere fact that one Latin verb has two senses. Someone who argues at length for the claim that whatever a particular verb applies to in one sense, that same verb applies to in the other sense, has not committed a simple fallacy of equivocation.

It is certain — it is obvious to the senses — that in this world some things are in process of change.

But everything that is in process of change has that change initiated in it by something else.

We cannot go on to infinity in this line,

The curious indentation in this text is not merely a freak of the editor or the printer: it is aimed at clarifying the structure of the argument to some extent. The sections which begin at the left-hand margin are the main premisses of the argument, and its conclusion. Those sections indented are subsidiary arguments introduced in support of what has gone immediately before. Beyond this, there are sections yet further indented which are yet more subsidiary arguments adduced in support of elements in the first-level subsidiary arguments.

The first questionable premiss, then, is premiss 2, everything that is in process of change has that change initiated in it by something else, and there is a long and difficult argument to support it. It stands in need of clarification, indeed, as much as support. At first sight, it looks like a straightforward denial that there are self-movers, things that initiate change in themselves. This claim appears ridiculously false: the dog which begins to bark would seem to be an obvious counter-example. St Thomas, who follows Aristotle in holding that one of the things that marks out animals is their capacity for self-movement — "the originating principle is in themselves" — does not in fact hold anything so false. His claim is rather that no material thing is a self-mover in a carefully-defined sense: there is nothing which as a whole initiates change within itself as a whole. In an animal, let us say, what happens is that one part initiates change in another part.

This caveat is important if we are to make much of the argument which he adduces in support of his premiss 2. Skipping the indented subsidiary arguments, as before, the structure is as follows:

It is true that if this second principle were true, Aquinas would have achieved his point, and made it clear that whatever is in process of change has that change initiated in it by something else, since nothing can both be F and not-F at the same time, so if an initiator of F-ness has to be F, and something that comes to be F has to be not-F, then nothing can initiate change in itself. But it is hard to believe that he can be meaning to offer us such an obviously defective argument.

It is worth noticing, at least, that while the tone and the choice of example suggest that Aquinas may indeed wish to say that only what is F can make something F, the actual text of the argument does not say this. The conclusion which is indeed warranted by the premisses, and is explicitly drawn, is that only something that is actually something or other can make something else to be F.

The reason for this might be as follows. The only requirements on that which is made to be F, so far as the process of "being made to be F" is concerned, are that it should be not-F, and that it should be able to be F, that it should be F in potentiality. No doubt there are plenty of other determinations which it has, plenty of other descriptions which are true of it, but from the fact that it is a subject of a change to being F, all we are entitled to conclude about it is that it is non-F and can be F, is F in potentiality. Neither of these two descriptions suffice to give us a rationale or explanation for its coming to be F. Neither being non-F nor being capable of becoming F offer us any explanation of the fact which we have to explain, which is that it becomes F. If there is to be any explanation of its becoming F, it must be at the very least in virtue of some other aspect, given by some other true description: a description which is not purely negative, such as " — is not-F", nor purely potential, such as " — can become F".

This does not get us any further than the claim which Aristotle and Aquinas would make for the obvious self-movers such as animals, that when the dog begins to bark, for example, it is a case of one part or aspect of the dog initiating a change in another part. We seem to have some justification for the claim that nothing as a whole initiates change in itself as a whole.

It remains to be seen whether this claim will serve us in the development of our account of the First Way: but before we examine that question, it may be worth while examining whether even the more modest claim just made can be justified. That it is justified on Thomistic grounds, and on the basis of the text of the First Way seems clear; and the parallel passage in the Summa contra gentes, which lists a variety of different kinds of change, seems to justify our taking the text of the First Way in this sense, and abandoning the obviously false suggestion that only what is F can make something else F. But there may also be considerations which we can draw from more contemporary views on the nature of change and explanation which may help us to provide a more convincing justification.

We can begin by saying that every change is a beginning of existence. Not that the dog’s starting barking is the beginning of existence of the dog, of course, but it is at least the beginning of the existence of the barking. And we can also insist, pace Hume, that every beginning of existence has a cause. Anscombe, in the latter of the two articles cited, points out that the only way we can identify a beginning of existence as such, as being a genuine beginning of existence, is by attributing to it a cause. To claim that the dog’s beginning to bark is indeed a beginning of existence of the dog’s barking, and not, for example, the mysterious invisible and untraceable arrival to the dog’s throat of a barking which may have existed for centuries elsewhere, is only possible through an identification of the cause, which we may take to be some kind of stimulation in the dog’s brain.

This seems to get us no further than the account we gave of what St Thomas actually said. We are, perhaps, forced to admit some cause or explanation of each change that is something different from the change itself. We can admit, too, that when the change is from not-F to F, from being able to be F to being F, it is not the being not-F as such, not even the being able to be F as such, which is the cause of the change. The cause, we may be willing to admit, must be something actual, some real existent, we might say, using the terminology we used earlier in discussing St Thomas’s doctrine on existence.

But this does not get us very far. We have to admit, perhaps, that there is something in the dog other than its barking which explains its beginning to bark; something in the dog which explains its beginning to bark, which is also other than its previous non-barking, and other than its previous being able to bark. Neither of these two latter features are any kind of cause or explanation of the dog’s beginning to bark; rather they are its logical presuppositions. But can we get from here to the general thesis St Thomas wishes to reach, that everything that is in process of change has that change initiated in it by something else?

It may be possible to bridge the gap. If every change is, under a different description, a beginning of existence, and every beginning of existence requires a cause, it may be possible to argue that every change requires a cause, under some description. And it may be that possible answers to the question, "under what particular description?" may be subject to certain limitations.

A crucial feature in the philosophical analysis of any change is the identification of the subject of the change. This point has been known at least since the time of Heraclitus, who told us that upon those who go down into the same rivers, other and other waters ever flow. An interesting feature of our shifting from the consideration of the dog’s starting to bark in the night as being a change in the dog, to our consideration of the same phenomenon as the beginning of existence of the dog’s barking, was that when we shifted our consideration, and thus the description we gave of the phenomenon, we found ourselves obliged to make a simultaneous shift from the dog as subject of the change to the barking as subject of the beginning of existence; or (in the possible alternative description envisaged) to the barking as the subject of a mysterious and invisible journey through what science-fiction writers like to call "hyperspace". What difference does this shift of subject make?

Interestingly enough, it seems to make no difference to the argument. We know that the dog began barking because of some stimulation in its brain, and because we know this — because we can attribute a cause or explanation — we reject any possibility that the barking was really started some time earlier by some other dog and transferred to our Rover. If this is so, it seems that we can generalise from "every beginning of existence must have a cause" to "every initiation of a process of change must have a cause". We can go one step beyond this, even. Every process of change (as opposed to an instantaneous change) is itself a continually repeated series of initiations of a process of change. To deny this would seem to take us into the territory of Zeno, in which we are to suppose a series of static instants at which no change takes place. We should rather say, with the medieval Aristotelians, that every instant is to be considered either as the last instant of the preceding state or as the first moment of the succeeding state. If this is so, then every process of change requires a cause which is external to that process of change itself. We do not have to deny the existence of self-movers such as animals, but it seems that we do have to accept the Aristotelian account of them: that in the self-movement of animals it is strictly one part of the animal which moves another part, which initiates a process of change in another part. Clearly, too, "one part" and "another part" need not be understood in terms of spatially extended parts, though one of the arguments which Aristotle gives for this thesis seems most naturally taken in the sense of spatially extended parts. The mature Aristotelian doctrine actually suggest that in an animal it is the soul that moves the body, and here there is no idea of the spatial location of the soul. Leaving aside the soul and the body, we can instead at least admit that animals and other obvious self-movers, initiators of change in themselves, do not count as self-movers in the strong sense, as St Thomas would say: it is not the whole animal which initiates a process of change in the whole animal.

This still does not yet seem to have brought us clear reasons to accept the thesis that everything which is in process of change has that change initiated in it by something else. We seem to have been brought to the conclusion that every process of change has to be initiated by something other than the process of change itself, but we still seem far away from St Thomas’s conclusion. St Thomas’s conclusion seems to be that every process of change has to be initiated by something other than the subject of the process of change, and all we seem to be able to conclude is that every process of change must be initiated by something other than the process of change itself.

It seems that it will be necessary to bring forward the use of the lumping-together move, to which we would normally attribute a later place in the First Way as in the other Ways. Our premiss 3 says: "But if that which initiates the change is itself in process of change, then it too must have its change initiated by something else: and so on." This is a point which can surely be made, mutatis mutandis, for the conclusion we seem to have reached. If we admit the conclusion that every process of change has to be initiated by something other than the process of change itself, we can perform at this stage a low-level lumping-together move, and consider the process of change and that which initiates it as a whole. Is this whole itself in process of change, or not?

There seems little reason to deny that every material thing acts on other material things — initiates changes in it — while being itself in process of change; indeed, in virtue of being itself in process of change. However many intermediate steps, aspects, and mechanisms there may be in the material subject of any process of change, we will eventually have to consider that material subject of a given process of change as itself the subject of a process of change, a process of change which can no longer be explained as being initiated by something else within that subject. It begins to look as if for the material universe, at least, the principle "Everything which is in process of change has that change initiated in it by something else" will hold good. I do not know whether Aquinas would regard this as sufficient: it may be that he thinks the principle holds good for e.g. the angels, in which he had such a great interest. But I take it that since Aquinas’s project is to prove the existence of the invisible God from the things we see, the principle need not apply more widely than the material universe, for the First Way to work as he wishes it to work.

Even if this account of Aquinas’s defence of the principle "Everything which is in process of change has that change initiated in it by something else" is sound, there is a difficulty with it. It seems to make the First Way indistinguishable from the Second Way, as we shall see in the next chapter. The Second Way works from considerations of efficient causality in general, and though the above discussion has been couched as far as possible in the terms of "process of change" which are appropriate to a consideration of the First Way, it has made essential use of the principle "Every beginning of existence has a cause": a principle which to apply indifferently to all cases of efficient causality. But we can consider this later; in any case, St Thomas clearly regards the First and Second Ways as closely linked, and perhaps to assimilate one to the other is not too serious a fault of exegesis.

Perhaps we need to take a step back at this stage and take a broader view. We began by alleging that the feature X identified by the First Way is "being in process of change". This seems fairly undeniable. For this feature to be an adequate feature X for the purposes of the argument, according to the structure which we claimed would be followed, it needs to be a feature such that whatever displays that feature requires an explanation in terms of a relation to something else. The subsidiary argument St Thomas gives, which we have been examining, is an argument aimed at proving that this is indeed the case. We can perhaps legitimately make two claims. One is that St Thomas’s subsidiary argument does indeed seem to work provided that we are restricted enough in our identification of the subject of the change — which again, of course, implies some kind of qualification (perhaps an unwelcome one) to the answers we might be entitled to offer to the question "What counts as being a ‘something else’, in relation to which whatever displays feature X requires to be explained?" The second claim is the one made in the last paragraph but one, that whatever we may think of the principle "Everything that is in process of change has that change initiated in it by something else" as a universal metaphysical principle, there seem to be good reasons for accepting it as true of all bodies. And perhaps this is all we need to show.

It is all we need to show at this stage in the argument: for it should by now be obvious that we are still very far short of proving the existence of God. One of the more curious features of the garbled versions of the Five Ways which can still sometimes be met with in books of apologetics is that they (mis-)represent, say, the first half of one of St Thomas’s arguments and then pretend to have proved the existence of God. The argument so far given would bring us to God only if the "something else" which explains the possession of feature X by some parts of the world must necessarily be something which does not itself display feature X. This is because Aquinas takes it that "God" means something that initiates change and is not itself in process of change. But a number of the considerations we have adduced about material things show that this direct jump to God is one which we have no right to make.

It looks, indeed, as if the whole material universe is something which displays feature X as a whole; it looks as if the whole material universe is perpetually and in every part in process of change. I do not know whether this is a contingent fact or a necessary one, but it is probably the fact which makes St Thomas think that this "way" is the clearest of all the five. But this fact is irrelevant to the argument. "Being in process of change", besides having the mark proper to a well-chosen feature X, of requiring an explanation in terms of a relation to something else, also has the second mark proper to a well-chosen feature X, of being generalisable from a part to a whole. If a system has a part which is in process of change, then that system is in process of change. We do not need to claim that every part of the system is in process of change, though in fact this seems to be the case with the material universe. If there is change in any part, there is change in the whole, just as if there is (say) colour in any part there is colour in the whole: i.e. the whole is not monochrome, pure black and white, even if most of it is.

Thus whether it is every part of the world or only some part of it which is in process of change, "being in process of change" is something which can be said truly of the world. The world, as a lumped-together system, displays feature X. And, by the account given of feature X, anything which displays feature X requires an explanation in terms of a relation to something else: in this case, to something which initiates change. Here we can begin a very tedious set of suppositions, imagining, as it were, outside each limit of the world as we know it, some initiator of change. But we would then be able to ask of that initiator of change, whether or not it is in process of change. If not, then we have reached God already; but if it is in process of change then we have wearily to lump this new initiator of change in with the world as previously delimited, and ask about this (newly expanded) world, "What is it that explains its possession of feature X?" and so on, and so on.

This is where the next step which Aquinas makes may come as something of a relief: "we cannot go on to infinity in this line". As already commented, Aquinas in the Summa contra gentes calls this premiss "doubtful" or "questionable". Some of the points we have already made are relevant here, since this step is a part of the "lumping-together" move, and to give an adequate account of the first questionable premiss we found ourselves having to begin to make the lumping-together move somewhat prematurely. The point at issue here is that if we grant that a system of things which display feature X itself displays feature X, and therefore requires an explanation in terms of a relation to something else, it does not matter how large the system is. If a system of things in process of change is itself in process of change, then that change must have been initiated in it by something else. The same is true for any initiator of change in the system which is itself in process of change, and it doesn’t matter how many of them their are. To return to the homely example, it doesn’t matter how many Inuit there are, if the only explanation of the presence of any is his or her relation to another: the presence of the group as a whole requires an explanation, even if the group is infinitely large.

St Thomas here is sometimes taken as saying that the universe could not be infinitely large, or could not have existed for ever. This is a mistake. St Thomas certainly thought that the universe, though incalculably large — and I mean just that, incalculably: the medievals knew that the distance to the fixed stars was so great that the whole diameter of the Earth, which they knew quite accurately, was infinitesimally small by comparison — was finite in extension. He would also have had difficulties with the notion of an infinite spatial extension, difficulties which seem to me very reasonable. But though he also held that the earth had not existed for ever, he held that it might well have done. Only faith in God’s revealed word is sufficient to inform us that "In the beginning God created the heavens and the earth". Our own reason is only sufficient to tell us that the world, even if it has existed as long as God has existed, i.e. for ever, has for ever been God’s creation. Thus St Thomas has no difficulty with the idea of the everlasting existence of the world, and thus no difficulty with the notion of an infinite series of initiators of change which are themselves in process of change. He gives the example later on in the Summa: if an eternal blacksmith had been making horseshoes for all eternity, he would have produced an infinite number of horseshoes and used, worn out and broken an infinite number of hammers and of anvils. There is nothing wrong with this kind of "going on to infinity".

Thus Aquinas’s point is not that the series of initiators of change, themselves in process of change, could not go on into infinity. He was sure, from his knowledge of the Bible, that in fact it didn’t, but he is equally sure that it could have gone on for ever. His point is rather that even if such a series goes on for ever, it fails to explain the existence of feature X. In his own words:

Go to Second Proof

Return