Let's turn now to Scotus's attempt to prove that the infinite being actually exists. Read Opus Oxoniense book I, d.2, q.1, Hyman & Walsh, p.564, down to the middle of p.565, to "Thus in the first article there are nine conclusions".
Some comments. The editors have omitted the preliminary arguments pro and con; we begin with Scotus's explanation of the organisation of his argument. The first two sentences turn on a contrast between what the translator calls "adequately grounded demonstration" and "demonstration of fact", in Latin, between demonstration propter quid and demonstration quia. The distinction comes from Aristotle, Posterior Analytics, I.13. It is the difference between a demonstration that conveys an understanding of why the thing is and must be so, and one that merely shows that it is so. (Some translators use the terms "demonstration of the fact" and "demonstration of reasoned fact".) An argument sometimes establishes a fact without giving any understanding of it. According to Aristotle, the middle term of a syllogism that conveys understanding must be the cause or principle or reason why the predicate of the conclusion inheres in the subject. The cause or principle is something prior to or higher than the terms it connects. Hence a demonstration propter quid is a demonstration a priori, from something prior; a demonstration quia is a demonstration a posteriori. In the present case the terms, the subject and predicate of the conclusion to be proved, are "infinite being" and "actually existent". Scotus says that there is no higher or prior middle term knowable to us by which we could understand why these terms are necessarily connected. To us: in themselves the terms may be connectable by some higher principle, but we can't know it. God may understand why he exists, but the best we can hope for is to prove that he exists, not by recourse to a higher term but by recourse to something lower, namely creatures. "But as far as we are concerned, it is demonstable by a demonstation of the fact, from creatures." The point Scotus makes here was made also by Thomas Aquinas, PHIL252 Readings, pp. 104-5: the proposition is self-evident in itself but not to us, to us it can be demonstrated in a demonstration quia.
The third sentence contrasts God's relative and absolute qualities: in relation to creatures God is cause, creator, end, etc.; in himself he is good, wise, etc.; in relation of creatures (who are also good, wise, etc.) he is preeminently good, wise, tc.; or first good etc. Scotus is saying that it is easier to demonstrate God's existence as cause etc., or as preeminent, or as first, than from absolute properties.
In the second paragraph the terms should be familiar. Recall from PHIL252 Supplement, p.118, Aristotle's catalogue of four causes (or becauses); God is not the formal or material cause of any creature, so his causal attributes relatively to creatures are efficient and final (recall that the "final cause" is the goal or end, the efficient cause is the productive cause). The exemplar is the model copied, the Form in Plato's sense--not in the thing produced, but the model in imitation of which it is produced.
In the list of conclusions note: "Second, I show that the same which is first according to one primacy, is also first according to the other primacies". It was a defect in Thomas Aquinas's "Five Ways" that he did not show that the five beings concluded to were one and the same being: Scotus will argue this point.
Number the paragraphs as follows:
At the beginning of the long paragraph that begins just below the middle of p.565 , at "The first of these conclusions is this", write "1.111."
Over the page, at the first new paragraph on p.566, beginning "The third conclusion about the first causally effective being is this", write "1.113."
On the same page, at the second last paragraph, beginning "The first conclusion is", write "1.121." At the next paragraph, write "1.122," at the next write "1.123."
At the third paragraph after that, beginning "The first conclusion is this: Some eminent nature "etc/. write "1.131." At the next, "1.132," at the next "1.133."
At the paragraph beginning "As to the second article", write "1.2."
Read the next two paragraphs, down to the first omission dots on p.566. The first objection, which Hyman and Walsh omit, is that an infinite regress is possible. In reply Scotus invokes the difference between essentially and accidentally ordered causes; see Duns Scotus, Philosophical Writings, pp.40-44, a lengthy discussion.
The second objection is something new: A demonstration, according to Aristotle, shows that something must be so, so how can its premisses be matters of contingent fact? Scotus points out that he argues from the proposition that some being can be an effect--see paragraph 1.111, line 3, "Some being can be an effect". We know that this premiss is true, since some beings are effects--from contingent fact we can infer possibility. But it is a necessary proposition that if effects are possible causes are possible. That the possible first efficient cause is actual is proved in 1.113.
Section 1.112 is omitted (it argues that the first efficient cause must be uncaused, incapable of being caused).
The last stage of the argument rests on the premiss: if a thing is possible, that must be either because it already exists of itself, or because it can be caused. But this second alternative is ruled out in this case, because a possible first uncaused cause can't be caused. Therefore, if it is possible it must be actual.
At the end of the first paragraph of 1.113, "The fifth argument given for A" (i.e. for proposition A) refers to a passage in what would have been 1.112, which is omitted.
Comment. The "five arguments similar to those which were given for the first conclusion" are in a passage omitted at the bottom of p.565, in the answer to the objection that an infinite series is possible. Scotus is saying there that the arguments to prove that the series of efficient causes producing efficient causes cannot be infinite can be adapted to show that there cannot be an infinite series of things sought as means to ends sought as means to further ends sought as means to still further ends and so on.
Notice the point that not only is the "first end" without a final cause--i.e. a further end for the sake of which this one is sought--but it is also without an efficient cause. This is because (according to Aristotle) an efficient cause always acts for an end: if the "first end" had an efficient cause, there would have to be some further end for the sake of which its efficient cause brought this ultimate end into existence--which is contradictory (if something is ultimate there is nothing further). You may object that whenever we bring into existence something we value for itself, we are effecting an end that does not yet exist, and effecting it for its own sake (not for a further end). But Scotus would reply that speaking more accurately what we aim at is to enjoy or possess such a thing, and bringing it into being is for the sake of enjoying it or possessing it. The objection he answers is one that mightn't occur to you: namely that the final end might be brought into existence incidentally or per accidens. "Intrinsic cause" translates causa per se, which is the opposite of causa per accidens. Scotus argues that in no order of causation, efficient or final, can a per accidens cause be primary. The ultimate end cannot be a by-product.
Para. 1.123 refers to 1.113.
The gist of the argument of 1.131 is that the series of natures arranged in order of perfection cannot stretch upwards infinitely: there must be a best or most perfect nature.
In the argument of 1.132 the crucial premiss is: what can have an end is excelled in goodness and perfection by that end. Scotus is talking of an end for the sake of which the thing exists, not of some incidental goal. At some time today your goal may have been to get to some place at a certain time, but this is not the reason for which you exist, for the sake of which your productive cause brought you into being. To be at a certain place at a certain time is not something better and more perfect than a human being.
But the end for the sake of which a being exists, for the sake of which its productive cause brought it into being, is better than it, at least from the viewpoint of the productive cause. If you tidy your papers as a first step in writing an essay (and let's suppose you wouldn't do it otherwise), then you value the essay more than you value the tidiness of your papers.
Aristotle's universe is teleological: each being or kind of being exists for some purpose or end, and the end is better than the thing that exists for its sake. Now look again at 1.132. The best and most perfect being can't have an end, because then it would not be the best and most perfect being; and if it can't have an end, then it can't be produced by an efficient cause, because an efficient cause always products its effects for the sake of an end.
Compare 1.132 with 1.122. The corresponding argument about the first efficient cause, that it cannot be caused, was omitted (a third of the way down p.566, at the omission dots). It argued that the first efficient cause can't itself have an efficient cause; 1.122 argues that the first (or ultimate) final cause (or end) can't itself have either a final cause or an efficient cause (because an efficient cause acts in view of some end); 1.132 argues that the first being in the order of perfection and eminence can't be excelled by anything, so can't have a final cause, so can't have an efficient cause.
Now compare the three third conclusions: paragraphs 1.113, 1.123 and 1.133. The first contains the essential argument: If something is possible, then either it already exists actually of itself, or it can be produced by some efficient cause. The first in each of these three orders can't be produced by an efficient cause (this is the second conclusion in each case); therefore it must actually exist.
A comment: Scotus is talking about quiddity or nature. Scotus will later argue that there is only one individual of this nature, but in this paragraph he argues that there is just one highest nature, which is highest in each of the three orders; he is not yet arguing that there is just one individual. Suppose the nature of horse was the noblest nature, the final cause of the universe, and the efficient cause of all other animals and things: this would not mean that there was just one horse.
To my mind the first proof in this article needs a supplement to justify the statement "since nothing other than itself could be its end". Why not? In the modern critical edition this stage of the argument is expressed differently: "Now the first efficient cause does not act primarily or ultimately for the sake of anything distinct from itself"... (this is Wolter's translation in Scotus Philosophical Writings) The reason given is: "If it were to act per se for the sake of any end other than itself, then something would be more noble than the first efficient cause". But this reason supposes that the first efficient is also the most noble or most eminent, which is not yet proved--in fact it is the next point.
The argument for the second conclusion of this article also seems to have a hole. Why is an equivocal cause nobler than its effects? (I should explain what an equivocal cause is. He doesn't mean an equivocal term; we're talking here about things, cause and effect. An equivocal cause produces effects different from itself in quiddity or nature. Human parents producing children are univocal causes. God producing through animal parents the various sorts of animals is an equivocal cause.) Is an equivocal cause necessarily nobler than its effects because it is equivocal? Perhaps it is, but the point needs explaining.
Read 1.31, and then 1.32. Comment: "Nothing can be non-existing unless something positively or privatively incompatible with it can exist". I don't know what he means by positively and privatively; but why say that nothing can be non-existing unless something in some sense incompatible with it exists? This seems to invoke what A.O. Lovejoy in The Great Chain of Being calls the principle of plenitude, that everything possible exists unless there is some reason why not. But there is a less heroic assumption on which this argument might proceed: Remember that in 1.113, 1.123, 1.133 he proved the actual existence of an entity first in each of the three orders. So perhaps what he means here is that the non-existence of something that actually (already) exists is possible only if there can be something incompatible with its continued existence: a sort of principle of existential inertia. You and I already exist, but our non-existence at any future moment is possible because there are things existing that could kill us. Notice that a few lines down he refers to destruction, which presupposes prior existence.
The minor premiss is that there cannot exist ("cannot be" in the sense of cannot exist) anything incompossible (by compossible he means possible together with, at the same time) with what exists of itself, anything capable of destroying the existence of a thing that exists of itself. The argument for this minor premiss is a dilemma: either p or q: if p then...; if q then.... So the argument has two branches. Let's call "A" the being that exists from itself, and "B" a being incompossible with it. Either B also exists from itself, or it exists from another.
First branch: If B exists from itself and is incompossible with A (which also exists from itself), then (and now comes another dilemma--a dilemma within a dilemma) either two incompossibles coexist, which is impossible (since incompossible means not possible at the same time), or neither will exist; and if neither exists, then B does not exist, which is the point to be proved--that there cannot exist something incompatible with A, the thing that exists from itself. We can add that since it was proved earlier that the first being does actually exist, it is not the case that neither A nor B exist; so we must close off the first branch alternative, that B exists from itself.
Look now at the second branch of the dilemma. Suppose B exists from another, say C. Scotus says: "The objection is that no cause (in this case, C) can destroy any being (A) through repugnance of its effect (B) to that being (A) unless the cause (C) gives to its effect (B) a more perfect and intense existence than is the existence of the one to be destroyed (A)". So on this hypothesis, the being of B must be more perfect and intense than the being of A. What is wrong with that? He continues: "But no derivative being (B) has from its cause (C) a more noble existence than the existence of what is from itself (A), since everything caused (B) has dependent existence, but that which is from itself (A) has independent existence". There is an assumption here that perfec tion and intensity are a function of dependence/independence.
So that seems to be the argument spelt out fully. One feature of it that bothers me is that Scotus speaks of existence here as if it were a force, with degrees of intensity. This seems excessively metaphorical.
Now read 1.321. The argument here seems confused and unintelligible. Wolter's better text makes better sense, in Duns Scotus Philosophical Writings, p.51. Here is his translation:
If two necessary natures existed [call them alpha and beta], some reality proper to each would distinguish one from the other. Let us call these real differences A and B. [This is like the specific differences between species of the same genus. Remember that alpha and beta are natures; so alpha is like the genus "necessary being" as differentiated by A, and beta the genus "necessary being" as differentiated by B]. Now [here is a difference with Hyman and Walsh] either A and B are formally [i.e. essentially] necessary or they are not. [Again the argument is a dilemma. Let's look at the first branch.] If we assume them [i.e. A & B] to be necessary, then each necessarily existing nature [i.e. alpha and beta] will possess two formal reasons for its necessary existence, for in addition to A and B [which in this branch of the argument we are assuming to be necessary], each [of alpha and beta] is formally necessary by reason of that part of its nature in which it is like the other [i.e. the generic part "necessary being", to which A & B are added like specific differences]. [Wolter's version from this point agrees substantially with Hyman and Walsh]. Now this is impossible, for since neither of the two reasons of itself includes the other, if either be excluded the being would still exist necessarily in virtue of the other. In such a case the being would exist necessarily in virtue of something which, if eliminated, would still leave the nature existing as necessarily as before.
This is the impossibility. The argument seems plausible: If for either of two necessary reasons which are independent of one another something is so, then it is so for just one of those reasons, and the other is not, after all, necessary--if either reason by itself is sufficient they are not both necessary. However, there may be an equivocation here: if we say that something exists necessarily by virtue of some reason, the necessarily can go with "by virtue of", "exists necessarily by virtue of some reason", which makes the reason necessary; or the necessarily can go with exists--"exists necessarily, by virtue of this reason (or that reason or some other reason or some combination of reasons)". What if alpha exists necessarily by virtue of the combination of A with whatever is common to alpha and beta? Is this impossible?
Let's read on; we come to the second branch of the argument. Hyman and Walsh say "Indeed", etc. Wolter says:
On the other hand, if neither nature [alpha or beta] is formally necessary in virtue of these real differences, then the latter [A & B] are not of the essence of necessary existence and consequently neither is included in a necessary being. For any entity which is not of itself necessary being is only possible being. Nothing merely possible, however, is included in what exists necessarily.This is the end of Wolter's version. The differences with the Hyman and Walsh version based on the older edition may not be due to corruptions in the older text, but to the fact that Scotus was a great reviser and polisher: the better text may represent his own efforts to get his argument clear and cogent. Let's review this argument. In 1.1 Scotus argued for the actual existence of a first being in each of three orders: such a being is possible, and since it can't be caused it can only be possible because it actually exists. In 1.2 he argued that a first being in any of these three orders must be also a first being in each of the other two. In 1.3 he is arguing that there can only be one first being (or, at least, only one first nature). The argument so far is that since a first being can't be caused, it must be not only actual but also necessary. Now if there were two necessary natures something would have to distinguish them (two indistinguishable natures are just one nature). But the distinguishing differences can't both be necessary for necessary being, because each of the two supposed necessary natures manages to be necessary without one of them (namely the one that is distinctive of the other); and they can't both be unnecessary, because nothing unnecessary can be part of a necessary being. So both branches of the dilemma lead to a dead end. We must negate the original supposition, that there are two first natures: there can be only one.
Read 1.322. This argument rests on the authority of a text from Aristotle. To evaluate it properly we would have to read the relevant part of Aristotle's Metaphysics. Let's pass on.
Read 1.323. Scotus assumes that there must be one universe, one ordered whole. The word universe implies totality: two distinct universes, totalities, is a self-contradictory concept. But why must the totality form a single order of means to ends?
Next comes a set of omission dots. Go back to the top of p.565, and notice the distinction between absolute and relative properties. "Hence I shall first show the existence of certain relative properties of infinite being, namely primacy and causality". That is what we have been looking at so far, the first article which had three sub-articles (1.2 begins the second sub-article). The things discussed in this first article are certain relative properties of the highest nature, relative to other natures, namely its being their efficient and final cause, and its being nobler and more eminent than any other nature. Reading again from the top of p.565: "And second, from these [relative properties] I shall show the existence of infinite being, since these relative properties belong to infinite being alone". This is a quia argument, inferring the existence of something from its properties in relation to other things. Infinity, which is to be inferred in the second article, is an absolute property: see p.564, second new paragraph.
So the omission dots at the top of p.569 cover the beginning of article 2, in which the infinity of the necessary being is to be inferred from its relative properties of being first efficient, first final, and most eminent. The introductory paragraph of article 2 in Wolter's translation reads:
I proceed as follows: First I show that the first efficient cause is endowed with will and possesses such intelligence that this cause understands an infinity of distinct things, and that its essence, which indeed is its intelligence, represents an infinity of things. Second I go on from this to infer the infinity of this being.Hyman and Walsh omit the first (amounting to 10 pages in Wolter's edition), and resume at the second. We will pick this up in the next cassette.
This is the end of cassette 2.
With the first mailout you should have received a handout headed Scotus on the Existence of God. If you look at p.10 you will see the beginning of an appendix of extracts from Scotus's book De Primo Principio, On God as First Principle. Notice that the second paragraph begins "3.8". The translator has used decimal numbers to indicate the structure of the argument. In the eight page summary I've provided, whenever you see a decimal paragraph number you can find it in the appendix of extracts. Perhaps you should separate pp.10-13 and keep them on the table beside the summary as you read it. On p.2 of the summary, the fifth line, you'll see the paragraph reference "3.8", and there are others whenever there are matching extracts. The summary follows De Primo Principio.
Go back to 2.2, "These preliminaries having been shown". Let me tell you what these preliminaries are (in what would have been 2.1, which Hyman and Walsh omit): the first being has intellect and will; the knowledge and volition of this first being is identical with its essence; its intellect knows everything that can be known with a knowledge that is eternal, distinct, actual, necessary and by nature prior to the existence of the things known. Sub-article 2.1 second main article corresponds to the sections of the Summa theologiae in which Thomas Aquinas discusses the external operations of God (PHIL252 Readings, p.122-143). You will recall that Aristotle and the Arabian philosophers had seemed to deny, or had actually denied, that God has distinct and particular knowledge of things other than himself. Scotus, like Thomas Aquinas, and like the Islamic theologians, says that God has eternal distinct and particular knowledge of every actual and possible thing and event. This is argued for in the omitted ten pages.
Now read to the end of 2.21. A comment. The word "coloured" is like our expression "touched up"; Scotus needs to touch up Aristotle's principle. Aristotle held that the universe, including things other than the first cause, existed eternally; according to him, therefore, the movement caused by the first cause is infinite. Scotus does not hold the eternity of the universe, so he says that the first cause can, could, has power to, produce an infinite movement, whether he actually does or not. The fact that the first cause is not dependent on anything else implies that its productive ability is not limited. Scotus goes on in a discussion that occupies 6 pages in Wolter's translation, omitted by Hyman and Walsh, to elaborate and defend this argument.
Now read 2.22. That is clear enough, except for the word "enthymeme" near the end: This means a syllogism of which one of the premisses, or the conclusion, is left unexpressed: usually an argument with a tacit premiss.
Now read 2.23. The reference to Augustine gives a clue to the background. You may have heard Augustine's remarks in the Confessions: "You have made us for yourself, O Lord, and our hearts are restless till they rest in you". Augustine's Confessions records a restless search for happiness, in which every achievement or possession fails to satisfy. We all know the feeling. Behind Augustine is of course Plato: Plato's doctrine of eros, as in the Symposium and (especially) the Phaedrus. The human soul has a desire for the perfect. Scotus is saying here that no finite good satisfies. The first or ultimate end, whose actual and indeed necessary existence is supposed to have been proved in the first main article of the question, must be infinite.
Read 2.24. The second paragraph, "Another argument to the same effect" Scotus regards as a reformulation of the first. The Latin text more literally translated says: "To this conclusion it is argued in another way, and it is the same". But it is argued more elaborately. Notice that Scotus thinks that if there were any in compability between infinity and being we would be able to see the incompatibility without argument: "Just as contradictories contradict from their own characteristics, and this cannot be proved from anything more manifest [there is no way of proving that S is P and S is not P are contradictories and cannot both be true], so non-repugnance is from characteristics" etc. The argument of 2.24 might remind you of Anselm: Scotus is saying that the most eminent being must be infinite, because if it were finite it could be excelled--assuming that a being could be infinite, that infinity and being are not incompatible.
Near the end of the paragraph you've just read, at "I expound it popularly" write 2.241, three lines down at "The proposed position can also be persuasively supported" write 2.242. Now read to the end of 2.245. These are five short dialectical arguments in support of the proposition Scotus thinks is manifest and can't be proved from anything more manifest, namely that a being can be infinite, that infinity and being are not incompatible. At the end of 2.244 the reference is to 1.113--if an uncausable being is possible, it must be actual, because its possibility can't consist in being causable.
Now read 2.25. Some comments. "Coloured" again means "touched up", so as to make the argument work. Scotus thinks that Anselm needed to modify his definition of God: God is that than which, being thinkable without contradiction, nothing greater can be thought without contradiction. You can't prove the existence of something that is contradictory in its concept. Such a thing is not really thinkable, so Anselm could have said: God is the thinkable being than which no greater is thinkable. The point is perhaps the same as the one made by Gaunilon, PHIL252 Readings, p.29.
In the second paragraph of 2.25 notice the contrast between quidditative being and being of existence. Like Avicenna, Scotus thinks of possible beings as having already a kind of being, though not yet existence: the reality of a self-consistent nature. When a possible becomes actual, something (not a thing) that already had quidditative being acquires the being of existence. Behind this is Parmenides' doctrine that what is not cannot be thought: if we can think of possible beings they must have at least quidditative reality, the being of a possible nature.
In the third paragraph of 2.25 notice the distinction between intuitive and abstractive cognition. Abstractive knowledge, according to Scotus, is knowledge that leaves you uncertain whether the thing actually exists now--e.g. your present knowledge of someone you saw an hour ago does not guarantee that that person still exists.
Anselm's argument as retouched amounts to the argument of 2.24, in the second formulation. We must think of the most eminent being as infinite and as having the being of existence, because this is thinkable without contradiction, and if we think of this being as not infinite or not actually existing then we can think of another that is infinite and existing, and the so-called most eminent being would then not be most eminent, because this other being would be.
Now read the Solution to the whole question, in which Scotus sums up what he thinks he has proved.
Scotus re-wrote this long and complex argument in a little book called De Primo Principio, translated by A. Wolter, A Treatise on God as First Principle. It is in our library at B765.D73.D43.
Obviously there is a possible essay topic or two in Scotus's proof that an
infinite being exists. You would need to read the whole argument in Duns Scotus
Philosophical Writings, and also if possible, De Primo
Principio. You might look at the question: Does Scotus succeed in proving
that an infinite being exists? Or you might try to pry some part of it loose,
e.g. whether Anselm's argument as coloured by Scotus is successful? Another
possible topic might be found in the use made of a similar argument by Descartes
in Meditation V, and by Leibniz, in Leibniz, Philosophical
Papers and Letters, ed Loemker, vol.2, pp.634-5 and vol.1, p.354.
Go
to Medieval
Philosophy