Commentary
Commentary
Chapter 1 The Divisions and Types of Essential Orders 1.1(1)–(2) Opening prayer. Scotus's opening prayer places his own efforts in philosophical theology squarely within the tradition of St. Anselm, who began one of his own arguments for God's existence with the prayer, "Lord, you who grant understanding to faith, grant that, insofar as you know it is useful for me, I may understand that you exist as we believe you exist, and that you are what we believe you to be."1 Scotus and Anselm were in turn inspired by St. Augustine's method of addressing God in prayer as he sought
God in thought (e.g., Confessions 1.1.1).
I Am Who I Am. In the Bible, in the book of Exodus, Moses encounters a burning bush in the desert. God speaks to Moses through the bush, instruct ing him to liberate God's chosen people from slavery in Egypt. Moses asks for God's name, and God replies, "I am who I am" (Exodus 3:14). The English is a literal translation of the Latin, which Scotus directly quotes from the commonly used Latin Bible of the day, St. Jerome's Vulgate—which in turn accurately translates the original Hebrew. Scotus's invocation of this "blessed name" sets the stage for his own approach to demonstrating God's existence.
You are true existence. The precise meaning of the expression, "I am who I am," is elusive. But it is on the basis of this name that Scotus immediately addresses God as "True Existence" and "Total Existence." As Scotus under stood the divine name, it represents God as totally unlimited. Scotus in fact is the great philosopher of God's infinity, devoting a substantial portion of Chapter 4 to various proofs that God is infinite (4.67(46)–87(74)). God's infinity in turn helps us understand what Scotus means by "true existence" and "total existence." God is "true existence" not in a sense of "true" that implies all other existence is fake. Instead, as Scotus explains much later on, in 4.92(85), God is "true existence" in the sense that he is the ideal or stan dard against which everything else is measured. Our thoughts are true by conforming to reality; but God's thoughts are true on their own, and the rest of reality is true by conforming to God's thoughts. Likewise, God is "total existence" not in the sense that everything that exists is God or a part of God.
1. Anselm, Proslogion 2 (Williams, 81).
Commentary Instead, as Scotus also explains in 4.92(85), God's existence is "total" because
God is the exemplar of every possible way of being.
The being you have predicated of yourself. The conclusion of the opening prayer proposes a starting point for coming to know God: being. The justification
Scotus offers for the method is that God predicated being of himself: I am who I am.
1.2(3)–3(5) Introducing Essential Order 1.2(3) Properties of being. Having alighted on being as the starting point of his rational route to know God, Scotus then specifies that starting point more precisely. He claims that being has properties, that essential order is among these properties, and that essential order is a "more fruitful" way to proceed.
So the inquiry will focus not on being in general, but on the essential order of being.
The properties of being (passiones entis) are those properties or attributes that pertain to being as such. The most well known of these properties of being are goodness, truth, and unity. Properties like these are called transcen dentals, because they are found across all categories of being (e.g., substance, quality, quantity, etc.), and can pertain even to God, who does not belong to any category of being. On Scotus's understanding of transcendentals, a transcendental property of being is the sort of property that belongs to no genus—it is not "contained" under a genus, as Scotus puts it—and hence is the sort of property that can pertain to any being, just because it is a being— in Scotus's words, a transcendental property has no "predicate above it except for being."2 This requires some unpacking. Consider that for something to be blue, it must also be colored, and for something to be colored it must also be extended. Being colored here stands as genus with respect to being blue, and being an extended thing stands as genus with respect to being a colored thing. Or again, consider that for something to be human, it must also be an animal, and for something to be an animal it must also be a substance. Here being an animal stands as genus with respect to being human, and being a substance stands as genus with respect to being an animal. Then these two properties, being blue and being human, are "contained" under higher genera and so are not transcendental properties. By contrast, for something to be good, it need not be anything besides a being. Of course, for a human to be good it must be a human, and so an animal, etc. But for a thing to be good, it
2. Duns Scotus, Ordinatio 1.d.8.p.1.q.3.n.114 (Vatican 4:206).
Chapter 1 need not be contained under any genus. The same goes for all transcendental properties of being.
One of the most innovative features of Scotus's metaphysics is his doc trine of the disjunctive transcendentals. A disjunctive transcendental is a prop erty of being that comes in pairs of disjuncts, for example, finite or infinite, and is such that any being is one or the other: in this case, finite or infinite. A being is characterized by a disjunctive transcendental just in case it is char acterized by at least one of its disjuncts. Some disjunctive transcendentals are exclusive, such that if a being is characterized by one of the disjuncts, then it is not characterized by the other: finite and infinite, again, is a good example of just this sort of exclusive disjunctive transcendental. But other disjunctive transcendentals are inclusive. For example, cause or effect is a disjunctive tran scendental, but one and the same being may be both a cause of an effect, and an effect of some cause.3
Essential order is one of the disjunctive transcendentals. The two disjuncts of this disjunctive transcendental are essentially prior and essentially posterior.
(Scotus begins to say what he means by these disjuncts in 1.3(5)). Therefore, any being whatsoever is either essentially prior or essentially posterior. But essential order is one of the inclusive disjunctive transcendentals. One and the same being may be prior to one being but posterior to another being. In fact, Scotus thinks that most of the beings we know about are both prior and posterior. God, Scotus goes on to argue, is prior to all other beings in at least three types of priority; and there is nothing to which God is posterior.
Essential priority and posteriority. As the name suggests, essential order is an order due to essence. Not all orders are due to essence: for example, spatial and temporal ordering is rarely essential. If I am prior to you in a line (say, at a concert or theme park), or was born prior to you in time, then you are indeed posterior to me in these two ways. But these sorts of orders are not relevant to Scotus in the Treatise. What exactly essential order is becomes clearer as we go on.
1.2(4) Explication of a division. So far we have seen Scotus's focus move from being, to properties of being, to one particular (disjunctive) property of being: essential order. Now the focus gets even more fine-grained, zooming in on essential order in order to reveal several different types of essential order. In the remainder of Chapter 1, Scotus divides essential order four times.
Repugnance. Two things resulting from a division (the dividentia) are repug nant to each other, in the relevant technical sense of "repugnant," when they
3. Duns Scotus, Ordinatio 1.d.8.p.1.q.3.n.115 (Vatican 4:206–7).
Commentary are really different from each other, and one is neither included in nor follows from the definition of the other, and it is not possible for something to have or to be both dividentia. For example, the genus fruit divides into apples and bananas (etc.), which are really different from each other, and being an apple neither includes nor entails being a banana, and nothing can be an apple and a banana simultaneously.
Scotus uses this concept of repugnance frequently in the Treatise, where it has more general application. Two things (in the broadest possible sense of "thing") are repugnant when they are really different from each other, and one is neither included in nor follows from the definition of the other, and it is not possible for something to have or to be both.
1.3(5) Both posterior and prior are ordered. Before beginning these divisions, however, Scotus asserts that in essential orders, if there is an absolute first in some type of order, that absolute first is rightly said to be prior to whatever is posterior to it. At first glance, this looks like a trivial point: essential order clearly implies relations: the relations of priority and posteriority. However, it was common for medieval philosophical theologians to deny that God has any real relations to creatures. The basic reason for denying real relations of God to creatures is that relations are a kind of accidental property, but God has no accidental properties; whatever God is, he is essentially. Scotus is therefore anticipating an objection to his method, an objection that holds that essential order cannot really apply to God, because God cannot have a real relation of priority to anything posterior to God. Scotus is unmoved by the objection. He asks us to attend just to the "common sense" meanings of posterior and prior: if x is posterior to God, then God is prior to x.
1.4(6)–(8) First Division of Essential Order Eminence and Dependence 1.4(6) The first division of essential order yields the two orders of dependence and eminence. Of the seven total types of essential orders Scotus explicates, the first is the essential order of eminence and the remaining six are types of essential dependence.
1.4(7) Essential eminence: absolute ranking. Scotus's understanding of emi nence is meant to be an absolute ranking: not better or worse relative to some particular kind of standard, but better or worse absolutely speaking.
This absolute ordering would rank not just beings belonging to the same or similar kinds, but belonging to any kind whatsoever, such as humans and
Chapter 1 tarantulas, stars and giraffes. Scotus argues that beings really can be ordered in this absolute way, in 3.39(35)–(36).
1.4(8) Essential dependence: correcting Aristotle. Scotus subdivides the essential order of dependence three times over the remainder of the chapter. Before considering these divisions it is important to focus on an example Scotus draws from Aristotle to make his point. Aristotle implies that one thing,
A, depends on another thing, B, just in case B can exist without A but A cannot exist without B.4 Whatever Scotus might have thought about this as a description of some instances of dependence, he does not think that this is adequate as a definition of priority in the order of dependence, and still less of essential dependence.
Before giving his own definition of essential dependence, Scotus criti cizes Aristotle's definition of dependence. Suppose B necessarily causes A. In such a circumstance, B could not exist without A; nevertheless, B would not thereby depend on A for its existence—an effect depends on its cause and not vice versa. Thus, Aristotle's definition of dependence is inadequate.
Scotus then offers his own definition of essential dependence. If there is no contradiction between "B exists" and "A does not exist," but there is contradic tion between "A exists" and "B does not exist," then B is essentially prior to A in an order of dependence, and A is essentially posterior to B in the same order.
The full significance of Scotus's definition of essential dependence cannot be fully appreciated until 3.26(6), in which Scotus tells us that in his arguments for the First Principle he will reason from premises about what is possible, rather than what is actual. To foreshadow a bit more: Scotus will assert that if we abstract from actual existence and consider just possible beings, we can see that some possible beings can exist only if something else causes them to exist. From here he will mount an argument for the conclusion that there is a possible being that, if it exists, is not caused by anything—and go on to argue that this possible being exists in fact. Granting his reasoning, it follows that there really would be a contradiction in supposing that an essentially caused thing exists but that an essentially uncaused thing does not.
1.5(9)-(12) Second Division of Essential Order Causal and Non-Causal Dependence Causal and non-causal dependence. The second division is between types of essential dependence, causal and non-causal. Non-causal essential
4. Aristotle, Metaphysics 5.11.1019a1–4 (Barnes, 2:1609).
Commentary dependence itself divides into two, and this is the third division, explicitly discussed at 1.6(13). For clarity's sake I describe both types of non-causal essential dependence under the third division, immediately below.
1.6(13)-(14) Third Division of Essential Order Two Orders of Non-Causal Dependence Scotus discerns two ways in which beings may be non-causally essentially ordered in such a way that one is essentially dependent on the other: what we will call proximate non-causal dependence and remote non-causal dependence.
Proximate non-causal dependence. In proximate non-causal dependence, two effects, A and B, which are effects of one and the same cause C, are essentially ordered to each other if C can produce A only if C first produces B. Scotus offers a plausible example later on in 2.18(36): if an efficient cause produces a material substance that has a certain quality, say, being blue, that efficient cause could not make that substance blue unless it first made it a body hav ing some extension. This is because a thing must be extended in order to be blue. Therefore, even if both the extension and the color of the substance are dependent on one and the same cause, there is an essential ordering between the extension and the color: the color depends on the extension. Let the extension of the substance be B, and the color be A. Then we can say that A essentially depends on B, even though B is not the cause of A. Figure 1 visualizes this non-causal dependence:
C B A Figure 1 In the figure, the solid arrows represent causal relations, and the dotted arrow represents the proximate non-causal dependence relation. B is above A to rep resent the direction of the essential dependence: A essentially depends on B.
Remote non-causal dependence. In remote non-causal essential dependence, A essentially depends on B, and A and B have a common cause, D. But D produces A by first producing an intermediate effect C, which is also a cause of A. In this type of essential dependence, therefore, A is more remote from the common cause it shares with B. The idea here is that in order for D to
Chapter 1 produce A, it must produce both C, which is one of A's causes, along with B, which is not one of A's causes. In this case, A is remotely non-causally essentially dependent on B (whatever additional essential dependences it may have on C and D). Unfortunately, Scotus does not give any examples of this rather elaborate type of essential dependence. Figure 2 helps illustrate the schema Scotus has in mind:
D C B A Figure 2 Again, solid arrows represent causal dependence, whereas the dotted arrow represents the remote non-causal essential dependence of A on B.
B and C are above A in order to represent the fact that in order to produce A, D must first produce C and B—even though B is not a cause of A.
1.7(15) Fourth Division of Essential Order Four Orders of Causal Dependence The fourth division divides causal essential dependence into four types of causes: final, efficient, formal, and material. Scotus has a lot to say about these types of causes throughout Chapter 2, so I will not discuss them here.
1.8(16)–(17) Conclusion Harvesting the fruits. In this concluding section Scotus sums up the four divi sions and the several types of essential order that he has distinguished by means of these divisions (see Figure 3 below). Here he counts the two types of noncausal essential dependence as one type of essential order. One explanation of this conflation is that, as he himself admitted, it is not so easy to see how noncausal essential dependence really is divided up into these two distinct types.
Thus, while he technically distinguished seven total types of essential orders, here at the end of Chapter 1 he says he has distinguished just six.
Commentary
ESSENTIAL ORDER
DIVISION 1
Dependence Eminence
DIVISION 2
Causal Dependence Non-Causal Dependence
DIVISION 3
Proximate Non-Causal Dependence Remote Non-Causal Dependence
DIVISION 4
Final Causality Efficient Causality Formal Causality Material Causality Figure 3
Chapter 2 The Relations Between the Types of Essential Order 2.9(1) In this chapter Scotus proves various theorems about the types of essential order and the relationships between these types. Most of these the orems are used in Chapter 3 when Scotus finally turns his attention to argu ing for the existence of a nature that is first in the orders of efficiency, finality, and eminence.
2.9(2)-9(7) Description of Essential Order (Theorems 2.1-2.3) Scotus never offers, in this book or elsewhere, a definition of essential order.
Instead, in Chapter 1 he says that essential order is a disjunctive property of being whereby natures are ordered by priority and posteriority, and then goes on to distinguish seven types of essential orders. Then, here at the beginning of Chapter 2 he argues that all essential orders have three additional features: they are irreflexive (Theorem 2.1), non-circular and therefore asymmetrical
(Theorem 2.2), and transitive (Theorem 3.3). These do not define essential order, however, because other types of order have these properties but would not count as essential order. Consider, for example, temporal order (before and after), and some kinds of spatial order (e.g., in front of and behind). So it is best to understand Scotus here at the beginning of Chapter 2 as adding further description of essential orders, rather than a definition.
2.9(2)–(3) Theorem 2.1: Nothing whatsoever is essentially ordered to itself.
Scotus thinks this is self-evident. If something could be essentially ordered to itself, then it could be either more eminent than itself, or dependent on itself (or both). But the consequents are absurd, so nothing can be essentially ordered to itself.
2.9(4)–(5) Theorem 2.2: In any essential order a circle is impossible.
If essential order is irreflexive (by Theorem 2.1) then it is also non-circular.
Consider an essentially ordered triple: C, B, and A. If C is essentially prior to B and B is essentially prior to A, then C is essentially prior to both B and A.
Commentary But if C is identical with A, it follows that C is both essentially prior and essentially posterior to itself—which is contrary to the already established
Theorem 2.1. No matter how many additional things there are in an essen tially ordered series, nothing that is posterior to C can be prior to C in that same series.
2.9(6)–(7) Theorem 2.3: What is not posterior to the prior, is not posterior to the posterior.
The gist of the theorem is that essential orders are transitive. It is perhaps easier to understand through its contrapositive: what is posterior to the pos terior is posterior to the prior. Intuitively, if A is posterior to B, and B is posterior to C, then A is also posterior to C. Note that transitivity, along with non-circularity (asymmetry), only holds across priority and posteriority of the same type: for example, if C is an efficient cause of B and B is an efficient cause of A, then C is an efficient cause of A; but if C is, say, the material cause of B and B is an efficient cause of A, it does not follow that C is the material
(or efficient) cause of A.
2.10(8)-17(33) Fourth Division of Essential Order 2.10(8) Comparing essential orders. In the remainder of Chapter 2, Scotus proves various theorems about the relationships between different types of essential orders. He organizes these theorems according to the four divisions of Chapter 1, but follows a reverse order, starting with theorems of the fourth division and moving down.
The four causes. Theorems 2.4–2.9 concern the fourth division, that is, the four causal types of essential dependence. The four types of causes are the famous
Aristotelian causes: material, formal, efficient, and final.1
Here is a brief review of the four causes. Where there is some material object, there is (i) something it is made out of (material cause), (ii) its form
(formal cause), (iii) someone or something else that made it (efficient cause), and (iv) that for the sake of which it was brought into existence (final cause).
This pair of scissors, for example, is stainless steel (material cause) shaped into two handled blades fixed together (formal cause). Someone or some thing machined or forged these parts and put them together (efficient cause), and did so with a goal or end in mind (such as cutting), an end that moves or motivates the efficient cause to produce the scissors, and this end is the final cause of the scissors.
1. Aristotle, Physics 2.3.194b16–195a3 (Barnes, 1:332–33).
Chapter 2 2.11(9) Theorem 2.4: What is not ordered to an end, is not an effect.
Whatever has an efficient cause also has a final cause. Scotus gives three arguments for this theorem.
2.11(10) First proof. Everything produced by a per se efficient cause is pro duced for the sake of an end. Every effect is produced by a per se efficient cause. Therefore, every effect is produced for the sake of an end.
Per se and per accidens causes. In order to understand the argument, we need to know what Scotus means by a per se efficient cause, and we need to know why he thinks that every effect has at least one per se efficient cause. Scotus himself refers to Aristotle's Physics 2 twice in this proof, so we do well to look there. Adapting Aristotle's own example, consider a man who is a house builder, a musician, and pale-skinned. Then, of one of the houses he built, we could say correctly that (i) a housebuilder built the house, (ii) a musician built the house, and (iii) a pale-skinned thing built the house. (i)–(iii) are all true, but (i) is a much more sensible thing to say than (ii) and (iii): being a housebuilder is more relevant to building a house than being a musician or pale-skinned. There is nothing about musicianship or paleness that offer any explanation of how a house comes to be. Being musical or pale are, we might say, accidental to the housebuilder's ability to build houses. By contrast, housebuilding is the feature of the man that explains how he was able to build the house.
With all this in mind, we can see a little better what Scotus means when he talks about per se efficient causes. A per se efficient cause may be con trasted with a per accidens efficient cause. A housebuilder builds a house per se—that is, through what housebuilding itself is—while a musician and a pale-skinned thing build a house per accidens—that is, because they are qualities attached to someone who also happens to be a housebuilder. The housebuilder therefore is a per se efficient cause of the house, while the musi cian and the pale-skinned thing are per accidens efficient causes of the house.
By now it should not be difficult to see why Scotus thinks every effect has a per se efficient cause. A good efficient causal explanation of the existence of some caused thing will pick out something that has the right sort of charac ter to be the efficient cause of that effect, and the right sort of character will have something to do with the power to bring about effects of just that kind.
Acting for ends. Scotus also claims that every effect of a per se efficient cause is produced for the sake of an end. And from this it follows that every effect of a per se efficient cause has a final cause, since the end of a thing (in a strict sense of "end" distinguished at 2.12(16)-(17)) just is that thing's final cause.
So why does Scotus think that every effect of a per se efficient cause is pro duced for the sake of an end? He does not offer a clear answer, but I suspect
Commentary he thought it was self-evident—that is, true by the meaning of the relevant terms—that a per se efficient cause acts for the sake of an end. In order to state what a per se efficient cause is doing, at the moment of its causing, we must state the end, or goal, or purpose of the activity—without such an end
(e.g., the finished house), the activity (e.g., housebuilding) is unintelligible.
2.11(11) Second proof. An end is a final cause precisely insofar as it "met aphorically" moves an efficient cause to produce its effect for the sake of that end. Every effect has at least one efficient cause that is "metaphorically moved" by an end. Therefore every effect has a final cause.
Scotus has a distinctive understanding of the nature of final causation, which he elaborates in the two corollaries to Theorem 2.5 (discussed below).
On this understanding, the final cause is prior to what it causes only insofar as it is an end that is loved by some agent that, out of its love for that end, efficiently causes its effect. An agent's love for an end explains the sense in which an agent is "metaphorically moved" by an end: we often talk about desire or love moving or motivating us to act, but since our acts involve a will, this motion is not strictly speaking causal (the way in which a locomotive engine moves its train of cars or the way in which fire heats up the water in the pot); hence the motion is metaphorical. As Scotus explains in the second corollary to Theorem 2.5 (2.12(18)), the final cause is the final cause of the effect of the efficient cause, and is not any kind of cause of the efficient cause's causing of its effect.
2.11(12)–(13) Third proof. Every efficient cause produces its effect for a rea son. The reason an efficient cause produces its effect is the final cause of its effect. Therefore every effect has a final cause.
2.12(14) Theorem 2.5: What is not an effect, is not ordered to an end.
Whereas Theorem 2.4 holds that whatever has an efficient cause has a final cause, this theorem says that whatever has a final cause has an efficient cause.
2.12(15) Proof. Here is a heavily reworked version of the proof, offered for clarity's sake. By definition, the end (finis, telos) of a thing is the final cause of that thing. When a cause and the thing it causes are essentially ordered, the cause is essentially prior to what it causes. Therefore the final cause of a thing is essentially prior to that thing. And therefore the end of a thing is essentially prior to that thing. But the end of a thing is essentially prior to that thing only if the end is loved by something that efficiently causes that thing for that end. Therefore the end of a thing is loved by something that efficiently causes that thing for that end. Therefore everything that has an end has an efficient cause.
Chapter 2 2.12(16) First corollary: on final causes. Here Scotus rejects what he considers to be a "false opinion" about the nature of the end and final cause. This opin ion holds that the final cause of a thing is its natural end, that is, whatever it is that is achieved through the properly functioning, characteristic activity of a thing. For example, if all goes well for the acorn, it becomes an oak tree.
The mature living oak is thus, on the view Scotus rejects, the final cause of the acorn.
The problem with this view is that the end of the acorn is achieved too late to be a final cause. Remember, for Scotus, every final cause is essentially prior to what it causes. Since the mature oak is temporally posterior to the acorn, it cannot be essentially prior to the acorn.
Assuming that this really is as problematic as Scotus thought it was, there are at least three choices: reject final causes altogether, or grant that some thing temporally posterior to x could indeed be essentially prior to x, or come up with some alternative way of understanding final causes, according to which they are essentially prior to what they cause. Scotus takes this third option.
His solution involves distinguishing two senses of "end." Strictly speaking, the end is the final cause, and so is essentially prior to its effect. But more loosely speaking, the end is the state at which the characteristic activity of a thing is aimed. In the loose sense of end, therefore, the mature oak tree is the end of the acorn, but (according to Scotus) it does not follow from this that the mature oak is also the final cause (that is, the end in the strict sense of "end") of the acorn.
An end, in the strict sense, can be a final cause only if it is essentially prior to what it causes. And this fact seems to have inspired the most innovative aspect of Scotus's understanding of the final cause: an end, insofar as it is loved, can be essentially prior to the thing that has that end. Imagine a car penter with a table in mind, who, before he makes the table, has a desire to make the table. In this case, the table as loved by the carpenter exists before the table. Therefore an end, insofar as it is loved, can be a final cause.
From this view of final causes it follows that the only things that can be finally caused are things with at least one efficient cause that is personal, that is, capable of thinking and loving, and so loving the end for the sake of which it brings about its effect. This personal efficient cause might not be the most immediate efficient cause of some effect, but somewhere along the essentially ordered chain of efficient causes that bring about the effect, there has got to be at least one personal efficient cause.
It also follows from this view of final causes that the final cause, strictly speaking, is whatever end the efficient cause loves such that, for the sake of that end, the efficient cause produces its effect. This end need not be the
Commentary same as the loose-sense end of a thing, that is, the "last operation" of a thing or what its natural activity aims to achieve. Scotus recognizes this possibility.
2.12(17) Aristotle and eternal dependence. Scotus then offers some interpre tation of Aristotle. Aristotle thought that the celestial intelligences exist eternally and have a final cause (the First Mover). Some think it follows from the intelligences' eternal existence that they have no efficient cause.
But according to Theorem 2.5, everything that has a final cause also has an efficient cause. So if these other interpreters of Aristotle are correct, then
Scotus contradicts Aristotle. But Scotus does not understand his own view to be in contradiction with Aristotle's. According to Scotus, something can exist eternally but still be efficiently caused. From his reading of Avicenna,
Scotus distinguished between two sorts of priority: priority according to time, and priority according to essence or nature. In many instances of effi cient causation, the efficient cause exists temporally prior to its effect. Since the celestial intelligences exist eternally, clearly they do not have an efficient cause that is temporally prior to them. But on Scotus's understanding of essential order, when an efficient cause and its effect are essentially ordered, the efficient cause's causal activity is simultaneous with its effect. So imagine an efficient cause that is eternally causing an effect that is essentially ordered to it. Then its effect eternally exists and exists simultaneously with the causal activity of its efficient cause. In such a case, the efficient cause is essentially prior to its effect, even though it is not temporally prior. And in just this way,
Scotus thinks, Aristotle held (or perhaps would have held) that the First Mover is both the final and the efficient cause of the celestial intelligences.
The paragraph about what Aristotle held, or would have held, ends with an important distinction about the possible: the possible is that which is opposed to the impossible, but also that which is opposed to what is neces sary from itself. For Avicenna (and Scotus who followed him) some things are necessary from another, and such things are not opposed to the possible.2
2.12(18) Second corollary. Here Scotus clarifies that when an end as loved is a final cause, it is the final cause of the efficient cause's effect, and not the final cause either of the efficient cause itself or the efficient cause's causal activity.
Scotus elaborates on this in the proofs of Theorem 2.9 at 2.17(33), below.
2.13(19) Theorem 2.6: What is not an effect, is not made of matter.
This theorem says that whatever is not an effect of an efficient cause, is not made of matter—or, to state the contrapositive, whatever is made of matter
2. Avicenna, The Metaphysics of "The Healing" 1.6.1–10 (Marmura, 29–33).
Chapter 2 is an effect. This implies that there are no uncaused material things. Scotus offers three proofs.
Before explicating these arguments, note that this theorem does not say or imply that matter itself is not or cannot be uncaused. Instead the focus is on things made of matter. If it is possible for matter to exist all on its own, without any form at all, then as far as this theorem goes, matter could exist and not be the effect of any efficient cause. But later on, at 2.16(32), Scotus clarifies that matter itself is caused.
2.13(20) First Proof. Everything made of matter is composed of matter and form. Matter has "contradictory potency," that is, it can make a composition with any form whatsoever. But matter cannot cause itself to make a compo sition with any form. Therefore if matter composes something with form, something besides matter is the cause of that composition. A form itself can only be the formal cause once matter and form are united; therefore some thing besides form is the cause of that composition. Therefore there is an efficient cause of every composition of matter and form.
The argument clearly relies on Scotus's understanding of hylomorphism, that is, the view that material objects are composites of matter and form.
The key premise is that matter cannot cause itself to make a composition with any form. This is due to the nature of matter's potency to form: it is
"contradictory potency," that is, considered in itself, it can be united with any form, but likewise considered in itself, there is no more reason why it should be united with one kind of form than another. Thus, if it is united with a form, there is nothing about the nature of matter that explains why there is this particular composition of matter and form. So something besides mat ter must explain this. Form cannot explain it, either. So Scotus reaches for some other cause, an efficient cause, which unites matter and form into a composite. If you wonder why he picks the efficient cause instead of the final cause, note that he has already shown both that anything which has an efficient cause has a final cause, and that anything which has a final cause has an efficient cause. So either way he gets his conclusion—everything made of matter has an efficient cause.
Consequence. In this section Scotus uses the word "consequence" (consequen tia) for the first of many times. As he uses the term, "consequence" refers to a relation between sentences or propositions, namely, that one follows logically from the other. It is, we might say, the following-from relation between a sentence that functions as a premise and another sentence that functions as a conclusion. Do not confuse "consequence" with "consequent" (consequens):
"consequent" refers to a statement in a conditional sentence (the part that follows the "then" in an "if/then" sentence)—or perhaps, more generally,
Commentary a consequent (consequens) is simply the conclusion of any consequence
(consequentia).
2.13(21) Second proof. The argument can only make sense if we suppose there is an implicit assumption that whatever is made of matter has a final cause.
Then we can reason that whatever has a final cause has an efficient cause (by Theorem 2.5). And for anything that has a final cause and an efficient cause, any other causes it has are posterior to its final cause and efficient cause.
Anything made of matter has a material cause. Therefore anything made of matter has an efficient cause.
Scotus offers no support here for his key premise that if something has a final cause and an efficient cause, then any other causes it has are posterior to its final cause and efficient cause. This premise follows from the second proof of Theorem 2.4 (on the priority of the final cause to the efficient cause) and the first proof of Theorem 2.6 (on the priority of the efficient cause to the material and formal causes), but then the second proof is not independent from the first proof of this theorem.
2.13(22) Third proof. Anything made of matter is a composite of matter and form. A composite of matter and form is itself one entity that is not reducible to its matter and form—in colloquial terms, the whole is greater than the sum of its parts. Anything that is one entity yet composed of several parts, is caused to be one by something besides its parts. Matter and form, as parts of the composite, cannot cause the composite to be one. Therefore the cause of the composite's being one is neither its material cause (matter) nor formal cause (form). Therefore the cause of the composite's being one is at least one of its extrinsic causes: its efficient cause or its final cause. Anything that has a final cause also has an efficient cause (Theorem 2.5). Therefore anything made of matter has an efficient cause.
Extrinsic and intrinsic causes. Scotus here uses the expression "extrinsic cause" for the first time. Efficient and final causes are "extrinsic" because they are not parts of the thing they cause, whereas material and formal causes are
"intrinsic" because they are parts of the thing they cause.
2.14(23) Theorem 2.7: What is not composed of matter, is not composed of form; and what is not composed of form, is not composed of matter.
Essential parts. Scotus offers one proof for this theorem, according to which whatever is composed of matter is also composed of form, and vice versa. In the proof, the key idea is essential part. Whereas "integral" parts are what we might think of as ordinary parts of a composite such as the paws, fur, and teeth of a mole, "essential" parts are the matter and form(s) of a composite,
Chapter 2 in this case prime matter and the various forms that compose the substance of the mole.
2.14(24) Proof. By its nature, matter is a potency to receive form and thereby to be the matter of a composite substance. So for some actually existing matter, if it is a part of a composite then there is at least one form that is the actualization of matter's potency. By some miracle, matter might exist all on its own without form, and form might exist all on its own without matter, but in these cases matter and form would not be parts of a composite. Thus, wherever one is part of a composite, the composite has the other as a part too.
2.14(25) To supplement the proof, Scotus here invokes an intuitive principle about the nature of parts and wholes, namely, that there is only a genuine or proper part of a whole when there are two or more parts of the whole. Noth ing can be composed of just one thing. So if matter composes a substance, then that substance has some other part. The sort of part that, with matter, composes a substance, is form.
When Scotus appeals to Theorem 2.1 in support of his claim that some thing existing from just one element would not really be an element, I suggest that he understands being an element to imply being a part; then, if a part is essentially ordered to the whole, and by Theorem 2.1 nothing can be essen tially ordered to itself, then something existing from an element—that is, being nothing more than an element—would not in fact be an element at all.
2.15(26) Theorem 2.8: What is not caused by extrinsic causes, is not caused by intrinsic causes.
The theorem says that if something lacks a final cause and an efficient cause, then it also lacks a material cause and a formal cause. The theorem follows from the conjunction of Theorems 2.4, 2.6, and 2.7, since by 2.6, whatever lacks an efficient cause lacks a material cause, and by 2.4, whatever lacks an efficient cause lacks a final cause, and by 2.7, whatever lacks a material cause also lacks a formal cause. This is why Scotus says that Theorem 2.8 is "clear enough from the preceding four theorems." Nevertheless, he adds two addi tional proofs.
2.15(27) First proof. It is probably not possible to make sense out of this proof. In the first two sentences Scotus says that the causality of the extrin sic causes is more perfect than the intrinsic causes. In the second sentence
Scotus says that the extrinsic causes are prior to intrinsic causes in their caus ing, as the perfect is prior to the imperfect. Now the difficulty here is that the several uses both of perfection and causality make it unclear whether Scotus is relying on the notion of essential order of eminence or essential order of
Commentary dependence. If dependence, then it is hard to see how this first proof differs from the second proof immediately following it. But, if eminence, then it does not seem that the argument can succeed.
2.15(28) Second proof. Intrinsic causes are essentially posterior to extrinsic causes in the order of dependence. Thus, whatever has intrinsic causes has extrinsic causes. Therefore, whatever lacks extrinsic causes must also lack intrinsic causes.
2.16(29) Theorem 2.9: The four types of causes are essentially ordered in their causing of one and the same thing.
2.16(30) The theorem holds that the four causes are essentially ordered in their causing of one and the same thing. It can be proved from Theorems 2.4–
2.8 and their proofs, in several different ways. Here is one way: by Theorem 2.7, whatever has a material cause also has a formal cause; by Theorem 2.6, whatever has a material cause also has an efficient cause; and by Theorem 2.4, whatever has an efficient cause has a final cause. Thus, every material thing is caused by all four types of causes. In addition to this general appeal to earlier theorems, Scotus offer two proofs.
First proof. In the first, Scotus reasons from the unity, or oneness, of the thing caused by all four causes, to some principle of unity among the four causes, insofar as they jointly cause whatever they cause. Recall Scotus's third proof of Theorem 2.6, at 2.13(22), which maintained that whatever is made of matter and form must have an efficient cause, on the grounds that matter and form alone are insufficiently unified to be the cause of the unity of the com posite. Here Scotus tells us that all four causes must be sufficiently unified in order to cause an effect that is itself unified. If several things produce one thing, the several things must be unified in some way, and the sort of unity
Scotus has in mind for the four causes is the unity that things have insofar as they are essentially ordered. Here and elsewhere (3.36(26)) Scotus says that it is in this sense of unity that the entire universe is one thing.
2.16(31) Second proof. Here Scotus relies on three earlier theorems of the fourth division: by the second proof of Theorem 2.4 (2.11(11)), the final cause is the first among the causes in their causing; by the second proof of Theorem 2.6 (2.13(21)), the efficient cause is second among these causes; and by Theorem 2.8, the material and formal causes are posterior to the effi cient cause.
Chapter 2 2.16(32) Next, Scotus tells us that he is not interested in determining the right ordering of the material and formal causes in this unity of essential order that the four causes have in their causing. Instead, it suffices simply that the material and formal causes are posterior to the final and efficient causes. Nevertheless, he offers some brief remarks, speculating that matter is prior to form in a type of non-causal essential order of dependence, while form is prior to matter in the essential order of eminence.
2.17(33) Finally, Scotus offers a very important distinction. It is one thing, he tells us, for causes to be essentially ordered in their causing, and another thing for the things that are causes to be essentially ordered. When Scotus says that the four causes are essentially ordered in their causing, he does not wish to imply that the final cause is a cause of the existence of the efficient cause, or again that the efficient cause is a cause of the existence of the mate rial and formal causes. Instead, what he means is that the efficient cause depends for its causality on the causality of the final cause, and again that the material and formal causes depend for their causality on the causality of the efficient cause.
2.18(34) Third Division of Essential Order The third division divides the non-causal essential order of dependence into two types: proximate and remote. See 1.5(11)–6(14). Scotus has nothing much to add here.
2.18(34) Theorem 2.10: When two things are related to the same cause, that cause is either proximate or remote.
This theorem, offered without proofs, asserts that when two things are essentially ordered to the same cause, where neither of these two is a cause of the other, then one of the two is more proximate to the cause than the other. Intuitively, relative proximity or remoteness is determined by how many links in a causal chain separate a cause from what it causes. For exam ple, imagine a cause that initiates two independent causal chains, 1 and 2.
In Chain 1, let's say there are three links: D1, C1, and B1; and in Chain 2 there are four links, D2, C2, B2, and A2. Figure 4 below makes these chains easier to visualize:
Commentary CAUSE D2 D1 Chain 1 C2 C1 Chain 2 B2 B1 A2 Figure 4 In the figure, B1 is more remote than C1, C1 is more proximate than
B1, and D1 is more proximate than both C1 and B1. Likewise, D1 is more proximate than C2, C2 more proximate than B1, and so on.
Scotus would agree with this intuitive understanding of proximity and remoteness, according to which relative proximity is simply a matter of counting links in a chain. But when it comes to essential orders, he rec ognizes an additional way in which one thing can be proximate to a cause relative to another thing. In non-causal essential orders of dependence, one thing essentially depends on another thing, but is not caused by that thing.
Suppose, for example, that D2 essentially depends on D1 but is not caused by D1. As far as counting links goes, D2 is neither more nor less proximate than D1. Nevertheless, in this case, Scotus thinks that D1 is more proxi mate to the cause than D2, precisely because D2 essentially (non-causally) depends on D1. Figure 5 illustrates this in the following way:
CAUSE D1 D2 Figure 5
Chapter 2 Here the dotted arrow represents the essential (non-causal) dependence of D2 on D1.
2.18(35)-19(40) Second Division of Essential Order The second division divides the essential order of dependence into two: causal and non-causal essential orders of dependence. See 1.5(9)–(12). Scotus includes two theorems under this division.
2.18(35) Theorem 2.11: It is not the case that each thing [B] more proxi mately caused by a cause [C] is itself a cause of a more remote thing [A] caused by that same cause [C]. Therefore, it [B] is something caused prior
[to A], but it is not causally prior [to A].
This theorem says that two things can be more or less proximate to their common cause, without themselves being causally related. Or, more pre cisely: where two things, A and B, are caused by the same cause, C, and where B is more proximate to C than A (and A more remote from C than B), it does not for that reason obtain that B is a cause of A. Therefore, B might be caused prior to A, but need not for that reason be a cause of A.
Consider Figure 4 again. There, D1 is not a cause of C2, yet D1 is more proximate to the cause that is common to D1 and C2.
2.18(36) Scotus offers as an example a quality's dependence on quantity. For commentary on this example, see the entry under 1.6(13)-(14) above. Con sider Figure 6:
CAUSE QUANTITY QUALITY Figure 6 Here quantity and quality are represented as immediately caused by one common cause. The dependence of quality on quantity is represented by the dotted arrow.
Scotus begins to introduce a proof of the theorem ("This is proved also by reason"), but the actual proof has been omitted—by Scotus himself or by a later scribe, we do not know. But see 1.5(12), not exactly for a proof of the Theorem but for an elaboration of what he has in mind.
Commentary 2.19(37) Theorem 2.12: A thing essentially depends only on a cause or on something caused that is more proximate to some cause.
Scotus tells us that this theorem is advanced in order to show the sufficiency of the second division. In this division he has divided essential dependence into just two types: causal and non-causal. The division is "sufficient" only if there can be no more types of essential dependence than these two. As we have seen in the third and fourth divisions, the two types of essential depen dence distinguished in the second division themselves divide into further types. The point, however, is that if the second division is sufficient, any type of essential order of dependence will be either causal or non-causal.
Put this way, the answer is obvious, since "non-causal" is simply the nega tion of "causal." Yet these are my words, not Scotus's. In the theorem itself
Scotus puts it like this: a thing essentially depends only on a cause or on something caused more proximately by a cause. By "something caused more proximately by some cause," Scotus has in mind just the sort of essential dependence discussed under Theorem 2.11 above, namely, a sort of essential dep endence that obtains when one thing non-causally depends on another, and when both share a common cause (however remote). To show the sufficiency of the division, then, he needs to show that there cannot be essential depen dence where the posterior non-causally depends on the prior but the prior and posterior have no cause in common.
2.19(38) Proof. The goal of the proof is to show that in every instance of non-causal essential dependence the dependent thing and that on which it depends share a common cause. (To avoid confusion, it should be said that while in his earlier use of "A" and "B" at 2.6(14), A depended on B, here at 2.19(38) and following, B depends on A; I have followed Scotus's shift in the commentary.) We can explicate it as follows. Suppose for reductio that B essentially depends on something, call it A, which is neither a cause of B nor is a more proximate effect of one of B's causes. If such a case were possi ble, A would be in a causal order totally independent of B, and yet B would nevertheless essentially depend on A. Thus, if A does not exist then B does not exist. Now, suppose A does not exist. In such a circumstance, everything in B's causal order—both B's causes and the things more proximate to B's causes on which B essentially depends—would together be the sufficient cause of everything immediately prior to B, yet B would not exist, because A does not exist. But, Scotus asserts, whatever is sufficient to cause the essen tially prior is sufficient to cause the essentially posterior. Therefore B would exist, even if A does not exist.
The argument might be faulted for begging the question, specifically when it holds that B's causes and the things more proximate to B's causes on which B essentially depends are sufficient to cause B. After all, this seems
Chapter 2 to be precisely what is disputed by anyone inclined to reject the theorem.
Whatever the merits of the argument, the gist seems to be that there is simply no need to posit anything beyond B's causes and the things more proximate to B's causes on which B essentially depends. To do so is theoret ically wasteful.
2.19(39) First objection. Here Scotus imagines someone objecting that if A itself could be caused by B's causes, then B could depend on A.
Reply. But Scotus points out that ex hypothesi, A is something independent of the order of B's causes. Thus the objection is futile.
2.19(40) Second objection. Scotus considers a second objection. Natural effi cient causes do not cause prime matter to exist. Yet the composite effect of the efficient cause is made of matter and so essentially depends on matter. So here it seems we have something on which a composite essentially depends but which is not in the order of causes that cause the composite.
Reply. Scotus responds that while some particular efficient cause might not produce prime matter, some more remote efficient cause does produce prime matter. Moreover, prime matter is not a good candidate for A, because prime matter is caused and is therefore part of the total network of causes that brings about B, and also (though Scotus doesn't say this), prime matter itself is a (material) cause of the composite (that is, the substance composed of matter and form) and so is one of its causes.
2.20(41)-23(50) First Division of Essential Order Theorems 2.13–2.16 finish off the chapter and concern the first division, that is, the first division of essential order into the orders of dependence and eminence.
2.20(41) Theorem 2.13: Not everything that is excelled essentially depends on that which is more eminent than it. Therefore the first member of the first division does not entail the second member.
This theorem says that being posterior to something in the order of emi nence does not entail being posterior to it in the order of dependence.
2.20(42) Proof. Inductively, many things are evidently independent of what is nobler than they are. For example, arguably, a human is nobler than a grasshopper, but grasshoppers do not essentially depend on humans either for their existence or their activity.
Commentary 2.20(43) Theorem 2.14: Not every dependent thing is excelled by that on which it depends.
2.20(44) This theorem says that being posterior to something in the order of dependence does not entail being posterior to it in the order of eminence.
Proof. The key example here in the proof is that a composite substance is prior in eminence to its matter, but is posterior to its matter in the order of dependence—it depends on it as its material cause.
2.21(45) Theorem 2.15: Plurality is never to be posited without necessity.
This theorem is simply an assertion of the common Aristotelian methodol ogy that plurality is not to be posited without necessity.3
2.21(46) As he makes clear, Scotus posits this theorem in order to show the sufficiency of the first division: he discerns no need to divide essential order beyond the orders of eminence and dependence. By the same token, he claims to see no need to posit more than six total types of essential order
(one of eminence, one of non-causal dependence, and four of dependence).
This claim is confusing, since the official tally is seven: one of eminence, two of non-causal dependence, and four of dependence.
2.22(47) Theorem 2.16: Everything ordered to an end is excelled.
Theorems 2.13 and 2.14 maintained the mutual non-entailment of eminence as such and dependence as such. But in Theorem 2.16 Scotus claims that in one type of essential order of dependence, namely final causality, there is an entailment: whatever has a final cause has something prior to it in the order of eminence. Scotus offers two proofs for the theorem, then considers an important objection.
2.22(48) First Proof. Scotus claims that the end, call it A, is prior in eminence to that which is for the sake of the end, call it B. Then he offers a proof for this claim. He does this in a roundabout way, by proving that A is neither less good than B nor equally as good as B. Suppose B were equally as good as A.
Then B would be equally as lovable and desirable as A. But then an efficient cause would have no more reason to produce B for A's sake than to produce
B for B's sake. But, Scotus alleges, this would violate Theorem 2.1: since B would be the final cause of itself. Similar reasoning establishes that A is not less good than B. Therefore A is better than B, that is, prior in the order of eminence.
3. Aristotle, Posterior Analytics 1.25 (86a34–36).
Chapter 2 2.22(49) Second proof. Here Scotus reasons by way of a double analogy: as the end is to that which is for the sake of an end, so the principle of artisanal knowledge (e.g., the relevant knowledge and skill) is to works of art, and so the premises of a practical syllogism (e.g., the relevant principles of practical reasoning) are to the conclusion. Since the principle of artisanal knowledge is better than works of art, and since the premises of a practical syllogism are truer than its conclusion, so the end is better than that which is produced for the sake of the end.
2.23(50) Objection. The objection is aimed at the theorem itself and not to either proof in particular. It reasons from bad actions. Frequently we bring something about for the sake of a bad end. In such cases it's easy to see how the end can be less good than that which is brought about for its sake.
Suppose I create a musical masterpiece, but do so entirely to make my rival jealous. Then the end of my action (making my rival jealous), is less noble than that which I bring about for the sake of that end (the masterpiece).
First reply. Scotus's reply relies on the notion that final causes, like efficient causes, are transitive. An individual bad agent acts for the sake of ends that, in some real sense, are under its control (e.g., it is my fault that I acted for the sake of making my friend jealous), but no individual agent except God has the final say about what additional, higher, ends any action is ordered to. Scotus seems to be saying that my bad end does not offer a genuine counterexample to the theorem, provided that there is some higher end that moves some higher efficient cause to bring about my musical composition for the sake of that higher end. For example, while I might have an evil end, God (let's say) intends me to compose the masterpiece for the sake of making many people happy. Thus, despite my bad end, there is some end (in this case, God's) which is nobler than that which is brought about for the sake of the end.
Second reply. Scotus considers, but then rejects, a second sort of response to the objection. According to this response, the inferior end is only the end for the sake of which the efficient cause acts in a qualified way. There is some additional end for the sake of which the efficient cause acts, and it is more fundamental to an overall explanation of what the efficient cause is doing.
Yet Scotus rejects this response, because he sees no reason to suppose that, in every case, an agent acting for a bad end is really acting for a good end.
Chapter 3 On the Triple Primacy of the First Principle 3.24(2) This opening prayer announces Scotus's goal to show by reason what he believes by faith: that there is a unique, actually existing nature that has
Triple Primacy: it is first in the essential orders of efficient causality, final causality, and eminence, where "first" here means posterior to none and prior to all others, if there are others.
3.24(3) In Theorem 3.6, Scotus argues that there can be only one nature that is first in the essential order of efficiency. In Theorem 3.15 he proves that the unique first in the order of efficiency is the same as the unique first in the order of finality, which in turn is the same as the unique first in the order of eminence. In other words, one and the same nature is first with respect to all three. But it is not until the end of Chapter 4, in Theorem 4.11, that Scotus argues that God alone has this nature (by identity with it).
Each primacy argument follows the same basic structure: for each of the three essential orders, (i) establish the possibility of a first in each order, then
(ii) establish the actuality of a first in each order.
The arguments for a first in efficiency and a first in finality are not arguments for a first efficient cause and a first final cause. To be a first cause (in either order), something else must exist. Therefore being the first cause of everything else is the sort of feature a thing can have only if other things exist. "First cause" therefore does not belong to the divine nature as such—even though only the divine nature could be the first cause. But being the first in the essential orders of efficiency and finality does not require that there be anything else but the first. In Theorem 3.2, Scotus tells us exactly what he means by "first" in the essential order of efficiency: a first "cannot be produced, and can produce its effect without the power of anything else." Similar formulations hold, mutatis mutandis, for a first in the orders of finality (Theorem 3.8) and eminence (Theorem 3.12).
Commentary 3.25(4)-37(26) The Primacy of Efficiency 3.25(4) Theorem 3.1: Among beings, some nature can produce an effect.
By "beings" and "nature," Scotus does not here mean actually existing things. He explains what he does mean two paragraphs below, at 3.26(6).
For now, suffice it to say that he means something like possible beings. So here in the theorem he is saying that at least one possible being can pro duce an effect. But of course, he does not mean that, as a merely possible being, it can produce an actual effect. Nor does he mean that, as a merely possible being, it can produce only a possible effect. Instead, he means that there is at least one possible being, or nature, such that, if actual, it can produce an effect. The shorthand way of expressing this sort of complicated modal claim is simply to ascribe properties to natures, properties like being able to produce an effect.
3.25(5) Proof. Something can be produced. Therefore, something can pro duce. The consequence—that is, that the conclusion follows from the prem ise—is established by the nature of the "correlatives," producible (effectibile) and able to produce (effectiva). Note that something still can produce an effect even if it never in fact produces an effect. Similarly, something could have, necessarily and by its very nature, the ability to produce an effect, while its actual production of an effect (if it produces an effect) could be totally con tingent. Eventually we will see that Scotus thinks that the first in the order of efficiency is just such a nature: it is by nature the first effectiva—the first thing able to produce an effect—and it actually exists necessarily, but it is totally contingent whether it in fact produces any effects; thus it would have been first in the essential order of efficiency even if it had never produced any effects.
Scotus then offers two arguments in support of the claim that "something can be produced," which is the "antecedent" of his one-premise proof.
First argument for the antecedent. Something is contingent. What it is to be contingent is to be able to exist after not existing. Nothing can produce itself
(by Theorem 2.1). Therefore whatever can exist after not existing, if it exists, is produced by another. Therefore whatever is contingent is producible. Note here that for Scotus "contingent" does not mean or imply actually existing. It just means being able to exist after not existing. The argument fails if "contin gent" means or implies actually existing. Scotus is here talking about possible beings, not actual beings.
Second argument for the antecedent. Something, x, can lack a perfection it can have. Therefore x can undergo change (i.e., the acquisition of the perfection x lacks). Therefore the thing resulting from the change can begin to exist (i.e.,
Chapter 3 the acquired perfection or, better, x so perfected). Whatever can begin to exist, if it exists, is produced by another. Therefore whatever can undergo change is producible.
3.26(6) From premises about the possible. Here Scotus makes explicit an argu mentative strategy he has already been using. He will argue, he says, from premises about the possible rather than the actual. He gives an example of an argument that reasons from a premise about the actual: something is effected
(effecta); therefore something is effective (efficiens). This is a type of argument familiar to readers of Aquinas' First and Second Ways: each begins with a fact that is "evident to the senses," and so actual: respectively, some things undergo change and some things begin to exist.1 Scotus does not criticize this type of argument. Instead, he offers three reasons why he prefers to argue from premises about the possible. (i) Possibility is inferred from actuality, but not actuality from possibility. (ii) Premises about actual things are contingent while premises about possible things are necessary. (iii) Premises about actual things are about existent being, while those about possible things are about being as such. I'll say something about each of these in turn.
(i) If Scotus can establish his own arguments for God's existence from necessary premises, then his efforts supplement, do not compete with, and are logically independent of, other arguments for God's existence that reason from premises about actual things.
(ii) Aristotle had made the necessity of the premises a criterion for suc cessful demonstration.2 The knowledge we can derive through syllogistic reasoning stands on a firmer basis if the premises are necessary rather than contingent. Also, and I think more fundamentally, by reasoning in this way
Scotus will show, if he is successful, that being first in the orders of efficiency, finality, and eminence, does not at all depend on there being any contingently existing things. God has this Triple Primacy just by what God is, and not at all by what God has done.
(iii) My speculation here is that this reason is important to Scotus because it makes the search for the First Principle truly an endeavor in metaphysics, which he says is the study of being qua being.3 That he thinks of the Treatise as metaphysics is clear toward the end of Chapter 4, where he explicitly says that his book has reasoned only from metaphysical claims (4.93(86)). Why metaphysics should matter so much to Scotus, in his efforts to prove God's
1. Aquinas, Summa theologiae 1a.q.2.a.6.corp. (Hause and Pasnau, 53).
2. Aristotle, Prior Analytics 1.12.32a6–14 (Barnes, 1:51).
3. Duns Scotus, Questions on the "Metaphysics" of Aristotle 1.q.1.n.160–61 (Etzkorn and Wolter, 1:58–59).
Commentary existence, may have something to do with Aristotle's own assessment of the superiority of metaphysics (the "divine science") to other sciences.4
3.27(7) Theorem 3.2: Something able to produce an effect is simply first— that is, it cannot be produced, and it can produce its effect without the power of anything else.
The theorem states that there is a first in the order of efficiency, and clarifies what being first means: it cannot be produced, and it can produce its effect without dependence on any other efficient cause. This clarification is import ant: at this point in the book, Scotus is not assuming that if there is a first in the order of efficiency then there is just one nature that is first. (He argues that there is just one in Theorem 3.17.) Also, here we are still in the possi bility phase of the argument: Scotus is not trying to establish the existence of such a nature, but rather that such a nature is possible, that is, that there is nothing incoherent, or "repugnant," about the conceptual components involved in being first in the order of efficiency. If the proof of Theorem 3.2 is successful, it establishes that there is in fact a nature the total intelligible content of which includes being first in the order of efficiency, but it does not establish that there is in fact an actual thing that has that nature.
Scotus offers one proof of the theorem, then offers an extended defense of one its key premises, namely, that there cannot be an ascending infinite series of essentially ordered causes.
3.27(8) Proof. Something—call it A—can produce an effect (by Theorem 3.1). A is either first or not first. If A is first, then the theorem is proved. If A is not first, then A essentially depends (in the order of efficiency) on some thing else—call it B—for A's existence or activity. B is either first or not first.
If B is first then the theorem is proved. If B is not first, then B essentially depends (in the order of efficiency) on something else. And so on. There cannot be an ascending infinite series of things able to produce an effect.
There cannot be a circle of things able to produce an effect (by Theorem 2.2).
Therefore there is a first in the essential order of efficiency.
3.28(9) Objection. The objection asserts that an ascending infinite series of causes is indeed possible. By "ascending infinite series" Scotus means a series such that for every member of the series there is another member of the series prior to it. A descending infinite series would be a series such that for every member of the series there is another member of the series posterior to 4. Aristotle, Metaphysics 1.2.982a3–982b10 (Barnes, 2:1553–54); Duns Scotus, Questions on the "Metaphysics" of Aristotle 1.Prol.n.16 (Etzkorn and Wolter, 1:6–7).
Chapter 3 it. (There is nothing here to suggest that Scotus would deny the possibility of a "descending" infinite series, that is, a series such that for every member of the series there is another member of the series posterior to it.)
Note, in keeping with Scotus's strategy of arguing from premises about the possible, that what is here at issue, strictly speaking, is not whether there can be an ascending infinite series of causes. Remember, he is prescinding from existence and considering the nature of being as such. So instead, at issue is whether there can be an ascending infinite series of things able to produce an effect, such that something able to produce an effect essentially depends on something able to produce an effect, which itself essentially depends on something able to produce an effect, and so on. That said, Scotus writes somewhat loosely in his response to the objection, framing the response in the objector's own term (cause) rather than Scotus's own (thing able to pro duce an effect). As he says in the first proof of Theorem 3.4, below (3.33(19)), all the arguments he gives here against the possibility of an infinite series of essentially ordered causes can be understood as concerning "nature, essence, and possibility."
3.28(10)-(11) Essentially ordered causal series. Now Scotus in fact thinks that there is a kind of causal series, a series ordered per accidens, which can have an ascending infinity, but that this kind of series is not at issue in the proof for Theorem 3.2, which denies that there can be an ascending infinite series of causes ordered essentially (or per se).
Before arguing against the possibility of an ascending infinite series of essentially ordered causes, he offers three characteristics of an essentially ordered series of causes—to which he will refer when arguing against the possibility of an infinity in such a series.
First feature of essentially ordered causal series. An essential order of causes orders the causality—the very activity of causing—of the causes so ordered.
It does not order their existence. Thus, if A, B, and C are three causes essen tially ordered, and if A is prior to B and B prior to C, then A is a cause of B's causing, not B's existence, and A and B are causes of C's causing, not C's exis tence, and so on, down to the effect produced by the whole series of causes.
Second feature of essentially ordered causal series. In an essential order of causes, the prior cause is both different in nature from and more eminent than, the pos terior cause. It is different in nature because no cause essentially depends for its causing on a cause of the same nature as itself. The point makes some sense if we understand it very narrowly (always a good hermeneutic while interpret ing Scotus): two substances of the same specific nature may well be essentially ordered in their causing, as man and woman are essentially ordered in their
Commentary production of a child,5 but of course the causality of the man is of a different nature from the causality of the woman. It remains a mystery to me why Sco tus also thinks that in an essential order of causes, the prior is more eminent than the posterior, especially in light of Theorem 2.14, which says that not every essentially dependent thing is excelled by that on which it depends.
Third feature of essentially ordered causal series. In an essential order of causes, the prior's causing of the posterior's causing is simultaneous with the poste rior's causing. That is, C's causing, B's causing of C's causing, and A's causing of B's causing of C's causing, all occur simultaneously. This is because Scotus really does mean that the prior causes the causing of the posterior. He doesn't mean that the prior causes the posterior to be able to cause its effect (say, later on). Thus, the only moment available for the prior to cause the posterior's causing is the very moment of the posterior's causing.
3.29(12) Against ascending infinite series of essentially ordered causes. To estab lish the crucial premise that there cannot be an ascending infinite series of things able to produce an effect, Scotus argues for three closely related claims, which he labels A, B, and C: (A) An infinity of essentially ordered causes is impossible. (B) An infinity of accidentally ordered causes is impos sible unless an essentially ordered series of causes comes to an end (i.e., has a first prior member). (C) If there is no such thing as an essentially ordered series of causes, an infinite series of causes is impossible.
3.29(13) Scotus offers five proofs of (A).
First proof of (A): An infinity of essentially ordered causes is impossible. The total ity (universitas) of essentially ordered caused things is itself caused; so it is caused by something that is not a member of that totality.
Scotus does not offer any support for the premise that the totality of essentially ordered caused things is itself caused. And in general it does not follow merely from the fact that all the parts of a whole have a property, F, that the whole has F too. Some readers might fault Scotus here for commit ting a fallacy of composition—that is, the fallacy of inferring that a whole has a property that all of its parts have, from the fact that all its parts have that property. But this would be hasty. He does not tell us why he supports the premise. In any case, it is not always unreasonable to infer a property of the whole from a property of the parts. (Every part of Michelangelo's David is made of marble, and so is David, and, plausibly, David is made of marble because every part of David is.) 5. Duns Scotus, Ordinatio 1.d.3.p.3.q.2.n.496–98 (Van den Bercken, 225–26).
Chapter 3 Another puzzle about this argument is that it is not at all obvious how it is supposed to support (A). This is because, at first glance, nothing in the argument rules out the possibility of an ascending infinite series of essen tially ordered causes. Here is why. Suppose, contra Scotus, that there is an ascending infinite series of essentially ordered causes. But also suppose, with Scotus, that the totality of (infinitely many) essentially ordered causes is itself caused. Then it follows that there is an uncaused cause of an infinite series of essentially ordered causes. Then the argument might help to establish that there is something that is first in the essential order of efficient causality, but it would not rule out the possibility of an infinite series of essentially ordered causes—as Scotus clearly intended it to do.
But I think that closer inspection reveals something plausible about how Scotus thought that this argument supports the conclusion that there can't be an ascending infinite series of essentially ordered causes. First, consider what it must mean for an uncaused cause outside the totality of caused things to cause that totality: the cause outside the totality causes that totality by imme diately causing at least one member of the totality. But it does not immediately cause every member of the totality, because then the totality would not be a causal series. Now let's make a leap, in a spirit of theoretical parsimony (Theo rem 2.15): suppose that the uncaused cause outside the totality of essentially ordered caused causes, causes that totality by immediately causing only one member of that totality. Then there is one member of that totality that has an uncaused cause. But then this one member would not have a caused cause.
Then this one member is the first caused cause, and its cause is the first cause.
But then the order of caused causes is not infinite (in an ascending order).
It seems to me that this extrapolation helps us see how this first proof of (A) is supposed to rule out the possibility of an infinite series of essentially ordered causes. Assuming the totality of caused things is itself caused, then the totality is not an infinite series. But it does not give us any reason to support the crucial premise that the totality of caused things is itself caused. I won der if Scotus thought that that premise is obviously true and not in need of argument—the sort of claim that one simply sees to be true whether or not one has a proof, as one sees that 2+2=4, or that a surface cannot be white and black simultaneously.
Second proof of (A). If an ascending infinite series of essentially ordered causes is possible, then an actual infinity of essentially ordered causes is possible.
An actual infinity of essentially ordered causes is impossible. Therefore an ascending infinite series of essentially ordered causes is impossible.
The antecedent of the first premise holds, as Scotus says, from the fact that in essentially ordered causal series, every member of the series causes simultaneously. So if there are infinitely many members of the series, then at
Commentary one and the same moment there is an actually infinite number of essentially ordered causes. Scotus follows Aristotle6 in rejecting the possibility of an actual infinity—except in the mind of God (as evidenced by his discussion of divine ideas in 4.68(48)-69(50), and at the end of 4.91(85)).
Third proof of (A). If something is prior in an essentially ordered series of causes, then it is closer to a beginning. If there is an infinite series of essen tially ordered causes, then there is no beginning. Assume for reductio that there is an infinite series of essentially ordered causes. Therefore, there is no beginning. If there is no beginning, nothing in the series is closer to a beginning than anything else in the series. Therefore nothing is prior in an essentially ordered series of causes. But something is prior. Therefore there is not an infinite series of essentially ordered causes.
This argument appears to beg the question in its first premise. The goal of the argument is to show that an essentially ordered series of causes implies a first in the series, that is, a beginning. It therefore seems illicit to assert that priority in such a series implies being closer to a beginning. I leave it to readers to find a more favorable reading of this third proof of (A), if one can be found.
Fourth proof of (A). The higher up a member in an ascending series of essen tially ordered causes, the more perfect it is in its causing (by the second fea ture of essentially ordered causal series offered at 3.28(11)). Therefore the infinitely higher in such a series is infinitely perfect in its causing. If some thing is infinitely perfect in its causing, then it causes whatever it immediately causes independent of any other cause's causing. Therefore it is an uncau sed cause. Therefore there is not an infinite series of essentially ordered causes.
The gist of the argument is this: an infinitely perfect nature would be infinitely perfect in its causing, and since infinite perfection in causing implies being able to cause without being caused, it follows that an infinitely perfect nature would be a first in an essential causal order. Since there is no repugnance between infinite and perfect and cause (assumed here but more or less argued for in the fifth proof of infinity given at 4.78(63)-79(65)), it follows that there is a nature infinitely perfect in its causing and so first.
Here Scotus does not offer support for the premises that the infinitely higher in such a series is infinitely perfect in its causing, and that if some thing is infinitely perfect in its causing, then it causes whatever it imme diately causes independent of any other cause's causing. But see Scotus's seventh argument for divine infinity, developed at length in 4.81(67)–85(72).
Fifth proof of (A). If being able to produce an effect can belong to a nature perfectly, then the nature to which it so belongs can produce an effect without 6. Aristotle, Physics 3.5.204a20–206a8 (Barnes 1:348–51).
Chapter 3 dependence on a prior cause. Being able to produce an effect can belong to a nature perfectly. Therefore the nature to which it so belongs can produce an effect without dependence on a prior cause.
The main idea here is that perfect ability to produce an effect is the sort of property that is possible for some nature to have. For Scotus this means that there is some nature—some possible being—that has that property. And having this ability, Scotus thinks, entails the ability to produce an effect inde pendent of any other cause—and so to be first in the order of efficiency.
But, strictly speaking, if such a thing actually exists, its existence alone does not rule out the possibility of an ascending infinite essentially ordered series: for suppose that being with perfect ability to produce an effect in fact produces nothing, but other causes, which depend for their causing on other causes, do produce effects. Then there could be a causally lazy being with the ability to produce effects uncaused by other causes, existing side by side with causally active beings all depending on other causally active beings for their causing. Scotus doesn't address this possibility. But see the proof of Theorem 3.19 (3.46(58)) for an implicit reply to this concern.
3.30(14) Proof of (B): An infinity of accidentally ordered causes is impossible unless an essentially ordered series of causes comes to an end (i.e., has a first prior member).
The gist of the argument is that a cause in a series ordered per accidens itself depends for its causing on a causal series ordered essentially. In a causal series ordered per accidens, a cause does not depend, for its causing, on its cause(s)
(i.e., its cause(s) in its per accidens causal series). But this does not mean that its causing is completely uncaused. Instead, each cause in a per accidens causal series is, at the moment of its causing, dependent on an essentially ordered causal series. So while a per accidens series might ascend (or descend) to infin ity, each and every causal link in the series is a terminus of an essentially ordered series originating in an uncaused cause (since, by the arguments for (A), there cannot be an infinite ascending series of essentially ordered causes).
For example, a family tree represents a per accidens causal series: grandpar ents cause parents who cause children, but parents do not depend on grand parents at the moment they cause their children. Nevertheless, parents do depend on other causes at the moment they cause their children; the whole causal series on which they depend at that very moment is essentially ordered, and terminates in a first. Thus, while a per accidens causal series may well extend infinitely up and down, it can only do so if the causal activity of each member of the series is linked up with a finite essentially ordered causal series.
3.31(15) Proof of (C): If there is no such thing as an essentially ordered series of causes, an infinite series of causes is impossible. Here Scotus argues that an infinite causal series ordered per accidens is possible only if an essentially
Commentary ordered causal series is possible. The argument begins with Theorem 3.1 as a premise: some nature (call it N) can produce an effect. Suppose there are no essentially ordered causal series. Then, when an individual with nature N
(call it an N-individual) causes its effect, it does not depend for its causing on something else's causing. An N-individual is either caused to have N, or not caused to have N. If there is an N-individual that is not caused to have N, then it has its ability to produce its effect independent of any cause giving it that ability—therefore there is no infinite causal series through which it has N. If there is an N-individual that is caused to have N, then it is dependent on at least one cause for its having N. Then the whole series of causes on which it depends for having that nature, is itself either finite or infinite. But it is not infinite, because there cannot be an infinite series of causes ordered per accidens if there is no finite series of causes ordered essentially (by the argument for (B), above), and ex hypothesi there are no essentially ordered causal series. Therefore the series ordered per accidens is finite. Therefore, on the denial of essentially ordered causal series, there can be no infinite causal series.
The upshot of (A), (B), and (C) taken together is that if there are essen tially ordered causal series then they are finite and so have a first member, and if there are no essentially ordered causal series but only accidentally ordered causal series, then the latter are finite and so have a first member.
3.32(16) Theorem 3.3: That which can produce an effect in a way that is simply first is uncausable, because it cannot be the effect of anything else and it can produce an effect independent of anything else.
The theorem states that a nature that is first in the order of efficiency— that is, a nature that cannot be produced and can produce its effect without depending on anything else—is uncausable. As Scotus notes in the short proof, the theorem follows from Theorem 3.2.
3.32(17) The theorem is not trivial, as it might at first appear, because The orem 3.2 states merely that what is first in the order of efficiency cannot be produced, that is, cannot be the effect of an efficient cause. But Theorem 3.3 adds that a nature first in the order of efficiency is not causable in any way: it has no final cause (by Theorem 2.5), no formal cause (by Theorem 2.7), and no material cause (by Theorem 2.8).
3.33(18) Theorem 3.4: Something able to produce an effect in a way that is simply first exists in actuality, and some actually existing nature can produce an effect in this way.
Here we move from possibility to actuality.
Chapter 3 3.33(19) Proof.
1. For any nature, if it can exist then either it can exist by another or it can exist of itself.
2. A nature first in the order of efficiency can exist (by Theorem 3.2).
3. Being first in the order of efficiency is repugnant with being able to exist by another (by Theorem 3.3).
4. Therefore, a nature first in the order of efficiency can exist of itself (by premises 1–3).
5. A nature first in the order of efficiency cannot cause itself to exist (by Theorem 2.1).
6. A nature first in the order of efficiency cannot be caused to exist by another (by premise 3).
7. Therefore, if such a nature does not in fact exist, it cannot exist (by premises 5 and 6).
8. But it can exist (by premise 2).
9. Therefore a nature first in the order of efficiency actually exists (by premises 7 and 8).
Suppose that a nature first in the order of efficiency does not exist. Since nothing can cause itself to exist, and since a first in the order of efficiency cannot be efficiently caused to exist, nothing can bring it about that the nature in fact exists. So it can't exist. But it was supposed to have been shown that such a nature can exist, in Theorem 3.2. So we can derive a contradiction that such a nature both is and is not possible. Given the contradiction, we have a choice: accept with Scotus that such a nature actually exists, or deny contra Scotus that such a nature is possible.
3.33(20) Corollary. The first in the order of efficiency does not just happen to exist and be prior to everything else in the order of efficiency; it is contradic tory that anything should be prior to it in the order of efficiency.
3.34(21) Theorem 3.5: What is uncausable exists necessarily from itself.
Here we move from actuality to necessity.
Specifically, Scotus argues that an uncausable possible nature has neces sary existence "from itself." Necessary existence "from itself " is to be distin guished from necessary existence "by another." For Avicenna, from whom
Scotus gets the distinction, a being is necessary by another when it cannot fail to exist, but is caused to exist (by a cause that could not fail to cause it and could not fail to cause what it causes).7 Everything necessary in this sense
7. Avicenna, The Metaphysics of "The Healing" 1.6.4–6 (Marmura, 31–32).
Commentary ultimately depends on something necessary from itself, that is, something that could not fail to exist and is uncaused. Scotus himself denies that there is anything that has necessary existence by another, because he denies that any creature exists necessarily.
3.34(22) Proof. If nothing incompatible with the first in the order of effi ciency can exist, then the first in the order of efficiency cannot fail to exist.
Nothing incompatible with the first in the order of efficiency can exist.
Therefore the first in the order of efficiency cannot fail to exist. Therefore it is necessary—and not by another, but from itself.
Scotus defends the antecedent in roughly the following way. Suppose that something incompatible with a nature first in the order of efficiency can exist. Then, if it exists in fact, it exists either of itself, or by another. But not of itself, since the first in the order of efficiency exists of itself and is incom patible with it. But not by another, since no caused thing is powerful enough to cause, by its coming into existence, the non-existence of the uncausable.
So nothing incompatible with the first in the order of efficiency can exist.
Therefore it cannot fail to exist.
3.35(23) Theorem 3.6: Existing necessarily from itself pertains to only one nature.
Here we move from necessity to uniqueness.
Having established that there is at least one nature that is first in the order of efficiency and that exists necessarily, Scotus now argues that there is and can only be one nature that is first in the order of efficiency. He offers three main arguments, all of which rely on Theorem 3.5, that the first in the order of efficiency has necessary existence from itself.
3.35(24) First proof. If more than one nature has necessary existence (from itself, not by another), then necessary existence is a common (i.e., universal, shareable) nature. Then, in each necessarily existing nature, there will be at least two essential features: the shareable nature they have in common and in virtue of which they have necessary existence, and some distinguishing features (formalitates) in virtue of which they are distinguished from each other. Two impossibilities follow from this:
The two impossibilities. Both impossibilities involve reasoning about neces sary existence as a shareable property (or common nature, as Scotus puts it). If necessary existence is a property shared by (at least) two individuals, then necessary existence is insufficient for something with necessary existence to exist in fact; its necessary existence must be further specified by its dis tinguishing features. But then—the first impossibility—necessary existence
Chapter 3 would not be the most actual feature of a thing; instead, necessary existence would be in potentiality to distinguishing features. And also—the second impossibility—it contradicts the meaning of necessary existence that it be insufficient on its own for some necessarily existent thing to exist.
3.35(25) Second proof. Here Scotus assumes that for any two natures that share something in common, one is more eminent than another. But if two natures had necessary existence in common, it would be insufficient to make one better than the other (since they both have it). So some other property would make the better nature better than the other. But nothing can make something better than necessary existence can make it. So necessary exis tence cannot be a shareable nature.
3.36(26) Third proof. If two natures shared necessary existence, neither would essentially depend on the other. Therefore neither would be essentially ordered to the other. But the unity of the universe consists in the essential order of things in the universe. So one of the two natures with necessary existence would not be a part of the universe—but there is no need to posit more than one universe, as Scotus argues below in 3.37(26).
Objection to the third proof. There could be an essential order of eminence between them, and this would be sufficient for both to be parts of the universe.
First reply. Nothing is more perfect than necessary existence. Therefore two natures with necessary existence would not be essentially ordered with respect to eminence. See also 3.34(25) above.
Second reply. If there were two natures each of which is a first in the order of efficiency, then what is immediately essentially posterior to one is either
(i) immediately essentially posterior to both, or (ii) immediately essentially posterior just to that one. If (i), then any subsequent essentially ordered pos teriors will also be essentially ordered to both firsts. Then there will be a redundancy of essential order throughout the universe. If (ii), then either the second first has nothing essentially posterior to it, or it does and it composes a second universe with whatever is essentially posterior to it.
3.37(26) Second reply continued: against both (i) and (ii). There is no theo retical need to postulate two firsts, or two universes, or redundant essential orders in one and the same universe. So we should assume there is just one first and just one universe with just one order.
Commentary 38(27)-38(34) The Primacy of Finality The arguments for a first in the order of finality (Theorems 3.7-3.10) are extremely compressed, relying (as Scotus tells us) on much of the work done to establish a first in the order of efficiency (Theorems 3.1-3.6). In the argu ments for efficiency, the key relata were effectiva (can produce an effect) and effectibile (can be produced). Here, the key relata are finitiva (can be an end) and finibile (can be ordered to an end).
3.38(27) Theorem 3.7: There is some nature among beings such that it can be an end.
Some nature among beings can be an end. Note that here we begin with possible beings, just as in Theorem 3.1. To be ordered to an end is to be the causatum—the finitum—of a final cause. And to be an end is to be the causa finalis itself—provided we understand being an end in the strict sense that Scotus discusses in Theorem 2.5. Therefore to be able to be ordered to an end is to be finibile, that is, able to be a finitum; and to be able to be an end (in the strict sense) is to be finitiva, that is, able to be a final cause.
3.38(28) Proof. Something can be produced (by the proof of Theorem 3.1).
Therefore, something can be ordered to an end (by Theorem 2.4—whatever has an efficient cause has a final cause). In the text, Scotus does not actually draw the conclusion, which is implied: since something can be ordered to an end, therefore something can be an end—as the theorem states.
3.38(29) Theorem 3.8: Something that can be an end is simply first, that is, it cannot be ordered to some other end, and not by anything else's power is it the end of other things.
Some nature is first among natures that can be an end. Here, too, as in Theorem 3.2, Scotus clarifies just what being first means: it does not mean that there is some actually existing thing that is in fact the first final cause.
Instead, to be first in the essential order of finality means to be a nature that cannot be ordered to an end, and that is independent of anything else for its being the end of other things.
3.38(30) Scotus tells us the theorem is proved by five arguments similar to those offered for Theorem 3.2. He is most likely referring to the five argu ments offered in support of proposition (A) in 3.29(12); (A) says that an ascending infinity of essentially ordered causes is impossible. These argu ments were offered in turn in order to support the proof of Theorem 3.2. It is reasonable to suppose that the proof itself is implied when Scotus here tells us to look to the five arguments of (A). Briefly, the five arguments for (A) are:
Chapter 3 (i) from the need for a cause outside the totality of caused things; (ii) from the impossibility of a simultaneous actual infinity; (iii) from the conceptual connection between priority and a beginning; (iv) from infinite causal per fection; (v) from the possibility of independent causing. Each argument can be formulated as an argument against the possibility of an ascending infinite series of essentially ordered final causes, and each argument, mutatis mutan dis, would have the same strengths (or weaknesses, as the case may be) of the original versions.
3.38(31) Theorem 3.9: The first among beings that can be an end, is itself uncausable.
What is first in the essential order of finality is uncausable.
3.38(32) What is first in the essential order of finality has no final cause, by the meaning of first. It has no efficient cause, by Theorem 2.4. Also, see
Theorem 3.3 and its proof. Note, as in Theorem 3.3, we are still in the possi bility stage of Scotus's argument for the actual existence of a first in the order of finality.
3.38(33) Theorem 3.10: The first among beings that can be an end, actually exists, and this primacy belongs to some actually existing nature.
Here we move from possibility to actuality.
3.38(34) See the proof and the corollary of Theorem 3.4.
Scotus does not provide additional theorems that take us from actual ity to necessity and from necessity to uniqueness, as he did in the theorems about the order of efficiency. Plausibly, however, it is implied that suitably modified versions of those Theorems and their proofs establish the necessity and uniqueness of the nature that is first in the order of finality. In any case,
Theorem 3.15, which establishes the sameness of the first in efficiency with the first in finality and the first in eminence, makes additional theorems and proofs unnecessary.
3.39(35)-39(42) The Primacy of Eminence 3.39(35) Theorem 3.11: Among the natures of beings there is something that excels.
By now the strategy should be familiar. The theorem states that among (pos sible) beings, something excels or exceeds (excedens). This is the analogue in the order of eminence of effectiva in the order of efficiency, and finitiva in the order of finality.
Commentary 3.39(36) Proof. The correlative of excelling (excedens) is being excelled
(excessa). That some nature is excelled follows from Theorems 3.7 and 2.16.
3.39(37) Theorem 3.12: Some eminent nature is simply first in perfection.
Here, again, we argue from the fact of priority within an essential order (i.e., something can produce, something can be an end, something excels), to a first in that order, all the while remaining in the possibility stage—considering natures as such, prescinding from actual existence.
3.39(38) Proof. Scotus again refers back to Theorem 3.2 and the five argu ments against the possibility of an ascending infinite series of essentially ordered causes. But since the order of eminence is not a causal order, the first of these five arguments is not relevant. The fourth and fifth arguments can be made relevant with some heavy modifications.
Recall that the fourth argument reasoned that an infinitely perfect nature would be infinitely perfect in its causing, and since infinite perfection in causing implies being able to cause without being caused, it follows that an infinitely perfect nature would be first in an essential causal order. Modified to show that there is a most eminent nature, the argument could go like this: an infinitely perfect nature implies being unsurpassed in eminence and sur passing all other natures in eminence. Therefore an infinitely perfect nature is first in the order of eminence. Both the original fourth argument and the modified version rely, it seems to me, on an implied premise, namely that infinite and perfect are non-repugnant, such that there is an infinitely perfect nature.
The fifth argument is rather similar to the fourth, both in its original and modified versions. The original argument asserted that being able to produce an effect does not imply any imperfection, and thus some nature has this ability perfectly. But perfect ability to produce an effect implies being inde pendent of any other cause. Therefore that nature in which being able to produce an effect is perfectly realized is a nature that is first in the order of efficiency. To modify this argument for eminence, we start with the following premise, which is practically trivial: being eminent does not imply any imper fection. Thus, some nature has eminence perfectly. But perfect eminence implies being unsurpassed in eminence and surpassing all others. Therefore the nature that is perfectly eminent is first in the order of eminence.
For additional support, Scotus claims that forms are related to each other as numbers are related to each other, citing Aristotle's Metaphysics 8 as the source of this comparison.8 I have not been able to make a judgment about 8. Aristotle, Metaphysics 8.3.1043b33–1044a14 (Barnes, 2:1648).
Chapter 3 what Scotus hoped to illustrate through the analogy. From what he says else where, larger numbers depend on smaller numbers, and all numbers depend on the number one. So while numbers "ascend" to infinity, they do have
(literally) a first. Yet he also says that the number one is less perfect than larger numbers.9
3.39(39)-(40) Theorem 3.13: The Supreme Nature is uncausable.
Remaining in the possibility stage, Scotus argues that the first in the order of eminence is uncausable.
3.39(40) First proof. Scotus argues exclusively from previously established theorems that the nature that is first in the order of eminence is uncausable.
Since, by Theorem 2.16, an end is better than that which is for the sake of an end, the first in eminence cannot be for the sake of an end, and so can not have a final cause. But then it cannot have an efficient cause, either (by Theorem 2.4), and therefore it cannot have a material or formal cause (by Theorem 3.3).
Second proof. A second proof refers the reader back to the proof of (B) in Theorem 3.2 (see 3.30(14)). This is the argument for the claim that an ascending infinite series of causes ordered per accidens is possible only if an ascending infinite series of causes ordered per se is impossible. In that argu ment Scotus implies that every caused cause in a per accidens series depends for its causing on at least one efficient cause to which it is ordered per se. Thus, nothing can be an effect of a per accidens series unless it is always an effect of a per se series. But the first in eminence cannot be an effect of a per se series (by Theorem 2.4). So it cannot be an effect of a per accidens series, either.
3.39(41) Theorem 3.14: The Supreme Nature is something actually existing.
Here we move from possibility to actuality.
3.39(42) Proof. See the proof of Theorem 3.4. The corollary holds that the actually existing most eminent nature doesn't just happen to be the most eminent thing, but is unsurpassably eminent in the strongest sense: it is con tradictory that something else should be more eminent than it. The corollary foreshadows his later reference, in 4.79(65), to Anselm and his clarification of Anselm's famous argument that reasons from a description of God as "that than which nothing greater can be conceived."
9. Duns Scotus, Reportatio 1-A d.2.p.1.q.1–3.n.33 (Wolter and Bychkov, 1:124).
Commentary 3.40(43)-40(44) The Triple Primacy 3.40(43) Theorem 3.15: To one and the same unique and actually exist ing nature belongs the triple primacy within the three types of essen tial orders discussed in this chapter, namely, efficiency, finality, and eminence.
In Theorem 3.6, Scotus argued that there could be only one nature that is first in the order of efficiency. He had already argued, in Theorem 3.4, that this nature actually exists. Here he argues that one and only one nature is first in all three orders: efficiency, finality, and eminence. Since he has already argued for the actual existence not only of the first in efficiency but also the first in finality (Theorem 3.10) and eminence (Theorem 3.14), it follows that the unique nature with Triple Primacy actually exists.
3.40(44) First proof. As noted, Scotus offers a proof of necessary existence (from itself, that is, not from another) only for a first in the order of efficiency. Here he makes explicit what has been implicit, namely, that from the actual exis tence and uncausability of the first in finality and the first in eminence, their necessary existence follows. But, by Theorem 3.6, necessary existence belongs to just one nature. Therefore one and the same nature has the Triple Primacy.
Second proof. Each First Nature is uncausable. There cannot be a multitude of uncaused things. Therefore one uncausable nature is first in all three orders.
Scotus evidently thinks it is obvious that there cannot be a multitude of uncaused things, since he states the premise as a rhetorical question—how will a multitude exist of itself?
3.41(45)-45(56) The Uniqueness of the First Nature 3.41(45) The intensely pregnant theorem. Describing Theorem 3.15 as "intensely pregnant," Scotus claims that it implies six additional conclusions, which he discusses (in a disorderly manner, it must be said) in the remainder of Chapter 3. These six conclusions cannot be identified with the remaining
Theorems (16-19) of Chapter 3, since there are only four remaining theorems.
Scotus himself uses the term conclusio throughout the Treatise both for the theorems that structure the treatise, and also for conclusions of arguments offered as proofs of the theorems. This is rarely confusing, but it is confusing here, especially because Scotus nowhere says what exactly these "six conclu sions" are. The reader is made to guess.
The six additional conclusions. The first three conclusions (1-3), he tells us, are about the unity, or oneness, of the nature that is first in each of the three
Chapter 3 essential orders. But this is ambiguous: it could mean (i) that there is one and only one nature that is first in all three orders (as Theorem 3.15 says) or it could mean merely (ii) that each order has one and only one first (which leaves it open that each order has its own first and none of these three firsts is iden tical with any of the others). In favor of (i) is that this is exactly what Scotus has just shown, in Theorem 3.15, so we might expect him to elaborate on this unity in a more fine-grained way. Nevertheless, I favor (ii), for the following reasons. First, the proofs of Theorems 3.17 and 3.18, which he offers in sup port of 1-3, make sense on (ii), but not on (i). Second, (ii) fits better with what
I take to be the best way to understand the second three conclusions (4-6).
About 4-6, Scotus says that they concern the identity of the nature that is first in one order with the natures that are first in the other orders. In these conclusions Scotus seems to return to what was established in Theorem
3.15—that one and the same nature has the Triple Primacy—but strength ens the view; while he does not argue for divine simplicity until Chapter 4
(Theorems 4.1 and 4.10), here he takes a first step toward simplicity.
We can enumerate 1–6 on Scotus's behalf in the following way:
1. Only one nature is first in the order of efficiency.
2. Only one nature is first in the order of finality.
3. Only one nature is first in the order of eminence.
4. The nature first in efficiency is identical with the nature first in finality.
5. The nature first in finality is identical with the nature first in eminence.
6. The nature first in eminence is identical with the nature first in efficiency.
Some disorganization. The final two sentences of 3.41(45) do not flow well with the first part of this section. It is indeed true that Theorem 3.15 follows from Theorem 3.6, but why Scotus says this here is unclear. Also, while it is indeed befitting to state the major premises of the arguments for 1-6, he never states them. It is possible that these two sentences represent mere notes and therefore some unfinished business in the Treatise. Be that as it may, as confusing as Theorems 3.16-18 are for the reader who is looking for clear statements of and arguments for the "six conclusions," there is nothing fundamentally incomplete or unsubstantiated in the remainder of Chapter 3. To this remainder I now turn.
3.42(46) Theorem 3.16: It is impossible for one and the same thing to depend essentially on two things in such a way that its dependence terminates wholly in each.
The first two conclusions. The purpose of the theorem is to add support for 1 and 2 (listed under 3.41(45), above). It says that one and the same thing
Commentary cannot essentially depend on two things in the same order of dependence in such a way that its dependence "terminates wholly" in each. Where B depends on A, for B's dependence to "terminate wholly" in A is not for B to depend on nothing besides A, but instead for A to be the last term, or ter minus, of B's dependence; there is nothing prior to A in an essential order of dependence, on which B depends.
3.42(47) Proof. Scotus first argues that nothing can depend on two causes in the same order of causation (e.g., two efficient causes) in such a way that its dependence wholly terminates in each. This is because, according to Scotus, causation cannot be redundant or overdetermined. He then applies this prin ciple of non-redundancy to each type of essential dependence.
In Theorem 3.6 Scotus had established that there could be no more than one first in the order of efficiency; he established this through necessary existence (itself established by Theorem 3.5). Then, in Theorem 3.15, the oneness of the first in the order of efficiency was extended to the first in the orders of finality and eminence. So strictly speaking there is no need for Sco tus here to offer an alternative demonstration that there can only be one first in the orders of efficiency and finality. Nevertheless, good arguments love company. Also, Theorem 3.16 establishes the oneness of these firsts through the nature of dependence itself, rather than through the nature of necessary existence; there is therefore something more direct about the proof here than in Theorems 3.6 and 3.15.
3.43(48) Theorem 3.17: To only one nature belongs each primacy of any type of extrinsic cause.
The first three conclusions. Explicitly, the purpose of this theorem is to add sup port for 1 and 2 (listed under 3.41(45), above), the two conclusions about the "extrinsic" causal essential orders. Nevertheless, somewhat confusingly, the fourth proof of the theorem is about the essential order of eminence, which is not a causal order, and the first proof works just as well for eminence as for efficiency and finality.
The theorem as stated is ambiguous: it could be either a repetition of Theorem 3.15: one and only one nature is first in all three orders; or it could be saying something weaker: only one nature is first in efficiency, only one nature is first in finality, and (implicitly, given the fourth proof) only one nature is first in eminence. But the first proof makes clear that this sec ond, weaker, reading is intended.
3.43(49) First proof. Suppose there were two firsts in an essential order and suppose that each first had at least one thing posterior to it. Call all the things
Chapter 3 posterior to either first "the posteriors." Then for every posterior, it is poste rior to both firsts or only to one. But not to both firsts (by Theorem 3.16).
But not to only one, either, since it would follow that there would be two distinct essential orders and so two distinct universes. But there is no need to posit more than one universe. So there is not more than one universe
(by Theorem 2.15). So there are not two firsts in an essential order. The argument works for each of three types of essential order under discussion.
Therefore: (1) only one nature is first in the order of efficiency, (2) only one nature is first in the order of finality, and (3) only one nature is first in the order of eminence.
Additional probable proofs. Scotus then tacks on five brief "probable" proofs in support of 1-3. The second proof supports 1-3; the third supports 1 and 2; the fourth supports 3; the fifth supports 2; and the sixth supports 1-3.
3.44(50) Second proof. Here he reasons that since ascending essential orders advance from plurality to paucity, it is plausible that each terminates in just one first.
3.44(51) Third proof. The higher up a cause in a causal essential order, the more things it can cause. Plausibly, then, there is a first in each causal essen tial order, which can cause all other things. If so, there is just one first in each causal essential order.
3.44(52) Fourth proof. Natures are like numbers (see 3.39(38)): for each pair of natures, one is greater than the other. If there were two firsts in the order of eminence, one would not be greater than the other. Therefore there is just one first in the order of eminence 3.44(53) Fifth proof. It is unintelligible that there is no single final end in which everything comes to rest. Therefore there is just one first in the order of finality.
This is a startlingly hopeful view about what we might call the unity of goodness: while the ends pursued here and now often appear irreconcilably competitive, the reality is that they all point to the same ultimate end.
3.44(54) Sixth proof. If, for any of these three essential orders, more than one nature is first in it, then one of the firsts in that order does not virtually contain the perfection of the other firsts in that order. But the unsurpassably perfect nature is such that it virtually contains the perfection of every other nature. Therefore there is just one first in each order.
Commentary A "perfect being" argument. It seems to me that this argument makes most sense if we take it as establishing a stronger conclusion than Theorem 3.17: not just that there is one first for each primacy, but that one and only one nature has all three primacies. (Of course, Theorem 3.17 follows from the stronger conclusion.) If there is at least one most perfect (i.e., eminent) nature, it would virtually contain the perfection of every other nature. For a nature, N, to "virtually contain" the perfection of every other nature is for N to be such that, for every perfection a nature can have, either (i) N has that perfection, or (ii) N can cause something else to have that perfection, or both (i) and (ii). There is a most perfect nature (by Theorem 3.12). Being first in the order of efficiency and the order of eminence are perfections a nature can have. But these perfections are not causable (by Theorems 3.3 and 3.9). Therefore the most perfect nature has these two perfections and cannot cause anything else to have them. Therefore there is only one first in each of the three orders and the first in one order is also the first in the other two.
3.45(55) Theorem 3.18: The First Nature that can produce an effect is most actual because it virtually contains every possible actuality.
The first in finality is the best, virtually containing every possible good.
The first in eminence is most perfect, eminently containing every possi ble perfection.
The last three conclusions. In this theorem, Scotus turns his attention to 4–6 (listed under 3.41(45), above). In supporting these last three "conclusions," he bolsters support for Theorem 3.15. Yet Theorem 3.18 as stated does not obviously have anything to do with 4–6. So this requires some unpacking.
Within the theorem, Scotus lists each type of order, and for each type, attributes a superlative to it and offers a reason for the superlative: the first in efficiency is the most actual, the first in finality is the best, and the first in eminence is the most perfect.
The first in efficiency is the most actual because it has the causal power to produce whatever can be produced. The first in finality is the best because all action of an agent acting per se is for the good, and the first final cause is, ultimately, that for the sake of which all per se action is undertaken. The first in eminence is the most perfect because it lacks no perfection possible for a nature to have.
3.45(56) Proof. Scotus then explores conceptual connections between these superlatives and finds them unified by the concept of goodness, or goodness in the highest degree. Each superlative expresses an aspect (ratio) of goodness in its highest degree: the most actual (the first in efficiency)
Chapter 3 expresses the highest communicability (i.e., ability to share what it has with others), the best (the first in finality) expresses the highest lovability, and the most perfect (first in eminence) expresses the highest integrity or wholeness. The highest goodness, Scotus says, has each of these insepa rably from the others—inseparably in the strong sense that the nature of the highest goodness entails being all three. Therefore whatever nature has one of these aspects is the highest goodness and has all three aspects.
Therefore (4) the nature first in efficiency is identical with the nature first in finality, (5) the nature first in finality is identical with the nature first in eminence, and (6) the nature first in eminence is identical with the nature first in efficiency.
3.46(57)-48(63) Everything Is Ordered 3.46(57) Theorem 3.19: One actually existing nature is first in the three aforementioned orders with respect to any other nature, such that any other nature is posterior to that First Nature in these three ways.
Theorem 3.15 had already established the actual existence of a nature that is first in all three orders. Theorem 3.18 established again that one and the same actually existing nature is first in all three orders. Here in Theorem 3.19
Scotus repeats these previous theorems but adds that anything besides this one triply First Nature is essentially posterior to this nature. Instead of prov ing it, he introduces a powerful objection to the theorem and then offers several replies to the objection.
3.46(58) Objection. Scotus introduces an impudent person who might object that even if there is a unique nature that is first in all three orders, it does not follow that every other nature is posterior to this First Nature. This echoes an objection I myself raised to the fifth argument for (A), in 3.29(12).
3.46(59) First reply. Scotus replies by appealing to Theorem 3.6, which holds that only one nature has necessary existence from itself. By Theorem 3.15, the one nature that has necessary existence from itself has the triple primacy.
Thus, any other nature is posterior in all three orders to this unique First Nature. Scotus's own route to the conclusion goes through the uncausability theorems for each primacy argument (Theorems 3.3, 3.9, 3.13). His idea is that being uncausable (and possible) entails being necessary from itself, and since there can only be one such nature, any other nature is either necessary from another or not necessary at all—either way, any other nature is essen tially posterior to the First Nature.
Commentary Similar to Theorem 3.17, Scotus goes on to add five brief replies, perhaps, like those of Theorem 3.17, also offered as "probable proofs." Each focuses on one type of order and aims to show that there can be nothing besides the first in that order that is not essentially posterior to the first.
3.47(60) Second reply. What is neither an end nor ordered to an end exists in vain. Therefore everything is either an end or ordered to an end (and so belongs to the essential order that terminates in the first final cause). I know of no way to make the notion of "existing in vain" philosophically precise that doesn't appeal to the notions of being an end or being ordered to an end, so I am not sure what exactly is supposed to be philosophically problematic about existing in vain. But the fourth reply, which appeals to the fact that anything with an efficient cause has a final cause and vice versa, offers a promising way forward: if there were anything that is neither an end nor ordered to an end, it would be neither the first in finality or efficiency nor—since nothing without a final cause has an efficient cause—something essentially posterior to anything in the order of efficiency. But this is impossible, given the fourth reply, below.
Third reply. This proof assumes that everything has some degree of eminence, that is, nothing is wholly bad—only what is not a being at all has no good ness. But what is neither the most eminent nor excelled by something else has no degree of eminence. Therefore everything is either the most eminent or is excelled by something (and is therefore essentially posterior to the most eminent nature).
3.47(61) Fourth reply. Whatever has a final cause has an efficient cause (by Theorem 2.4), and whatever has an efficient cause has a final cause (by Theo rem 2.5). What is first in the order of efficiency is first in the order of finality
(by Theorem 3.18). By the second proof of Theorem 3.19, everything is either an end or ordered to an end. Therefore everything is either an efficient cause or an effect (and so belongs to the essential order that terminates in the first efficient cause).
3.47(62) Fifth reply. Everything is either the most eminent or is excelled by something (by the third proof of Theorem 3.19). But whatever is ordered to an end is excelled (by Theorem 2.16), and whatever is ordered to an end is an effect. Therefore everything is either the first in efficiency or posterior to it.
Sixth reply. It is irrational, because theoretically unnecessary (see Theorem 2.15), to postulate a being with no essential order to the unique triply First
Nature. Therefore there is no such thing. See the first proof of Theorem 3.17
Chapter 3 and the second response to the objection to the third proof of Theorem 3.6
(3.36(26)).
3.48(63) Having argued that every actual or merely possible thing is either the First Nature or essentially ordered to it in the orders of efficiency, finality, and eminence, Scotus praises God. Here, implicitly, Scotus identifies God with the Supreme Nature uniquely picked out by its possession of the triple primacy. In Chapter 4 Scotus proves various theorems about attributes of the Supreme Nature; through these theorems the divinity of the Supreme
Nature comes more and more into focus.
Chapter 4 On the Simplicity, Infinity, and Intellectuality of the First Being Concerning the First Nature. The focus of Chapter 3 was the necessary existence of the Supreme Nature, triply first in the orders of efficiency, finality, and eminence. Chapter 4 concerns the attributes or perfections of this nature. Specifically, Scotus argues that the Supreme Nature is simple
(Theorems 4.1, 4.2, and 4.10), has all pure perfections in the highest degree (Theorems 4.2 and 4.3), has intellect and will (Theorems 4.4-4.8), is infinite
(Theorem 4.9), and is unique (Theorem 4.11). In Chapter 4 Scotus most fre quently refers to the Supreme Nature, or the nature that has triple primacy, as the First Nature. Except for his prayers, Scotus does not start using the word "God" to refer to the First Nature until Theorem 4.4, on the intellect and will of the First Nature.
4.49(1)-(2) Scotus begins with simplicity, and returns to simplicity much later, after the long section on infinity. Here he tells us that the reason for breaking up the arguments for simplicity in this way is to avoid a circular argument. This suggests that there is a danger of proving infinity through simplicity, and then proving simplicity through infinity. In fact he does prove simplicity through infinity (Theorem 4.10), and infinity through simplic ity (the fourth proof of Theorem 4.9). But the arguments for simplicity in Theorem 4.1 do not rely on infinity, and none of the other arguments for infinity relies on simplicity. So Scotus avoids the circle.
4.50(3)-51(6) Simplicity 4.50(3) Theorem 4.1: The First Nature in itself is simple.
4.50(4) The theorem states that the First Nature is simple "in itself," and here
Scotus says that by this he means that the essence itself is simple. Therefore Theorem 4.1 does not rule out the possibility that the First Nature has acci dents, that is, properties that are not essential properties and so, if possessed by a simple nature, would imply a composition of essence and accidental
Commentary properties in the First Nature. Scotus does not attempt to argue that the First Nature cannot have accidents until 4.89(80).
First proof. The First Nature is uncaused (by Theorem 3.3); therefore it is not composed of matter and form (by Theorems 2.6 and 2.7).
Second proof. The First Nature exists necessarily from itself (by Theorem 3.6 and 3.15). If the perfections of the First Nature were really distinct from each other, then the First Nature would have necessary existence through at least one of its perfections.
But it cannot have necessary existence through only one of its perfections, for two reasons: first, if it did, then the rest of the perfections of the First
Nature would exist necessarily only derivatively from the perfection through which the First Nature has necessary existence; second, the First Nature itself would exist necessarily only derivatively from the perfection through which it has necessary existence. The problem with this is that either some part of the First Nature or the First Nature as a whole would be dependent for its necessary existence on its property of necessary existence—and so the First Nature would fail to be truly first (i.e., dependent on nothing and that on which everything else depends).
And it cannot have necessary existence through more than one of its per fections, for two reasons: first, if it did, then it would have several necessary existences—but this is redundant; second, the various necessary existences could not form a unity—since, as necessary existences, none would be in potency to another and would not be in potency together to compose some thing else.
Therefore its perfections are not really distinct from each other. Where some perfections belong to one and the same thing and are not really distinct from each other, they are really the same as the thing to which they belong, and really the same as each other, even if they are formally distinct from the thing to which they belong and formally distinct from each other.
Formal distinction. Scotus is well known for this subtle type of distinction.
In his most mature description, formal distinction occurs where there is real identity but also difference in ratio (i.e., content, account, concept, notion, meaning).1 As Blander puts it: "x is formally distinct from y [if and only if ]
(a) x is really identical to y; (b) the account [ratio] of what it is to be x is not the same as the account of what it is to be y; and (c) the account of what it is to be y is not included in the account of what it is to be x."2
1. Duns Scotus, Reportatio 1-A d.33.q.3.n.63 (Wolter and Bychkov, 2:330).
2. Josh Blander, "Same as It Never Was: John Duns Scotus' Paris Reportatio Account of Identity and Distinction," British Journal of the History of Philosophy 28, no. 2 (2020): 236.
Chapter 4 4.50(5) Corollary. The First Nature is not in a genus. To belong to a genus is to have a nature that belongs to a species and can therefore be defined. A spe cies is constituted by its genus (e.g., animal), and difference (e.g., rational), and a definition expresses the species (e.g., "rational animal," the classical defini tion of man). In a real definition, the difference term cannot express content that is expressed by the genus term. If it does—that is, if the genus term expresses any of the content that the difference term expresses—then a defi nition has "vain repetition." To borrow one of Aristotle's examples,3 which Scotus discusses elsewhere,4 "biped" cannot be a difference of the genus of animals with feet, since "biped" itself means "with two feet," resulting in the vainly repetitive definition "animal with feet and two feet." By contrast, in the definition "rational animal," the difference and genus terms express totally diverse content.
Scotus's point here is that the reasoning of the second proof, above, shows that no two perfections of the First Nature are totally diverse: there is not one perfection by which the nature exists necessarily from itself, and another by which it is wise, and yet another by which it is powerful. Instead, the First
Nature is all these things (and more!) by itself alone. Therefore it cannot be defined—that is, no well-formed definition could apply to it.
4.51(6) Objections to the second proof. As Scotus himself would grant, the First
Nature has various perfections that are not formally the same as each other: for example, divine wisdom is not formally the same as divine power. So the perfection through which the First Nature has necessary existence is not formally the same as every other perfection of the First Nature. The sec ond proof argues that there cannot be really distinct perfections in the First
Nature, but the objection asserts that the argument offered against really distinct perfections in the First Nature works equally well as an argument against formally distinct perfections in the First Nature.
Moreover, in Christian doctrine, God is one divine essence in three divine persons. Each divine person shares exactly the same divine essence. So one divine person cannot be distinct from another divine person by the essence they share. So they are distinct by at least one property that is peculiar to each divine person. So a divine person has two (formally distinct) constitu ents: the divine essence and the personal property. Applied to this Trinitar ian context, then, the objection is that since the distinguishing property of a divine person is formally distinct from the divine essence, and since the 3. Aristotle, Metaphysics 7.12.1038a19–24 (Barnes, 2:1639).
4. Duns Scotus, Questions on the "Metaphysics" of Aristotle 7.q.19.n.7–10 (Etzkorn and Wolter, 2:310).
Commentary divine person has necessary existence through the essence and not through the personal property, a divine person would have its necessary existence derivatively.
Scotus's reply relies on his understanding of the nature of infinity, which he does not discuss until Theorem 4.9—see the third and fourth proofs of infinity in 4.75(56)–77(62). But, briefly for now, he denies the objector's claim that his second proof works equally well against the view that there can be formally distinct perfections in the First Nature. He denies this on the grounds that, where formally distinct perfections belong to one and the same thing, and at least one of these perfections is infinite, then these for mally distinct perfections belong to that one thing by identity rather than by composition (i.e., parts to whole).
With respect to the divine persons, then, since the divine essence is for mally infinite, the divine essence and the personal property belong to the divine person by identity, not composition.
4.52(7)-54(11) Pure Perfections 4.52(7) Theorem 4.2: Whatever is intrinsic to the highest nature is the highest such thing.
Intrinsic perfections. The First Nature has intrinsically that which it has just by virtue of what it is, independent of the existence of anything else. For example, being wise is an intrinsic perfection, but being the creator of the world is an extrinsic perfection, since nothing can be a creator unless there is some creation.
Note, however, that the three primacies proved in Chapter 3 are all intrin sic perfections: Scotus proves that there is a unique first in the order of effi ciency (which does not imply that it has any actual effects), not that there is a unique first efficient cause (which would imply that it has actual effects); similarly for the other two primacies.
Theorem 4.2 says that whatever perfection the First Nature has intrin sically, exists in the First Nature in the highest possible degree to which that perfection can exist. So if the First Nature is intrinsically wise, then its wisdom is the most perfect wisdom possible. And so on for all the intrinsic perfections of the First Nature.
4.52(8) Proof. Assume an intrinsic perfection of the First Nature could be greater than it is. Each intrinsic perfection of the First Nature is the same as the First Nature (by Theorem 4.1). Therefore the First Nature could be greater than it is. But the First Nature is unsurpassably great (by Theorems
Chapter 4 3.14 and 3.15). Therefore no intrinsic perfection of the First Nature could be greater than it is.
4.53(9) Theorem 4.3: Every pure perfection belongs to the highest nature necessarily and in the highest degree.
Here Scotus argues that the First Nature has every "pure perfection," a type of perfection he picks up from Anselm.5 But before arguing for the theorem, he adds precision to Anselm's description of pure perfection.
4.53(10) Anselm on pure perfection. Anselm had said that a pure perfection is that which, in anything whatsoever, is better than what is not it. Scotus takes issue with this description, for two reasons. The first concerns the ambiguity of the negation (i.e., what is not it). Strictly speaking, the negation of wise is not-wise, and it is impossible to tell whether being wise is better than being not-wise, since not-wise includes everything that is not wise, such as love and justice but also caninity and furriness. The second concerns the compatibility of a proposed pure perfection with the subject that is, by Anselm's descrip tion, supposed to be better by having that perfection. Being wise might be better than any property of a dog, but a dog cannot be wise, so being wise cannot be good in a dog, and hence it cannot be better for a dog to be wise than not-wise.
Correcting Anselm. These objections to Anselm's description of pure perfec tion lead Scotus to add some precision. He settles on this: a pure perfection is a perfection that is better than anything incompossible with it. Thus, being wise (let's suppose) is incompossible with being a dog or being furry, but compossible with being loving or just. Therefore, so far as we can tell from these examples alone, being wise, loving, and just are still candidate pure perfections (because they are compossible), but caninity and furriness are no longer candidate pure perfections (because they are not compossible with being wise and being wise is better than being canine and furry).
Denominative predication. Aristotle says that denominative (or paronymous) predication occurs "when things get their name from something, with a dif ference of ending," as the grammarian gets his name from grammar and the brave from bravery—here grammarian and brave are the denominative terms.6 Thus, "Socrates is a grammarian" predicates grammar of Socrates
5. Anselm, Monologion 15 (Williams, 21–22).
6. Aristotle, Categories 1.1a.13–15 (Barnes, 1:3); Duns Scotus, Questions on the "Metaphysics" of Aristotle 9.q.3–4.n.19 (Etzkorn and Wolter, 129).
Commentary denominatively, and "Socrates is a wise (thing)" predicates wisdom of Socrates denominatively.
With this in mind, we can make a little sense of what Scotus means when he says that the incompossibility in his revised definition of pure perfection
"ought to be understood according to denominative predication." Anselm originally said that a pure perfection is that which is better than what is not it, in anything whatsoever. This led to the puzzle about wisdom in a dog: wisdom seems to be better than any property of a dog but also incompossible with some essential canine properties. So Scotus's new definition scraps the "in anything whatsoever" and simply has "better than anything incompossi ble with it," with the instruction to take the incompossibility to hold between denominatives (i.e., the things denominated by denominative terms). Thus, the fact that a dog can't be made better by wisdom shouldn't make us wonder whether wisdom is a pure perfection. Instead, on Scotus's procedure, we sim ply compare wise thing and canine thing and ask which is better to be.
Importantly, however, Scotus expresses some skepticism at 4.58(22) about our ability to intuit which perfections are pure perfections—even using wis dom (or, taken denominatively, wise thing) there as an example of a per fection we cannot know (independently of other argumentation) is a pure perfection.
4.54(11) Proof. With respect to anything incompossible with it, a pure per fection is prior in the essential order of eminence (e.g., a wise thing is more eminent than a dog). The First Nature is first in the order eminence (by The orems 3.14 and 3.15). Therefore no pure perfection is incompossible with it
(by the definition of pure perfection). Therefore the First Nature has all pure perfections.
Scotus then argues that the First Nature has all pure perfections nec essarily (and so essentially), rather than contingently (and so accidentally).
The First Nature is first in the order of eminence. If being necessary were incompatible with pure perfections, then the First Nature could have no pure perfections. But if it could have no pure perfections it would be excelled (by the definition of pure perfection). The First Nature cannot be excelled
(by Theorem 3.15). Therefore it can have pure perfections. It does have pure perfections (by the proof of Theorem 4.3). Therefore it has them most per fectly (by Theorem 4.2). It is better to have a pure perfection necessarily than to have it contingently. Therefore the First Nature has all pure perfections necessarily. Therefore it has them essentially. Therefore it has them by iden tity (by Theorem 4.1).
Scotus here assumes that what the First Nature has necessarily, it has essentially. In general it does not follow that if a thing has something neces sarily it has it essentially. But in the case of the First Nature it does: the First
Chapter 4 Nature exists necessarily (by Theorems 3.5 and 3.15), and the First Nature is simple (by Theorem 4.1). So whatever belongs to that essence belongs to it necessarily.
There is a possible quibble here. Theorem 4.2, strictly speaking, was about the degree of the First Nature's intrinsic perfections (namely, the highest degree). But here the argument requires that the First Nature's way of having its intrinsic perfections is in the highest degree. On the surface it is con ceivable that something could have a perfection in the highest degree, but not have in the highest degree a perfection (where "in the highest degree" modifies "perfection" in the former but "have" in the latter). For example, suppose something is maximally wise but not identical with wisdom: then it has the highest degree of wisdom but doesn't have in the highest degree that perfection. But the nature that is first in eminence surely not only has its intrinsic perfections in the highest degree, but has in the highest degree its pure perfections. So I don't think the quibble should be a matter of concern.
Scotus does not offer support for the premise that it is better to have a pure perfection necessarily than to have it contingently. But the premise is plausible: if it's a pure perfection, how could it be better to be able to lose it than to be unable to lose it?
4.55(12)-66(45) Intellect and Will 4.55(12) Theorem 4.4: The first efficient cause understands and wills.
This first of five theorems about the intellect and will of the First Nature states that the first efficient cause understands and wills. Since the First Nature is first in the order of efficiency (and therefore the first efficient cause of anything that exists besides itself), to show that the first efficient cause thinks and wills is sufficient to show that the First Nature thinks and wills.
4.55(13) First proof. The first efficient cause is a per se agent (see the proofs of Theorem 2.4). Every per se agent acts for the sake of an end (by Theorem
2.4). A natural agent (that is, an agent lacking intellect and will) acts for the sake of an end only if it is essentially dependent in the order of efficiency on an agent with intellect and will loving that end. The first efficient cause is not essentially dependent on anything. Therefore it has intellect and will.
See the two corollaries to Theorem 2.5 (2.12(16)-(18)) for Scotus's rea sons for thinking that a natural agent acts for the sake of an end only if it is essentially dependent in the order of efficiency on an agent with intellect and will (and see the commentary on the same).
Skepticism about final causes? In this paragraph, Scotus seems to entertain the possibility that a natural (non-intelligent) agent might be an efficient cause
Commentary and yet not act for the sake of an end: such an agent "would act just as it does even if it were independently acting for the sake of no end at all." On the surface, this admission appears to express some skepticism about Theorem
2.4 itself, which says that what is not ordered to an end is not an effect—or, everything that has an efficient cause has a final cause too.
Yet I do not think a skeptical interpretation of this puzzling passage is warranted. So what could Scotus mean? Recall that Scotus distinguishes two senses of "end": end as final cause, and end as that species-relative perfecting state that an agent seeks by nature (2.12(16)-(17)). My suggestion is that when Scotus says here that a natural agent would act just as it does even if were acting for no end, he means "end" in the strict sense. Then we can read his comment in this way: a natural agent, whether or not it is finally caused, acts for the sake of whatever ends are built into it by its nature. Reading it this way, then, whatever puzzle remains has to do with Scotus's reasons for thinking that ends can only be causes insofar as they are loved by intelligent agents—for that, see 2.12(16)-(18).
4.55(14) Second Proof. (This proof begins in 4.55(13) and then spills over to 4.55(14).) The first efficient cause is a per se agent (see the proofs of Theorem
2.4). Every per se agent acts for the sake of an end (by Theorem 2.4). The end moves the first efficient cause to act, and the first efficient cause moves its effect to its end, either (i) as loved by an act of will, or (ii) naturally. But not (ii). Therefore (i). Whatever is capable of loving an end, and moving an effect to its end by an act of will, has both intellect and will. Therefore the first efficient cause has intellect and will.
To see why Scotus denies (ii), consider that everything acts for the sake of its final end—even the First Nature. But the First Nature is dependent on nothing else for its perfection (by Theorem 3.15), and lacks nothing that would increase its perfection (Theorems 4.2 and 4.4). So if this First Nature efficiently causes something besides itself, it does not do so in a natural (and therefore necessary) pursuit of its final end—it already has whatever it would seek. So here Scotus says that if the end for the sake of which the first efficient cause acts, moves it to act naturally, it would not in fact efficiently cause any thing—it would simply go on being and enjoying itself. But then it wouldn't in fact be the first efficient cause (it would simply be the First Nature). So assuming this First Nature is the first efficient cause, it efficiently causes something, and therefore does so by an act of will, freely, non-necessarily. The first efficient cause wills its final end in all it efficiently causes, but there is nothing about the First Nature that determines it to efficiently cause anything.
4.56(15) Third proof. The proof has two "consequences": from the first to the second premise and the second premise to the conclusion. First consequence.
Chapter 4 Something causes contingently; therefore the first cause causes contingently.
Second consequence. The first cause causes contingently; therefore it causes by willing. Scotus offers short proofs for both consequences.
Scotus thought that it is obvious that something is contingent. Echoing Avicenna, he tells us that those who deny contingency should be "beaten until they confess that it is possible for them not to be tormented."7 He thought it almost as evident that the will is the power of contingency. There are two basic types of powers, he says: a natural power, which produces its effect necessarily, and a rational power, which produces its effect contin gently. On his understanding, a rational power just is a will. 8
Proof of the first consequence. The first efficient cause causes the causing of every secondary efficient cause. So if it causes necessarily, they all do. But they don't all cause necessarily. So it causes contingently. Proof of the second consequence. Will is the explanatory principle of contingency.
4.56(16) First objection. If will is the explanatory principle of contingency, then even if the first efficient cause causes necessarily, any created will causes contingently, just insofar as it is will. So we need only posit created wills to account for contingency. Second objection. According to Aristotle, we could account for contingency by holding that while God moves necessarily, con tingency arises from the plurality of effects far down the causal chain. Third objection. Contingency can be explained as a natural motion impeding the natural motion of something else.
4.56(17)–(19) Reply to the first objection. Scotus grants that the will explains contingency, but here denies that there could be wills at all if the first effi cient cause causes necessarily. The pith of the reply is in the last sentence: it is impossible for the will to cause necessarily what it causes by willing. So where the objector asserts that there could be other wills and so contingency, even if the first cause is not a will, Scotus counters that if the first cause is not a will there could be no contingency and so no other wills. This reply leaves Scotus and his objector at a stalemate. Reply to the second objection. Scotus denies that a plurality of diverse effects, all coming ultimately from necessary causes, could give rise to contingency. Reply to the third objection. An impediment to a natural motion cannot explain contingency because the impediment itself is caused by necessary causes.
7. Duns Scotus, Reportatio 1-A d.39–40.q.1–3.n.30 (Wolter and Bychkov, 2:473; Avicenna, The Metaphysics of "The Healing" 1.8.11–13 (Marmura, 42–43). Avicenna in turn may be echoing Aristotle, Topics 1.11.105a3–7 (Barnes, 1:175).
8. Duns Scotus, Questions on the "Metaphysics" of Aristotle 9.q.15.n.22–25 (Wolter, 371–73).
Commentary 4.57(20) Fourth Proof. A natural cause (i.e., a cause lacking intellect and will), causes to the full extent of its causal power—impediments, not the cause itself, always explain why some particular cause fails to produce effects to the full extent of its power. But the first cause would have no impediments.
So we would expect a first cause lacking intellect and will to make things as perfect as they can be. But the universe evidently is not as perfect as it can be.
Therefore its lack of perfection cries out for personal explanation: a cause with intellect and will could make a choice to make the universe less perfect than it could be. Note the limited scope: this is not a proof of God's goodness, just his intellect and will.
Matter does not obey. The point of this quick rejoinder is to offer an alternative explanation of why there is badness in the world. Scotus's reply is that a suf ficiently powerful agent could keep matter in line, and therefore a first cause without intellect and will—acting to the utmost of its power—would create a world in which matter is never unruly. Elsewhere, Scotus says that matter is in "obediential potency" to a sufficiently powerful agent.9
4.57(21) Fifth proof. The First Nature is the most eminent nature (Theorem 3.15). Intelligent things are better than all other non-intelligent things. If the First Nature were non-intelligent, then any intelligent thing would be better than it in some respect. But what is first in eminence cannot be surpassed in eminence. Therefore the First Nature is an intelligent thing.
The argument is odd, since at first glance it treats "intelligence" as a pure perfection and attributes it to the First Nature on that basis. But Scotus crit icizes this method of arguing for attributes of the First Nature immediately below, in his objections to the sixth proof. Now, strictly speaking, the fifth proof as glossed here does not make use of the concept of pure perfection.
Nevertheless, it seems to me to succumb to the same sort of criticism Scotus offers against the sixth proof. So I'll return to the fifth proof after consider ing Scotus's reasons for rejecting the sixth.
4.58(22) Sixth proof. Here Scotus formulates but rejects a sixth proof. Under standing, willing, wisdom, and love are pure perfections. The First Nature has all pure perfections (by Theorem 4.3). Therefore it has understanding, willing, wisdom, and love.
But Scotus doesn't see how we can know that understanding, willing, wisdom, and love are indeed pure perfections, independent of other rea sons for thinking that these are perfections of the First Nature. So he rejects
9. Duns Scotus, Questions on the "Metaphysics" of Aristotle 9.q.12.n.11 (Etzkorn and Wolter, 2:552).
Chapter 4 the sixth proof. His reasoning goes something like this. We know that the First Nature is the first in eminence (Theorem 3.15) and therefore contains every pure perfection (Theorem 4.3). The most eminent nature contains no imperfection, cannot be surpassed in perfection in any way, and all of its perfections are compatible with each other. Therefore every perfection of the First Nature is a pure perfection. But we don't know which perfections are pure perfections and which are not, except those perfections we have independently established belong to the First Nature—by, for example, the first five proofs of Theorem 4.4, which Scotus does accept. So we can't just assume that some perfection is a pure perfection and then attach it to the First Nature via Theorem 4.3.
Wisdom and the first angel. Consider Scotus's first pair of examples from this section: wisdom and the nature of the first angel. Comparing a wise thing
(qua wise) to the first angel (qua first angel), we cannot know which is better.
But we can know that the First Nature is incompossible with the nature of the first angel (since the latter is by nature causable and finite). So we know the nature of the first angel is not a pure perfection—if it were, then the First
Nature would lack a pure perfection, which is impossible by Theorem 4.3.
Since no perfection that is not a pure perfection can be better than a pure per fection, it follows that we cannot tell whether or not wisdom is a pure perfection (merely by comparing wisdom and the first angel).
Scotus moves on to consider several attempts to show that we can know, by comparing it with other perfections, that wisdom is a pure perfection.
These attempts try to show that if some perfection such as wisdom would make anything better if it had it, then it is a pure perfection. But Scotus is unmoved, for reasons already given in his correction of Anselm's definition of pure perfection considered in 4.53(10). Wisdom cannot make a dog bet ter, because wisdom is incompatible with canine nature. Granted, if a dog could be wise, it would be better if it were wise. But then again, if a dog could be the first angel it would likewise be better. For that matter, if the angel could be a dog (remaining what it is), it would be better, since being a dog is a perfection. The point seems to be that we cannot identify what the pure perfections are simply by thinking about what would (even if per impossibile) make a thing better. Instead, to know whether some perfection,
P, is a pure perfection (and so belongs to the First Nature), we need to know both (i) that to have P is better than to have any other perfection incom patible with P, and (ii) P is compatible with the First Nature. And (ii) is the hard part.
The fifth proof reconsidered. The fifth proof does not invoke pure perfections, but it is an argument from eminence, or a "perfect being" argument, since it attributes intelligence to the First Nature on the grounds that it is better to
Commentary be intelligent than otherwise (and the First Nature is most eminent). In an argument from eminence of the sort given in the fifth proof, we judge that something that has some perfection, P (e.g., intelligence) is, precisely because it has P, better than everything that lacks P. For example, an intelligent thing, precisely because it is intelligent, is better than everything that is not intelli gent. Then, since the First Nature is most eminent, we infer that it is intelli gent. The problem, as the objection to the sixth proof suggests, is that unless we already know that the First Nature is intelligent, we don't know that each intelligent thing is better than every non-intelligent thing. For suppose the First Nature is non-intelligent. Then, since it is the most eminent nature, at least one non-intelligent thing is better than every intelligent thing.
In light of the vulnerability of the fifth proof to the criticism of the sixth, readers may wish to consider whether Scotus might have intended the crit icism of the sixth proof to apply to the fifth proof as well, notwithstanding that the most natural way to read the text is that Scotus accepts the fifth proof but not the sixth.
4.59(23) Theorem 4.5: The first cause causes contingently whatever it causes.
4.59(24) First proof. Theorem 4.4 established that the First Nature causes by willing rather than by the necessity of its nature. The third proof of The orem 4.4 asserted a link between willing and contingency: there is contin gency if and only if there is willing. So just as we can infer contingency from the fact that something comes about by willing, so too we can infer willing from the fact that something comes about contingently. This first proof of Theorem 4.5 may be seen as an extension of this line of reasoning: however far down a causal chain you move away from that first, contingent causing, every cause in the chain causes contingently whatever it causes. Remember, in essentially ordered causal series, all causes in the series cause simultane ously (see 3.28(11)); therefore the contingency of the first cause in such a series imparts contingency to the whole series.
4.59(25) Second proof. God loves himself necessarily as his final end. But he succeeds in loving himself whether or not he causes anything. There fore he need not cause anything. So if he does cause anything he does so contingently.
4.59(26) Objections to the second proof. First objection. The first objection rea sons that since the First Nature's existence is necessary (by Theorem 3.5) and since that nature is simple (by Theorem 4.1), its willing is the same as its necessary existence and so is itself necessary, not contingent.
Chapter 4 Second objection. The second objection concedes for the sake of argument that the First Nature's willing is contingent, but then infers that since the First Nature's willing determines what every created will wills, no created will wills what it wills contingently. Third objection. The third objection draws on the conclusions of the first two objections. If, as the first objection argues, the First Nature's willing is necessary, and if, as the second objection argues, no created will wills contingently, then how does it come about that people sin? It seems that a most perfect first cause would not determine people to will sinfully. Fourth objection. The fourth objection is what we might call the problem of too much contingency.
Scotus concedes these are difficult objections. Rather than answer them, he refers us to his discussion of God's knowledge of future contingents10— possibly the Reportatio version (Scotus's latest Sentences commentary), for the following reasons. What we might call the "standard view" of Scotus's solution to the problem of God's knowledge of future contingents is this:
God knows future contingents by willing the truth value of propositions that, by the meaning of their terms alone, are neither true nor false. So God knows that the Antichrist will come because he wills the proposition "The Antichrist will come," to be true. But God's willing is contingent. So it is contingent that the Antichrist will come. But in the Reportatio Scotus came to see that his standard view could not account for God's knowledge of sinful actions—since God cannot will evil.11 Since the third objection to Theorem
4.5 raises the problem of God's causation of sinful actions, there is reason to think he was aware of the tension in his standard view when he wrote this section of the Treatise, which at least raises the possibility that the Treatise was composed after the Reportatio examinata.
4.60(27) Theorem 4.6: The First Nature is the same as its love for itself.
4.60(28) First proof. The gist here is that the First Nature's love for itself is not the sort of thing that can be caused; it is the ultimate reason why any thing else is caused. But what is uncausable and exists, exists necessarily (by Theorem 3.5), and there can only be one necessarily existing thing (see the first proof of Theorem 3.15 and the second proof of Theorem 4.1). Therefore the First Nature and its self-love are identical.
10. Duns Scotus, Lectura 1.d.39 (Vatican 17:481–510); Ordinatio 1.d.38–39 (Vatican
6:303–8 and 401–44); Reportatio 1-A d.38–40 (Wolter and Bychkov, 2:448–491).
11. Duns Scotus, Reportatio 1-A d.46.q.1–2.n.21 (Wolter and Bychkov, 2:562–63); Gloria Frost, "John Duns Scotus on God's Knowledge of Sins: A Test-Case for God's
Knowledge of Contingents," Journal of the History of Philosophy 48, no.1 (2010): 15–34.
Commentary 4.60(29) Second proof. As Scotus says, this proof is similar to the first.
4.61(29)–(30) Third and fourth proofs. These proofs invoke aspects of Aristotle's discussion of the First Mover's intellectual activity in Metaphysics
12. Aristotle held that the First Mover not only has itself as the object of its intellectual activity (since it is the best thing to think about) but is identical with its intellectual activity—if it were not, it would be dependent on some thing besides itself for its continual achievement of the best state to be in, and so would not after all be the best a substance can be.12 Likewise, if it were not identical with its intellectual activity but in a "potency of contradiction" to it (that is, it could either be or not be engaged in that activity) then it would take some work to continually achieve the best state—and this work would detract from it in some way.13 Note that Scotus calls the fourth proof a merely probable argument, not a demonstration.
The third and fourth proofs are about the First Nature's intellectual activ ity, not its love for itself; therefore they function better as proofs for the sameness of the First Nature and its self-understanding, than the sameness of the First Nature and its self-love. The corollaries of the theorem (4.62(32)) show that Scotus means to prove both here.
4.61(31) Fifth proof (rejected). The First Nature's power (to love) and the object of that power (itself) are the same; hence the act of that power (lov ing) is the same as the First Nature. Scotus rejects the proof because it doesn't generalize to other wills, both angelic and human.
4.62(32) Some corollaries. First corollary. The first and second proofs argued for the identity of the First Nature and its love for itself. Its love for itself is an act of the will. Therefore the will itself—that by which it wills—is also identified with the First Nature. Second corollary. Identity holds between the First Nature and its act of understanding itself, since understanding is a pre condition for loving and so, as the First Nature's self-love is uncausable (and so necessary and so identical with the First Nature), its self-understanding is uncausable (etc.). Third corollary. The First Nature is the same as its intellect, for more or less the same reasons that it is the same as its will. Fourth corol lary. When the First Nature understands itself, it does not understand itself through or by means of some additional entity such as a concept. If it did so, it would be dependent for its self-understanding on this additional entity.
Since it is not dependent in any way, it must be the same as that by which it understands itself.
12. Aristotle, Metaphysics 12.9.1074b18–21 (Barnes, 2:1698).
13. Aristotle, Metaphysics 12.9.1074b28–30 (Barnes, 2:1698).
Chapter 4 4.63(33) Theorem 4.7: No act of understanding can be an accident of the First Nature.
4.63(34) First proof. The First Nature can efficiently cause whatever can be caused, things like oaks and ants and planets and plums. Since God effi ciently causes his effects by his willing (by the third proof of Theorem 4.4), and since willing entails knowledge, prior to willing, of what is willed (by the second corollary of Theorem 4.6), God has knowledge of oaks, ants, and so on, prior to causing them. In the first proof Scotus gives us a partial expla nation of how God knows things like these prior to creating anything. He knows them by knowing himself: God's knowledge of what can be caused is somehow included in his self-knowledge.
4.64(35) Second proof. Here Scotus reasons that since, by Theorem 4.6, one of the First Nature's acts of understanding is the same as the First Nature
(and so not accidental to it), all of the First Nature's acts of understanding (if there are more than one) are the same as it. Perhaps Scotus's thought is that it is theoretically simpler to suppose that an intellect is related to all of its acts in the same way (e.g., by identity, by subject to accident, etc.) than to suppose that it might be related in one way to some and another way to others.
4.64(36) Third proof. This proof seems to rely on a general principle that the more noble cannot be in potency to what is less noble. Alternatively, it might be relying merely on a less general premise that one act of understanding simply cannot be the subject of an additional act of understanding.
4.64(37) Fourth proof. The key claim here is that the most perfect act of understanding—that act of understanding than which there could be none more perfect—will be the act of understanding of everything that can be understood. Since the First Nature has an act of understanding (by Theorem
4.4) and since whatever perfections it has, it has perfectly (by Theorem 4.2), it follows that it has this perfect act of understanding. Since it is the same as this act of understanding, it follows furthermore that none of its acts of understanding is accidental to it. God has exactly one act of understanding.
The theorem holds for willing as well as understanding. We could construct argu ments analogous to the four proofs just given that show that no act of willing can be an accident of the First Nature. It can will whatever can be willed, prior to causing anything to exist (first proof); one of its acts of willing is non-accidental to it (second proof); one act of will cannot be the subject of an additional act of will (third proof); and the most perfect act of will is that act of willing everything that can be willed (fourth proof). Probably we could
Commentary construct analogue arguments that Scotus would be happy with—with the exception of the fourth, since willing everything that can be willed does not follow from perfect willing, the way that understanding everything that can be understood does follow from perfect understanding.
4.65(38) First rejected proof. The attempted proof says that the intellect of the First Nature is "the same for everything." I take it that this means that its intellect stands in no sort of potency to any intelligible thing: it cannot be affected or modified by the things it understands, as our intellects are (e.g., when they passively receive sensory information). The proof then infers from the identity of the First Nature's intellect with its act of understanding, that its act of understanding likewise is "the same for everything."
Fallacy of accident. Scotus objects by offering a counterexample. The First Nature's will and intellect are the same (by Theorem 4.1). But it does not fol low from this that what the First Nature understands it also wills. After all, it understands the proposition "There are humans on Venus" but evidently does not will there to be humans on Venus. So too, from the premises that the First
Nature's intellect is the same for everything, and that the First Nature's intellect is the same as its act of understanding, it does not follow that the First Nature's act of understanding is the same for everything. The fallacy in both the proof and the counterexample is the fallacy of accident, which is, roughly, the fallacy of inferring that a property of a property also belongs to the subject of the lat ter property, or as Aristotle puts it, a fallacy of accident occurs "whenever any attribute is claimed to belong in like manner to a thing and to its accident."14
4.65(39) Second rejected proof. The proof reasons that since the First Nature understands everything by one act, any additional act would simply duplicate what is already understood—and such duplication is not in fact possible. But the soul of Christ sees all things in God and yet can have additional acts of understanding; so too can the blessed in heaven. Still, one might won der whether these additional acts in the case of the soul of Christ and the blessed are due to limitations imposed by their finite natures, and therefore whether these can really serve as counterexamples to what otherwise looks like a pretty good argument.
4.65(40) Third rejected proof. This rejected proof echoes some of the reason ing of the fourth proof, a proof that Scotus accepts. The fourth proof rea soned that a most perfect act of understanding would encompass everything that can be understood, and since God has understanding in the most perfect
14. Aristotle, Sophistical Refutations 5.166b29–36 (Barnes, 1:281).
Chapter 4 way, it follows that God has this perfect act of understanding. The problem with the rejected proof might be that it does not define or describe "the maximum perfection of understanding" explicitly in a way that yields the conclusion—as Scotus does describe "the most perfect act of understanding" in the fourth proof.
4.66(41) Theorem 4.8: (i) The intellect of the First Nature actually understands—always, necessarily, and distinctly—all intelligible things, and (ii) its understanding of these is naturally prior to their actual existence.
Scotus's proofs for this theorem take the theorem piecemeal: two proofs for (i) and two for (ii).
4.66(42) First proof of (i). The intellect of the First Nature can understand everything intelligible. But it cannot have an act of understanding that is not the same as itself (by Theorem 4.7). Therefore by that act it does understand everything intelligible—since, given the premises, if it did not it could not.
Scotus's support for his premise that the intellect of the First Nature can understand everything intelligible is generalizable to any intellect; there is nothing about being the intellect of the First Nature that makes it able to understand everything intelligible. Instead, the proper object of intellect as such is being as such, and (implicitly) being as such is intelligible, such that there can be no unintelligible beings.
4.66(43) Second proof of (i). Here Scotus draws on the venerable image of God as a craftsman. Since anything that exists other than the First Nature, exists because the First Nature caused it to exist, and since the First
Nature creates, if it creates, by intellect and will, the First Nature is like a craftsman with respect to anything that can be made. But if it is like a crafts man with respect to everything that can be made, then it must understand everything that can be made.
Scotus raises a brief objection to this proof and does not reply to it. The objection is that a craftsman need not have knowledge of every single arti fact he can make—a general skill suffices. Scotus considers a similar sort of objection in the second objection to the first proof of infinity in Theorem 4.9; see his reply below.
4.66(44) First proof of (ii). By "naturally prior" here, I take Scotus simply to mean "essentially prior"—that is, just the sort of priority at issue in Chapters
1-3 of the Treatise. The First Nature's understanding of all intelligible things is the same as the First Nature and so necessarily exists from itself. Whatever necessarily exists from itself is naturally prior to anything that does not.
Commentary 4.66(45) Second proof of (ii). Everything that exists, besides the First Nature, is made to exist by the First Nature willing it to exist. So it must know what it wills to exist prior to the existence of what it wills to exist.
67(46)-87(74) Infinity Scotus devotes more words to proving the infinity of the First Nature than he does to any other of its attributes. This emphasis on infinity is also on dis play in the relevant questions of his Sentences commentaries, which overlap significantly—but not completely—with the proofs of infinity given here.15
It is of limited helpfulness to offer detailed comparison of the proofs given here with those given elsewhere, so I will not do so. Two details are worth mentioning, however: in the Treatise, but not in the Sentences commentaries,
Scotus offers a proof of infinity from simplicity. Also, while in each Sentences commentary Scotus leads off with the proof of infinity from efficiency, in the Treatise this proof is the seventh and last proof he offers (not counting the eighth and ninth proofs, which he rejects).
4.67(46)–(47) Theorem 4.9: You are infinite existence and are incompre hensible by what is finite.
4.68(48)–69(50) First proof of infinity: first proof from intellect 4.68(48)–(49) The core of the proof occurs in the first two sentences. The premise is that the intellect that understands everything (namely, the intel lect of the First Nature), understands an actual infinity of intelligible things.
The conclusion is that only an infinite intellect could understand that many things. The proof is, as Scotus says, an enthymeme, that is, a syllogism with a missing or implied premise or, as Peter of Spain puts it, "an incomplete syl logism . . . in which a conclusion is brought in hurriedly."16 To complete the proof, Scotus offers a proof of the premise—here called the antecedent—and then a proof that the conclusion really does follow from the premise—here as elsewhere called the consequence.
Proof of the antecedent. Again, the antecedent is the single premise of the proof: the intellect of the First Nature understands an actual infinity of 15. Duns Scotus, Lectura 1.d.2.p.1.q.1–2.n.64–86 (Vatican 16:134–42); Ordinatio
1.d.2.p.1.q.1–2.n.111–47 (Vatican 2:189–215; PW, 62–76); Reportatio 1-A d.2.p.1.q.1–
3.n.44–80 (Wolter and Bychkov, 128–40).
16. Peter of Spain, Summaries of Logic 5.3 (Copenhaver, 199).
Chapter 4 intelligible things. A potential infinity of things is actually infinite if all those things are actual simultaneously. Intelligible things are a potential infinity with respect to created intellects. The First Nature understands everything and it understands everything it understands simultaneously (by Theorem
4.8). Therefore it understands an actual infinity of intelligible things.
Proof of the major premise of the proof of the antecedent. The major premise of the above argument is that intelligible things are a potential infinity with respect to created intellect. The proof of the premise is simply that its denial entails that, given enough time, a created intellect could eventually under stand everything. Scotus asserts that no created intellect could come to understand everything, from which the premise follows.
Proofs of the consequence. Here Scotus offers two proofs that the conclusion of the original proof (only an infinite intellect could understand an actual infinity of things) really does follow from the premise of the original proof— the intellect of the First Nature understands an actual infinity of intelligible things. The first proof goes something like this: the more things an intellect understands at once, the greater the power of that intellect. The First Nature understands an actual infinity of things at once (by Theorem 4.8). Therefore it is an intellect of infinite power of understanding.
The second proof of the consequence is somewhat similar to the first: it holds that the more things an intellect understands distinctly by one and the same act, the more perfect (i.e., complete) is that act. The intellect of the First
Nature understands distinctly an actual infinity of things (by Theorem 4.8).
Therefore its act is infinitely perfect.
4.69(50) Objections. Scotus considers two closely related objections to these proofs. A greater number of things understood does not entail a greater power of understanding, since many things might be understood through one and the same concept, or by one and the same intellectual act.
Replies. The gist of Scotus's replies is that while many things of the same type might indeed be understood through one and the same concept, or by one and the same intellectual act, this is irrelevant. After all, according to his doc trine of haecceity (literally, thisness), each individual thing has some feature in virtue of which it is the very individual it is—some feature not shareable with any other thing.17 Thus, the one act of understanding by which the First
Nature understands every intelligible thing is an act that extends not just to an actual infinity of things, but an actual infinity of types of things.
17. Duns Scotus, Ordinatio 2.d.3.p.1.q.1–6 (Spade, 57–113).
Commentary The proof from numbers and shapes. Scotus concludes with a brief argument, drawn from Augustine, according to which there are infinitely many num bers and shapes, and the First Nature understands them all, therefore, etc.
4.70(51)-74(55) Second proof of infinity: second proof from intellect 4.70(51) For any cause, C, its causality could be strengthened by the assis tance of an additional cause if and only if C is finite. But the causality of the First Nature cannot be strengthened by the assistance of any additional cause. Therefore the First Nature is infinite.
Intuitive cognition. To show that the causality of the First Nature cannot be strengthened by the assistance of any additional cause, Scotus considers the source or ground of the First Nature's knowledge of whatever it knows.
Here Scotus introduces a sort of knowledge he calls "vision." But by "vision," he means not the sensory power of sight but a type of cognition he calls intuitive cognition and describes as "knowledge precisely of a present object as present and of an existing object as existing," and also as "an intuition of a thing as existing and present."18
The feature of intuitive cognition most relevant to this proof is that— in finite intellects—the intuitively cognized object is a proximate cause of the knowledge of that object, in whatever intellect intuitively cognizes that object. Thus, the knowledge or understanding of the intuitive cognizer is made more perfect, or is "strengthened," by the causality of the intuitively cognized object.
The First Nature's intuitive cognition is uncaused. But the First Nature does not depend on the causality of any object for its knowledge of that object, since by its self-understanding alone it has perfect understanding of everything that can be known (by Theorems 4.2 and 4.8). Therefore no additional cause strengthens the First Nature's understanding in any way. And therefore its understanding is infinite. Since the First Nature's understanding is the same as the First Nature itself (by the third corollary of Theorem 4.6, and Theorem
4.8), the First Nature itself is infinite.
4.71(52) First objection. This objection invokes a familiar theme of Aristote lian science, namely that the highest knowledge of a thing includes knowl edge of its cause(s).19 The objector then reasons that if the First Nature's
18. Duns Scotus, Quodlibet 6.19 (Alluntis and Wolter, 136–37); Lectura 2.d.3. p.2.q.2.n.285–90 (Vatican 18:321–23).
19. Aristotle, Metaphysics 1.1.981a24–27 (Barnes, 2:1553); Posterior Analytics 1.2.71b9–
12 (Barnes, 1:115).
Chapter 4 knowledge of (e.g.) the stone is not caused by anything, then the First Nature does not know the stone through any causes, and so is less perfect than it might be.
Second objection. This objection argues that the conclusion of the proof is too strong. Suppose that the First Nature by knowing itself knows everything at least as well as it would if its knowledge of other things were caused by those things. But, since each of these things is finite, it may well be that the First
Nature can know them all by knowing itself and yet not be infinite.
Third objection. This objection concedes that the First Nature's knowledge of (e.g.) the stone is at least as perfect as it would be if the stone caused that knowledge, but asserts that the existence of the stone would indeed amount to some addition in perfection: instead of just the First Nature by itself, there would be the First Nature plus the stone, and that pair would be overall more perfect than the First Nature by itself. But if the perfection (e.g., of the total ity of existing things) could be increased by the addition of the stone, then the First Nature is not infinite after all.
4.72(53) Reply to the first objection. The First Nature does indeed have perfect knowledge of, for example, the stone, even though it depends neither on the stone's causes nor on the stone itself for its knowledge of the stone. To see this, grant (as the objector does) that the First Nature really can know the stone without the causal input of the stone itself. The question then is whether, in not knowing the stone through the stone's causes, the First
Nature's knowledge of the stone is less perfect than it would be if it did know the stone through the stone's causes. And Scotus's answer, of course, is that the First Nature's knowledge would be no more perfect. The reason for this answer is that while the stone's causes explain the stone's existence, they do not explain the stone's causing the "vision" (i.e., intuitive cognition) of itself in a knowing subject—"that vision would exist even if the thing seen were uncaused by such a prior cause," as Scotus says.
Still, the reply might seem like a dodge. When the objector says that a thing's causes can produce knowledge of the thing (e.g., the stone) that is more perfect than the knowledge that the stone alone can produce, surely what is meant is that the intellect that has knowledge both of the stone and the stone's causes, has more perfect knowledge of the stone than the intellect that is ignorant of the stone's causes. Understood this way, Scotus's reply— that the stone's causes don't contribute anything to the stone's causing the "vision" of itself—seems irrelevant. At the same time, understood this way, the objection is toothless, since the First Nature would know each of the stone's causes in the same way it knows the stone—through its one act of self-understanding.
Commentary 4.73(54) Reply to the second objection. According to the second objection, if the First Nature's knowledge of things is not caused by those things, then all we can conclude is that the First Nature eminently possesses the perfections of the things it knows—we cannot conclude that it is infinite.
Formal and eminent perfection. Scotus's reply makes use of a distinction between the formal and eminent possession of a perfection. To have a per fection, P, formally is to be really characterized by that perfection. To have
P eminently is (i) to have a perfection, P*, that is more perfect than P, and (ii) on account of P*, to be able to cause something else to have P formally. Sco tus claims here that among finite things—and only among finite things—a cause that has P eminently becomes greater or more excellent when it has an effect that has P formally. But the First Nature is not made greater or more excellent in any way by the existence of any of its effects (since, by Theorem
3.14, it is first in eminence). Therefore it is infinite.
Now the reply itself is actually unclear about what exactly is made greater by the addition of an effect that formally possesses a perfection. As far as the text itself goes, it could be either that the universe is better by such an addi tion, or that the perfection is made better, or that the cause is made better. In the explication of Scotus's reasoning I offered in the preceding paragraph, I went with the cause because it is the only option that makes any sense in light of the original proof and the second objection.
First caused nature. At the end of the reply Scotus offers an example that he himself would reject but which is "postulated by some." The "some" is at least
Avicenna and the view is that God does not directly create the whole world, but emanates an Intellect (which emanates Soul, and so on, down to the familiar creatures).20 So Intellect is, as Scotus puts it, the first caused nature.
According to Scotus's gloss, this nature eminently possesses every possible perfection that a creature can formally possess. But the first caused nature produces effects that formally have the perfections it has eminently, and the universe as a whole is more perfect thereby.
4.74(55) Reply to the third objection. I cannot claim to understand this reply.
Here is a hard-won guess about at least part of what is going on. The First Nature can produce in creatures every perfection that creatures can have.
For any perfection, P, which the First Nature can produce in creatures, the First Nature can produce it either immediately or through secondary causes
(where at least one of its secondary causes possesses P formally). In neither case does created perfection add anything to the First Nature—neither its
20. Avicenna, The Metaphysics of "The Healing" 9.4.5–6, 11 (Marmura, 328, 330).
Chapter 4 causality nor its being. The reason seems to be that P could not exist formally in any creature unless the First Nature possesses P formally or eminently.
But then the First Nature's causality is infinite, since, for any perfection, P (and there are infinitely many such perfections), no created cause makes the First Nature any more able to produce P than the First Nature is all on its own. For example, suppose there is an infinite ascending series of per acci dens ordered causes, in which a horse is produced by horses, which itself is produced by horses, and so on, ad infinitum. Then, for every horse that exists or has existed, it was produced by horses. Suppose that not even God could make a horse all on his own; suppose, that is, that for any horse God wants to make, God can make it only by causing the causality of another horse.
Then it turns out that there is a perfection, being a horse, which can only come about through other creatures that have the same perfection. So at first glance it looks like horses add to God's power to create horses, and so God's causal power is not infinite. But Scotus denies the inference, because even in this circumstance, nothing at all could have the perfection of being horse unless God were causing the universe (including its horses) to exist, at every moment of its existence, as its essentially or per se ordered first cause. Thus,
God would still be the ultimate source of the perfection of being a horse.
4.75(56)-76(59) Third proof of infinity: third proof from intellect Theorems 4.6 and 4.7 yield that the First Nature's act of understanding is a substance. This proof tries to show that if an act of understanding is a sub stance then it is infinite.
To make some headway into this proof, start with the idea that all acts of understanding other than the First Nature's belong to a certain species of accident. An accident, in Scotus's Aristotelian vocabulary, is a non-essential property of a substance. In this context, the relevant substance is a person
(e.g., a human being or an angel) and the relevant species of accident is an act of understanding, a species of accident in the genus of quality (as opposed to quantity, relation, etc.). A species includes both a genus and a specific difference.
So, let us say that the species intellectual act in the category of quality has qual ity for its genus, and intellectual act for its difference. In Scotus's terminology, the specific difference contracts, or limits, the genus. If we consider the genus on its own—quality—it is indifferent or open to any species of quality. So in a species it is the specific difference that does the job of contracting or limit ing the genus to the particular species that the species is.21
21. Duns Scotus, Ordinatio 1.8.p.1.q.3.n.100–109 (Vatican 4:199–203); Lectura 1.8.p.1. q.3.n.99–105 (Vatican 17:33–36); Questions on the "Metaphysics" of Aristotle 7.q.19.n.51–53
(Etzkorn and Wolter, 2:322–23).
Commentary In the proof Scotus is asking us to consider the species, intellectual act in the category of quality, and isolate that difference, intellectual act. What would it take for that difference to belong to substance rather than to quality?
Scotus answers that what it would take is for the substance to be infinite.
To see how he gets there, consider his claim in the first sentence of the second paragraph of the proof: "things that agree in the formal nature from which their specific difference is received, agree also in genus, if each formal perfection is finite." The "formal perfections" here are the genus and the spe cific difference. Where each of these is finite, he says, then wherever there is some particular specific difference, intellectual act, that specific difference contracts one and the same genus—in this case, quality. On this principle, there couldn't be an intellectual act in the category of substance, where sub stance is finite. Where genus and specific difference are both finite, they are bound to each other. Thus, if an act of understanding is finite, then it is an accident; but the First Nature's act of understanding is not an accident (by Theorem 4.7); therefore it is infinite.
This principle might seem ad hoc—Scotus doesn't offer any reason for accepting that a finite specific difference can only contract one and the same genus. And strictly speaking the proof only works if this principle holds water.
But there is a deeper implicit principle at work here, which does seem to be well motivated and which at least helps us see what Scotus is driving at: genus is always finite, and therefore the specific difference of a genus must also always be finite. To see this, consider that to be a genus is to be that which can be con tracted by specific difference. Limitation and so finitude are built into what genus is. A specific difference of a genus by definition contracts a genus into a species, and so is bound to be limited to whatever range of powers are built into the genus. Thus, intellectual act, just insofar as it is the specific difference of the species creaturely act of understanding, is itself limited by the nature of its genus.
But if we can understand intellectual act not as a specific difference but in itself, according to its formal nature or absolute nature, then we can begin to see how that very nature, or formality, is not intrinsically limited by any genus.
And this is exactly what Scotus wants to exploit. If we can detach intellectual act from its role as the specific difference of a genus, then we can consider it on its own. And what we find, or at least what we have no reason to reject, is that there is nothing about intellectual act that expresses any limitations at all. But where an intellectual act is a substance (as is the case with the First
Nature according to Theorems 4.4, 4.6, and 4.7), the intellectual act is not a difference of any genus and so is unlimited, that is, infinite.
77(60)-77(62) Fourth proof of infinity: from simplicity 4.77(60)-(61) Scotus says this fourth proof is similar to the third, and so it is. The First Nature belongs to no genus (by Theorem 4.1) and so whatever
Chapter 4 formal features it possesses—such as intellectual act, as discussed in the pre vious proof—it possesses not as specific differences contracting a genus, but in some other way. By simplicity (Theorem 4.1, again) the only way that the First Nature could possess these formal natures is by identity: one formal nature in the First Nature is not formally the same as another formal nature in the First Nature, but it is really the same as it. And where a formal nature is not limited as a genus or a difference, it is infinite.
4.77(62) This paragraph briefly sums up the four proofs of infinity Scotus has already offered: three drawn from the First Nature's intellect, and one from simplicity. The next three are taken from the three types of essential orders central to the Treatise: eminence, finality, and efficiency.
4.78(63)-79(65) Fifth proof of infinity: from eminence 4.78(63) Being first in eminence entails that there is no nature such that it is or can be more eminent than the most eminent nature. But for any finite nature, it can be surpassed in eminence. Therefore the most eminent nature is infinite.
Two arguments for the minor premise. Scotus offers support for this refresh ingly tidy proof by giving two arguments for the minor premise—that is, that any finite nature can be surpassed in eminence. According to the first argument, the infinite is not repugnant to being, that is, it is compatible with being, that is, infinite being is possible. But the infinite is greater than any finite thing. Any finite being therefore can be surpassed in eminence, and so fails to be first in eminence. According to the second argument, anything that can be infinite is as perfect as it can be only if it is infinite. Infinity is not repugnant to being, and there is a most eminent being. Therefore, the most eminent being is infinite.
4.78(64) Defending the minor premise of the arguments for the minor premise.
Scotus then suggests that the minor premise of this pair of arguments—that "the infinite is not repugnant to being"—stands in need of no proof. He then explains why the premise does not need to be proved: where repugnance cannot be found, non-repugnance should be assumed. We can detect noth ing in being as such that entails that every being is finite. Therefore we should assume that being and infinite are non-repugnant, or compossible. And if so, the minor premise holds.
Scotus then offers three additional defenses of the premise that the infinite is not repugnant to being, which together elaborate the core reason for accepting the premise: we have no good reason to think that there cannot be infinite being; therefore we should think there can be.
Commentary 4.79(65) Enhancing Anselm. Anselm described God as "something than which nothing greater can be thought" and then developed an argument to show that such a thing exists not only "in the understanding" but actually exists.22 Anselm's argument is frequently interpreted as relying on some ver sion of the claim that it is greater to exist in reality than in the mind alone. On some interpretations of Anselm, existence in reality is a property that makes a thing that has it greater than it would be if it lacked that property, and this in turn is supposed to function as a premise in an argument for God's existence.
But Scotus offers a rather different Anselmian argument. First, he adds a little precision to Anselm's formula, to suit his purposes. God is something thought without contradiction than which nothing greater can be thought without contradiction. The point of these additions is to assert the possibility that God exists. As the proof of infinity from eminence instructs, where we find no repugnance or contradiction, we should assume possibility.
Second, rather than focusing on existence in the mind and existence in reality, Scotus focuses on esse essentiae (the being of essence) and esse existen tiae (actual existence). Here, Scotus's claim that God has esse essentiae just means that there is an essence—for example, the divine essence—which is coherent and therefore lacks contradictions. When you think about the essence of God you can't find repugnance or contradiction in it; Scotus wants you to infer from this that it is possible that such an essence actually exists.
So, assuming that it is possible that God exists, we can go on to ask whether God actually exists. Now, by Theorem 3.3, it belongs to the nature of God (insofar as he is first in the order of efficiency) to be unable to be caused. Thus (as the proof of Theorem 3.4 argues), if God does not exist, then it is not possible that God exist. But it is possible that God exists. Therefore
God actually exists.
Because God actually exists and is something thought without contra diction than which nothing greater can be thought without contradiction, it follows that God is greater than anything that exists only in the mind—that is, something that has esse essentiae and whose nature does not entail that it exists. But this assertion that God is greater than anything existing in the mind alone does not, for Scotus, function as a premise of his Anselmian argument.
A second enhancement. Scotus offers a second way of enhancing Anselm's argument, a way that relies on his doctrine of "vision" or intuitive cognition.
The aspect of intuitive cognition most relevant here is that there can only be an intuitive cognition of something that actually exists. By contrast, the other kind of cognition, abstractive cognition, is such that an intellect could have
22. Anselm, Proslogion 2 (Williams, 81–82).
Chapter 4 abstractive cognition of something that does not exist—as you might have if you learned enough about dodo birds, which do not exist anymore. Scotus's point here is that if something can be both intuitively and abstractively cog nized, then it is more thinkable than something that can only be abstractively cognized. Scotus seems to think that if a thing is something than which nothing greater can be thought, then it is more thinkable than anything else that can be thought. I'm not sure this follows. But suppose it does. Then, if something is more thinkable if it can be both intuitively and abstractively cognized, and if a thing's being able to be intuitively cognized implies that that thing actually exists, then something than which nothing greater can be thought actually exists.
4.80(66) Sixth proof of infinity: from finality The First Nature is the first in finality (Theorem 3.15) and so is the ultimate end of everything. So it is the ultimate end of our willing. This ultimate end is either a finite or an infinite good. But no finite good can fully satisfy our willing. But it is contradictory that the ultimate end should be unable fully to satisfy our willing. So the ultimate end is an infinite good. So the First
Nature is infinite.
4.81(67)-85(72) Seventh proof of infinity: from efficiency This argument, adapted from Aristotle, reasons that only a nature with infinite power is capable of efficiently causing an infinite number of effects simultaneously. The First Nature is capable of this. Therefore it is infinite in power.
4.81(67) Infinite motion. The relevant texts from Aristotle assert that the first mover produces its effects over an infinite duration, and infer from this that the first mover is infinite in power. To move "with infinite motion" in this context does not mean or imply that the first mover itself is in motion, or is moved by something else. Instead, it moves only in the sense that it moves other things. Scotus "enhances" the inference in two ways, and then qualifies these enhancements, before producing his own version of the argu ment in 4.84(70).
First enhancement. The argument does not need to suppose that the First Nature does in fact move with an infinite motion; it need only suppose that it can move in this way.
Second enhancement. If motion is infinite then there is an infinite number of possible effects. Therefore the First Nature has power to produce an infinite number of effects and therefore is an infinite power.
Commentary 4.82(68) Objection to the first enhancement. The objection says that the degree of power required for producing one effect over a short duration is the same degree of power required for producing an infinite number of effects over an infinite duration. Therefore, from the fact that the first mover can produce an infinite number of effects over an infinite duration, it does not follow that it has infinite power.
Objection to the second enhancement. Similarly, this objection grants that a first mover might be able to produce an infinite number of individuals, but asserts
(following Aristotle) that there is merely a finite number of species. And the ability to produce individuals of a merely finite number of species does not entail infinite power.
4.83(69) First objection to the objections. In 4.68(48)-69(50), Scotus argued that since the First Nature, by understanding itself, understands an infinite number of intelligible things, the intellect of the First Nature is infinite.
Here in the objection, the rhetorical question implicitly asserts a fortiori that the nature with the ability to cause the existence of an infinite number of things has infinite power.
Second objection to the objections. Similarly, in 4.70(51)-74(55), Scotus argued that the First Nature's intuitive cognition (or "vision") of any intelligible thing does not depend in any way on anything outside the First Nature.
He inferred from this that the First Nature is infinite in intelligibility and infinite in being. A fortiori, then, the nature that can cause not merely the being seen but the existence of an infinite number of things is infinite in power.
The nature proximate to it. This enigmatic qualification at the end of the objec tion is another reference to Avicenna's "first caused nature." See commentary on 4.74(54).
4.84(70) Reply to the first objection to the objections. In the argument offered in 4.68(48)-69(50), a key premise is that the intellect of the First Nature understands an infinity of intelligibles simultaneously, not successively. So here: on the assumption that the First Nature can produce an infinite num ber of things simultaneously, not merely successively, it follows that the First
Nature has infinite power.
Objection to the reply. The problem, however, is that an infinite number of things cannot exist simultaneously—the "nature of the effect does not per mit it," as the objection says. The objection makes sense if we recognize that "infinite number of things" implies "every intelligible thing whatsoever." Then it is easy to see that an infinite number cannot be produced simultaneously: after all, it is intelligible that this ball is white all over at such and such a time
Chapter 4 and place, and it is equally intelligible that the same ball is black all over at the same time and place. But the ball cannot be white all over and black all over at that time and place. So the existence of the white ball excludes the existence of the black ball, and therefore at least one intelligible thing cannot be produced simultaneously with all others.
But this impossibility, or "defect," on the part of the effects (namely, that they cannot all exist simultaneously), does not entail that the First Nature does not have sufficient power to bring them about simultaneously, were they able to exist simultaneously.
Reply to the objection. But showing that the impossibility of producing an infinite number of things simultaneously is due to the nature of the effects and not due to the nature of any causal power, is not enough to show that the First Nature's power is infinite. For suppose that it has power sufficient to produce an infinite number of things simultaneously, but only by producing at least some things together with a secondary cause. In this case the second ary cause would add to the First Nature's power, and therefore (according to 4.70(51)) the First Nature's power would not be infinite.
Reply to the reply. In reply the imaginary interlocutor distinguishes between two ways of possessing a causality (or causal power): eminently and formally.
This refers back to 4.73(54), where Scotus distinguishes between possess ing a perfection eminently and formally. Recall that to have a perfection, P, formally is to be really characterized by that perfection. To have P eminently is (i) to have a perfection, P*, that is more perfect than P, and (ii) on account of P*, to be able to cause something else to have P formally.
Having a causality, or causal power, is one kind of perfection. We can distin guish as many different types of causalities as there are possible effects. So cau salities are distinguished from each other based on their objects. Thus, there is the causality of wisdom (i.e., the causal power to make something wise) and the causality of caninity (i.e., the causal power to make a dog), and so on.
To possess a causality, P, formally, then, is to be able to produce the object of that causality—for example, to be able to make dogs or make things wise.
To possess a causality, P, eminently, however, is (i) to have a causality, P*, that is more perfect than P, and (ii) on account of P*, to be able to cause some thing else to have P formally.
The claim here in this objection to the reply is that it has been well proved that the First Nature has every causality (at least) eminently. Since there is an infinite number of causalities, it follows that the First Nature is infinite in power.
Scotus's own solution. Scotus endorses this last reply and offers his own final "enhancement" of Aristotle's original argument. If the First Nature has
Commentary formally every causality in such a way that it could produce every effectible thing at once (if it were possible for every effectible thing to exist at once), then it is infinite in power. The First Nature has eminently every causality in just this way. If the First Nature is not infinite in power, then it is less perfect than it would be if it had every causality formally. But to have a cau sality eminently is more perfect than it is to have it merely formally. There fore, a fortiori, because it has every causality eminently in such a way that it could produce every effectible thing at once, the First Nature is infinite in power.
4.85(71) Infinite power and omnipotence properly speaking. Here Scotus clar ifies that by proving infinite power he has not thereby proved omnipotence properly speaking. To have omnipotence in this strict sense is to be able to be the total cause of every effectible thing—that is, to be able to produce everything that can be produced without any secondary causes.23 Nothing that Scotus has said here in this seventh proof of infinity yields that much stronger conclusion. To see this, suppose that the First Nature has some cau sality merely eminently, and not formally. Call the object of this causality,
A. Then, in order to produce A, the First Nature must produce something else that formally has the causality to produce A. Then the First Nature is not the total cause of A. Instead, Scotus is here content to reason that, since
(i) the First Nature can cause anything producible (either by its power alone or together with a secondary cause), and (ii) there are infinitely many pro ducible things, then assuming that (iii) the First Nature has sufficient power to cause these simultaneously (despite the fact that due to the natures of these possible effects they cannot all exist simultaneously), therefore (iv) the First Nature has infinite power.
Omnipotence properly speaking does not add to perfection in power. Scotus then offers two reasons for the following claim: if the First Nature "had all at once what was required to be a total cause, it would be no more perfect than it now is when it has what is required to be the first cause." The gist of the claim is that even if the First Nature could be shown to be omnipotent in the strict sense, it would not thereby be shown to be more perfect in power than this seventh proof of infinity has shown the First Nature to be. The first reason is that if the First Nature does require a secondary cause to produce some effect, this requirement is not due to any deficiency in the First Nature's power but rather to a deficiency in the nature of the effect.
Second, an agent is more perfect if it has a perfection eminently than if it has it formally.
23. Duns Scotus, Ordinatio 1.d.42.q.un.n.9 (Vatican 6:343).
Chapter 4 4.85(72) Reply to the second objection to the objections to the enhancements.
This reply reaches back to the objection offered at 4.83(69), which drew on Scotus's earlier argument that the First Nature does not depend on other things for its intuitive cognition of everything that can be intuitively cog nized. The crucial premise of that argument is that an essence is infinite if no finite thing can add to its perfection. Since no finite thing can add to the per fection of the First Nature's intuitive cognition—because it does not depend on anything for its intuitive cognition of whatever it intuitively cognizes—it follows that the First Nature's intellect is infinite. But being the first cause of the existence of all finite things does not, on its own, establish that no finite thing adds to the perfection of the First Nature. Therefore being the cause of the existence of all finite things does not, on its own, establish that the First
Nature is infinite in power.
As an illustration of his point, Scotus refers to the "highest created nature," another reference to Avicenna's "first caused nature." (See commen tary on 4.73(54).) Suppose the First Nature produces everything it produces by immediately producing this highest created nature, which in turn goes on to produce everything else. Then the First Nature is the total cause only of one thing, this highest created nature. But from this fact alone, that the First
Nature is the total cause of just one finite thing, we cannot infer that the First Nature is infinite in power. And the same reasoning holds if we suppose that God has created any finite number of things.
86(73) Eighth proof of infinity (rejected): from creation This is the first of two rejected proofs of infinity. Here the argument seems to be this: the extremes of creation are non-existence and existence. Creation is in some sense like a change from non-existence to existence. But there is an infinite distance between non-existence and existence. Therefore only an infinite power can, so to speak, bring something from non-existence into existence.
Scotus rejects this argument both here and in his latest Sentences com mentary.24 He does not deny that God creates ex nihilo, and therefore in some sense causes things to exist, which once did not exist. But he denies that it follows from this that there is an infinite distance between non-existence and existence. This is because if there is anything like "distance" between a non-existent and an existent thing, it is merely a finite distance. (By analogy, the distance from zero to one is one, not infinity.) Thus, it cannot be inferred that only an infinite power could bring something from non-existence into existence.
24. Duns Scotus, Reportatio 1-A d.2.p.1.q.1–3.n.58–62 (Wolter and Bychkov, 1:131–33).
Commentary 87(74) Ninth proof of infinity (rejected): from the lack of intrinsic causes
This second rejected proof reasons from the lack of intrinsic causes of the First Nature, to its infinity. Intrinsic causes are material and formal causes.
Since the First Nature has no efficient cause, it has no material cause (by Theorem 2.6), and since it has no material cause it has no formal cause (by Theorem 2.7). Aquinas had argued that form and matter are mutually lim iting: form limits matter because by its nature matter is indeterminate with respect to any form; matter limits form because by its nature form is share able by potentially infinitely many things.25 Therefore, the argument goes, something that by nature has neither a material nor formal cause is infinite.
Scotus rejects this argument, here and in two of his Sentences commentar ies.26 He points out that on Aquinas' own view, an angel is by nature imma terial but is not therefore infinite.27 Scotus adds a few more examples to reinforce the general point: being finite is not ultimately due to something extrinsic to a nature but rather to what a nature essentially is. Therefore, from the fact that the First Nature lacks the limiting principles of matter and form, we cannot determine whether or not it is infinite.
The rejection of the ninth proof of infinity marks the end of this long succession of proofs of Theorem 4.9.
4.88(75)-90(83) Proving Simplicity from Infinity 4.88(75) Theorem 4.10: Every kind of simplicity follows from infinity.
As promised in 4.49(2), Scotus now returns to the simplicity of the First Nature, arguing first that it has no parts at all, then arguing that it has no quantitative parts, and finally arguing that it has no accidents.
4.88(76) First type of simplicity: no parts If the First Nature is composed of parts, its parts are either finite or infinite.
If finite, then the First Nature is finite. But the First Nature is infinite (by Theorem 4.9). Therefore its parts are not finite. If its parts are infinite, then a part of the First Nature is "not less than the whole." But every part is less than the whole of which it is a part. Therefore its parts are not infinite. There fore the First Nature is not composed of parts.
25. Thomas Aquinas, Summa theologiae 1a.q.7.a.1.corp. (Hause and Pasnau, 85).
26. Duns Scotus, Ordinatio 1:d.2.p.1.q.1–2.n.140–44 (Vatican 2:211–13; PW, 74–75);
Reportatio 1-A d.2.p.1.q.1–3.n.52–57 (Wolter and Bychkov 1:130–31).
27. Thomas Aquinas, Summa theologiae 1a.q.50.a.2.corp.
Chapter 4 This laconic argument is not fully satisfying. First, it seems true that no finite number of finite parts can compose a whole that is infinite, but per haps an infinite number of finite parts could compose a whole that is infinite.
Scotus does not consider this possibility.
Second, it is not at all clear that an infinite part would be "not less than the whole" of which it is a part. Suppose that the Infinite Nature's several per fections are parts, and that each is infinite. Then compare one of the parts— infinite wisdom, say—with the First Nature as a whole. The First Nature would have infinite wisdom along with several other infinite perfections.
Plausibly, then, it would be greater than any one of its parts.
But perhaps in this case the First Nature would not be greater than the sum of its infinite parts. Since, according to third and fourth proofs of infin ity (4.75(56)–4.77(61)), nothing infinite can enter into the sort of compo sition in which one part is determinable and another determining (potency and act, respectively), were the First Nature composed of infinite parts, these parts would not make up a genuine unity; they would be a metaphysical
"heap" or mere sum of parts.
4.88(77)-(79) Second type of simplicity: no quantitative parts 4.88(77)-(78) Scotus here entertains the idea that the First Nature, infinite in perfection, is or has a magnitude. If so, then it has quantitative parts. Ulti mately, as the conclusion of 4.88(78) makes clear, the First Nature cannot have or be a magnitude because magnitude requires a material subject and the First Nature is immaterial (see Theorem 2.6). But Scotus first proceeds by considering magnitude as such, in abstraction from matter, and arguing that infinite perfection cannot exist in a magnitude and so has no quantita tive parts.
Infinite perfection cannot exist in a finite magnitude. Logically, a magnitude is finite or infinite. So showing that infinite perfection cannot exist either in a finite or infinite magnitude is sufficient to show that it cannot exist in a magnitude at all. Start with finite magnitude. Scotus arrives dialectically at the conclusion that infinite perfection cannot exist in a finite magnitude. The first effort to establish the conclusion goes like this: any finite magnitude can increase in size. Plausibly, as a magnitude increases, its total amount of perfection increases. But infinite perfection cannot be increased. So it cannot exist in a magnitude.
Scotus is not satisfied by this argument. He considers the possibility that infinite perfection, if it were in a magnitude, would exist completely in each part of the magnitude, rather than spread out across the magnitude. (It would be "whole in the whole and whole in each part," rather than "whole in the whole and part in each part.") If so, then an increase in magnitude would
Commentary indeed result in an increase of the magnitude that has infinite perfection, but it would not result in an increase of the perfection of the magnitude.
But Scotus is not satisfied by this argument, either. He reasons that while an increase in magnitude may not increase the perfection of the magnitude, it would increase its "power of efficiency," that is, its total efficient-causal power. By analogy, a small fire has the total (finite) perfection that fire is, but a large fire has greater burning power than a small fire. So a finite magnitude, however perfect, could always increase in power. But the First Nature has infinite power of efficiency (by the seventh proof of Theorem 4.9). Therefore it does not exist in a magnitude.
Infinite perfection cannot exist in an infinite magnitude. This is because an infinite magnitude cannot exist. Aristotle's own view is that magnitude is potentially but not actually infinite. It is (potentially) infinitely divisible and (potentially) infinitely increasable.28 But a potentially infinite magnitude is actually finite. Since infinite perfection cannot exist either in finite or infinite magnitude, it cannot exist in magnitude at all, and therefore cannot have quantitative parts.
4.88(79) The immateriality of understanding. Scotus offers an additional, very brief argument that the First Nature has no quantitative parts. The act of understanding is immaterial and not extended and so has no quantita tive parts.29 The First Nature is understanding. Therefore it has no quantitative parts.
4.89(80)-90(83) Third type of simplicity: no accidents Aristotelian accidents. An "accident," for Scotus and the Aristotelian tradition, is a feature a thing can have that is not a feature of the essence of the thing.
Beulah the Cow is essentially bovine but only accidentally brown. Here Scotus approves two arguments for the conclusion that the First Nature has no accidents, but rejects an additional four arguments for the conclusion.
4.89(80) First proof. To be able to be the subject of an accidental property is to be able to be perfected in some way. But the First Nature is infinite in perfection and so cannot be perfected in any way. Therefore it cannot be the subject of an accidental property.
Second proof. Accidents are either material accidents or immaterial acci dents. Material accidents are accidents of things with quantitative parts. But 28. Aristotle, Physics 3.6.206a.14–17 (Barnes, 1:351).
29. Duns Scotus, Quodlibet 9.26–30 (Alluntis and Wolter, 225–28).
Chapter 4 the First Nature has no quantitative parts, as the arguments above show
(4.88(77)-88(79)). Therefore it has no material accidents. Immaterial acci dents are acts of will and understanding. But the First Nature's act of will and understanding is essential to it, not accidental. Therefore the First Nature has no immaterial accidents. Therefore it has no accidents at all.
4.89(81) Third through fifth proofs (rejected). Scotus offers three very brief enthymemes, the third through fifth proofs, in support of the conclusion that the First Nature has no accidents. He rejects all three on the grounds that while each shows "that an accident does not belong to the essence" of the First Nature, none shows "that an accident does not accidentally belong" to the First Nature. There is an implicit distinction here between an accident belonging to the essence of a thing, and an accident belonging acciden tally to a thing. Consider Beulah the Cow. Her essence is bovine and brown ness is one of her accidents. Notice that the accident of brownness belongs to Beulah, not to bovinity. Scotus's point in his objection is that none of these arguments rules out the possibility that the First Nature—considered here not as an abstract nature, divinity, but as a concrete thing, God—has an accident.
4.90(82) Sixth proof: from pure perfection (rejected). The sixth and last argu ment for the conclusion that the First Nature has no accidents reasons from the pure perfections. I myself cannot see how the proof is supposed to work, but, implicitly, the proof holds (i) that each pure perfection is "the best from itself" and (ii) that there cannot be several unqualifiedly best things. Scotus agrees with (ii) (see Theorem 3.15) but not with (i), as becomes clear in his objections to this proof. (Why anyone would hold (i), I'm not sure.) But, given (i) and (ii), the argument seems to go like this: If the pure perfec tions of the First Nature are not the same as the First Nature, then (iii) no pure perfection would be the best from itself and (iv) there would be several unqualifiedly best things. But not-(iii) and not-(iv). Therefore the pure per fections of the First Nature are the same as the First Nature. But that which is the same as another thing, is not an accident of that thing. (Also implicit: every feature of the First Nature is a pure perfection—this is never proved but see Theorem 4.2 and 4.3.) Therefore the First Nature has no accidents.
Pure perfection is not unsurpassable perfection. Now Scotus himself agrees that the First Nature's pure perfections are the same as the First Nature (this is entailed by simplicity, for which Scotus has at this point offered and endorsed several arguments). But he does not think that this can be inferred from the nature of pure perfection. A denominative pure perfection (e.g., a wise thing rather than wisdom) is better than anything incompossible with it (e.g., a
Commentary dog), but it is not thereby better than anything whatsoever. For example, perhaps it is not better than other pure perfections with which it is compos sible, or perhaps it is not better than the First Nature itself, with which it is compossible. And so on for all other denominative pure perfections. Thus,
Scotus here leaves it open that the First Nature both has all pure perfections accidentally, and is better than each pure perfection. Nothing about this sce nario contradicts the nature of pure perfection as such. (Of course, for other reasons Scotus thinks that in fact the First Nature is the same as its pure perfections and so is not better than any of them.)
4.90(83) Here Scotus sums up: he thinks the first two proofs are successful, but the last four are not.
4.91(84)-93(86) In Praise of God This beautiful metaphysical prayer is the rhetorical crescendo of the Treatise.
It can stand mostly on its own, without comment. But a few points are worth making.
4.91(84) Eternity. Here Scotus praises the First Nature for its eternity, which entails that it is "without the potential to exist successively." The First Nature's existence does not proceed from one moment to the next, but is enjoyed all at once. Here Scotus offers something like a reason for thinking this about the First Nature: there cannot be succession where there is nothing caused.
Scotus thus seems to derive the First Nature's eternity (or timelessness, as contemporary philosophers of religion would put it) merely from the fact that the First Nature is uncaused. This is a striking, but sadly undeveloped argument. Perhaps there is an assumption or intuition in the background, according to which, if something exists from one moment to the next, then there must be some causal explanation of how it gets there. Where a thing exists but has no cause, no causal explanation can be given. Therefore an existing, uncaused thing must not exist successively.
Immutability. Scotus says that he has discussed immutability earlier in the Treatise. But, like eternity, he does not discuss it explicitly. Probably he has in mind 4.89(80), in which Scotus infers simplicity from the fact that the First
Nature is infinitely perfect and so lacks any potency. From this it follows that the First Nature cannot change and so is immutable.
4.92(85) Truth and falsity. Here Scotus says that the false is "what is not as it appears to be." Falsity occurs when a thing looks one way, but is really some
Chapter 4 other way. Falsity thus implies a duality: the thing itself, and some other thing that explains why the first thing looks the way it does. Scotus's point here is that when there is nothing else to distort the appearance of a thing, there cannot be falsity. Since the intellect of the First Nature has just one source for all its knowledge (namely, its own essence), and nothing can dis tort that intellect's self-understanding, it cannot be the subject of any falsity.
That Greek and Platonic word. Scotus emphatically affirms the classical divine ideas theory according to which there is no need to postulate Platonic Forms or Ideas, since God's essence alone is the exemplar of everything intelligible.
4.93(86) Metaphysical claims. The Treatise has not appealed to revealed doc trines but only metaphysics, the science of being qua being. Recall the open ing prayer (1.1(2)) in which Scotus asks for God's help as he seeks to discover what can be known about God simply from the nature of being.
The subsequent treatise. There was once some speculation that the treatise to which Scotus refers is part of the work known as Theoremata. There is now authoritative consensus that this speculation is unfounded.30 If Scotus ever got around to writing this companion treatise, it has since been lost. But, given the late composition of the Treatise, it is probable that he never wrote its companion.
4.94(87)-97(93) The Uniqueness of God 4.94(87)-(88) Theorem 4.11: You are one God, besides whom there is no other, as you have said through the Prophet.
Why uniqueness? In Theorem 3.15 Scotus established that a nature with the Triple Primacy actually exists. In Theorem 3.19 he established that there is just one nature with the Triple Primacy. Then, throughout Chapter 4, Scotus derived various attributes of this one nature, sufficient for it to be God. At first glance, then, Theorem 4.11 might seem superfluous. But it is not. After all, for all Scotus has said so far, there could be several actual things with this one nature, just as there are many actual human beings sharing one human nature. If so, there would be several Gods. So here in the final theorem of the Treatise Scotus argues that there is and can be exactly one actual thing that has this nature. He offers five short proofs for the theorem. Each proof focuses on an attribute of the First Nature and argues that exactly one actu ally existing thing can and does have it.
30. Wolter, "Introduction" (Wolter, DPP, xiv–xvi); Kluxen, "Kommentar" (Kluxen, 252);
M. Dreyer, H. Möhle, and G. Krieger, "Prologomena" (Opera philosophica 2:578–79).
Commentary 4.95(89) First proof of uniqueness: from infinite intellect If there were two infinite intellects, A and B, A would understand B and B would understand A. It has already been established that an infinite intel lect understands everything it understands through itself, either as formally or eminently containing everything intelligible (see the second proof of infinity under Theorem 4.9, and especially 4.70(51) and 4.73(54)). But an infinite intellect cannot eminently contain an infinite intellect, because with respect to intellect they are equally perfect. Therefore A cannot understand
B through what A eminently contains, and vice versa. Neither can A under stand B through what A formally contains (and vice versa). There are several reasons for this. First, since each is infinite intellect, A does not depend on B for A's understanding of B (or vice versa)—see again the second proof of infinity, especially 4.70(51). But then, second, A's understanding of B would include only what A and B share in common, and would not include that by which A and B are distinct from one another (and vice versa). Thus, neither would understand the other perfectly. And third, since A's understanding of B would include only what they share in common, A would understand B
"under a universal" and therefore abstractively (and vice versa). But intuitive understanding is more perfect than abstractive understanding. But A cannot understand B intuitively, because A neither contains B eminently, nor con tains every feature of B formally, nor depends on B for its understanding.
Thus, again, neither would understand the other perfectly. Therefore each would lack perfect understanding and so would not be infinite intellects after all—which is a contradiction. Therefore there is only one infinite intellect.
4.96(90) Second proof of uniqueness: from infinite will The proof from infinite will assumes for the sake of argument not just that there are two infinite wills, A and B, but (implicitly) that there are two highest (infinite) goods (i.e., that which is "most lovable"), and that each infinite will is, necessarily, one of the highest goods. This implicit assumption makes the proof rather puzzling when considered alongside the fifth proof of uniqueness, which reasons explicitly from the nature of infinite goodness.
But with respect to his ultimate goal of proving the uniqueness of God, this is not problematic, since he has already shown that the highest good (i.e., the first in the order of eminence) is also infinite will (see Theorems 4.4 and 4.9).
The highest good is most lovable. An infinite will loves most what is most lovable. So A is the most lovable and B is the most lovable. But A cannot love both A and B the most. Scotus offers two reasons for this. First, A nat urally loves itself more than B. (He offers no reason for this claim and it is hard to see why it should be accepted, on the assumption that there are two highest goods.) Second, if B were destroyed, A would be no less happy. To
Chapter 4 see this, consider that one highest good is sufficient for the perfect happiness of every lover (see 4.80(66), the sixth proof of infinity), including an infinite will. So A's beatitude does not consist, even in part, in enjoyment of B (if it did, A's happiness would be diminished by B's destruction). But it contradicts the nature of infinite will that it should fail to enjoy the highest good. There fore there are not two highest goods. Therefore (by the implicit assumption of the proof) there are not two infinite wills.
4.97(91) Third proof of uniqueness: from infinite power This proof relies on a key feature of essential dependence, namely that it is dependence in the natures of things. But there is a unique nature that has infinite power (by Theorem 3.16 and the seventh proof (from efficient cau sality) of Theorem 4.9). So suppose that there are two things having this one nature, A and B. Then anything essentially dependent on A is also essentially dependent on both A and B, since they share the same nature (and essen tial dependence is dependence due to nature). But it cannot be essentially dependent on both A and B, since A and B would be co-firsts in an essen tial order of dependence, each contributing equally to their joint effect. The problem with this is that each would be sufficient for producing their joint effect (since each is infinite power), and neither would be dependent on the other—but then we would have a redundant dependence, which is ruled out by Theorem 3.16.
In closing, Scotus entertains again the possibility that something could essentially depend on just one of these infinite powers (already ruled out by the nature of essential dependence). He asks a rhetorical question meant to assert implicitly that there would need to be some sort of explanation of why, given two natures of infinite power, something essentially depends on one but not the other. There cannot be a deeper or more authoritative power among natures to arbitrate. So the two infinite powers would, so to speak, need to work it out for themselves which is to be the essentially prior. But since each is infinite power (and also, perhaps, since none is more eminent than the other) there would be no rational way to divide the labor. It would be, if possible, entirely random. And Scotus seems to reject the possibility that there could be total randomness at the top of an essential order of dependence. Therefore there is and can be only one infinite power.
4.97(92) Fourth proof of uniqueness: from necessary existence If there could be more than one necessarily existent nature, then there would be infinitely many of them. This is because, in general, a shareable property is infinitely multipliable (e.g., just considering human nature on its own, there
Commentary is no reason why there cannot be infinitely many humans). And necessary existence—necessary existence from itself, that is—is such that it is uncaus able (see Theorems 3.3, 3.4, and 3.5). Since everything that does not exist and is possible only comes to exist by a cause, it is impossible that there is a nature with necessary existence from itself that does not actually exist (see
Theorem 3.5). Therefore if there can be more than one thing with necessary existence, there are infinitely many such things. Now Scotus clearly intends this as a reductio ad absurdum—of course there can't be infinitely many neces sarily existing things! But, from the nature of necessary existence alone, it is not clear why it should be absurd that there are infinitely many such things, especially since Scotus has already granted the possibility of an actual infinity of things (see the first proof of Theorem 4.9).
To see why it is absurd that there should be infinitely many necessarily exist ing things, we need to look a little deeper. Scotus did argue, in Theorem 3.6, that necessary existence from itself belongs to just one nature. But the unique ness of the nature is not at issue here. Instead, we're concerned with whether or not there can be more than one actual thing with that nature. So Theorem
3.6 on its own does not offer much help. One plausible suggestion is that the principle of parsimony (Theorem 2.15) offers good, if far from demonstrative, grounds for supposing there is just one necessarily existing thing. From the proof of Theorem 3.1, we gather that the starting point of Scotus's proof of God's existence is the fact that something can be produced. Theoretically, what turns out to be needed to explain this fact is something Triply First that has all the features Scotus argues it has, including necessary existence. More than one necessary existent, therefore, is theoretically superfluous.
If more than this is required, in order to meet Scotus's goal of a demon stration of God's existence (see 3.26(6)), I suggest we need to consider what necessary existence is supposed to entail. The first proof of Theorem 3.15 argues that since the first in efficiency, the first in finality, and the first in emi nence all have necessary existence, it follows that there is just one nature that has all three primacies. So the unique nature that has necessary existence is first in these three ways. But from the Triple Primacy Scotus proved (over the course of Chapter 4) that there is just one nature that is infinite intellect, will, power, and goodness. And Scotus offers (more or less) independent argu ments that only one actually existing thing can have these features. The sug gestion, then, is that the impossibility of more than one necessarily existing thing can only be established by considering what else is supposed to follow from the necessarily existent nature. Admittedly, this suggestion weakens the argumentative force of this fourth proof of God's uniqueness, since it makes it logically dependent on the other proofs.
Chapter 4 4.97(93) Fifth proof of uniqueness: from infinite goodness Here Scotus offers two short proofs. First proof. Two infinite goodnesses would be better than one infinite goodness. But nothing can be better than infinite goodness. Therefore there is just one.
Second proof. By the nature of what it is to be a will, the will is fully "at rest"— satisfied, happy, blessed—by infinite goodness. If there could be two infinite goodnesses, then the will of a person aware of this possibility could not be fully at rest by one infinite goodness. But a will can be fully at rest in infinite goodness. Therefore there is just one infinite goodness. And so one God.
4.98(94) Closing Prayer Here ends the commentary on the treatise on the first principle by John
Scotus.