Secretum secretorum - De Mirabilibus Mundi - Universals
Background for Ars Magica sagas

Common Natures

Avicenna, Aquinas and Scotus

This is just our old friend, the problem of universals, back again with some new terminology and some new twists. On the topic of this section, the best single thing you could possibly read is the following paper: Joseph Owens, "Common Nature: A Point of Comparison between Thomistic and Scotistic Metaphysics," Mediaeval Studies 19 (1957), pp. 1-14.

Both Aquinas and Scotus have their point of origin on the question of common natures in some passages of Avicenna.


We start with Avicenna's Logica, where he is worried about predication. Recall the way Abelard had set up the question of universals. For him it was a question of "What is predicated of many?" If you say "things", then you are a realist; if you say "only words", then you are a nominalist. (This is not the way Boethius had set up the problem. Boethius says nothing about predication in this connection.)

When the question is framed this way, Avicenna is a realist, and for that matter, so will Aquinas and Scotus be realists. But they are realists with a difference. It is not the straightforward kind of realism found in Boethius' De trinitate, or in William of Champeaux's first theory.

Some Passages from Avicenna on Common Nature

The first two passages below are translated from the Latin text in Avicenna, Opera, Dominic Gundissalinus, tr., (Venice: Bonetus Locatellus for Octavianus Scotus, 1508). Unless there is a modern critical edition, this 1508 edition is the one usually cited. The last three passages are translated from the Latin text in Avicenna Latinus: Liber de philosophia prima sive scientia divina, S. Van Riet, ed., 2 vols., (Louvain: E. Peeters, and Leiden: E. J. Brill, 1977, 1980). For these last three passages, I have also given the folio reference to the 1508 edition.

(1) Logica, III, fol. 12ra:

Animal is in itself a certain something, and it is the same whether it is sensible or is understood in the soul. But in itself it is neither universal nor is it singular. For if it were universal in itself, so that animality from the fact that it is animality would be universal, it would be necessary that no animal be singular. Rather, every animal would be universal. If, on the other hand, animal from the fact that it is animal were singular, it would be impossible for there to be more than one singular, namely, the very singular to which animality is due. And it would be impossible for another singular to be an animal.

(2) Ibid.:

Now animal in itself is a certain something, understood in the mind that it be animal. And according as it is understood to be animal, it is nothing but animal only. If, on the other hand, beyond this it is understood to be universal or singular, or anything else, then beyond this, namely, that which is animal, there is understood a certain something that happens to animality.

(3) Metaphysica V, 1, Van Riet, ed., II, p. 228 lines 31-36; 1508 ed., fol. 86va:

Horsehood, to be sure, has a definition that does not demand universality. Rather it is that to which universality happens. Hence horsehood itself is nothing but horsehood only. For in itself it is neither many nor one, neither is it existent in these sensibles nor in the soul, neither is it any of these things potentially or actually in such a way that this is contained under the definition of horsehood. Rather [in itself it consists] of what is horsehood only.

(4) Ibid., Van Riet, ed., II, p. 231 lines 74-81; 1508 ed., fol. 86vb: Hence, if someone should ask whether the humanity that is in Plato, insofar as it is humanity, is other than that which is in Socrates, and we say no, as we must, we will not have to agree with him when he says, "Therefore, this one and that one are the same in number", because the negation was absolute and we understood in it that that humanity, insofar as it is humanity, is humanity only. But insofar as it is other than the humanity that is in Socrates, that is something extraneous. On the other hand, he did not ask about humanity except insofar as it is humanity.

(5) Ibid., Van Riet ed., II, p. 233 lines 36 - p. 234 line 44; 1508 ed., fol. 87ra:

Now animal can be considered by itself, although it is together with what is other than

Given this realism of predication, let us look at passages (1) and (2) of the text above. "Animal", Avicenna says, is in itself a certain something, but in itself it is neither singular nor universal. ('Singular' here means "individual".) This is passage (1). Here is the argument for this claim:

(a) If animal were in itself universal, then it could never be predicated of singulars. Anything that is an animal would be universal. There could be no singular or individual animals.

(b) If on the other hand animal were in itself singular, then it could not be predicated of many individuals. There would be only one animal.

So the facts of predication, that animal is predicated of singulars, and that it is predicated of several of them, require that all by itself it be neither singular nor universal. Passage (2) of the text tells us that, instead, universality and singularity are things that are accidental to animal.

What on earth is going on here? Let us look at passage (3), for some clarification. This is a passage not from Avicenna's Logica but from his Metaphysics. In this passage, he talks about horsehood (equinitas, "equinity") rather than animal, but the principle is the same. So let's continue to put it all in terms of animal, for the sake of uniformity.

What belongs to animal by itself, it seems, is exactly what is included in the definition of animal, and whatever else is entailed by that definition. No more, no less. Hence, A by itself is B if and only if B is entailed by the definition of A. ('A'' and 'B' are to be replaced by general terms here; we're talking about the problem of universals, after all.)

Now the definition of animal is 'sensate, animate, corporeal substance'. And that's all. There is no mention, for instance, of singularity or individuality. That is not built in or entailed. There is nothing there about this sensate, animate, corporeal substance. Neither is universality built into the definition. There is nothing said there, or implied there, about that.

It is animal, as just described, that is "predicated of many", this animal that in itself is neither singular nor universal. In fact, only it can be predicated of many. And it is what later authors will term a common nature. (The actual term 'common nature' is apparently not in Avicenna.)

Let us pause now to head off two objections, and thereby to clarify what is going on here.

(1) First objection: Step (a) of Avicenna's argument is fine, but step (b) is just a fallacy. If animal is in itself just whatever is included in or entailed by the definition of animal, then if animal were in itself universal, it would indeed follow that it could never be predicated of singulars. This follows on the general ground that if B is predicated of A, then the definition of B, and the various parts of the definition of B, are also predicated of A.

But no such consideration will support step (b) of Avicenna's argument. If animal is in itself singular or individual, it does not by any means follow that animal cannot be predicated of many individuals, as Avicenna says. All that follows is that every one of the many individuals of which animal can be predicated will be a singular. And that is hardly objectionable.

Reply ad (1): This looks like a good objection at first. Hence it is all the more instructive to see why it fails. The objection gets its plausibility, I think, from regarding predication as exclusively a matter of language, so that 'predicate' means 'predicate-term'. But that is not so for Avicenna. For him, the animal that is "predicated of many" is not just a piece of language; it is a piece of the world. Predication is a metaphysical relation primarily, of which predication in language is just a reflection.

For Avicenna, to say that animal, for instance, is predicated of Socrates, is to say that there is a certain metaphysical something out there (we know that it will later be calle d a "common nature"), and that this something enters into a certain intimate metaphysical relation with Socrates. The details of that relation remain to be seen, but we can say this much at least already: it is in virtue of this metaphysical something that Socrates comes to have all the features that are built into or entailed by the definition of animal. The reason Socrates is "sensate" (has sensation), for instance, is just that this feature is brought to Socrates, contributed, by the entity animal that we say is "predicated" of Socrates and that Socrates "has", in some sense yet to be explained. In other words, the reason Socrates is "sensate" is just that animal itself is "sensate".

The same thing holds, of course, for all the other features built into or entailed by the definition of animal. They are not just features of other things, of which we then say that animal is "predicated". They must also be features of animal itself. If animal did not itself have these features, then the presence of animal in Socrates and Plato and Browny the ass would hardly be enough to guarantee, as it does, that they have these features. That is, after all, where they get them.

What this means is that not only are Socrates and Plato and Browny the ass "sensate, animate, corporeal substances"; so is animal itself. In other words, animal itself is an animal! In general, common natures are self-predicable. They have to be, if they are going to do their job.

We are now in a position to see why part (b) of Avicenna's argument does hold after all. If animal were in itself singular, that is, if singularity were built into or entailed by the very definition of animal, then in virtue of self-predication, animal itself would be a singular. And if there is one thing we can say about singulars, it is that they are not "predicated of many". Remember how Abelard quoted Porphyry on this point.

The key to all of this is to think of "predication" in metaphysical terms, not as a relation confined exclusively to the realm of language.

(2) Second objection: I raised and responded to the first objection mainly to remove an obstacle to the correct understanding of this theory. But the second objection will lead into the heart of the theory itself. It goes like this:

Earlier, we saw that it is animal in itself, with neither universality nor singularity built in, that is "predicated of many", and that in fact it is the only kind of thing that can be "predicated of many". On the other hand, you may object, I thought Aristotle had defined the universal as what is "predicated of many". (Abelard has a quotation here too.) So doesn't it follow that the "common nature" is universal after all, even though we just said that it isn't?

Reply ad (2): Sure it is universal - no doubt about it. But it is not universal in itself. In other words, its being predicable of many is not something built into the definition, and it doesn't follow from what is built into the definition either. But that doesn't mean that animal is not predicable of many; it just means that it is not in itself predicable of many.

I suppose we had better take a closer look. The common nature can exist, Avicenna says, in two modes:

(a) In singular things - in particulars. Insofar as singulars have definitions at all (they do not properly have definitions in an Aristotelian context), such a singular animal would be defined as "this sensate, animate, corporeal substance", or something like that. Singularity does accrue to the common nature insofar as it exists in external things.

(b) In the soul - in concepts. This, of course, is the Aristotelian theory according to which the nature of the external object is in the mind of the knower. And in concepts, the common nature is abstracted from all singularity, all individuating conditions. In the concept, therefore, the common nature is universal.

The common nature animal by itself is neither singular nor universal. By itself it is just sensate, animate, corporeal substance, and whatever all that implies - but no more. The common nature animal, however, is singular - only not just in itself. That is something added on. It is singular in individual animals. Although the definition of animal does not mention singularity, the definition of this animal (if it had one) does.

Likewise, the common nature animal is universal - only not just in itself. That is something added on. It is universal in concepts. Although the definition of animal does not mention or entail universality, the definition of the concept "animal" does.

Now look at passage (4). The question of unity or multiplicity simply cannot be sensibly asked about the common nature taken by itself. Recall, on Boethius' realist view, the view of William of Champeaux's first theory and the view of Clarenbald of Arras, the humanity of Socrates and the humanity of Plato were one humanity. On Boethius' view in the Commentary on Porphyry, William of Champeaux's second theory and Gilbert of Poitiers' view, they were two humanities.

Avicenna is saying that, if you ask about the humanity that is in Socrates - insofar as it is humanity, and not insofar as it is in Socrates - there is nothing to make it any different from Plato's. There is nothing singular about it. But neither is there anything in the definition of humanity that says it is one in Socrates and Plato. There is nothing universal about it either.

Hence the common nature in itself has no unity. This is not to say that it lacks unity, in the sense that it is a multiplicity or plurality. It does not have that either. The question of unity or multiplicity simply doesn't arise at that level. It arises when we ask in what mode (see the two "modes" distinguished above) that common nature is taken as existing. But then we are not asking about the common nature in itself any more, but rather about the common nature as in singulars, or as in a concept.

Nevertheless, look at passage (5). Although common natures have no unity (or for that matter plurality) all by themselves, they do have a kind of being all by themselves. Although they exist only in individuals or in the mind, nevertheless they have some kind of being of their own. This is not the being of existence (esse existentiae, as people later put it). That they do not have all by themselves. They only get that insofar as they are in individuals, or alternatively, in the mind.

Rather this being that belongs to the common nature all by itself is a kind of lesser being - what came to called esse essentiae or the being of an essence. (This terminology is not Avicenna's, but was used by later Scholastics who got the basic idea ultimately from Avicenna.) The idea is that the common nature is an entity in its own right. And while by itself animal is neither singular nor universal, it certainly is all by itself a different entity from the common nature stone, for instance. Each common nature has an integrity of its own, as a kind of metaphysical block - and that is what we call its lesser being. It has some "ontological status", as it used to be fashionable to say; it is not absolutely nothing all by itself, even though all by itself it does not fully exist.

So we have a curious state of affairs when we put all these passages together. Common natures have a kind of lesser being, but they have no unity at all. (This is not to say, of course, that they are multiple or plural.) All this, of course, is with respect to the common nature in itself.

Now what do you suppose happened when you took this doctrine and injected it into Latin philosophy, which was by now thoroughly Augustinian, and very conscious of the Augustinian equation of being with unity? (This was pretty much accepted even by those Latin philosophers who were not otherwise especially Augustinian in their outlook. It was a kind of metaphysical axiom.) Avicenna is not quite so clear about this equation. In some places he seems to accept it, but here he does not. Apparently it was not an important matter for him. But it was an important matter for the Latins.

They could not tolerate Avicenna's doctrine here just as it stood: a lesser being but no unity at all. For them, being and unity went hand in hand. Hence, they could go one of two ways:

(a) They could take seriously Avicenna's denial of unity to common natures in themselves, and conclude that they have no being by themselves either, not even a lesser being, contrary to what Avicenna said.

(b) Or they could take seriously Avicenna's claim that common natures have a lesser being all by themselves, and conclude that therefore they have a lesser unity all by themselves too, despite Avicenna's denial that they do. On this alternative, you would distinguish kinds or grades of unity to go with kinds or grades of being.


According to Owens in his paper "Common Nature", Aquinas took the first road, Scotus the second. Let us look at Aquinas first. I am going to present the interpretation of Aquinas that Owens gives, because it represents what has come to be a more or less standard - or at any rate common - interpretation. I think there are serious difficulties with this interpretation as a philosophical theory, although as an interpretation it may be perfectly correct. But I will tell you about that later.

For Aquinas, on Owens' interpretation, the only kind of being a thing has is its act of existing, its esse or what later would be called its esse existentiae. Aquinas has no room for Avicenna's "lesser being".

In part, this is because of his theory of predication. Aquinas does not think that true judgments in general correspond part for part with the world. But he does think this is so in some cases - for example, in the judgment 'Socrates is a man'. Just as the predicate 'man' is there bound to the subject 'Socrates' by the copula 'is', so too the judgment so formed is true because in reality the common nature man is bound into the individual Socrates by an esse. If the common nature already had its own esse by itself, Aquinas thinks, then this would prevent that kind of composition. It would get in the way.

Let's look a bit more closely at Aquinas on common nature. With one small addition, this discussion can be turned in to a partial commentary on the notion of form in the very obscure Chapters 2 and 3 of On Being and Essence. Here is the addition:

Aquinas in Ch. 2, section 12, pp. 43-44 of On Being and Essence (see also p. 31, n. 7 of the text) distinguishes what he calls the form of the part from the form of the whole. The form of the part is the same as the substantial form, without matter. (The omission of matter is what makes it partial; it is only part of the essence.) In man, for instance, the form of the part is the rational soul. The form of the whole, on the other hand, is the whole essence or quiddity, including both the substantial form and prime matter. In the case of a man, the form of the whole is man or humanity. It is a bit perverse, perhaps, to call this a "form", since it includes matter, but that is what he calls it. (There is in fact some motivation for calling this a "form", but we needn't stop over that here.)

In an immaterial substance, of course, the distinction between the form of the part and the form of the whole vanishes. The substantial form just is the essence in that case. There is no matter involved.

Henceforth, we are concerned only with the form of the whole, the entire essence, Avicenna's common nature. I mention the form of the part only to set it aside.

Now these common natures may be thought of or considered from various points of view. These give rise to different concepts, expressed by different terms. Let us approach it like this:

In the case of each term T, let us ask ourselves what other terms T* we can join to T as a predicate in such a way that the resulting sentence 'T is (a) T*' (I put the indefinite article 'a' in parentheses, because Latin doesn't have one, and because English doesn't use it when T* is an adjective) is (a) necessarily true (in that case we shall say that T* is "included" in T); (b) necessarily false (and in that case we shall say that T* is "excluded" from T); (c) contingent (and in that case we shall say that T* is neither included in nor excluded from T). (Note: This is my way of setting things up. Aquinas doesn't do it this way. Rather, this is intended to be an explanation of what he does do.)

Now, given all that, we can say that an essence or common nature or form of the whole can be considered:

(1) As a part of the individual, abstracting with precision from all other ingredients of the individual - that is, from its accidents, from esse, from everything else. This means that we don't just leave those other ingredients or parts out of consideration; we positively exclude them. We cut them off (praescindere), which is what is "precise" about the concept. The word that expresses the common nature considered in this way is 'humanity'. Grammatically, it is the abstract form of the noun. Using the machinery just above, we can say that humanity then (i) includes just exactly what is contained in the definition of 'man' (and what is entailed by that - hereafter let's just leave this clause to be tacitly understood), and (ii) excludes everything else.

(2) The essence or common nature or part of the whole can also be considered as a whole, that is, abstracting without precision from everything else. The other ingredients of the individual thing, its accidents, its esse, are left out of account, but are not positively ruled out. The term to express the common nature considered in this way is just 'man'. Grammatically, it is the concrete form of the noun.

Item (2), the form of the whole considered as a whole, comes in three subdivisions. It can be considered

(a) Absolutely. We don't have any special terminology reserved to express this in every case, but when it is in subject position, we frequently express it by 'as-' or 'qua-' talk. For example, 'man as man', or 'man qua man'. (Or maybe 'man as such'.) The form of the whole considered as a whole and absolutely in this way includes exactly what is in the definition of 'man', and excludes only what is incompatible with the definition of man (what is incompatible with what it includes).

Everything else is neither included nor excluded. The form of the whole considered as a whole can also be considered:

(b) According as it has esse, and this in two subcases:

(i) Esse in this or that individual. (It is a bit hard to see how this is said to be abstract at all.) Considered in this way, it includes whatever belongs to the individual, and excludes everything else. (I would say that it includes whatever is in the definition of the individual, and excludes everything else. But for Aquinas, and Aristotle, individuals don't have definitions, strictly speaking.)

(ii) Esse as a concept. Considered in this way, it includes whatever is in the definition of the concept, and excludes everything else.

Avicenna is sometimes undecided about whether to use the abstract or the concrete form of the noun (for instance, he uses 'animal' but also 'humanity'), I think (1) above is what he means by the common nature taken in itself. That is, he means what Aquinas calls "the form of the whole considered as a part". It is neither singular nor universal - both must be added externally. For Avicenna this is what is predicated.

For Aquinas, (1) has no being. The term 'humanity', for instance, is simply a non-denoting term. It prescinds from - that is, excludes - everything that is not included in the definition or essence of man. But that of course means it prescinds from or excludes esse, which is "included" only in the "definition" or essence of God. (Avicenna of course would say it prescinds only from the esse that is accidental to it, not from its own lesser being, but Aquinas is not buying any of that.) Since esse is positively excluded, of course humanity cannot exist.

Humanity, then, does not exist. Only the composite of humanity plus various differentiating accidents - for instance, the composite Socrates - exists. (This commits us to the perhaps odd view that an existing whole has non-existing ingredients or "parts" - and indeed, non-existing essential parts. For Aquinas, you cannot weasel out of the oddity of this by allowing that while this "part" doesn't really exist, nevertheless it's still out there, with at least some minimal kind of lesser reality. Many people will find this too much to take. It is worth considering whether they are just engaging in philosophical prejudice, or whether there is really any good argument to rule out this kind of theory as impossible.)

Furthermore, given some funny business with Aquinas' theory of predication, I think it follows that for him, (1) above - that is, the form of the whole considered as a part - is not what is predicated, contrary to Avicenna's view. In order to be predicated at all on Aquinas' theory, the common nature must be taken in a way that doesn't positively exclude esse. Hence, we need some form of (2) above. In order to be predicated of many, it must not include this or that esse, either in individuals or in the mind. That of course would preclude its being predicated of anything else. Hence, for Aquinas only (2a) remains. Only that is what is predicated of many.

Putting it another way, (1) cannot be predicated of many because it excludes too much. (2b), in either of its alternatives, cannot be predicated of many because it includes too much. (2a), however, is "just right", and it is that that is predicated of many.

Now, so far we have the common nature or essence considered in these various ways, and in some cases expressed by special terminology to reflect these various ways of conceiving it. What we want to know now is what these various ways of conceiving or considering the common nature have to do with ontology, with what is really out there. In each case - (1), (2a), and (2b) in both alternatives - our way of considering the common nature is a correct representation of what is going on, in the sense that we are not including anything that isn't there. But under what circumstances is our way of conceiving the common nature an exact representation of what is going on? When we have included everything? In short, according to which of the above divisions and subdivisions does a common nature exist in exactly the way we consider it? Obviously, the answer is: only (2bi) and (2bii). The common nature taken as in (1) does not exist, as we have already seen, since it excludes esse. Similarly, the common nature taken as in (2a) does not exist just like that, since, just like that, it takes no account of esse at all. Taken absolutely, it is just too indeterminate to exist.

Let's put this another way. For Aquinas, there is a being (in the participial sense) that answers to the term 'the man Socrates', and another that answers to the term 'the concept man'. But there is no being that answers to the term 'humanity' (or even to the term 'the humanity of Socrates'), although there are beings of which humanity is a constituent part. And there is no being answering to the term 'man' taken absolutely, or as we sometimes put it, 'man qua man'.

Thus, what is predicated of many has no being in reality. There is none of this "lesser being" stuff that Avicenna talks about. Neither does what is predicated of many have any unity of its own. (Aquinas agrees with Avicenna here.)

Let us go back to the experiment we have performed several times before. Pull the accidents off Socrates and Plato, until you get down to the common nature. How many do you have left? Boethius' strongly realist theory in the De trinitate, William of Champeaux's first theory and Clarenbald of Arras all say: One. Boethius' theory in the other passages discussed earlier, William of Champeaux's second theory and Gilbert of Poitiers all say: Two. Avicenna says that you indeed have a common nature left, and it has its own proper being, but the question how many you have is simply inapplicable when you frame the question this way.

What is Aquinas' answer? ZERO!

Although you can talk about doing this kind of thing, what you end up with verbally or conceptually is just the common nature taken either according to (1) or according to (2a). And nothing answers to that.

On this interpretation of Aquinas' theory of common natures, there are two mistakes to avoid:

(i) Common natures do not exist. That's wrong. They do exist, for Aquinas. They exist in individuals and in the mind. But they do not exist in abstraction with precision or in abstraction without precision and absolutely - that is, according to (1) or (2a). They derive no being at all from themselves.

(ii) Since sameness and difference do not pertain to the nature taken in abstraction without precision and absolutely - that is, (2a) - or for that matter with precision - that is, (1) - therefore the common nature man in Socrates and that in Plato are not diverse and are not the same. They are neither the one nor the other, but rather some kind of indeterminate third something in between. That is wrong too. They are diverse. The nature in Socrates is all tied up with Socrates' matter, with Socrates' differentiating accidents and with his esse. The one in Plato is all tied up with Plato's. They are quite diverse. They are just not diverse (or, for that matter, the same) considered in abstraction with precision or in abstraction without precision and absolutely. It is diverse in diverse things - which is to say, not taken absolutely or with precision.

(Incidentally, note that (i) and (ii) address the question of being and unity, respectively.)

Well, that is Owens' interpretation of Aquinas - or rather my way of putting that interpretation. It is a very influential interpretation, and many people accept it - both as the correct interpretation and as the correct theory in its own right. And, it must be said, there is some pretty strong textual support for this interpretation.

Nevertheless, I think there are some insuperable objections to it. This is not to say that Aquinas didn't really hold this theory. But if he did hold it, he shouldn't have held it.

What is the problem? Well, I think there are two problems. They are distinct, but closely related.

(a) First, how on earth is this theory ever going to account for the community of the "common" nature? For Aquinas, if the interpretation sketched above is correct, the nature taken in abstraction without precision and absolutely - that is, (2a) above - is what is predicated of, and so is somehow metaphysically in, Socrates and Plato, and for that matter also "in" the mind. On the other hand, on this same interpretation, there is nothing answering to the nature so taken, as we have just seen. Nothing at all! So it looks as if Aquinas is heading straight for nominalism. There is nothing out there that is really common. But if that is the case, then why do we say we have a man in all these cases? Why do we group them together under that one word? Isn't it just arbitrary?

(b) There is an altogether similar problem with the Aristotelian theory of knowledge, which Aquinas by and large accepts. (He accepts enough of it to give us this problem.) If the nature in the mind is not somehow the same as that in the thing, then how do I ground the objectivity of knowledge? I thought the whole point of this Aristotelian approach to knowledge was to say that were the same.

On the interpretation given above, Aquinas turns out to be unable to make up his mind. Sometimes he sounds like a nominalist, sometimes like a realist. If you stress the claim that knowledge and common predication is grounded in the common nature, then he looks like a realist. But if you stress the fact that the common nature, taken in the only sense in which it can do these things, doesn't exist, then he certainly sounds like a nominalist.

You see, on this interpretation, although Aquinas doesn't want to grant any reality at all to the common nature in the relevant sense, he nevertheless wants it to do work for him. He refuses to give it any metaphysical rights, but yet he demands that it take on epistemological and predicational duties. It is supposed to ground the objectivity of knowledge; it is supposed to be the justification for the fact that we predicate the same term 'man' non-arbitrarily of Socrates and Plato. These are not trivial tasks. And yet we are supposed to entrust them to a complete non-entity? Doesn't this violate a basic philosophical principle, a kind of "Principle of Fair Play": No philosophical duties without corresponding metaphysical rights? But perhaps, you may say, such a "Principle of Fair Play" is just philosophically wrong! Maybe you can get away with this. Maybe, in fact, this is the only way you can end up with an adequate theory of common natures, which is to say, an adequate "solution" to the problem of universals. After all, I don't have any knock-down argument that you can't make complete non-entities do philosophical work for you.

Nevertheless, however you stand on that, and however I stand on it, it seems to me that Aquinas is committed to accepting what I called a "Principle of Fair Play" at least in this case. That is, the particular kinds of jobs Aquinas is asking the common nature to do are jobs that, on his own grounds, cannot be done without granting them some kind of being.

What are these jobs? They are unifying jobs. Look at problems (a) and (b) above. The common nature in Socrates and the common nature in Plato must somehow be the same nature, or else there is no real basis for Socrates' and Plato's both being men. What is it, then, that is the same in both Socrates and Plato? Or, to put it in terms of predication, what is it that is predicated of both? On the grounds we have seen above, it can only be the nature taken as a whole and absolutely - that is, in sense (2a). That is what is supposed to account for the fact that Socrates and Plato are both men, since that is what is predicated of many. But it can only do this if the nature, taken in that way, is the same in Socrates as it is in Plato. On the other hand, to say it is the same is just to say it is one. (Aquinas explicitly accepts the identification of sameness with unity in many texts.) And to say it is one is just to say that it does have being after all, on the basis of the Augustinian identification of being with unity, around which Owens built his entire paper!

In other words, to say that the same one common nature in sense (2a) above is what is predicated of many, and then to say that this same one common nature has no being at all, is to deny the Augustinian equation of being and unity, not to uphold it in the face of Avicenna's theory, as Owens makes it out to be. And since I think there is incontrovertible evidence that Aquinas adhered to the Augustinian equation, I also think there is good evidence that either Owens' interpretation of Aquinas' theory is wrong, or else Aquinas himself has a hopelessly inconsistent theory.

The same thing can be said with respect to problem (b) raised above. If there is no sense whatever in which we can say that the common nature that exists mentally in cognition and the common nature that exists externally in Socrates and Plato is the same one common nature, then the entire Aristotelian account of knowledge collapses, and with it the Aristotelian guarantee of the objectivity of our knowledge. That account depended crucially on our being able to say that there is no inference needed to be sure that our concepts match reality, since our concepts are formally identical with the realities out there. But if we say that they are the same, then we are committed, on the basis of the Augustinian equation, to saying that they have some degree of being too, contrary to the theory.

Notice that it does no good to say that, yes, the common nature does have being; it has one (mental) being in the mind, and another being in Socrates, and yet a third being in Plato. Now we have three beings, and so three unities. And none of them is what we want. Each of them is confined to its own little realm, to the mind, or to Socrates or to Plato, and none of them is common. Hence none of them is available to take on the duties Aquinas requires. The only way to do that, as far as I can see, is to give the common nature in sense (2a) some degree of being on its own. And that is why I think Owens' interpretation, while it may or may not be a good interpretation, is certainly a bad theory.

Let us now recall a very important principle appealed to, for instance, by Avicenna in his famous "Suspended Man" passage (I'm paraphrasing):

If it is possible to conceive x without y, then x and y are really distinct. Each has its own being, and x is independent of y.

It is this principle that allows Avicenna to claim the real distinction of mind and body, and also allows Avicenna to claim that common natures have their own lesser being. I can think of "man", after all, without thinking of Socrates or Plato, so that "man" must be something distinct from individual men, something with its own ontological integrity.

I know of no place where Aquinas directly attacks this principle but, if Owens if right, then in effect Aquinas can say that Avicenna has confused two things:

(a) it is possible to conceive x without conceiving of y, and

(b) it is possible to conceive of EM>x-without- y. That is, it is possible to conceive of x separated from y.

('Conceive' here means something like "consistently conceive".) Now Aquinas will accept Avicenna's principle in sense (b), but not in sense (a). If he were to accept it in sense (a), then we would have common natures with their own lesser being, which is not so bad, from my point of view. But if Owens is right, then even though I can consistently conceive of common natures absolutely - that is, in sense (2a) above - it does not follow that they have any being of their own in that absolute way.

Notice, incidentally, that sense (a) just above is exactly what Aquinas himself uses in On Being and Essence on p. 55, where he argues that esse is really distinct from essence (the business about knowing what a man or a phoenix is, but not knowing whether one is). But earlier I said that I thought this was a bad argument on other grounds, so I can hardly appeal to it here as evidence against Owens' interpretation.

Duns Scotus

For Avicenna, remember, common natures in themselves have no unity, but they do have a lesser being. For Aquinas, on the interpretation outlined above, they have no unity or being at all. For Scotus, they have both a lesser being and also a lesser unity.

Avicenna had distinguished kinds or degrees of being, but not of unity. That is Scotus' contribution.

For Scotus, natures are always individuated when they exist in reality. We can also conceive a nature universally, in the mind. And between these two conditions, there is the nature just in itself.

Scotus seldom uses the term 'common nature'. He simply speaks of 'natures'. Sometimes he also calls them 'formalities', or just 'realities'. But he never calls them 'res', "things". For Scotus, a res is always an individual.

Natures are indifferent in themselves to esse or existence, just as they are for Avicenna. This is the doctrine of the "neutral essence". Yet they have their own being, a kind of esse quidditativum, a "quidditative being", or "essential being". In addition to having their own kind of being, natures just in themselves have a kind of unity of their own too. The nature stone has a kind of unity of its own, and the nature man also has a unity of its own. And the two are in some sense distinct. We can say all of this without ever considering the question of existence. But we could not do this if they did not have some unity and being of their own.

For Scotus, there are three levels, so to speak: the level of the individual (res ), the level of the concept, and the level of the nature or formality just in itself, which can exist in either individuals or concepts. For each of these levels, we get a kind of unity (sameness) and a kind of diversity or distinction. This is a rather sticky but very important part of Scotus' doctrine. It is sticky because it is not clear exactly how to define these various notions. And Scotus seems to have changed his mind on some of them over the course of his career. But it goes roughly like this.

(1) At the level of the individual or thing (res), the unity is called numerical unity. Socrates is numerically one. This is the kind of unity an individual thing has. Logicians today like to treat this "unity" as a relation, and call it "self-identity". The distinction that goes with this level is called the real distinction. It occurs whenever you have numerically more than one thing. Tentatively, the criterion for real distinction is this:

x is really distinct from y if and only if it is possible for x to exist without y or else for y to exist without x (or both).

The word 'possible' here refers not to natural possibility - that is, the laws of nature - but to the power of God. Roughly, for Scotus, this amounts to "consistently conceivable". Thus, Socrates and Plato are really distinct, since it is possible to have either without the other. Similarly, God and creatures are really distinct, since while it is not possible to have creatures without God, it is possible to have God without creatures. Likewise, an individual substance is really distinct from its accidents. Not only can you have the substance without its accidents, that is, its separable accidents. You can also have the accidents without the substance - at least by the power of God. Many people thought that his is exactly what happens in the Eucharist.

(2) At the level of the concept, the unity that is appropriate is the kind of unity the concept of man has as one concept. Scotus has no special term for this. The kind of distinction that goes here is the (mere) distinction of reason. This is a very low-grade distinction. Tentatively, the criterion is:

x and y are distinct by a mere distinction of reason if and only if the concepts of x and y are just two concepts for the same item.

Note: I do not say the same thing (res). It can also be the same nature or formality. If it is the same thing, then the sameness or unity is numerical unity. If it is the same nature, then the sameness or unity is the unity that natures have. (See below). As an example, Cicero and Tully are distinct only by a distinction of reason; they are numerically one. So too, man and rational animal are distinct by only a distinction of reason; they are one with the kind of unity natures have.

As far as I can see, this whole level of concepts is not very important for the theory of unity and distinction in Scotus.

(3) Between (1) and (2), we have the level of the nature or formality. Here is where Scotus makes his big contributions. The unity that goes with this level is called real less than numerical unity or real minor unity. ('Minor' is just Latin for "less".) It is real - that is, it is not just a product of our minds, it is grounded in reality, on the side of reality, ex parte

rei. I don't just make it up. And yet it is less than numerical - that is, it does not amount to singularity. For then the nature would be an individual.

The distinction appropriate to this level is the celebrated formal distinction. (And unlike most uses of the word 'celebrated', I can well imagine a real celebration of the formal distinction, with noisemakers and funny hats. It is well worth such a celebration!) As a tentative criterion, let us try:

x and y are formally distinct if and only if x can be conceived without y's being conceived, or vice versa (or both).

This is a distinction that applies not in the order of individuals - that is, of actual existence. It is not the real distinction, in the technical sense introduced above (although of course it is "real" in the looser sense that it is "on the side of reality" - we do not make it up). Neither is it in the order of concepts or understanding - which is the other kind of act that can accrue to natures. Rather it belongs to the order of essence or quiddity.

Since essences are what is conceived, the criterion can be put in terms of conceivability, as above. But since essences or formalities have their own quidditative being, the distinction is one that is on the side of reality. It has a basis in reality.

(Notice, once again, that this is not called a real distinction, even though it has a real basis. The term 'real distinction' is reserved for something else. On the other hand, less than numerical unity of the kind that belongs at this level is called a "real" unity. That is because the opposite of the real distinction is not called "real" unity, but "numerical" unity. So the term "real unity" is not already spoken for, and is free to be used.)

For instance, there is a formal distinction between:

(a) Socrates and the common nature man. I can conceive the latter without the former.

(b) man and animal.

(c) man and ass.

(d) God, God's will, God's intellect. This is a curious and important application of the doctrine of the formal distinction. Some people thought this threatened to destroy the divine simplicity, since the formal distinction is after all based in reality.

Now let us ask some questions and make some observations about these kinds of identity and distinction:

(1) Are Socrates and Plato formally distinct? Are humanity and asininity really distinct? Both pairs satisfy the respective definitions, given above. For that matter, while Socrates and the common nature man are formally distinct, are they also really distinct? After all, I can have the nature (in Plato, say) without having Socrates.

The answer to all of these questions is supposed to be: No. The criteria will have to be adjusted in that case. The real distinction is meant to be a distinction between individuals only - or at least between individuated items. The formal distinction must have at least one side among natures - for instance, Socrates and his nature, or man and animal (in the latter case, of course, both sides are natures).

(2) There is nothing wrong with having numerical unity combined with formal distinction. For example, God, his will, his intelligence - all these are numerically one, but formally distinct.

(3) There is nothing wrong with having real distinction combined with real minor unity. The common nature man that is present in both Socrates and Plato is one with a real less than numerical unity, and yet the nature as it is in Socrates is really distinct from the nature as it is in Plato.

(4) For Aquinas, this entire middle realm is rejected. There is no real minor unity, and there is no formal distinction. Consider the following text from Aquinas' Summa contra gentiles, I, Ch. 26: "What is common to many is not anything over and above the many except by the reason alone."

That is, the distinction between the common nature and the individuals that have it - man and Socrates, for instance - a formal distinction for Scotus, is only a "distinction of reason" for Aquinas. We have two distinct words or concepts, "man" and "Socrates". But ex parte rei, there is nothing to answer to the common nature man except just individual men, Socrates and all the rest. The distinction between Socrates and his nature is merely one of reason. They are numerically the same.

(Note that this text is strong evidence in favor of Owens' interpretation of Aquinas. Nevertheless, the theoretical difficulties I raised against that interpretation still stand. If what is "common" to Socrates and Plato is nothing over and above the individuals, then in what sense is it "common"? In my own view, there is equally strong textual pressure to interpret Aquinas as granting some kind of reality to natures in themselves: every passage in which he affirms the Augustinian equation of being and unity is such evidence. But, however you stand on that, is is quite clear that Aquinas does not have the vocabulary to allow different grades of identity and distinction on the side of reality. That comes with Scotus.)

(5) Recall Avicenna's "Suspended Man" criterion of real distinction, above. (Avicenna of course does not use that term, since he does not distinguish degrees of unity, as he does of being.) Recall the ambiguity I noted above in that criterion:

(a) x can be conceived without conceiving of y, or vice versa;

(b) it is possible to conceive of x-without- y, or vice versa.

Aquinas accepted (b) as a criterion of real distinction, but denied that (a) is a criterion for any distinction at all on the side of reality. (At least, he in effect denied this.) Scotus agrees that (b), not (a), is the criterion of the real distinction. If one can exist without the other, then I can consistently conceive of that's occurring, and vice versa. But Scotus thinks that (a) is the criterion of a distinction on the side of reality too: the formal distinction. Avicenna blurred the distinction between (a) and (b). Aquinas accepted one criterion but did not accept the other as a criterion of any distinction on the side of reality. Scotus accepts both criteria, but as criteria of different kinds of distinction.

In Scotus, we must carefully distinguish two notions: community, as opposed to universality. (Actually, you have to make the same distinction with Avicenna too.) Community is what the nature in itself has. Universality is what the concept has.

What is the difference? Scotus says that for Aristotle, the universal is one in many, and predicable of many things. The ability to be in many things is what Scotus calls community, and amounts to Boethian universality. The ability to be predicated of many is what he calls universality, and it amounts to universality in the sense Abelard had talked about.

The formality or nature in itself has community. It can be in many things at once; it is not "repugnant" to it to be so. But it is not by itself able to be predicated of many. In order to be able to be predicated of many, it has to be thought - and this is what constitutes it as a universal. (I take it the idea here is that "predication" is not something that happens in the logical or semantical abstract. Predication is an act that human beings perform, and they do it with their minds. So nothing can be predicated without being thought of or conceived. This seems to be perhaps the implicit reasoning here.)

So for Scotus it is the concept that is predicated of many, not the nature in itself. Recall, however, that for Avicenna, the basic starting point of this whole business was the notion of predicability. He needed to have some real nature that was neither of itself singular nor of itself universal in order to allow for predication. This was what the real common nature was for. But for Scotus, the common nature in itself is not what is predicated. Hence, whatever motive Scotus had for positing these real common natures or "formalities", it cannot have been Avicenna's concern for predication.

Well, what were Scotus' motives? Before we answer that, we need to fill out the point just above. Recall that Avicenna had an argument why the universal could not be what is predicated. How does Scotus get around this argument?

The argument was that if what is predicated were of itself universal, then whatever it was predicated of would be universal too. It could not be predicated of singulars. It was just the converse of the argument about singularity.

When I predicate 'man' of several people - say, of Socrates and Plato - what I predicate is just what is contained in the definition. I predicate "rational, sensate, organic, material, corporeal substance" of them. I do not predicate "this rational, sensate, organic, material, corporeal substance" of them. For I could say that only of one thing. (See above on why this is not a fallacy.) Neither do I predicate "universal rational, sensate, organic, material, corporeal substance" of them, since I cannot say that of any singular.

Note the assumption: that singularity and universality accrue to the nature like accidents in the quidditative or intelligible order. That is, when you add these accidents, you change or add to the intelligible content of what you started with. That's why they get in the way of predication.

Not so for Scotus. Singularity and universality for him are not further determinations of the nature. Rather, they are what he calls "modes". This notion of a "mode" is characteristic of the Scotist school.

A "mode" for Scotus is whatever can be added to a nature without changing its intelligible content. For instance, the intensity of white light is a mode of that light. (What does this notion of "mode" have to do with the notion of a "mode" in Descartes and Spinoza later on?)

Note, incidentally, that the introduction of modes allows a new kind of distinction: the modal distinction. The light and its intensity are only modally distinct. It is not a real distinction, between two res, or a formal one between two formalities or between a formality and a res(since formalities determine), or a mere distinction of reason (since it seems to be on the side of reality). (Question: Is there a special kind of unity to go with the modal distinction?)

Existence, it turns out, is a mode of a nature. Hence Scotus rejects Aquinas' real distinction between essence and esse. (There is a problem trying to figure out just what the criterion of a real distinction was for Aquinas. It does not appear to be the "separability" criterion that Scotus uses.)

Singularity and universality are also modes of the nature. They do not change its intelligible content, they do not determine the nature any further in the quidditative or intelligible order. Hence they do not get in the way of predication. And since Scotus has independent reasons for wanting the universal to be what is predicated, he is now free to do so.

Now, what are Scotus' grounds for positing a real common nature? Basically, he gives two "metaphysical" reasons and three epistemological ones. Note: All along, I have been maintaining that the basic motivations behind realism were epistemological, not metaphysical or ontological ones. Hence, it is extremely interesting to see Scotus giving metaphysical reasons for his special brand of realism. They must be examined very closely, to see if they really do the trick. Here are his metaphysical reasons:

(1) Real common natures are needed to account for real relations, and in particular to account for real relations of sameness. There are relations of "sameness" in three of the Aristotelian categories:

(i) Substance, where it is called identity or sameness proper (in Latin 'sameness' and 'identity' are the same word, 'identitas', from 'idem' "same"), and where "sameness" means belonging to the same genus or species, or being the same individual. (Digression: Numerical identity was not generally treated as a relation in the Middle Ages. They regarded relations as always involving at least two distinct terms. There was no reflexivity in the mediaeval theory of relations. They would not say that Socrates is identical with himself, but rather that Socrates is numerically one. That is, they treated numerical identity as a "property" (in our loose, present-day sense) rather than as a relation. You can tinker with this, and see whether it makes any difference.)

(ii) Quality, where it is called similarity.

(iii) Quantity, where it is called equality.

(Get used to these terminological fine points, and observe them. Mediaeval authors did not say "equal" where they meant "identical", or vice versa.)

In any case, in order to account for real relations of sameness, Scotus thinks there must be something in each case on the part of reality that joins the relata. This something cannot be numerically one of itself, or else it could not be in both relata at once to join them. Therefore, it must have a real unity, but a unity that is less than numerical. It must be real, because real relations of similarity, equality and identity are not just the products of the intellect.

In a real relation of this kind, you have three things involved:

(a) Two res that are related, the so called "relata". Each of these is numerically one, and the two are really distinct from one another.

(b) A foundation for the relation - that is, something that is shared, a common nature, a formality. This has a real less than numerical unity.

(c) The relation itself, relating the two relata and founded on the common nature. For instance, similarity, equality, generic or specific identity.

For Scotus, the relation itself is a third thing, a third res. It is really distinct from the two relata. The foundation is only formally distinct from the relata. If A and B, for instance, are both white, then A and B are really distinct. That is, A can exist without B (or vice versa). But if that happened, then the similarity would vanish. So the relation is really distinct from the relata. So too, I think Scotus thinks you can have the similarity - that is, the relation itself - without having these particular relata, but I'm not sure how this goes. (Is it just that this same relation can be grounded in two other such relata instead?) Question: What about the distinction between the foundation and the relation? I suspect they are probably only formally distinct, but I am not sure.

(2) Real common natures are needed to explain causality. Between the cause and the caused there is (generally) a nature shared. For example, fire causes fire. (This is so in the case of what is called "univocal" causality. People also recognized an "equivocal" kind of causality, when the cause and the effect did not share a common nature - for instance, in the case of God and creatures. But Scotus' point only requires that some cases of causality be univocal.) This would be so even if there were no intellect. So this unity is not just a product of the mind, it is not a mental fabrication. It must be real. But it cannot be numerical unity, since nothing causes itself. Therefore, it must be a real but less than numerical unity.

Here are Scotus' epistemological reasons for positing common natures:

(3) We need them to provide a proper object for the intellect and for the senses. For the intellect - that is fairly clear. The idea is that the object of the intellect is the common. We frame our knowledge in terms of the common or general. We pigeonhole things. Now, unless the mind is to produce its own objects, these common natures must be real - that is, mind-independent. Knowing is supposed to be a discovery, not a making.

The case for the senses is rather more interesting. Recall the Aristotelian dictum: "Sensation is of particulars, but understanding of universals." (Sensus est particularium, intellectus autem universalium.) Scotus disagrees with both sides of this slogan. As for the side about the intellect, the disagreement is mainly verbal. The proper object of the intellect is not the universal, he thinks, but the common nature. Scotus is willing to accommodate the traditional terminology by calling the universal concept a "complete" universal or universal "in act", whereas the common nature all by itself is only an "incomplete" or "potential" universal. In any case, it is only the common nature, not the fully universal concept, that is the object of the intellect for Scotus.

But more important, Scotus disagrees with the part of the slogan about sensation too. For Scotus, intellect and sense have the same kind of proper object. (What then is the difference between intellection and sensation? That is a good question for Scotus.)

The argument for this interesting view is as follows. I hear sound, I see a color, and so on. The proper object of hearing is not this individual occurrence of the sound, and the proper object of vision is not this individual occurrence of the color. If it were, I could never perceive other occurrences. (The term 'proper object' leads to complicated business. Without getting too technical about it, just think of it as meaning something like "characteristic object".)

The proper object of a sensory faculty, therefore, is going to have a real unity, to be sure, but not a numerical unity. Therefore, it will be a real less than numerical unity. Sight distinguishes white from green, but not this occurrence of white from that one.

We do perceive individuals, to be sure, but we can discriminate individuals in perception only if we take into account the so called "common sensibles" - features that can be perceived by more than one sense faculty - for example, place and position. If we do not take those into account, then we cannot discriminate occurrences. The proper object of sensation is the common nature, and only it.

Suppose God created two physical objects of the same dimensions, color, and so on. The sense of sight could not distinguish them on the basis of proper sensibles, but only on the basis of position, which is a common sensible. If position is not at stake, then sight cannot discriminate between them at all. Suppose the entire field of vision were filled by a certain shade of red. Then it disappears and things get completely dark, and then the same shade of red returns and again fills the entire field of vision. What do we say? Do we say "There it is again" or do we say "There's another one just like the one before"? We don't know which to say, which just shows that sensation, when left to its proper sensibles, does not reach as far as individuals, but stops at the level of the common nature.

(4) Common natures are needed to explain predication. This does not mean of course that the common nature just by itself is what is predicated. Rather the universal concept is what is predicated. Instead, the common nature is needed to ground the objectivity of predication. If the universal were totally a work of the mind, rather than a work of the mind done to something real, then predication would be a fiction. Once again, knowing is discovering, not making.

(5) Common natures are needed to provide proper subjects for the sciences. According to Aristotle, science deals with the "universal". (See his Posterior Analytics.)

Here is a case where "authority has a nose of wax". Scotus interprets Aristotle's dictum here as being about the incomplete universal - the common nature. The basic idea here is that science is not about this individual or that one, but deals with types and classes. If these (incomplete) "universals" were not real, then science would be a pure product of the mind. Once again, the same point: Knowledge is discovery, not making.

Scotus argues, if the objects of a science were not real, but totally the works of the mind, then every science would deal with concepts. (They are the products of the mind.) Now the science that deals with concepts - that is, with what is predicated - is logic. Hence if there were no real common natures, all the sciences would reduce to logic. Several other people appealed to an argument like this too.

Now I want to weave together a few different themes:

On Scotus' view there is no need for a special divine illumination to ground certitude. The classical Augustinian reason for saying there was such a need was that the individual object was changeable and mutable, and so too was the intellect. Neither was firm enough to ground certain knowledge. This is a basically Platonic attitude.

For Scotus, it is simply not so. Material individuals are changing, to be sure, but they have immutable common natures in them. And so there is something in the individual, after all, that is sufficient to ground certitude.

This was just Aristotle's answer to Plato. Just as Aristotle doesn't need the theory of reminiscence, so too Scotus doesn't need the theory of illumination.

Aquinas, as we saw, was another one to make this move explicitly. He can do it too because he wants to find a fixed and immutable nature in things. But Aquinas' and Scotus' views on how the common nature in the individual works in knowledge will be quite different.

The difference concerns the role of the agent intellect. Aquinas accepts the Aristotelian dictum: "Sensation is of particulars, but understanding of universals". In fact, for Aquinas, this makes the distinction between sensation and intellection fairly clear.

Recall from the above that for Aquinas, there is no distinction on the side of reality between the individual and its nature. There is no real distinction, and there is certainly no formal distinction, since Aquinas simply has no such notion as a formal distinction in any case. Hence there is only a distinction of reason. (He actually does say this. See the quotation a little above, from the Summa contra gentiles, I, Ch. 26.) Therefore, when sensation gets through doing its job, the nature preserved in the sense image is still individual. It is not yet common.

What the agent intellect has to do then is to work on the sense image, to separate the common nature from the individuating conditions. Just how it does this is not very clear. Aquinas' descriptions are more a kind of program of what has to be done than an explanation of how it gets done. The agent intellect in Aquinas is a real black box. Think of it as somehow "warming up" the sense image, so that it mysteriously releases these vaporous natures. (The difficulties on this point are just another form of the difficulties I have discussed above with Aquinas' theory of common natures.)

Then, only after all that is done is the nature ready to be impressed on the possible intellect. The agent intellect does this chore too. So for Aquinas, the agent intellect has two jobs to do. First it must somehow get the common nature out of the particularized sense image, and then it must, second, stamp that on the possible intellect.

For Scotus, the agent intellect has much less to do. In the individual thing, the individual and its nature are distinct. There is a distinction for Scotus on the side of reality. It is more than just a distinction of reason. It is not a full-fledged real distinction in the technical sense of that term, but it is a formal distinction. Hence the nature present in the individual already has a community - that is, it already has the ability to be in many.

That is why Scotus can say that the sensible species or image is of the common nature and not of the individual, without needing to posit any kind of agent sense after an analogy with the agent intellect. Nothing has to to done to the nature to get it ready to be a proper object of sensation. (That's true for Aquinas too, but for a completely different reason - for Aquinas, the object of sensation is the individual.)

Likewise, for Scotus nothing has to be done to the nature to get it ready for the intellect. There is no "warming up" needed here. Scotus' agent intellect doesn't have that task to perform. All it has to do is to read off the common nature from the sense image, and then impress it on the possible intellect. For Scotus, the agent intellect is first a selecting device, and then a kind of press. That first task is quite different from what goes on for Aquinas. The agent intellect for Aquinas has to do more than just select. The common nature is not already there in the sensible species or image, just waiting to be read off. There is nothing to select. That is his whole point about the metaphysics of common natures.

The agent intellect has less to do for Scotus than it does for Aquinas. For Scotus, the agent intellect really acts, but it doesn't act on the sense image. It acts only on the possible intellect.

Remember how I said that Aquinas sounded like a realist if you pushed him one way, and like a nominalist if you pushed him another way? Here we see how Aquinas splits the difference. The difference between Scotus' and Aquinas' realism is in the role of the agent intellect. It has more to do for Aquinas than it has for Scotus, and to that extent Aquinas is less realistic, I suppose. (The more realist you are, the less work you will think the agent intellect has to do.)

William of Ockham will be a nominalist. He doesn't think we even have an agent intellect. And this is not because its task would be so easy that it is not needed. (There is nothing like an agent intellect, after all, on a strict Platonic theory.) Rather, it is because the job of abstracting on Ockham's theory would be so difficult as to be impossible. For Ockham we simply cannot form concepts by abstraction in the traditional way. He is a nominalist of the strict observance.

Aquinas distinguishes himself from Ockham's view insofar as for Aquinas, the agent intellect still has a possible task to perform. It is not asked to do the impossible. (At least, he thinks it isn't impossible; you, like Ockham, might disagree.)

Remember the two sides to the old problem of universals, the metaphysical side and the epistemological side. We can give a metaphysical and an epistemological formulation of both realism and nominalism, as they developed in the late-thirteenth and early-fourteenth centuries: Realism and Nominalism

If you concentrate on metaphysics, Aquinas sounds almost indistinguishable from Ockham. If you concentrate on epistemology, he sounds like a realist, and so a bit like Scotus. Only if you take both sides into consideration do you see what is going on in the large. Of course, there are still huge problems with Aquinas' theory, as I have indicated above. Here these problems show up in his theory of the agent intellect. Its performance is very mysterious in Aquinas.

Here are some further questions to keep in mind when thinking about Scotus:

(1) How to distinguish sensation from intellection. They both have the same object for Scotus, the common nature. What then is the difference between these two mental faculties? Why do we need them both? Or do we?

(2) How do we get knowledge of the individual? This is sticky in both Scotus and Aquinas. Aquinas has trouble saying how we can understand or intellectually know anything about the individual, since "sensation is of particulars, but understanding is of universals". For the same reason, of course, he has no trouble saying how we sense the individual. Scotus has trouble on both sides. Ockham will press Scotus on this point. We obviously do know things - both at the level of sensation and at the level of understanding - about individuals. We will discuss this later, when we talk about the theory of "intuitive cognition".

(3) Why is the agent intellect needed at all for Scotus? If the common nature is present in the individual and already capable of impressing itself on the sense without the aid of any agent sense, then why can it not impress itself on the possible intellect too, without the aid of any agent intellect? (A possible reply: The common natures are presented all together, all mixed up, in the sense image. Something is needed to select the one to be impressed on the possible intellect. But still, does that mean that the second job of the agent intellect distinguished above, its "impressing" function, is idle?)

Secretum secretorum index - De Mirabilibus Mundi index

Atlas Games - publishers of Ars Magica          Redcap - Ars Magica portal

Last modified: Sun Dec 27, 1998 / Jeremiah Genest