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

Abelard's Theory of Universals

All references are to the translation in Arthur Hyman, and James J. Walsh, eds., Philosophy in the Middle Ages: The Christian, Islamic and Jewish Traditions, Indianapolis: Hackett Publishing Company, 1973. 2nd edition, 1983., by page and line numbers; 'f.b.' means "from bottom".

Let us turn to Abelard's theory of universals in some detail. The passage I want to examine is in his Glosses on Porphyry, the first part of his big Logica ingredientibus. Porphyry had a significant role in the medieval problem of universals. The Logica ingredientibus is so called because its first word is 'ingredientibus'. The title should not be translated "Logic for Beginners", as is sometimes done. It is a logic for beginners, but that is not what the title means. It just means "The Logic with the First Word 'Ingredientibus'".

Note, before we get started, that on p. 170 Abelard cites Boethius' remark that logic is both a part or a branch of philosophy and at the same time a tool employed by the whole of philosophy. It is both a particular science, in the sense that it has its own subject matter and asks its own questions, and yet a general methodological tool for all the sciences. This should not be surprising if we remember that the hand is a part of the body and at the same time a tool of the body. This cute little simile is a medieval commonplace. It comes from Boethius.

Now let us do a close analysis of the text. Abelard begins his discussion of universals here (he discusses the problem also in other texts, but we are going to ignore them for the present) by referring to Porphyry's famous three questions, transmitted to Latin posterity by Boethius:

To these three classical question, Abelard adds a fourth one of his own (p. 171): (4) Would genera and species exist even if there were no individuals?

Recall how in Boethius' Commentary on Porphyry, there was what at first appeared to be an ambiguity with respect to the terms 'genus' and 'species' (and recall how we also said that perhaps that ambiguity was not really an ambiguity): On the one hand, he says that genera and species subsist in individuals - that is, the individual humanity of Socrates is his species. Yet he also says that species is the thought "gathered from the substantial likeness" of numerically distinct but similar individuals.

Abelard's doctrine is going to be very much like this. But Abelard is a little more explicit about his use of the term 'universal'. He is more careful. Abelard is not very much concerned with thoughts or concepts until later on in his discussion. But he is concerned with words. Now he takes it for granted that there are universal words. What he wants to know is whether in addition there are universal things corresponding to such words.

Abelard (p. 172) cites Aristotle's De interpretatione, 7, 17a37-40, where Aristotle says that a universal is naturally "apt" to be predicated of many. (It is "apt" to be predicated of many, although it does not have to be actually predicable of many. Recall Odo of Tournai's remarks on how there might just happen to be only a single individual in a species. This is the same idea here. A universal is still a universal even if it in fact happens to be predicable of only one thing, provided that it is "apt" to be predicated of many.) Aristotle then goes on to say that an individual is not like this - that is, is not predicable of many, not even "apt" to be. Strictly, of course, that leaves open two possible readings. It could mean that the individual is predicable of one thing only, or it could mean that individuals are the kinds of things that are not predicated at all. To break this ambiguity, Abelard cites Porphyry, who says that the individual is predicated of one thing only.

Now we have to ask, what kind of relation is this predication? What sorts of things do we say are predicated? Do we predicate things of things, or do we only predicate words of things (or of other words)?

This in effect is Abelard's way of setting up the problem of universals. Of course there are words that are predicated of many. Are there also things that are predicated of many? Do the facts of general predication in language accurately mirror an ontological relation in the world? An affirmative answer is realism; a negative answer is nominalism.

Notice that this is not the way Boethius set up the problem. We have then at least two versions of the problem of universals in the Middle Ages, an ontological version going back to Boethius, and a predicational version that can be found here in Abelard. The difference will be important, as we shall see.

Abelard cites some authorities for this kind of "predication" terminology. He is not engaging in neologism here. Aristotle and Boethius both allow that names are universals - that is, words are universals. Aristotle and Porphyry also attribute universality to things. Abelard will begin by considering the realist view, which has the authority of Aristotle and Porphyry, in several versions current in Abelard's own day.

Although Abelard does not say so explicitly here, his main target in this part of his discussion is his former teacher William of Champeaux, at one time Master of the cathedral school at Paris, and later of the monastic school of St. Victor just outside Paris. William held two theories of universals at different times in his career. Abelard tells us in his Adversities (pp. 16-17) that he so hectored William about his first view that William finally just had to give it up. William then adopted a revised, second view, and Abelard attacked that one too. Finally, he says, William had to give up on the question altogether, and his "lectures bogged down in carelessness".

In the present text, we first get an extended treatment of William's first view, and then of the second view in several variations. We have seen both of these views before, in Boethius. Once again, Boethius is the granddaddy of all theories of universals in the Middle Ages.

Abelard's attack on William's first view is a metaphysical attack. He thinks it is a metaphysically impossible position. And, as we would expect, that first theory was a realist one. Abelard's attack on the second theory is different. Basically, he thinks the second view is perfectly all right (although he does have some quarrels with the way it is expressed) - but it isn't a realism, as its proponents say it is.

Let us look first at William's first view (Hyman and Walsh, p. 172, Adversities, p. 16). It is generally conceded by many scholars - for instance, by Copleston, History of Philosophy, vol. 2, part 1, p. 168 - that Abelard's attacks on this view were devastating, that he thoroughly refuted William's brand of realism. Well, historically he may have caused William to change his mind, but I am going to maintain, on the contrary, that not a single one of the arguments Abelard gives here is conclusive against a realism of William of Champeaux's kind, although they perhaps are telling against a certain confused version of that theory. (Perhaps William himself was confused, but the point stands: There were ways he could have defended himself.)

William of Champeaux's First Theory

Consider the structure of an individual, say, of the individual man, Socrates. Well, Socrates is a substance, and so has substantiality. He is a body (that is, a physical object - although of course he is an animated physical object, but that comes later), and so he has corporeality. He is an organism, a living body, and so has life. He is a man, a rational animal, and so has rationality (and therefore humanity). He is a Greek, the teacher of Plato, snub-nosed, and so on. And so he has all those features too. Socrates is a kind of metaphysical layer-cake built up part by part in this way. We have a laminated view of individuals.

Now consider Plato. He has many of the same features as Socrates has. In particular, he has the same features up to and including Greekness. But he has some different features too; they are what differentiate him from Socrates.

Finally, let us consider "Brunellus" - that is, Browny the Ass, Abelard's favorite example of an irrational animal. Browny has many of the same structural features Socrates and Plato have: substantiality, corporeality, life, animality. But he does not have rationality and humanity. Instead he has irrationality and asininity.

Now - and this is the key point: If you start with Socrates and Plato, and take away all their features after humanity, what do you have left, one humanity or two? Is the humanity Socrates has his own, and the humanity Plato has his own, so that if you were to pull everything else off, so to speak, until you got down to the level of humanity, you would have two humanities - one for Socrates and one for Plato - or is there something one that is shared, some one humanity, so that when you pull off all the features that distinguish Socrates from Plato, you end up with one universal humanity? William of Champeaux's first view answered: You have only one humanity, and it is common to and shared by them both.

Similarly, if you go somewhat deeper and pull off rationality, so that you break up the humanity that you had, still you find something that is common to and shared by Socrates and Plato, on the one hand, and by Browny the ass, on the other: namely, animality. And similarly if we go yet deeper. You get the picture.

Now in each case where you find something common like this, it acts, so to speak, like matter with respect to the later forms - the so called "advening" forms of the text. And so, according to Abelard, William calls this the "material essence", as opposed to the "advening forms". See the terminology on p. 172, lines 16-10 from the bottom, in the passage in Hyman and Walsh.

Note several things about this theory and its terminology:

Well, this is William of Champeaux's first theory of universals. Now let's look at Abelard's objections to it. Consult the outline in Chapter \s7 above.

First objection

Contraries cannot be in the same thing at the same time. Indeed, that is more or less the definition of contrariety. For example, being white (all over) and being black (all over) are contraries. No one thing can be both white and black (all over) at the same time. But, the objection goes, on William's view, contraries would inhere in the same thing at the same time. For instance, both rationality and irrationality would inhere in animality at the same time. And rationality and irrationality are contraries. If they did not both inhere in animality at the same time, then you could not have the rational Socrates and the irrational Browny the ass existing at the same time. In short, this first view of William's, the objection says, violates the Law of Contraries, and so must be rejected.

First reply to objection 1

As it stands, this argument is not very compelling, and Abelard knows it. The obvious answer is that the Law of Contraries ("Two contraries cannot inhere in the same thing at the same time") was never intended to rule out this kind of thing. Rather it means that two contraries cannot inhere in the same individual at the same time. And William's first view does not violate the Law of Contraries, so understood. Animality may have both rationality and irrationality at the same time, but no one individual animal does.

Refutation of the first reply to objection 1

Abelard anticipates this reply and tries to counter it. He argues that, given William's view, it is not only the case that contraries inhere in the same universal thing at the same time. Abelard implicitly agrees that this would not after all violate the Law of Contraries. But things are worse. It turns out, Abelard thinks, that on this view contraries would inhere in the same individual at the same time. And that clearly does violate the law. Here is his argument.

(1) Socrates is identical with whatever is in Socrates other than the (advening) forms of Socrates. In other words, Socrates is identical with his material essence.

(2) So too, Browny the ass is identical with its material essence - that is, the ass is whatever is in the ass other than the (advening) forms of the ass.

(3) But whatever is in Socrates other than the forms of Socrates is the same as whatever is in the ass other than the forms of the ass. That is, the material essence of Browny is identical with the material essence of Socrates.

(4) Hence, Socrates is identical with Browny the ass.

(5) And since rationality inheres in Socrates and irrationality in the ass, if follows from (4) that they both inhere in Socrates. Q.E.D.

Plainly the argument is valid, but the premises need a little talking through. Let us begin with (3), which is the easiest. Clearly, there is some level at which (3) is true - for example, the level of animality. It is both a material essence of Socrates and a material essence of Browny, and is one and the same in each. So the force of this argument rests on the very strange premises (1) and (2). Why on earth would William ever want to say that Socrates is identical with his material essence animality, and the Browny the ass is likewise identical with his material essence, the same animality? The straightforward interpretation of William's view is that the individual is in every case to be identified with the sum total of all its forms, including, at any given level of analysis, the material essence together with all its advening forms. The individual is the product of this complete layering process, not some intermediate stage along the way.

But Abelard thinks he can block this interpretation (p. 173 bottom: "The truth of what we assumed above . ."). The individual can be identified with either:

(a) The matter (understand: the material essence, at some stage of analysis).

(b) The forms (understand: the advening forms, at some stage of analysis). Or

(c) Both (a) and (b) together, the sum total of all the forms, both those counted in the material essence and those counted as "advening" forms.

Alternative (c), of course, is the one that makes William's view most plausible. But Abelard wants to rule out (c), and also (b), leaving only (a). Here are his arguments:

Ad (b): In that case, accidents would be substance. That is, the individual substance, say, Socrates, would be identified with his accidents. This would of course follow only if William's "advening forms" are to be taken to be accidents. Now there is some evidence from another work of Abelard's, his Dialectica, that William did in fact talk this way. The key line here comes at the end (line 17), where Abelard claims that William wanted to say that differences inhere accidentally in their genera. That is, rationality is just as accidental to animality, in this sense, as being snub-nosed or being seated is to humanity. But the latter are paradigm instances of accidents. The evidence is hardly definitive, but if Abelard's remark here is correct, then Williamperhaps did regard his "advening forms" as accidents.

Now, in order to refute alternative (b) above, as Abelard wants to do, we must all agree that substances cannot be identified with accidents, as they would have to be on this alternative. Of course, if we are being Aristotelians, this follows simply from our standard terminological usage. Substance and accidents are in a sense opposites; they could scarcely be identified. It is not so clear that it follows, however, given William's own rather non-Aristotelian use of the term 'accident' - to stand for the "advening forms" of a strongly realist theory. But in any case, whether it follows or not, we don't have to worry about alternative (b). William certainly did not want to identify an individual with its advening forms. He would no doubt have been willing to grant Abelard that alternative (b) should be rejected.

Ad (c): This is plainly the alternative William would want to adopt. The individual is to be identified with the sum total of all its forms. But Abelard argues that, if this is so, then "body and non-body would be body". This seems to mean that, for instance, the material object Socrates would be identified with the sum of his "material" essence plus his advening forms which, with respect to his essence, would be "immaterial" (non-body). This sounds as if a material object would have non-material integral parts. And that, of course, would be a bad consequence. But the consequence simply doesn't follow; there is an equivocation on the word 'material'. The sense in which Socrates is a "material object" is not the same sense of 'material' in which we speak of his having a 'material essence'. The notion of "matter" involved in material essences is purely an analogy with the notion of "matter" involved in physical objects. The "material" essence is "material" only insofar as, like matter in the literal sense, it is indeterminate and underlies (advening) forms that narrow it down and specify it more fully. Hence we must conclude that Abelard's argument against alternative (c) fails.

But, if that is so, then why did Abelard ever come up with it? On the interpretation I have just given, the argument is an obvious fallacy. Is Abelard deliberately distorting William's view, or was William really so confused that he used the term 'matter' in the equivocal way on which Abelard's argument depends? Or is perhaps Abelard himself simply confused? Unfortunately, we don't really have enough surviving texts of William to be able to say with any confidence just what his doctrine was, and whether Abelard is distorting it or not. But no matter - William's doctrine can be saved, even if Abelard and William himself were both too confused to see it. Just clear up the equivocation.

In any event, Abelard thinks only alternative (a) above remains: the individual is to be identified with its material essence, to the exclusion of the advening forms. This gives him what he needs to defend premises (1) and (2) of his refutation of the first reply to objection 1 (see above).

There is still a problem, however. Abelard should have recognized that the notion of "material essence" on William's view was a sliding notion, relative to the depth of the analysis. There is no one form that is the material essence; it depends on how deep you go. Does Abelard's argument then force us to conclude that the individual is identical with all those forms that might, at some level of analysis, be called its "material essence"? That certainly seems hopeless. On the other hand, if not, then how are we to pick out one such form and identity the individual with it? Abelard might well reply: Don't ask me! It's not my theory.

Let me summarize the result of all this. I think the main problem here is that we have a terminology of matter, essence and accident that comes ultimately from Aristotle, and was known to the Middle Ages through Boethius's translations of and commentaries on Aristotle and Porphyry, and through his own independent logical writings. With this terminology comes a set of terminological conventions: substances cannot be identified with accidents. Material things don't have immaterial integral parts. And so on. However, the doctrine we are now dealing with just doesn't fit that terminology and those conventions very well. Abelard's objections so far apply only to a badly formulated version of Boethian realism. Such a realism can be defended against Abelard's attacks by simply getting straight on the terminology, and getting straight on what is and what is not implied by the doctrine.

Conclusion: the first reply to objection 1 is a good reply, and Abelard's refutation of it fails.

Second reply to the first objection

Abelard considers a second reply his objection. (The objection, recall, was the one about contraries' inhering in the same thing at the same time.) In effect, this second reply accepts Abelard's conclusion, that on William's first view two contraries are in the same thing at the same time, but says that this poses no problem. Animal - that is, the universal, animality - is indeed both rational and irrational, but in virtue of different forms. Animal is not irrational insofar as it has the form rationality, but insofar as it has the form irrationality. Neither is it rational insofar as it has the form irrationality, but insofar as it has the form rationality.

Refutation of the second reply to objection 1

Abelard's response to this is, "So what?" It still remains true that contraries are in the same thing at the same time. And what's so special about these "insofar as" considerations? There is nothing unique about contraries in this respect. Such considerations apply to all forms. It is not in virtue of redness that the apple is round, but in virtue of roundness. Neither is it in virtue of roundness that it is red, but in virtue of redness. Nothing is gained by the appeal to "insofar as" considerations, as though they made a difference.

Abelard's reply seems to me a conclusive refutation of a silly attempt to avoid his first objection. But the success of Abelard's reply here does not allow us to conclude that his original objection stands. The first reply is still open, despite Abelard's argument against it. Contrary forms are not in the same individual at the same time on William's first theory, despite Abelard's argument that they would be, and so the Law of Contraries is preserved intact.

Abelard's second objection

Abelard now begins a whole new line of attack. There would be, he says, only ten essences of all things if William's view were correct - the ten "generalissima" or ten Aristotelian categories. The argument is this: All substances are at bottom the same; they all share the common material essence substantiality. And that is indeed true on William's view. So too, all qualities are at bottom the same: redness, greenness, tallness, and the rest, all share the common material essence quality. And so too for all the other Aristotelian categories. Note that the categories are the bottom layer of the ontological layer-cake. They are "generalissima" - most general genera. They are the most material of material essences.

Here then is the problem: Take two things, say, Socrates and Plato. They have a selection of characteristics from each of the Aristotelian categories. They have characteristics from the category of substance, others from quality, others from quantity, and so on. Now, on William's theory, Socrates and Plato are supposed to differ from one another with respect to the peculiar combination of characteristics each has. But - and here comes the punch - any characteristic the one has turns out, as we have seen, to be at bottom (= penitus - "at bottom" is a pretty good translation of it, if we ignore the spatial connotations that are not in the Latin) the same as some characteristic the other has. Thus at bottom Socrates and Plato do not differ at all - not in the category of substance, not in the category of quality, and so on. Hence, Abelard concludes, all distinctions between things will vanish.

A possible reply William might make

(although there is no evidence whatever that he actually did)

Why think there is no more to things that their bottoms? We might very well accept Abelard's argument, up to and including the step where he says that Socrates and Plato, and indeed any individual, will be at bottom the same as any other in all the categories. But why conclude from this that Socrates is not distinct from Plato? They differ by their "advening forms", as we saw right from the beginning. There is a "too-fast" move in Abelard's argument here.

Abelard's probable reply to this hypothetical response

There's nothing "too fast" about it. I'm only arguing on your own grounds, Bill. You are the one who wants to hold not only the standard view that genera are divided by differences - for example, that substance is divided by "corporeal" and "incorporeal". You also want to hold the further thesis that difference-words - for example, the words 'corporeal' and 'incorporeal' - don't just refer to the advening form (corporeity, incorporeity), but to the combination of the advening form and the underlying genus or material essence. It follows, of course, that difference-words do the job of species-words too. That is, 'corporeal' not only refers to what is added to the genus "substance" to get the species "corporeal substance", but also refers to the underlying material essence "substance" itself, so that 'corporeal' means "corporeal substance" - which is the species. You want to hold this peculiar view, you say, because otherwise substance would be divided into accidents (that is, the genus or subject "substance" would be so divided). This would indeed follow. If "substance" is divided into "corporeal" and "incorporeal", and the latter referred only to the advening forms or accidents, then substance is indeed divided into accidents. Just why you think this is so bad, given your peculiar use of this terminology, I'm not prepared to say. But after all, it's your theory, not mine. (Once again, here we see inferences drawn, not from the theory itself, but from an ill-fitting terminology in which the theory is expressed.) All this is set out in the passage from Abelard's Dialectica. Now if what that passage says about William is not an out and out fabrication, Abelard has him. Here's why:

When you try to define a species, in the standard way, as a genus plus a difference, it turns out on William's view that the difference-word you use is also a species-word. That is, it contains a reference to the genus plus something else. (That's what a species is, after all.) That something else, of course, is the difference. But when you try to say just what this difference is, all you get is another reference back to the original genus plus something else.

Perhaps an example will help. Consider the definition of man as a rational animal. Here 'rational' is the difference-word and 'animal' is the genus-word. But on William's theory, as Abelard describes it, the word 'rational' not only picks out the difference - that is, what is added to the genus to give you the species (in this case, rationality). It also refers to the product or result you get when you do add the difference to the genus in this way. That is, the difference-word 'rational' here plays two semantic roles. (Whether William clearly distinguished these roles is not certain, since we have so little of his writings. Abelard's attack, if it is not a misrepresentation of William, suggests that he did not.) In addition to singling out rationality in some way, the difference-word 'rational' also bears some kind of semantic relation to man. After all, it is men who are rational animals. The difference-word 'rational' is predicated not of rationality but of people. Rationality isn't rational, men are - that is, rational animals are. Hence, the difference-word 'rational' really "means" rational animal. But if that is so, then when I define man as a "rational animal", I am really saying that man is a "rational animal animal", since 'rational' alone already has 'animal' built into it. And of course, 'rational animal animal' can be unpacked one step further, into 'rational animal animal animal', and so on.

The species, then, is ultimately the genus plus something else, which something else is itself the genus plus something else, which something else is again . . ., and so on, in infinitum. Hence, in the end all you have is the genus, over and over again, plus a promissory note that is never really cashed in. And since any genus that is not a most general genus (= a category) is also a species of a higher genus, any genus other than a category will run into these definitional problems. In short, the only forms there are that are not hopelessly involved in definitional infinite regresses are the ten Aristotelian categories. Hence, if every individual has characteristics from each of the ten categories, then it really does follow, as my argument claimed (Abelard is speaking again), that there is only one individual, and the distinctions among individuals vanish.

Once again, if this is not a distortion of William's view, then Abelard has him. But notice that the objection is not really against William's realism, but rather against his theory of the relation of genus and difference. William might easily revise that view to avoid Abelard's objection without giving up his realism. He can say, for instance, that differences do not belong to the genus they differentiate. They are not the original genus plus something else.

In this connection, you might compare Metaphysics B, 3, where Aristotle argues that "being" cannot be a genus, a sort of "supercategory", since then its differentiae, in order to do their job, would have to be beings, and so fall into the genus they are supposed to differentiate. The presumption, of course, is that differences do not fall into their genus in this way. All this is very difficult stuff in Aristotle, and in any case the Metaphysics was not yet translated into Latin, so that it could have had no direct influence whatever on the controversy between Abelard and William.

It seems to me that a more likely background for the dispute here is Anselm. Anselm wrote a De grammatico in which he discussed the semantics of "connotative" terms (although that is not what he called them). In that discussion, Anselm explicitly raised the kind of infinite regress considerations we have just seen, in order to argue that words like 'rational' do not contain a reference to their underlying genus. For Anselm, 'rational' does not mean "rational animal". Whether or not Anselm had any direct influence on the dispute between William and Abelard, it seems to me that this kind of semantic issue is the real basis for Abelard's argument. And, as far as I can see, William could well change his semantics without being thereby committed to giving up his realist metaphysics.

One last remark before I leave this second objection. The kind of definitional infinite regress Abelard has maneuvered William into has a technical name. It is called 'nugatio', which means "speaking nonsense" or "babbling". In Aristotle's Sophistical Refutations (which was only first appearing in translation in Abelard's own day, and did not yet have any influence), one way to refute your opponent was to reduce him to "babbling" - the Latin translation says 'nugatio'. This doesn't mean that you are supposed to get your opponent so confused that he starts gurgling and frothing at the mouth (although you could probably "win" that way too). It means you can push him into one of these definitional infinite regresses. The flavor of the technical use of 'nugatio' can perhaps be better captured by translating it 'stammering' - for example, "rational animal animal animal animal . . .".

Abelard's third objection

Abelard hurls another argument at his poor old teacher. On your view, Abelard says, we ought to call an underlying material essence "many" - in the sense that a universal is "many" (="common to many") - because of the several forms inhering in it. But then by the same token it seems we ought also to call Socrates "many" - and so a universal - because he too has many forms inhering in him. Your view, therefore, destroys all difference between universals and individuals.

A possible reply William might have made

(but once again there is absolutely no evidence that he did)

The two cases are not the same at all. The forms inhering in Socrates constitute him - result in Socrates. But the forms inhering in animality do not constitute animality, result in animality. They are added onto something already there; they "advene". The individual is the material essence plus the advening forms. Thus the relation of the advening forms to the material essence is quite different from their relation to the individual they constitute. It is the same as the difference between the relation of one addend to another addend, on the one hand, and the relation of either addend to their sum, on the other. This difference is enough to warrant our calling the material essence a universal without our being thereby forced to call the individual a universal too.

The fourth objection

Here comes another one. This objection is directly against the notion of individuation by accidents. Abelard argues that this would make individuals dependent on their accidents, metaphysically "posterior" to them. But the opposite is true. Accidents are "posterior" to and dependent on their individual substances. Accidents are ontologically parasitic.

William might reply (but again we don't know that he did)

Accidents may depend on their individuals for you, Abelard, and even for Aristotle and Boethius (in some passages), but not for. On my view, Abelard, accidents are ontologically prior to the individual. They constitute it, not the other way around. You have distorted my view. You have taken the notion of accident and interpreted it in an Aristotelian way that is not part of my theory. (William may in fact have talked in the Aristotelian way Abelard's objection presupposes. But if he did, he shouldn't have.)

Nevertheless, this fourth objection suggests a good argument against this form of realism - an argument that Abelard unfortunately never made. Aristotle's distinction of essence and accident, and his identification of the individual substance with its essence, so that the accidents depend on the individual substance and do not constitute it, all this was motivated at least in part by a desire to account for accidental change - that is, for cases in which we say that an individual changes color, say, but stays the same individual it was to begin with. The pin-cushion would stay the same, so to speak; the pins would come and go. Now we saw when we were discussing this realist view in Boethius that our real problem with it is that it rules out accidental change, freezes the individual. The change of a single accident would result in a whole new individual. This objection can be met as Leibniz did later on, by adding explicit time-specifications to the accidents (for instance, "red at time t). Nevertheless, that involves a rather major adjustment in the theory, and so the objection stands as a good objection to the theory as originally stated. But Abelard didn't make this objection.

Now let me summarize what we've seen so far. Some of Abelard's objections to William's first theory may hold against a confused form of realism that William perhaps actually held. But they do not refute this form of realism in general, although there is another argument that does refute it (the argument about accidental change), or at least requires one to change it in major ways. One wonders why William gave in to Abelard's arguments so easily.

Nevertheless, he did. As a result of Abelard's criticisms - at least to hear Abelard tell it - William abandoned his strong realism for a second view. And this is the second position Abelard discusses in the passage we are examining. We turn now to it.

William of Champeaux's Second Theory

On the first theory, if you take Socrates and Plato and strip off all the advening forms until you get down to the universal humanity, you get only one humanity, not two. The humanity of Socrates and the humanity of Plato are one humanity.

On the second theory (p. 174), this changes. You end up with two humanities, not one. The humanity of Socrates is not identical with the humanity of Plato. Each has his own. And if you ask what makes them distinct, the answer is "They just come that way." There is nothing - no advening form - that narrows down humanity to yield Socrates. There are advening forms, to be sure, but they do not individuate. Everything is individual to begin with.

Even on the first theory, the ten Aristotelian categories were distinct in this way. There is no one super-genus that is divided up by advening forms into the ten categories. The categories just are distinct all by themselves. They just come that way.

On this new theory, we can no longer say that Socrates' humanity and Plato's humanity are both "a substance essentially the same", as we did on the first theory (p. 172). Rather, the second theory says "they are one and the same not essentially but indifferently" (p. 175). The point is that here we are replacing a positive term by a negative one. There is no difference separating Socrates' humanity from Plato's humanity. The first view said positively, "They are the same". The second view says negatively, "They are not different". They do not differ in man or in humanity (note this phrase for future reference). Now, Abelard goes on, this lack of difference is to be spelled out in terms of similarity. The humanity of Socrates and the humanity of Plato are not identical, but they are alike. In addition to the discussion in his Glosses on Porphyry, Abelard sketches this second view very briefly in his Adversities, p. 17.

Here are some things to note about this second theory:

(1) While William's first view goes back to Boethius' remarks in the De trinitate about individuation by accidents, the second theory goes back to the view found in various forms in Boethius' Commentary on Porphyry, the Contra Eutychen, and also other passages in the De trinitate. One way of cashing out this view was the well-worked "signet ring" analogy, where the ring was the divine idea. Abelard doesn't develop the view in quite this way here, any more than Boethius did in his Commentary on Porphyry. Also, "matter" seemed to be the principle of individuation in the Contra Eutychen, although nothing is said about it in the Commentary on Porphyry, any more than there is here in Abelard. What William himself actually said about these points is anyone's guess.

(2) The key term 'indifferently' comes from Boethius' De trinitate, a passage that maintains a view roughly like William's second theory (but with the role of the divine ideas made explicit in Boethius). Boethius there uses the term 'indifference' for the basis of the unity of the Trinity. The three persons are not identical, but they are indifferent - there is no difference between them.

(3) We do possess a text of William of Champeaux in which he actually does hold this view. Most of William's writings are totally lost. But we do have this text. It is translated by Muckle in Adversities, p. 17 n. 12. (As far as I can see, there is no basis in this text for any suggestion of yet a third position, as Muckle says some people have thought.)

(4) William was not the only twelfth century author to hold a view like this, any more than he was the only one to hold a view like his first one. We also find this second theory, or a reasonable facsimile of it, maintained by the famous Gilbert of Poitiers (also called Gilbert de la Porée). Gilbert, like Clarenbald of Arras who maintained a view like William's first one, belonged to the School of Chartres. Bernard of Chartres seems to have been another one to hold a view like William's second one. The texts in the case of Gilbert are pretty obscure, but you get the gist. Also on Gilbert, see John of Salisbury's remarks in Hyman and Walsh, p. 169.

Basically, Abelard thinks this second theory of William's is perfectly all right. With certain adjustments, he will accept it himself. But he does not think it is a realism, as its proponents maintain. "And yet", he says, "they cling to the universality of things" (p. 175). That is, "they" think not only that words are universals, but also things. Thus Abelard's attack on this view is not that it is incoherent - although he does disagree with William's way of expressing it. Rather, his main question for this theory is "What is the thing that is supposed to be universal on this theory?" Recall the definition of a universal at the beginning of the passage we are now considering, the definition from Aristotle. What thing on this second view is predicated of many? Surely not Socrates' humanity, since that is predicated only of Socrates. And not Plato's humanity either, since that is predicated only of Plato. And, on this theory, there simply is no common humanity to be predicated at all.

Abelard considers two variations of this theory, variations that try to answer just this question, and so save the realism of the theory. (Just why people should be so concerned to insist that it is a realist theory is not very clear. Perhaps they thought that somehow they would save knowledge that way.)

Joscelin of Soissons' view

On this variation, the universal humanity is the collection (collectio) of all the individual humanities. All men taken together constitute the species "man", which is then predicated of them all. John of Salisbury (p. 169) is the one who tells us that this view was held by Joscelin, Bishop of Soissons, although there is no real evidence, I suppose, that Joscelin is the one Abelard had in mind. (There is no real evidence that it was anyone else, however.) Note that by 'collection', Joscelin surely did not mean a "set". Sets are extensional; they are defined by their members. If a set gains or loses a member, it becomes a different set. But the collection of men is presumably the same even though men are born and die. Or, if it isn't the same, then it is a very bad candidate for the universal "man". I'm not supposed to keep changing the universal human nature I share in just because other people come and go. One could, I suppose, think of Joscelin's "collection" as the set of all past, present, future (and perhaps merely possible) men, but that is surely an anachronistic way to view it. (In general, it is bad historical methodology to interpret old views in terms of set-theoretical structures. Sets were not invented - I use the word deliberately - until the nineteenth century.) On all this, see the fourth objection, below. On the whole, Abelard has no trouble with this "collection"-theory. Here are his objections:

(1) A collection is predicated only by parts. The collection of all men is not wholly present to each man in such a way that the whole collection enters into the very structure of each individual. But Boethius says that a real universal is not predicated by parts.

(2) Further, if you insist that there is a sense in which the whole collection is predicated of each of its parts (perhaps by taking predication as in effect just the converse of the relation of being a member of the collection), then whatever that sense is, it seems that Socrates may also be predicated in the same way of each of his parts, so that Socrates would be a universal too. This view then radically alters the dividing line between universals and individuals. The only true individuals on this theory would have no parts at all; they would be metaphysical atoms. Whether there are such things or not, they are certainly not the things we normally call individuals. This view ends up betraying the common-sense starting point with which it began. (The force of this objection seems to rest on taking the collection as an integral whole, a whole made up of parts. It is not clear whether this distorts the view or not. We simply do not know.)

(3) "Man" would not be a lowest species (infima species, species specialissima), as it is on the standard doctrine, a species containing only individuals under it, no subordinate species. For there are various subcollections of men - by race and nationality, for instance - and these subcollections would be universal (if universals are just collections), so that we would have subspecies.

(4) The fourth objection seems to arise from taking collections as extensional, so that the category (that is, the broadest universal) that contains all substances would be distinct every time a substance is destroyed (and for that matter, every time one is generated, although Abelard does not run the argument in this direction). So, over time, there would be several collections that qualify as the category substance.

Abelard mentions a cryptic reply Joscelin might make to this last objection. No collection included in the category, he might say, is itself a category or most general collection. I suspect this amounts in some way to rejecting the extensional view of collections, so that the collection that is the most general genus substance remains the same even though substances are destroyed (and generated). But, Abelard goes on, then what remains when a substance is destroyed would be just a species of the category substance, so that there would have to be another, coequal species, since a genus cannot be divided into a single species. (Recall Odo's remarks on this, and the difficulties with it.) But there are no other individuals, ex hypothesi, to constitute that other species; the others have all been destroyed. Hence that other species would have to be either (a) exactly contained in the first, so that it is really the same species, and then you have one and not two, or else (b) that other species would have to be a subspecies of the first, as rational animal is a subspecies of mortal animal. But in that case they would not be coequal. (I really cannot say I am confident about what is going on here, and so am not in a position to assess its exact worth. It may be better than I have made it sound.)

(5) On the realist view, individuals are composed of universals. Universals are the constituents of individuals, so that the individual is ontologically posterior. But Joscelin's view makes universals out to be composed of individuals, so that the universal is ontologically posterior to the individual, as a whole is posterior to (dependent on) its part.

(6) According to Boethius, who is being Aristotelian here, the species is the same - identically the same - as the genus. But on Joscelin's view they can't be, since the collection of all men is not the same as the collection of all animals. The sense in which the species is the same as the genus is this (for Aristotle, who on this point sounds a lot like Boethius' view in the Commentary on Porphyry and Contra Eutychen, which view William is adopting as his second position, and which view Joscelin is here defending): Socrates contains his own individual humanity, which is distinct from but similar to that of Plato. He also contains his own individual animality - and it is the same thing as his humanity. That is, by having his own humanity, Socrates has his own rational animality. His humanity is just a case of a special kind of animality - no more, no less. Conversely, the animality that Socrates has is not an irrational animality; it is a rational one, that is, a humanity. It is all the same thing. We will see more of this immediately below.

Walter of Mortagne's view

The considerations above lead naturally to the second variation on William's second view, which Abelard now considers. Recall, the question is, "What on this view is universal, what is predicated of many?" On this new variant, the individual Socrates is simultaneously an individual, a species and a genus. He is an individual insofar as he is Socrates, a species insofar as he is a man, and a genus insofar as he is an animal.

This view is based on the kind of consideration we have just seen in the sixth objection to Joscelin's view. What makes Socrates Socrates and a man and an animal is the same thing. It is his own animality, which is not irrational but a rational animality - that is, a humanity - and is not a Plato-style humanity - with brown eyes, say - but a Socrates-style humanity - with blue eyes. In Socrates, it is all the same thing. The different degrees of generality in our predication are just a result of our considering Socrates insofar as he is Socrates, insofar as he is a man, or insofar as he is an animal. That is, it all depends on how broad or narrow our focus is. But it is the same thing we are talking about in each case.

John of Salisbury (p. 168) tells us that a view like this one was held by a certain Walter of Mortagne, who later abandoned it for a view that said that the divine ideas were universals (although they couldn't be universals in the Boethian sense, unless one took the clause 'constitute the substance' in the Boethian definition of a universal in a very loose sense indeed). Bernard of Chartres also said the same thing about the divine ideas. This need not have involved anything more than a terminological change on Walter of Mortagne's part. That is, Walter, even after he changed his mind, could still have had a theory like William's second view, metaphysically speaking. The only difference would be in what he was willing to call a "universal" in this theory.

In any case, Abelard attacks this view too, this time with three arguments (pp. 176-177). Now, you may ask, if this view is based on the same kinds of considerations as those to which Abelard himself appealed in his sixth objection to Joscelin of Soissons, why is Abelard arguing against the view here? But of course he's not. He's only arguing that it is not a realism. Here are his arguments:

(1) On this view, humanity, insofar as it is narrowed down to Socrates, can no more be predicated of many than can Socrates himself, insofar as he is Socrates. Conversely, Socrates, insofar as he is a man can be predicated of many just as much as humanity can, insofar as it is just humanity. Hence, individuals and universals are predicable of exactly the same things in exactly the same senses. Whatever you may think of this argument, here is another one:

(2) Further (and this may also be argued as a simple corollary of (1)), this view destroys the distinction between individuals and universals. If Socrates is an individual, a species and a genus, then an individual is a universal, and vice versa. In general, one of Abelard's common strategies against other people's theories of universals is to claim that they destroy the distinction between individuals and universals.

William's own view

After refuting these two variants of William of Champeaux's second view, Abelard (p. 177) turns to face William's view itself. William wanted to say that Socrates and Plato agree in man, or agree in humanity (so did Walter of Mortagne). But he took these expressions negatively; recall his term 'indifferently'. They agree in man insofar as they do not differ in man, or in humanity. But Abelard objects that neither do they differ in stone, and yet we don't say that they "agree in stone", so that the term 'stone' could be predicated truly of them.

A possible reply

Abelard doesn't really fill in the details of any reply. But he does say that perhaps you might want to add some premise ('propositio', translated as 'proposition' in line 10 - the term frequently means 'premise') that allows the man case but not the stone case. That is, we want some premise to the effect that if A and B do not differ in man, then they agree in man. But we don't want it to work where you substitute certain other terms for 'man'. It doesn't really make any difference how you would argue for this special premise, since Abelard doesn't think it will help anyway. Here is why:

Refutation of reply

Abelard counters that it is not even true that Socrates and Plato do not differ in man. Hence, it could hardly be appealed to to explain anything. As I understand it, his analysis goes like this:

Since we are not talking about a universal man in the sense of William's first theory, we must gloss the statement 'Socrates and Plato do not differ in man' as follows: Socrates and Plato do not differ an "a" man, that is, in any (individual) man at all. The lack of an indefinite article in Latin allows this step to be made without calling for any special comment. All right, but now what does this new sentence mean? In general, how do we analyze a sentence of the form 'A and B differ in a C'?

Suppose we say that two things, A and B, do differ with respect to ("in") a color - that is, with respect to some color. Then we can parse this: Either A has some color that B doesn't have, or else B has some color that A doesn't have. (Perhaps one is colorless.) Hence, to say that A and B do not differ in a color is to say the denial of this: A has no color that B doesn't have too, and likewise B has no color that A doesn't have too.

Similarly, to say that Socrates and Plato do not differ in a man or in a humanity is to say that Socrates has no humanity that Plato doesn't have too and Plato has no humanity that Socrates doesn't have too. But on William's second view, that is just plain false! Each of Socrates and Plato has his own humanity, which the other does not have. Thus, far from saying that Socrates and Plato do not differ in a man, William should have said that they do. The claim that they do not encapsulates William's first theory, not his second.

This is a good example of Abelard's using his dialectical skill to dazzle his opponent. Later logicians would develop long treatises on how to treat the word 'differ' in contexts like this and various other contexts. (Their analyses looks pretty much like the one I have just given, which is why I think it is not implausible to interpret Abelard himself in this way.) Abelard here presupposes an already finely-honed analysis.

Unfortunately, I think it is probably unfair to William. William might well reply that Abelard's fine logic doesn't show that the view is wrong, but only that it was badly expressed. The basis point is unaffected: Socrates has his own humanity and Plato has his own humanity, and there is no difference between them - that is, no third entity that comes between them and is required to make them two humanities. They are two all by themselves; they just come that way. Their humanities don't have to be individuated by anything added on. They are already quite individual enough, thank you. There is a clear difference between this case and Abelard's case of the stone. Neither Socrates nor Plato has a stonehood.

Abelard can hardly object to such a theory, when properly stated. For, in the end, it disagrees in no respect with his own, except perhaps on some minor terminological points that don't really matter. But Abelard is quite clear that his theory is not a realism, as apparently William thought his own was.

This completes Abelard's attack on William's two views and their variations. It is important to notice just what he has done.

He thinks he has refuted William's first, strongly realist, view. Perhaps he has scored some points against William's own formulation. But realism can be touched up in such a way that every single one of Abelard's objections misses the mark. He does not raise the one objection that would have been conclusive, and would have required a major change in the theory: that the view makes accidental change impossible.

He has scored some points against William's way of expressing his second view. But his main success is elsewhere. He has pretty clearly refuted various attempts to explain how William's second view amounts to a realism. There is apparently nothing on that view, no non-linguistic entity, that is predicated of many. So if you are going to adopt a view like this, you might as well stop pretending you are a realist and admit that you are a nominalist.

Abelard's Own Theory

This is precisely the conclusion Abelard himself draws on p. 177: "it remains to ascribe universality of this sort to words alone". Only they are predicated of many. This passage marks a major point of articulation in the discussion. Abelard has now committed himself to nominalism. We have yet to see how the details work out.

On pp. 177-178, Abelard gives some preliminary explanations of the grammatical sense in which words may be called particular (individual) or universal. Then, on p. 178, he raises some questions that will serve to lead into the core of Abelard's own view. There are two questions, both having to do with the "signification" of those general or universal words. (McKeon, whose translation appears in Hyman and Walsh, translates 'significatio' as "meaning". We'll say some more about that in a moment.)

(1) First, there seems to be nothing for a universal word to name, "no subject thing", as Abelard says. For he has just argued that there are no universal entities in the realists' sense to be named by universal words. We'll see Abelard's answer to this in a bit.

(2) But second, there doesn't seem to be anything for universal names to signify either. Here we need a little lesson in the medieval terminology of signification. Aristotle, in De interpretatione 3, 16b19-21, says that verbs "signify" something just as names do, because he who uses a verb "establishes an understanding" - in Boethius' Latin translation the phrase is 'constituere intellectum', usually construed with the genitive. Hence, in general, terms signify what they establish an understanding of, or in more colloquial terminology, they signify what they make us think of when we hear them. (It follows from this basic notion of signification, which all medievals used, that signification is a species of the causal relation, and is just as transitive as causality is. If A signifies B and B signifies C, then A signifies C too. Some authors explicitly drew this conclusion. This, incidentally, is one reason why it is wrong to translate 'significatio' as "meaning". Meaning, whatever it is, is not transitive; signification is.)

In effect, therefore, the second of Abelard's questions is "What does a universal term make us think of when we hear it?" Certainly, it doesn't make us think of a universal thing, since he has just argued that there are no such things. But it doesn't seem to make us think of any individual thing either. When I hear the word 'man', I am not made to think of Socrates any more than I am made to think of Plato. And I cannot be made to think of all men, since I don't know all men. There are lots of people on the other side of the world (in fact, there are lots of them on the other side of the street) that I have never thought of in particular. Surely, when I hear the word 'man', I am not made to think of them - except in some very general sense that seems impossible to explain, since we have no general entities on this theory.

In short, there are two problems: (1) Universal terms don't seem to have anything to name, and (2) they don't seem to have anything to signify either.

Hence it looks as if universal terms cannot be sermones in Abelard's sense - that is, significant words or significant voces - since there is nothing for them to signify or name. It looks, therefore, as if Abelard's nominalism is committed to saying that universal terms are mere words without significance - mere flatus vocis, in Roscelin's phrase - with all the consequences that entails for our knowledge of the world. How is Abelard going to avoid this? How is he going to distinguish his view from Roscelin's?

Well, he does. He sketches his answer in the middle paragraph on p. 179 (before getting down to details). "But this is not so," he says.

First of all, what about naming? The objection here was that there is nothing for a universal term to name because there is no universal thing for it to name. Abelard's response is in effect, "So what?" Why can't the term name individual things?

Abelard's reply here is a bit obscure, but seems to run something like this. The objection works only if we think of naming as a kind of signification, and then argue that terms that signify more than one thing are equivocal (p. 177). On pain of equivocation, therefore, universal terms cannot name many things, if naming is a kind of signification, and since they cannot name just one universal thing (since there is no such thing), it follows that they cannot name at all.

Abelard's reply in effect severs the notion of naming from the notion of signification, in the sense of signification according to which a term is equivocal if it signifies several things. Abelard is willing to allow that there is a sense in which naming is a kind of signification. Universal terms, he says, signify "in a manner" by nomination (=naming). But that is not the kind of signification that is involved in equivocation.

It is not entirely clear to me why Abelard is willing to allow nomination or naming to be a kind of signification at all - even "in a manner". Anselm had already done something like this in his De grammatico, but the whole idea there is a bit strained. In any case, it is clear that naming is not signification in the sense that seems to be presupposed by the objection. I think, therefore, that Abelard has successfully handled the first objection. His answer proceeds by making the notion of naming in effect the same as the notion of predicating, and by distinguishing it from signification in the sense in which multiple signification would entail equivocation.

The second objection, however, is harder. What are we made to think of when we hear the word 'man'? In short, what is the link-up between our universal terms, our concepts, and the external world? How are we going to save Abelard's nominalism from the epistemological skepticism that it threatens to yield? This is the meat of Abelard's theory.

Well, here is where considerations about equivocation do apply. If a universal term is going to be universal, it must establish in us a single understanding, a single concept. ('Understanding' in these contexts does not mean the faculty of understanding, but either the act of understanding or else the object of such an act.)

That concept has to be a general concept. But what is it a concept of? Not of a general or universal thing, since there aren't any. Neither does the universal term make us think of individual things, for the reasons the objection states and Abelard apparently accepts.

And yet that general concept must somehow be grounded in those individual things, on pain of severing our thought from the world and reducing the theory to Roscelin's.

There must therefore, Abelard says (p. 179), be some common reason or cause why the universal term is "imposed on" the several individuals it names ("imposition" is the assigning of names to things), and so names the several individuals it does, and which links the name to the general concept we have when we hear the term. This "common reason or cause" is going to be the linkage between our concepts and the world that saves the objectivity of our knowledge.

Well, how does it work? We need to look at both sides of the question: at this mysterious "common reason or cause", and also at the "common concept" that it grounds.

We have seen that Abelard criticized William of Champeaux's second view for saying that, while Socrates and Plato had two distinct essences, nevertheless they agreed - "indifferently", to be sure - in man or in humanity. Abelard thought that this was just a verbal smoke screen. Abelard says instead that Socrates and Plato agree, or are alike in being a man, or in being man.

So what? What is the big difference here? Well, there is a big difference. A man is a thing - a 'res'. And there is no thing in which Socrates and Plato agree, no thing they can share, as Abelard has already argued.

Nevertheless, they must somehow agree, there must be some community between them, or else there would be no objective basis for our calling them both men, and we would be left with subjectivity and skepticism - and Roscelin's doctrine. The common predication of the word 'man' of both of them must be tied to reality somehow.

Well, Abelard bites the bullet. Since Socrates and Plato cannot "agree in" or share any common thing, and since they must nevertheless have some community, it follows that they must agree in or share some item that is not a thing - not a res.

They do not agree in man, he says, but they do agree in being a man (hominem esse), otherwise translated as "to be a man". Being a man, therefore, is not a thing.

This does not mean that being a man is nothing, that it isn't really out there. It is really out there. It has to be, since there is an important episte- mological job for it to do. Only it is not a thing - not a res.

Instead, it is what Abelard calls a "status" (fourth declension, so that the plural is also 'status'). Socrates and Plato agree in the status of man - that is, in being a man, or in to be a man (pp. 179-180). "We appeal," he says, "to no essence" (p. 180) - that is, to no thing.

Oh great, you may say. What on earth is going on? When it comes to the crunch, Abelard tries to fake his way out by appealing to some contrived and utterly mysterious kind of "non-thing" he calls a status. That's not a theory; that's a sign of desperation! Isn't Abelard in fact throwing the whole thing away? Isn't this mysterious status of man just the old realist universal man in disguise? Hasn't Abelard distinguished his own doctrine from realism by nothing more than a verbal subterfuge?

This is a particularly pressing question once we realize that Abelard needs the status for basically epistemological reasons, and epistemology has always provided the main arguments for realism.

In order to see what is going on, we must remember the way Abelard, following Aristotle, defined a universal in the first place. A universal for Abelard is that which is predicated of many. And while Abelard seems perfectly willing to speak in some passages in his writings of things being predicated of things, he is not willing to allow things to be predicated of many things, since "it remains to ascribe universality of this sort to words alone" (p. 177).

The universal man of William of Champeaux's first theory was a universal because it was supposed to be predicated of many. The status, however, which is just as objective, is not a universal because it cannot be predicated of many. To see this, we must look at the Latin.

What is the status of man? It is, he says, being a man. Now 'being' is ambiguous in English. It can be either (a) an adjective (a participle, in particular), meaning that which is, or else it can be (b) a noun (a gerund, in particular) meaning what that which is does - namely, be. (Compare the difference between 'The living and the dead' and 'Summertime, and the living is easy'.) Both of these senses come into play in the phrase 'Every being has being'.

In Latin, these are quite distinct verb-forms. What we have in the present case, where we are talking about being a man, is the gerund, the nominal form. And in Latin, the nominative of the gerund is the infinitive. So to make this perfectly clear, we can say that the status of man is to be a man (hominem esse).

Now the Latin sentence 'Hoc est hominem esse' (in English, 'This is to-be-a-man') is certainly odd, and perhaps even ill-formed. (In English, there is a reading of 'this is to be a man' that makes sense, namely, as amounting to 'This is going to be a man', or 'This ought to be a man'. But that is quite different in Latin, and involves the gerundive, not the gerund.)

Now I am not very concerned whether you understand all the grammatical fine-tuning here. But the general point is important. Although Abelard doesn't say so, I suspect that this grammatical business is the reason why the status of man - to be a man or hominem esse - cannot be predicated of many. It cannot be predicated at all. It is simply of the wrong form. It results in gibberish.

Syntactically, this is a matter of grammar. Metaphysically, it can be approached somewhat differently.

The theory of the categories may be regarded as providing a list of the basic kinds of predicates - whether we construe that linguistically or metaphysically. But the status of man does not fit into any of the ten recognized Aristotelian categories. Hence, it cannot be predicated, and so a fortiori cannot be predicated of many.

It follows of course that the status not only cannot be predicated of many, it cannot be predicated of even one. Hence the status is not only not a universal, it is not an individual either. I suspect this is what Abelard means when he says that the status is not a thing. Abelard quotes Aristotle (p. 172) as saying that some things are universal and others are individual.

Abelard disagrees about universals, of course, but the point is that since the status cannot be either an individual or a universal, it follows on the authority of Aristotle that it cannot be a thing at all.

Again, in another passage, later in the Logica ingredientibus (in his Glosses on the Categories), Abelard says that the categories signify the ten primary genera of things. Presumably then, since the status does not fit into a category, it is not a thing. But it is still out there.

To some extent, this is speculation. Abelard simply is not very informative about these non-things. But something like this must be going on, or else I do not see how everything Abelard says can be reconciled.

At this point, let us recall the two kinds of realism we have in circulation:

(a) Predicational realism, the view that there are real, non-linguistic things that are predicable of many.

(b) Boethian realism, the view that there are real entities - whether you want to call them "things" or not is up to you - that are common as a whole, simultaneously, and in a metaphysically intimate manner, to several things.

It now looks as if Abelard is a nominalist in the Aristotelian sense, but a realist in the Boethian sense. His doctrine of the status fits Boethius' definition of a universal exactly - or else it will do not do its job of grounding the objectivity of our knowledge. Please note this carefully.

It is the logical doctrine of predication, therefore, that is at the heart of Abelard's nominalism. He is a nominalist only in the "predicational" sense.

Now what about the second side of our problem, the question of the "common concept"? What does a universal term, after all, signify or make us think of? Now you might well ask, "What about the status?" Wouldn't that serve? Perhaps it would. But Abelard doesn't take that route, for reasons we shall see in a moment.

To ask the question as precisely as possible: What is the thing of which a universal term establishes an understanding or concept? There must be only one, under pain of equivocation.

Well, since it cannot be any real thing, as Abelard has already argued, it is, he says (p. 180), a "certain imaginary and fictive thing" - a "res ficta". That is, it is purely an intentional object, a thought object. It is in no sense real - not even in the sense in which the status is real even though it is not a "thing". The thought object is a thing, only it is a fake thing - a metaphysically impossible thing, if Abelard's arguments against William of Champeaux's realism are correct.

Later people will distinguish real being (esse reale), which all of us enjoy, from intentional being (esse intentionale), which thought objects enjoy. ('Intentional' here in the sense that thought intends or aims at, "tends towards", its object, whether the object is real or not.) This is the germ of the doctrine of intentionality that will play such a big role in modern phenomenology. People always say that Brentano got this notion from the Scholastics. Well, he didn't get it from Abelard; he got from later Scholastics. But we can already see the germ of it in Abelard.

There seems to be some obscurity in Abelard concerning the res ficta. On the one hand, he seems to think that this res ficta is a product of the activity of the mind, like dream images, and that it is this product we are made to think of when we hear a general or universal term (p. 180). On the other hand, later in the Logica ingredientibus, in his Glosses on the De interpretatione, he seems to say that such images or figments are not what we think of when we hear a general term, but rather mental products that are the means by which we think of what they are images of. We do not think of the images, we think of things through the images.

The terminology, at least, and perhaps the content of the doctrine as well, is fluid here. Perhaps the best way to view it is this. The universal term establishes in us an understanding, a concept, thought of as a kind of mental picture. That concept or picture is of a metaphysically impossible object. Since terms signify that of which they establish an understanding in us, and not that understanding or concept itself, it is the impossible intentional object that is signified by the universal term, not the concept or image of that object.

Just which of these - the impossible object or the understanding or concept of that object - is to be called the res ficta, I am not clear. But from the way Abelard introduces the term on p. 180, it seems to me that the res ficta is the intentional object, not the concept or image, and it is that res ficta that is signified by the term. But if I am wrong about this, and the res ficta is to be identified with the concept or image, it is easy to make the terminological adjustment. In that case, the term will signify the intentional object of the res ficta or image.

This theory is part of an account of all terms, not just of universal ones. Proper names as well as universal terms produce or establish in us a concept or image of an intentional object, which the term signifies. In the case of a proper name, of course, the intentional object is an individual and so may also be a really existing object. For universal terms, however, this cannot happen.

At the top of p. 181, Abelard suggests an exception to this. A proper name - say, 'Socrates' - need not produce in us a concept or image of Socrates, he says, provided that Socrates is present in person and I perceive him. For in that case we do not need the image in order to be made to think of Socrates; the reality suffices. But where Socrates is absent, I do need his image in order to think of him. (Note: Is Abelard here rejecting the Augustinian theory of representative perception?) Where the term is a universal term, however, there can never be a universal thing really present to my perception, since there are no universal things out there at all. Hence for universal terms we always need an image or concept in order to think of what they signify.

Before we get too far afield, let me point out something. In the penultimate paragraph on p. 180, Abelard considers a view that denies that concepts or images are intentional objects or products of the mind's activity, but instead identifies the concept or image with the very act of thinking itself. Abelard doesn't really argue against this view, but he says he disagrees with it. This is an interesting view, because after toying with a "fictum" theory very much like Abelard's, William of Ockham in the fourteenth century will opt for the act theory Abelard here rejects, the theory that concepts are identical with the mental acts themselves, not with the products of those acts, and certainly not with their objects.

So far, then, we have a fictum theory for Abelard, a theory that applies equally to general concepts and to particular or individual ones. What is the difference between these two kinds of concepts? The distinction is drawn on p. 181. Particular or individual concepts are mental pictures or images that represent one thing to the exclusion of others - for instance, the image of Socrates. General concepts are mental pictures or images that represent several things at once. They are in that sense "confused" concepts. This does not mean that there is anything wrong with them, or that they are stupid. It means that they "fuse together" a number of things. The concept "man", for instance, is equally a picture of every man. It is not any more a picture of Plato than it is of Socrates, or vice versa. But it is more a picture of those men than it is of anything that is not a man.

Well, this is troublesome, of course. At the bottom of p. 181, Abelard gives us an analogy. He says that we can paint a picture of a particular lion - "limping or mutilated or wounded by the spear of Hercules", as he oddly puts it. But we can also, he goes on, paint a picture of no lion in particular, but of a lion in general. This of course suggests that Abelard is leaving himself wide open to all the objections Berkeley would later raise against the notion of abstract general ideas - objections based precisely on the identification of concepts or ideas with mental images.

But let us not push this point. Let us suppose that Abelard can answer such objections - perhaps by pointing to the image of a "speckled hen". If you know the literature on Berkeley, you will catch the allusion. (Nevertheless, whether Abelard can answer such objections or not, the analogy with painting still seems to be a bad analogy.)

If we do not push the objection, then we can summarize Abelard this way:

(1) In the absence of individuals to the senses, a proper name, for instance 'Socrates', brings to mind the image or picture of Socrates, and signifies the individual Socrates. It does this whether Socrates exists or not.

(2) A universal term, for instance 'man', brings to mind a general picture or concept of no one man in particular, but of a man in general, and signifies the metaphysically impossible man in general that I think of through that concept or image.

Thus, terms as a whole signify a realist world, a world with universal things in it - only that realist world is a world of intentional objects, not of realities. Realism is the correct theory for the world we picture, the world we think of. It is, unfortunately, not the correct theory for the world that exists.

The doctrine of metaphysical realism makes the mistake of thinking that the world that is signified is the real world. Abelard thinks it is not. His distinction between naming and signifying properly speaking makes this outcome possible.

Boethius, in his Commentary on Porphyry (Chapter \s4 above), said that we can separate in the mind things that cannot be separated in reality. We can form abstract, general concepts although there is no such thing as a general or universal entity. Abelard is in effect just accepting this theory, and spelling it out in more detail. Unfortunately, he complicates matters by identifying concepts with images.

You may well think this outcome is a rather ironic one for a doctrine that is trying to preserve the objectivity of our knowledge. After all, the upshot of the whole thing is that the world we think about, and so the only world we could even have a chance of knowing, is not the world that exists, but rather a metaphysically impossible world populated with universals. What kind of objectivity does a doctrine like that preserve?

This brings us back to a question we put off a while ago: Why cannot the status serve as the significate of a universal term? Why cannot the term 'man' make us think of the status of man? It seems to be just the kind of significate we want. It is common to many, even though it is not predicable of many. And it would clearly ground the objectivity of our knowledge if it were the significate of the universal term. In that case, the universal term could make us think of something real, even if not strictly a thing, and not of some impossible universal thing predicable of many. The world that is signified, the world we think of or conceive, would then be just the real world after all, and knowledge would be saved.

The reason Abelard does not adopt this attractive approach is that he thinks we cannot form an image of a status.

Before I give you my reasons for this, I should mention one study that disagrees with me. It is Martin M. Tweedale, Abailard on Universals, (Amsterdam: North-Holland, 1976). On p. 208 of that book, Tweedale takes Abelard to hold that universal terms do signify the status. I find no text to support this claim, and indeed I think the whole business about painting pictures of general lions goes against it. But there is one passage in particular where I think Abelard clearly rejects this possibility. On p. 182, Abelard asks whether a universal term signifies the form to which the understanding is directed. He is talking here about the intentional object and not the image. And he goes on to say that this view is confirmed both by authority and by reason. (Note the interesting pairing here.) The 'by reason' seems to clinch it. The "authority" he cites is Priscian, the famous Latin grammarian (c. 500). Priscian, in a rather obscure passage given by Abelard, seems to suggest, Abelard says, that the divine ideas - that is, God's concepts - are concepts of the status of things.

It is easy to see why they would have to be, since the metaphysical structure of creatures is supposed to be patterned after the divine ideas. That suggests that the ideas are ideas of the status, which do enter into the metaphysical structure of things, rather than of metaphysically impossible universals, which do not.

Just how we get from this to the conclusion that universal terms signify the thought object is not clear. But in any case, Abelard seems to accept that conclusion; it is confirmed. he says, "by reason". What he does not accept - and here I think Tweedale misinterprets the passage - is the suggestion that the intentional object is the status.

That's fine for God, he says, but not for us. We cannot form an image or concept of the status. We have no way to picture accurately what it is to be a man. That is because the status cannot be sensed; it is not a sensible quality. We have no perceptual access to it, and therefore do not know how to picture it, we "men," he says on p. 182, "who learn things only through the senses." Notice how Abelard is here beginning the Aristotelian break with the Platonic/Augustinian tradition. That tradition would never have said anything like that. The remark is all the more surprising here, since Abelard didn't have all that much more of the Aristotelian texts at disposal than did the earlier medievals. (He did have a bit more of the Aristotelian logical works, which were just beginning to be translated in Abelard's day. In one passage, he confides breathlessly that he has actually seen a Latin copy of Aristotle's Sophistical Refutations.)

The only things we know about men are their sensible qualities, so that the picture we make to ourselves when we hear the term 'man' must be constructed out of those ingredients.

God's knowledge, however, is not confined to what can be sensed, so that God can picture the status of things. The connection between thought and reality is much closer for God's thought than it can ever be for our own.

Very well, but what does all this mean for the connection between our thought and reality in the final analysis?

Well, the status provides an objective non-arbitrary basis for the imposition of terms. The term 'man' is "imposed" to name all men because they are men, because they agree in being men. They share a common status.

The term cannot signify that status, however, because we cannot picture it. But we do the best we can, and form a kind of monstrous image of no one man in particular, but of an impossible man in general. This image is serviceable; it might, for instance, guide us in predicating the word 'man' correctly, since it is after all equally a representation of every man and of nothing quite so much as a man. It can serve that function. But it is too indefinite and indeterminate to be an exact image of any possible reality. And there is where the realists make their mistake.

Before we turn to an evaluation of this view, let us look briefly at the end of Abelard's discussion. On pp. 183-186 there is a further discussion of abstraction, taken from Boethius' Commentary on Porphyry. Then he turns to answer the four questions he raised at the beginning of the passage.

(1) With respect to Porphyry's first question, whether genera and species subsist, Abelard glosses this as: Do the words - general and specific words - signify something real, or are they purely mental, that is, do they not signify something real? Abelard's answer: Both. They signify by nomination - that is, they name or are truly predicable of - real things. But those real things are individual things, not universal ones. On the other hand, in the other sense of 'signify', the strict sense (what the terms make us think of), they signify nothing real, but only a fictive intentional object. That fiction nevertheless is not exactly empty, for the reason explained in his discussion of abstraction.

(2) As for Porphyry's second question, are genera and species corporeal or incorporeal, Abelard glosses this as: Do they signify corporeal or incorporeal things? And again Abelard's answer is: Both. And once again, he trades on the two senses of signification. By nomination, they signify corporeal things; in the strict sense they do not, but only signify fictions.

(3) With respect to Porphyry's third question, are genera and species in sensible things or separated from them, Abelard glosses this as: Do genera and species words signify things in sensible things or do they signify things not in sensibles? And once again, as you no doubt have come to expect by now, his answer is: Both. By nomination, they signify or name things in sensibles. For instance, 'humanity' signifies (names) the humanities in Socrates and Plato. But more strictly, they signify the common intentional objects, those impossible things that are not in a sensible object but are only fictions.

(4) Finally, what about Abelard's own, fourth question: Can universals continue to exist without any singulars? Again, he parses this: Can universal terms continue to signify as they do if there are no singulars that fall under the universal, no individuals for such terms to be predicated of? Again, his answer is: Yes and No. (This is not a man who wrote the Sic et Non for nothing!) They obviously cannot signify then by nomination, but just as obviously they can continue to signify in the strict sense; they can signify the common intentional object.

The distinction between nomination and signification in the strict sense is therefore the main vehicle for Abelard's answering the four questions that frame his discussion. But don't be fooled. There's a lot more than that going on in this passage.

Now, let us pause to evaluate this theory. It is subtle and deep, no doubt, but I think it just will not work. What, after all, leads us to form exactly the image we do when we hear the word 'man' - an image of an impossible man in general, but an image that is nevertheless equally if not exactly a representation of just exactly those individuals who share in the status of man?

What a coincidence! Isn't it odd that the fit should be so exact? It seems that this must be an extraordinary coincidence unless we are led to do this by somehow getting in touch with the status. The status must guide us. But how can that be? On Abelard's own account, we cannot sense the status and so cannot form any concept or picture of it.

It looks as though we can have no inkling at all of the status of things. And if that is so, then while the status may very well be out there, it can be of no epistemological use to us whatever - and the whole project breaks down.

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