Jump to ContentJump to Main Navigation
The Metaphysics of the Incarnation$

Anna Marmodoro and Jonathan Hill

Print publication date: 2011

Print ISBN-13: 9780199583164

Published to Oxford Scholarship Online: May 2011

DOI: 10.1093/acprof:oso/9780199583164.001.0001

Show Summary Details
Page of

PRINTED FROM OXFORD SCHOLARSHIP ONLINE (oxford.universitypressscholarship.com). (c) Copyright Oxford University Press, 2020. All Rights Reserved. An individual user may print out a PDF of a single chapter of a monograph in OSO for personal use.  Subscriber: null; date: 24 November 2020

Vehicle externalism and the metaphysics of the incarnation: a medieval contribution

Vehicle externalism and the metaphysics of the incarnation: a medieval contribution

(p.186) 10 Vehicle externalism and the metaphysics of the incarnation: a medieval contribution
The Metaphysics of the Incarnation

Richard Cross (Contributor Webpage)

Oxford University Press

Abstract and Keywords

This chapter analyses Duns Scotus's account of the various kinds of ‘tying’ relations required to unify the various properties exemplified by the incarnate Son, and his account of the various kinds of properties relevantly tied (universals in the case of the divine attributes, and something like tropes in the case of the human ones), explaining his reasons for needing such a complex analysis of the issue. The chapter tests his view for cogency, and offers it as a contribution to the modern debate on the metaphysics of christology.

Keywords:   Scotus, Aquinas, Hervaeus Natalis, relations, properties, tropes, compositionalism, concretism, dependence, causation, mereology, free will

In what follows I will use some insights from the medieval philosophers to develop an account of the incarnation that makes central the notion that Christ's human nature should be thought of as an instrument of the second person of the Trinity. Some medieval thinkers, especially those influenced by Aquinas, made this notion central to their doctrine of the hypostatic union. But the relevant accounts of instrumentality are developed with sufficient detail only by Duns Scotus. So I combine the insights of the (Franciscan) Scotus on the nature of instrumentality with the accounts of the incarnation proposed by some of his (Dominican and Thomist) opponents. Indeed, as I shall show, Scotus himself explicitly rejected the notion that Christ's human nature could be an instrument of the divine person, and had some powerful reasons for so doing. So I shall try to propose, on behalf of Scotus's opponents, replies that they could offer to Scotus's objections, in some cases using further insights derived from Scotus's inventive and creative philosophical theology.

The Metaphysics of the Incarnation

The thinkers I shall be interested in here all suppose that the incarnation is a relation between two individual substances: the second person of the Trinity, and a concrete human nature. They claim too that the relevant relation is one of dependence. But they differ on the relevant kind of dependence—specifically, whether the dependence should be thought of in specifically causal terms. Scotus, for example, thinks not:


The relation [between the human nature and the Word] is one of order or dependence of a different kind from every sort of dependence in the order of caused to cause…And although it is difficult to see that there is some such dependence, nevertheless it seems to appear somehow in subject and accident. For an accident has two sorts of relation to its subject: namely, of what informs to what is informed (and this necessarily includes imperfection in the informed subject, since [the subject has] potentiality in relation to a qualified (i.e. accidental) act); and [the accident has] another [relation] as of what is naturally posterior to what is naturally prior, on which it depends as on a subject, not as on some cause…If therefore these two relations of an accident to a subject are distinguished from each other, one is necessarily to the subject under the description of some sort of imperfection in the subject—viz. of potentiality—whereas the other does not necessarily posit any imperfection in it, but only natural priority and substantification in relation to the accident. The relation which is the dependence of the human nature on the divine person is most similar to this.1

The opposing view—that the relevant kind of relation is a causal one—is spelled out very clearly by Hervaeus Natalis, a slightly younger contemporary of Duns Scotus:

We should see the kind of mode in which the human nature is understood to be united to the Word and dependent on the divine suppositum (innixa supposito divino). It is impossible for a human being to explain this sufficiently. But among those modes, the union of soul to body is closest to this union: not as the body relates to the soul as a proper subject relates to its form, but as it relates to [the soul] as an instrument naturally joined [to the soul], through which the other [viz. the soul] in some way operates, even though the one relation is not without the other in an animate body. So it is necessary to imagine some such dependence of the humanity on the divine suppositum without its being the case that one is properly speaking a part in relation to the other, or that there is any kind of composition properly so-called. In order to express this relation, the relation of an instrument to that which operates through it is most appropriate—especially when the instrument is naturally united.2

The idea is that there are two ways in which soul and body are related. The soul is the form—the structure—of the body; and it uses the body in bringing about its consciously chosen activities: it is an efficient cause, and the body is its instrument. (The example Hervaeus goes on to give just after the passage just quoted is that of a human being and his hand—an example deriving from Aquinas).3 Scholastic accounts usually appeal to relations of the former (p.188) kind—formal, not efficient causation—to explain the unity of an item: hence Scotus's preference for seeing the human nature as something like an (accidental) form of the second person of the Trinity, rather than something that depends ‘in the order of caused to cause’. The second person of the Trinity and the human substance, for example, are supposed to be just one subsistent thing, and seeing the human nature as something like an accident provides some sort of account of how this might be: just as Socrates and his whiteness, for example, are one subsistent thing—namely, Socrates himself. So while it is easy to see how Hervaeus's account might give an account of Christ's intentional human activity, it is hard to see how it might give an account of unity—given that unity is customarily explained in terms of forms and their subjects.

In the next section, I consider an alternative account of unity, proposed by Duns Scotus—one that would have served Hervaeus very well. I develop, too, a fuller account of the efficiently causal relation between soul and body that would allow a more complete account of the hypostatic union—again, based on insights from Scotus. But first, I want to consider an objection to any account that sees the incarnation as a relation between two concrete substances. On the face of it, human persons are concrete human substances, and if Christ's human nature—a concrete human substance—is a person, it looks as though Nestorianism will be true. Medieval theologians from Scotus onwards generally denied Nestorianism by arguing that a necessary condition for a substance's being a person—what the medievals referred to in technical language as a suppositum (= hypostasis)—is failing to depend on some other substance. Hervaeus sets the position out as follows:

Suppositum’…means in reality that which has existence independently (non innixum) of any other suppositum: it is dependent on another neither by inherence (like an accident on a substance), nor as a part on a whole is (e.g. an arm on a body), nor somehow as something like a part on a whole is (e.g. the human nature dependent on a suppositum in Christ).4

Nestorianism is thus ruled out by fiat; but not in a wholly unprincipled way. Hervaeus, for example, claims that the human substance is an instrument of (p.189) the divine person, and there may be good reasons to suppose that a substance that satisfies all the requirements for being an instrument of the relevant kind lacks some feature that we might label ‘personhood’. (By ‘all the requirements’, I mean to include more than just the kind of instrumental relation highlighted by Hervaeus. I return to this issue later, when I have made clear what causal relations I think are necessary in this case.)

Instrumentality, Substantial Unity, and the Incarnation

Clearly, Hervaeus wants to use this causal relation to account for the hypostatic union. Thus, he needs some way to secure the fact that an instrument somehow becomes part of the person that uses it. Sadly, Hervaeus nowhere develops an account of instrumentality sufficient to guarantee this. But Scotus—who, as we shall see, explicitly rejects an instrumentality view of the incarnation—provides just the sort of account that would have proved useful to Hervaeus. Scotus claims that, in certain cases, a soul (or a whole body) can use a conjoined instrument in just the same way as it uses a part of the body. The context of the discussion is a particular problem in Scotus's theory of cognition. According to Scotus, when we call stored mental content to mind, the relevant occurrent cognitive act is caused jointly by the intellect and the stored content itself. Scotus labels such stored content an ‘intelligible species’. So his view is that the occurrent cognition is caused by the intellect and an intelligible species. But the occurrent cognition inheres in the intellect, so it seems that a partial cause of something inherent in the intellect—the occurrent cognition—is something else inhering in the intellect—the intelligible species. And this seems to raise problems about self-motion. Scotus's reply is that the inherence of the cause makes no difference to the causal story, and he illustrates his claim with a very striking example:

That [the species] perfects this intellect is accidental to the species, in so far as it is a partial cause with respect to the act of cognizing, concurring with the intellect as the other partial cause. For even if it perfects [the intellect], it does not give the intellect any activity pertaining to the intellect's causality. Example: the motive power in a hand can use a knife to cut up a body, in so far as [the knife] is sharp. If this sharpness were in the hand as its substance, then the hand could use it for the same operation, and nevertheless it would be accidental to the hand (in so far as the motive power is in it) that sharpness is in it, and vice versa, because the sharpness gives the hand no perfection pertaining to [motive] power. This is apparent, because the motive power is equally perfect without such sharpness, and it uses [the sharpness] in the same way (p.190) when it is in some other thing joined to the hand—such as a knife—as it would use it if it were in the hand.

So it is in the case at hand. If the species could exist in the intellect without inhering in it in the manner of form, and if by that mode of existence in [the intellect] it were or could be sufficiently conjoined to the intellect, these two partial causes, conjoined to each other, could [tend] to the same operation to which they now can when the species informs the intellect. This is also apparent if we posit some intelligible [object] present without a species. That object is a partial cause that does not inform the intellect (which is the other partial cause); but these two partial causes, close to each other without the informing of the one by the other, cause one common effect by their required proximity alone.

If this latter is posited, perhaps there is a reason for it to be impossible that an accident, which is an immanent and not a transient (transiens) principle, be sufficiently joined to the substrate unless it is in it subjectively—which is why it is called an accident. But surely the divine essence in the intellect of a blessed—which is neither immanent in the essence [of the blessed] nor in something of which it is a form—is a principle of intuitive cognition.5

The idea is that the knife and the body form something with a unity as tight as would obtain in the case that the blade was straightforwardly a part of the body (in the manner, say, of Johnny Depp's Edward Scissorhands character). The knife and the body become one subsisting thing. Presumably, the only significant difference is that the knife is easier to detach than (say) scissorhands would be. But if this is correct, it has a striking consequence for an account of substantial unity. For there is no unity in form between the knife and the body (even if there is in the case of Edward Scissorhands). So whether or not relations of formal causality sufficiently explain unity, they cannot be necessary conditions for unity. Unity can be explained equally by relations of efficient causality.

The application to the incarnation is obvious. By being an instrument of the Word, the human nature and the Word become one subsisting thing. Furthermore, the human nature and the Word become the subsisting thing that is the Word—just as the knife becomes (in effect) a part of the body. The body extends itself to include the knife; the Word extends himself to include the human substance.6 Hervaeus's position on the incarnation, then, marks a considerable advance in explanatory power over the one proposed by Scotus. (p.191) Scotus can claim that the union is like that of substance and accident; Hervaeus could claim (if he adopted Scotus's account of instrumentality) that the union simply consists in an instrumentality relationship. The nature's being an instrument of the divine person explains the union.

This view of the incarnation requires occasionalism to be false, since in effect it posits something like occasionalism in the case of the incarnation, and if occasionalism obtained generally, it would entail on Hervaeus's account that God is incarnate in the whole universe. Now, the medieval theologians all have strong accounts of God's primary causality, and of his causal concurrence with every creaturely act. Scotus, for example, claims that, barring special divine intervention, God and creatures are what William A. Frank has labelled ‘autonomous co-causes’ of creaturely effects.7 God is the primary cause of such effects, and creatures secondary causes: a creature acts ‘through a form that is proper and intrinsic to the agent, even though in acting through that [form the agent] is subordinated to a higher agent cause’.8

What Scotus has in mind is that the divine activity and the creaturely activity are necessary and jointly sufficient for the effect. Both God and the creature act by means of their intrinsic causal powers (‘through a form proper and intrinsic to the agent’). The picture is most easily grasped in the case that God fails to co-operate with the created cause—for example, the story of Shadrach, Meshach, and Abednego:9

There is an example in the case of the fire in the furnace, which did not act to bring about the destruction of the three boys—not because of some impassibility intrinsic to the boys, or because of some lack of passive potency [in the boys], or because of some intrinsic contrary impediment, but because God, through his will, did not cooperate in the action.10

The excluded options are intrinsic impassibility (the boys were simply such that they cannot be affected by anything external); lack of passive potency (the boys are such that they cannot be destroyed by fire); and an intrinsic contrary impediment (there is some kind of internal block on destruction).

The subordination claim, made in the last quotation but one, is not to be understood in any particularly strong sense. For example, it is not the case that God causes the creature to act. An example that Scotus often gives is that of the causal roles of father and mother in procreation:


If it is posited that a mother has active causal power in the generation of a child, that power and the active power of the father cooperate (concurrunt) as two partial causes—but ordered, because one is more perfect than the other. But the less perfect does not receive its causality from the more perfect cause; neither does the total causality belong more eminently to the more perfect cause. Rather, the less perfect cause adds something, to the extent that the effect can be [produced] more perfectly from the more and less perfect causes [together] than [it would be if it were] from the more perfect cause alone.11

The idea in the last sentence is that the proper causal structure of the world requires the cooperation of both causes; if just one of the two were the cause of the effect (for example, in the case of occasionalism), the causal structures would be violated.

So God and creature are two distinct and independent causes of any given effect; God contributes more (as the more perfect cause), but the secondary cause is an independent contributor: it is not moved by the primary cause. Scotus explicitly contrasts this with the case of instrumental causation. In this kind of causal cooperation, the lower cause is moved by the higher cause—it has a causal role in the production of the effect only because it is made to do so by the activity of the higher cause:

That which lacks an active form in its order, even when it depends on a higher [agent] in its activity, can be called an instrument. Rather, it acts only through the actual motion of some other mover, as is clear in the case of artificial instruments, such as an axe or a saw, and such-like, because if an instrument is posited to be active properly speaking (proprie) to some end result, it is necessary for it to have some intrinsic active form, either in esse quieto prior to any motion of another agent, or in fieri, when it is wholly moved by a higher agent. For if it has an active form in neither sense, then it would not in any way act properly speaking.12

A form active in esse quieto is a causal power in virtue of which something can act independently unless impeded by some external agent or circumstance (assuming God's primary causality, as just outlined, of course). In the case of a voluntary agent, the relevant power is such that the agent can contingently place itself in a situation where it causes unless prevented by something external.

Talk of esse quietum is a way of picking out the fact that the relevant cause is not moved by anything else; so too is talk of ‘an active form in its order’—we do not need to worry too much about these rather ugly bits of jargon. A non-instrumental cause has such a power. Instruments lack such a power. But they (p.193) do have causal powers or internal structures such that, when moved, they can have a genuinely causal role in the production of an effect. Scotus's technical term for such a power or structure is ‘a form active in fieri’: a form that has a causal role when moved. His reason for this is that, if the instrument had no such power, then the internal structure of the instrument would make no difference to the kind of effect that is produced, and (absurdly) any instrument would do for any effect:

For when a first act is the principle of some second act that has no per se active first act in that order, then it cannot bring about any second act in that order: otherwise anything could be said to be an instrument with respect to anything – and it could be said that God might create an angel through a fly as his instrument, which is nothing: for if it is repugnant to some nature that it is the active principle of certain actions, then it is not possible that, through any power whatsoever, it is the principle of them. For even if God can absolutely create a cold thing, he cannot [do this] by means of heat, such that heat is an active cause, in some order, in relation to coldness; or in some other example, where there is such repugnance to acting. From this it is clear that an instrument, if it is posited to act properly speaking to some effect, or [if it] has an active form in esse quieto prior to motion (as a secondary cause), or at least receives, in an actual motion, an active form by which it acts in its order of acting…then in this way this proposition is always true: ‘Every instrument, which is active properly speaking…attains the principal result through some intrinsic form, whether that intrinsic form precedes the motion or merely is in [the instrument] while it is actually moved.’13

(‘First act’ here refers to a causal power, and ‘second act’ to an effect of such a power, and the point of the awkward first two lines here is that a cause cannot use an instrument in causing an effect unless the instrument itself has some relevant causal power.) So not even God can use just any instrument in the production of any effect: there must be some structural feature of the instrument that enables it to be used in the production of such and such an effect. In the case of the incarnation, the intrinsic powers of the human substance are such that the divine person can do things that human beings can do—for example, touch things—that he cannot do unless incarnate.

What the knife example adds to all this is a stipulation that the instrument and the principal cause become one substance for the duration of their conjunction. Viewed in terms of Hervaeus's twofold account of the relation of body and soul—as subject and form, on the one hand, and instrument and agent, on the other—Scotus's view would be that the fact that the soul can use the body as an instrument is sufficient to guarantee their union as one person, (p.194) irrespective of the fact that the soul informs the body, or is the form of the body. So here instrumentality explains unity, irrespective of questions of inherence.

It would be an easy matter to give an account of the incarnation if we could show that every human activity fits this paradigm: that every human activity is such that, in the case of an incarnate divine person, it is caused directly by that person. I do not know what Hervaeus would say. Scotus would have a hard time accepting it, for as he sees it a necessary condition for being human is not merely being able to do the kinds of thing that humans can intentionally do, but also being causally dependent on certain internal causal structures and devices necessary for continued human functioning. Hervaeus's account of the incarnation does not posit any such dependence in the case of the second person of the Trinity. For example, according to Scotus a human substance is dependent on the continued functioning of its organs:

A part [is] an instrument, and properly active with respect to the whole. And being the principle of an operation pertains primarily to the part, and to the whole only through the part. And in this way the Philosopher calls the organs those parts according to which an operation pertains to the whole…[The organ's] active form belongs to it and to the whole: but it belongs to [the organ] primarily, and to the whole essentially (per se) but not primarily…The same power is participated in different ways in the whole and in the part, and…it can be conceded that such a part does not properly have…another action than the whole, acting through the part, [does]; but it has the same [action] in a different way…An example…: the organ of the nutritive power and the whole animal with respect to nutrition.14

The idea is that there are certain kinds of part—here the organs of bodily functions over which we do not habitually exercise conscious control—whose activities belong to the part and by piggybacking belong to the whole. But because the relevant organs are parts of the whole, they count as instruments of the whole. For example, the efficient causality of a heart is a partial cause of an animal's persistence. But because the heart is a part of the animal, there is clearly a sense in which the relevant efficient causality belongs to the whole animal as a (partly) self-sustaining thing: the animal, including all its relevant parts, has a necessary but non-sufficient causal role in its own persistence. That the whole stands in a certain causal relation to these various parts is necessary for the continued existence of the whole animal. Scotus seems right to think that there will be a difference in causal story between this case and that of (for example) using a knife: the instrumentality of my heart (in keeping me alive) is different from that of my arm holding a knife, or of my Scissorhands-style (p.195) modified arm. And the difference seems to be exactly the one he highlights: my heart is some kind of efficient cause of which my persistence is the effect; whereas in the knife case I am the efficient cause such that some external object is the effect.

Be that as it may, Scotus does not spell out the relevant conditions for the sort of parthood he has in mind here. He does, however, make it clear that the different organs have different forms (‘[The organ's] active form belongs to it…primarily’), and elsewhere he makes it clear that the organs count as part of one whole on the grounds that there is a form of the body, potential to the organic forms: just as the bodily form inheres in matter, so the organic forms inhere in the composite of matter and bodily form.15 But it seems to me that this inherence account is not the only possible account that would allow for the kind of organic instrumentality that Scotus defends here. For it is easy to imagine external vehicles having just the same causal role as bodily organs, but without such vehicles satisfying the inherence requirement. There are, for example, artificial devices that could perform the roles of the various organs: some are implanted (for example, artificial hearts such as the Jarvik-7), but some (such as incubators or dialysis machines) are fully external. And the causal story in both such cases will be much the same as obtains in the case of natural bodily organs: the organs/vehicles are in some way efficient causes of the persistence of the whole. If we adopt Scotus's (efficiently) causal account of substantial unity, we could say that such artificial devices become part of the substance that uses them. And these devices—unlike the knife—are necessary for the continued human existence of the person that uses them.

The discussion provides a way of fleshing out (as it were) Hervaeus's account of the incarnation. On the knife-style instrumentality that seems relevant to Hervaeus's account, we could provide what would be an account of the conscious activity of the second person of the Trinity: the person's using the human substance to perform certain actions freely chosen by the second person of the Trinity. And we could provide an account of the unity of person in the incarnation. But it is not so clear to me that this account would be sufficient for the divine person's actually being human (much as the knife-user is not a knife, even if the knife is as much a part of her as a Scissorhands-style blade would be). But we might posit that a person who is somehow dependent on a whole complex of human parts would count as a human person. On this view the Word would be passive relative to the relevant causal activity of the human parts, much as I am passive relative to the relevant causal activity of my organs: they cause my survival. So we could put the (p.196) theological issue more bluntly: the human substance in the incarnation is something like a cause of the continued human persistence of the second person of the Trinity: it performs a function, relative to the second person of the Trinity, analogous to that of (say) an incubator relative to a premature baby, or an artificial heart relative to someone with end-stage heart failure (bearing in mind that on Scotus's account these devices would count as parts of the person).

Now, I claimed above that the fact that a whole animal stands in a certain causal relation to its various parts (be they natural or artificial) is necessary for the continued existence of the whole animal. Clearly, it is accidental to the existence of the divine person that he exists (for a time) as a human being. So these parts are not necessary for his existence, even if they are necessary for his being human. But could we claim that having a human nature as the relevant kind of organ—one with an efficiently causal role in explaining the person's continued human existence—is sufficient for being human, and not merely necessary? I do not see why not. Consider a standard case. In such a case, possession of the whole array of relevant human parts is sufficient for being human, and the efficient causality of such parts is in some sense sufficient for human persistence. (For convenience, I am bracketing out environmental conditions, but the theory could take account of these too, mutatis mutandis.) It would be the same in the case of divine person: the efficient causality of the whole array of relevant human parts is sufficient for the human persistence of the divine person. The human life piggybacks onto the divine person: the divine person borrows it from the human substance.16 How might the human substance, or the whole array or relevant human parts, come to perform such a function? I suppose one obvious thought is that God, or a divine person, can put himself into whatever causal relations with the creation that he wishes. Certainly Hervaeus's straightforward instrumentality account requires something like this: God can use whatever he wishes as an instrument, and a fortiori can use a human substance as one. What I am suggesting is that, in addition to this active causal relation, God, or a divine person, can allow himself to be affected in whatever way he choose, and can begin to stand in a causal relation to a human substance analogous to that had by a human substance relative to an incubator.17

(p.197) This account has some important consequences for the metaphysics of personhood. I noted above that a substance that satisfies all the requirements for being an instrument of a divine person lacks some feature that we might label ‘personhood’. As Hervaeus might think of it, a person is a human substance that fails to be an instrument in the kind of total sense that I have been trying to outline. It is not the case that a person is always an instrument in the conscious intentional activity of another agent, and it is not the case that it is always an efficient cause of the human survival of some other substance. And this seems plausible enough. In this sense, a person is independent. But we see too that there are other senses in which persons are dependent: paradigmatically, the divine person is dependent, for his human survival, on the causal activity of a human substance; and all human persons are dependent on body parts. Furthermore, if the Scotus-style account of identity is correct, it makes no difference to the identity of such dependent persons whether their survival is contingent on internal (to the organic body) or external, or natural or artificial, causes.

Incarnational Instrumentality and Christ's Human Will

In Peter Lombard's Sentences, the discussion of Christ's human will occurs in distinction 17 of book 3. Book 3 of Hervaeus's Sentence commentary ends before a discussion of distinction 17, and thus lacks a treatment of Christ's human will. So it is not clear what his discussion might have included. But there seems to be an obvious objection to his instrumentality account. The human substance joined to the second person of the Trinity has, according to Christian orthodoxy, a will of its own. And this seems to be what Scotus calls a form ‘active in esse quieto’: the human substance is such that it can (contingently) put itself into situations where it causes unless prevented. Clearly, such a substance could be an instrument of another, and it could be so voluntarily; but it could also be so involuntarily, and in either case there would be a failure of instrumentality: the second person of the Trinity could not control the human will. A human substance is not the kind of thing that can be a total instrument of another substance in this sense—one whose actions are totally under the control of another substance. Scotus certainly argues as much—it is the key component in his rejection of Hervaeus-style christologies. When discussing Christ's human will, Scotus objects to the view that there could be such a will on the grounds that (in effect) the human substance is a total instrument, and nothing that is a total instrument could possess a will:


It is shown that there are not two wills in Christ…on the grounds that every will is the master of its act; but if there were two wills in Christ, one would not be a will, because it would not be the master of its acts. Proof: that power which follows the motion of another power is not the master of its act, but is subordinated to another in respect of its act. But if there were a created will in Christ, that [will] would follow in its action the motion of the uncreated will of the Word, because the Word would bring about the operation of the human nature. Therefore [there is no created will in Christ].18

Scotus accepts that there is a human will in Christ, since the alternative is unorthodox. So he argues rather, tollendo tollens, that the human substance cannot be a total instrument of the Word:

I concede the major premise, that every will is the master of its act. But when it is said in the minor premise that a will that follows the motion of another power is subordinate to that power, and not its own master, I say—as elsewhere, in the first question of this third [book]—that the Word has no causality over the act of the created will in Christ that is not had by the whole Trinity: and for this reason the created will in the Word is not deprived, on account of the union to the Word, of its mastery in relation to its acts any more than it would be if it were not united to [the Word].19

The conclusion is startling: the relation between human and divine wills is just a standard case of secondary causality.

Hervaeus's account of the incarnation is—arguably—an attempt to develop certain insights from Aquinas. Aquinas himself simply claims that human wills can be instruments, even total instruments:

The human will in Christ had a certain determinate mode from the fact that it was in the divine hypostasis: namely, that it was moved always according to the nod of the divine will.20

Aquinas's view of the will generally tends more to the compatibilist than Scotus's does: and presumably Aquinas could give an account of this total instrumentality by simply positing that God causes, whether directly or not, only those desires that eventuate in good actions. But the position for Scotus is a bit harder, since Scotus explicitly denies that desires—inclinations of the will—determine the will's decision.21 Self-mastery requires both the capacity (p.199) and the opportunity to choose between different desires. I return to this in a moment.

If total instrumentality seems hard to square with the presence of a will, Scotus's preferred option seems to tend radically towards the opposite error: if God has no more influence over Christ's human will than he has over mine, does it really make sense to think of the Word as a human person at all? Scotus, of course, has an easy reply: what makes the Word a human person is the fact that the human substance non-causally depends on him. As I noted above, this account lacks much explanatory power. But even if we accept it, Scotus's claims about Christ's human will seem to lead to a problem. Autonomous human substances have been known to sin, and if Christ's human will is subject to no more direct divine control than (say) mine, it seems hard to see how the human substance could be impeccable; and if the human substance is not impeccable, neither is the second person of the Trinity (since on Scotus's account the activities of that substance piggyback to the second person of the Trinity).22 Scotus attempts to reply to this without having to appeal to instrumentality:

The nature that [the Word] assumed, was of itself peccable and able to sin, because it was not beatified from the union alone, and had free will—and was thus able to aim either way (vertibilis erat ad utrumque); but it was confirmed from the first instant, through beatitude, such that it is impeccable, just as the other beati are impeccable.23

Given that, as he believes, Christ's human nature enjoyed the beatific vision, Scotus argues (parsimoniously) that the same account of impeccability can be given in Christ's case as can be given in the case of any of the blessed—supposing the generally accepted theological position that the blessed are impeccable. The basic idea is that such a substance retains the power to sin, but is (somehow) deprived of the opportunity to exercise this power:

It is clear that someone blessed is impeccable in the composed sense, i.e. he cannot simultaneously be blessed and sin. But the divided sense, ‘that someone remaining blessed does not have a power or possibility for sinning’ can be understood in two ways: either [that he does not have this power] because of something intrinsic to him that excludes this power, or because of some extrinsic cause that excludes from him the proximate potency [for sinning]: e.g. even if someone having sight has an intrinsic power for seeing any body, nevertheless it can be made, through some extrinsic cause, perpetually impossible, for him to see by proximate potency, as in the case that the cause brings about perpetual distance between the sight and the body: just as, if there were a perpetual obstacle between the heavenly empyreum and the eye of a damned (p.200) person, that eye could not see the heavenly empyreum (talking about proximate potency), and this because of some extrinsic cause perpetually preventing the power. But it could see by intrinsic and remote potency, such that there is no intrinsic cause of the impossibility.

So I say that in the will of Michael [the archangel], now blessed, there is no intrinsic cause through which the power for sinning at some other time is excluded, in the divided sense, for there is no intrinsic cause preventing this power from being reduced to act. But it is, in virtue of an extrinsic cause, impossible for this power to be a proximate [power] for sinning, viz. by the will of God going ahead of (praevenientem) the will such that it always continues its act of enjoyment, such that it can never reduce its remote power for not enjoying, or for sinning, to act: even though if the secondary cause were never led (praeventa) by a higher agent cause to one option, the proximate potency could go to the other option.24

The blocks on sinning are extrinsic to the agent: God somehow prevents it from sinning. Scotus does not explain what the relevant mechanism is, but presumably one obvious way would be for God to bring it about that the will is only ever subject to good desires, such that, whichever desire it acts on, the agent's activity will be good. God ‘goes ahead’ of the will by causing the good desires on the basis of which the will chooses. And we could think of this as analogous to the damned person's obstacle, preventing him seeing the sky. God prevents bad desires in the blessed, by causing only good ones.

This view, it seems to me, satisfies Scotus's requirement that a free agent has self-mastery. As he sees it, self-mastery amounts to an ability to choose between different desires; and this account does not require any commitment to a specific view of just how the relevant desires were produced. But what could we say on the assumption that the human substance is a total instrument of the second person of the Trinity? The point of this christology, of course, is to deny that the human substance has self-mastery. Given that orthodoxy requires that Christ have a human will, the question is whether it would be possible for there to be a human will—with libertarian freedom—that nevertheless was such that it was always necessitated by antecedent causal conditions (in this case the causal determination of the divine person). The idea is that the relevant human substance is such that it is always made to choose whichever alternative is chosen. If this situation—its always being made to choose—was a necessary feature of its, we would certainly want to say that the substance lacked human free will (understood in Scotus's incompatibilist sense). But perhaps the human substance united to the Word was not like this. It is the union that causes the human substance to be such that it is made to will such-and-such. But suppose it is a contingent matter (p.201) that this human substance is united to the Word—something that seems wholly plausible, since the union is simply a matter of the Word's putting himself into certain causal relations with a human substance. In this case, it seems that we should claim that the human substance indeed has free will, even if it is not in a situation in which it can exercise its autonomous freedom. De dicto, it is not possible that a human substance united to the second person of the Trinity act freely. But de re, it is possible, because the Word could cease to use the human substance as an instrument, and in this case the substance could begin to exercise autonomous agency.25

Does this instrumentality clash with the view that Christ's human substance felt the pull of human appetites incompatible with the divine will? Not at all; nothing about this view requires that Christ's human substance not have all sorts of such desires. All that the view requires is that the choice between the desires is made by the divine person. The case contrasts nicely with that of the blessed. In the blessed, God is the causal explanation of the desires, not of the choice between them; in Christ, the desires can arise in the normal way; the choice is made by the second person of the Trinity. Equally, it may well be the case that the human substance was not aware that it lacked ultimate autonomy. But this does not seem to be an insuperable problem. Even supposing that each of us has such ultimate autonomy, it is not clear that we can tell as much simply by introspection. For all we know, compatibilism is true; and this is so even if it seems to us, introspectively, that compatibilism is false. And the de re possibility of the autonomous activity of the assumed human substance in Christ is sufficient to secure the presence of a genuinely free will.

Incarnational Instrumentality and the Nature of God's Trinitarian Causal Activity

As we have already seen, Scotus holds that ‘the Word has no causality over the act of the created will in Christ that is not had by the whole Trinity’. His reason for this is that he holds the Augustinian axiom that ‘Father, Son, and Holy Spirit…operate inseparably’:26


The relation [between the human nature and the Word] is one of order in one of the related extremes. But it is not one of caused to cause, because that is common to the whole Trinity.27

Why think that the Augustinian axiom is true? Scotus holds that the divine persons share intellect and will, and he holds that shared mentality entails shared agency:

The formal and proximate basis for causation in God is the intellect or the will, or some act of intellect or will. But the intellect of the three persons is the same and the will [of the three persons] is the same, and consequently entirely the same act of understanding and willing is theirs, and the same object, whether primary or secondary. Therefore too the same formal basis of causing—even the proximate [formal basis]—is common to the three persons.28

So this depends on the view that the divine persons necessarily have shared mentality. One common medieval reason for holding this view is that, following Augustine, the medievals generally held that the only features that distinguish the persons are their relations of origin: the causal relations that they have to each other. If the only features that distinguish the persons are causal relations among themselves, it follows that they must have shared mentality: mental states are not relations of origin, and have some plausible claim on being monadic properties, not relations at all. Curiously, it is Scotus himself who has the resources for rejecting the relation theory of the constitution of a divine person, and thus for allowing there to be distinct mental lives and hence causal powers in the divine persons.

A standard reason for accepting the relation theory of the constitution of a divine person is that denying it entails the possibility that one person could exist without the other, and thus that (at best) tri-theism is true, given that one quick and easy way of distinguishing Trinitarianism from tri-theism is that in the former, but not in the latter, each person requires the existence of the other two. Thus Henry of Ghent:

If there is a suppositum in God, it is necessary that it is constituted by a respective ratio founded in the essence, which does not bring about any determination of the thing considered in itself (because it [viz. the relation] does not determine the essence), but [brings about the determination] only of the thing compared to another according to relative opposition, which is [a determination] of the related thing, which is itself constituted as it were from essence and relation, and itself determined and distinct.29

(p.203) Now, this view does not entail that the persons have no other distinguishing features, although Henry (along with all of the medievals) in fact assumes that this is the case. But be that as it may, Scotus disagrees with Henry's relation theory, and in so doing allows a fortiori for the possibility of there being non-relational distinguishing features such as intellect and will:

I say that if we posit that the persons are relational, it is necessary to posit that they are truly subsistents and that the same undivided nature is in them. This cannot be posited on account of some imperfection of the persons in [their manner of] subsisting, for they are posited to be as truly subsistent as they would be if they were absolute. Therefore it must be posited on account of the infinity of the essence which is in the subsistents. But the infinity of the essence would be the same if the persons were absolute. Therefore it would not be necessary for the nature to be divided in that case, just as now the [relational personal] property does not [divide the nature]. Therefore let this be proved: ‘Every nature, common to absolute supposita, is distinguished in them’. For this is true in creatures, but in the case at hand it begs the principal conclusion.30

The relevant requirement of orthodoxy is that tri-theism is false: that is, as Scotus sees it, that ‘the same undivided nature is in’ all three persons. Scotus does not see that satisfying this principle requires positing that the persons are distinct by relations, since it could just as well be the case that the same undivided nature were in all three persons on the assumption that what distinguishes the persons is some non-relational feature. As Scotus sees it, it is the infinity of the essence that entails its being numerically the same in all three persons. This does not mean that the persons might not also have necessary causal relations to each other—for example, to satisfy the inseparability requirement. The aim is merely to show that it is possible for the persons to have non-relational individual properties too. And developing this line of thought would allow for the possession of distinct causal powers, and thus for the possibility that Christ's human nature is genuinely an instrument merely of the second person of the Trinity. Again, Scotus's rich philosophical theology has resources sufficient to develop views that he himself might have found objectionable.


On the account that I am describing, the second person of the Trinity extends himself to include a human substance. He does this in two ways, according to (p.204) the two different accounts that I highlighted in Scotus's discussions of instrumentality. The first—an active causal relation—explains Christ's consciously chosen human activity; the second—a passive causal relation or capacity—explains Christ's organic bodily functioning. I take it that the second of these could be developed to include the divine person's relation to the mental states of the human substance: not just that he has access to these, but that these mental states count in some sense as his. After all, on the Extended Mind kind of view that I am describing, the human mind really becomes a part of the second person of the Trinity, and its thoughts are his thoughts. Whether the model can give a fully worked-out account of the causal mechanisms relevant here, I am not sure. But this is a problem not merely for the view that I am describing, but for all two-substance views.


(1) Scotus, Ordinatio [= Ord.], nn. 15–16, in Scotus, Opera omnia, ed. C. Balić et al. (Vatican City: Vatican Press, 1950–), IX, 6–7.

(2) Hervaeus, In quatuor libros sententiarum commentaria [= In sent.] (Paris, 1647), 283b.

(3) See Aquinas, Summa theologiae [= ST] III/1, 111b-112a. For this kind of instrumentality account of the incarnation, see too Giles of Rome, Lectura 3.1 (in C. Luna, ‘La Reportatio della lettura di Egidio Romano sul Libro III delle Sentenze (Clm. 8005) e il problema dell'autenticita dell’ Ordinatio', in Documenti e studi sulla tradizione filosofica medievale, 1 (1990), 181): ‘A human nature separated from the divinity is an efficient cause, and constitutes [a suppositum]; but the human nature joined to the divinity does not constitute a suppositum, but is as it were an instrument of the divinity, grounded in its hypostasis, through the mediation of which [viz. the human nature] the God-man does everything.’ For the development of the model in the later Thomist tradition, see Nieden (1997: 160–75).

(4) Hervaeus, Quodlibeta 3.6 (Paris, 1513), fo. 76rb. The locus classicus is Scotus, Ord., nn. 46–7 (Vatican, IX, 20–1).

(5) Scotus, Ord., nn. 500–1 (Vatican, III, 296–7).

(6) On the relevant notion of extension, see e.g. Clark and Chalmers (1998: 7–19). Clark and Chalmers talk of the self extending itself to include, for example, external memory devices. Scotus's suggestion here uses the notion of an extended self in contexts other than merely cognitive use, and I need this suggestion for my argument here. I owe to Anna Marmodoro, with gratitude, the idea that it would be possible to use the Extended Mind theory to model the metaphysics of the incarnation.

(7) See Frank (1992).

(8) Scotus, Ord. 4.1.1.un., n. 119 (Vatican, XI, 43).

(9) Daniel 3: 24.

(10) Scotus, Ord. 4.49.13, n. 9, in Scotus, Opera omnia, ed. L. Wadding, 12 vols (Lyon, 1649), X, 587.

(11) Scotus, Ord., n. 496 (Vatican, III, 294).

(12) Scotus, Ord. 4.1.1.un., n. 120 (Vatican, XI, 44).

(13) Scotus, Ord. 4.1.1.un., nn. 121–2 (Vatican, XI, 44–5); see too Scotus, Ord. 4.1.1.un., n. 44 (Vatican, XI, 19); Ord., nn. 118, 122–3 (Vatican, XI, 331, 332).

(14) Scotus, Ord., nn. 120–1, 123 (Vatican, XI, 331–2).

(15) On this, see Cross (1998: 68–9).

(16) It is a problem in certain sorts of christology that there are some cases of parthood in which the properties of the parts piggyback onto the whole, and some cases in which they do not. I do not know a principle that governs these various cases. For a good introduction to the christological application of the notion that a whole can borrow properties from its parts, see Stump (2003: 412–15).

(17) I ignore questions about divine passibility and passivity in such cases, because it seems to me that any doctrine of the incarnation has to allow that God can be affected by creatures, is passible, and is passive in relation to various creaturely activities and attributes.

(18) Scotus, Ord. 3.17.un., n. 2 (Vatican, IX, 563).

(19) Scotus, Ord. 3.17.un., n. 16 (Vatican, IX, 569).

(20) Aquinas, ST III/1, 114b–115a.

(21) See, e.g., Scotus, Quaestiones super libros metaphysicorum 9.15, nn. 21–2, nn. 31–2, 34, in Scotus, Opera philosophica, ed. Girard J. Etzkorn et al., 5 vols (St Bonaventure, NY: St Bonaventure Press, 1995–2006), IV, 680–1, 683–4.

(22) On this, see Cross (2002: 198–203).

(23) Scotus, Ord. 3.12.un., n. 13 (Vatican, IX, 383).

(24) Scotus, Ord. 4.49.6, n. 11 (Wadding, X, 455).

(25) In fact, the view that the human substance is only contingently united to the second person of the Trinity is, as is well known, one of the key innovations of Scotus's christology: on this, see Cross (2002: 299–300, 302–8).

(26) Augustine, De trinitate [= De trin.] 1.4.7, ed. W. J. Mountain, CCSL 50/50A (Turnhout: Brepols, 1968), 36; see too e.g. Augustine, De trin. 2.2.3 (p. 83); De trin. 5.14.15 (p. 223), which last Scotus cites at Quodlibet [= Quod.] 8, n. 3 (Wadding, XII, 205).

(27) Scotus, Ord., n. 14 (Vatican, IX, 6).

(28) Scotus, Quod. 8, n. 6 (Wadding, XII, 205).

(29) Henry of Ghent, Summa quaestionum ordinariarum 53.6, 2 vols (Paris, 1520), II, fo. 68rH.

(30) Scotus, Ord. 1.26.un., n. 82 (Vatican, VI, 37).