Augustine, Anselm and Duns Scotus Revisited
Interview by Richard Marshall
'The Confessions is more than a philosophical work, but it’s no less than that, and a lot of the philosophical content gets blurred (to put it very, very nicely) by translators who aren’t sensitive to that dimension of the work. The arguments about being and goodness in Book 7, for example, are razor-sharp in the Latin, but philosophically flaccid in a lot of English translations. There’s an extended analysis of voluntas, will, in Book 8, but you’d never know it if you’re reading it in the widely used translation that renders voluntas variously as “urge,” “whim,” “impulse,” “inclination,” and “volition.”'
'The problem with treating Anselm as a classical theist is that you can end up ascribing to him the views classical theists affirm, or even the views you think classical theists ought to affirm, when in fact Anselm said nothing about the issue either way, or even said the opposite. It’s a surprisingly frequent thing in the literature.'
'Can God make a stone so heavy that he can’t lift it? If you say yes, there’s something God can’t do: he can’t lift the stone. If you say no, there’s something God can’t do: he can’t make the stone. So either way, there’s something God can’t do. Although Anselm doesn’t consider the paradox of the stone (as it’s called) specifically, he does have an answer. That God is omnipotent doesn’t mean that any sentence that starts out “God can” comes out true.'
'In addition to the paradox of the stone, there is the lesser-known paradox of the burrito, from The Simpsons: can Jesus Christ microwave a burrito so hot that he himself cannot eat it? My philosophy of religion students usually see quite quickly that there’s no real paradox here: according to orthodox Christology, Jesus is fully human, so of course the answer is yes.'
'You can’t talk to a medievalist in philosophy for very long without hearing something about angels, though, for the record, that business about “How many angels can dance on the head of a pin?” doesn’t derive from the Middle Ages, but from benighted early modern folks who didn’t realize that angels functioned for the medievals in much the same way that thought experiments function for philosophers nowadays: as heuristic devices for bringing out the essentials of a philosophical problem.'
'The doctrine of analogy is often presented in what I think is a dangerously garbled way by people who say that all our language about God is metaphorical. It should be the business of every Thomist not to let such nonsense be uttered without immediate rebuke. “A mighty fortress is our God” is metaphorical – God is not in any literal sense a building – but “God is good” is perfectly literal.'
'Scotus thinks ...our words must apply univocally – with exactly the same meaning – to God and creatures. Why? Because the only concepts we have are concepts that derive from our experience of creatures; if we can’t apply those concepts to God, we have no concepts at all that we can apply to God. And that means we can say nothing about God. We could perhaps say what God isnot, but as Scotus says (in one of those wonderful moments that remind you that he was, after all, a Franciscan), “We do not have supreme love for negations.”'
Thomas Williams is a philosopher with interests in medieval philosophy and theology (especially Augustine, Anselm, Aquinas, and Duns Scotus) and the philosophy of religion. Here he discusses his new translation of Augustine's Confessions, Anselm and exigetical issues, his theistic proofs, Gaunilo’s Lost Island, the paradox of the burrito, Anselm on freewill, beauty and awe, Duns Scotus and natural theology, univocity, his argument for the existence of God, Pelagianism, his ideas about natural law impact on morality and God and his views on freewill and where he parts company with Aquinas.
3:16: What made you become a philosopher?
Thomas Williams: As a rising high-school senior I participated in a summer enrichment program at Vanderbilt University, and a couple of the faculty were philosophers. Before that point all I had heard about philosophy was that it was dangerous – hadn’t Paul warned us against it explicitly in Colossians 2:8? – but I found it pretty interesting. Not very interesting, because it was epistemology, which has never done much for me, but pretty interesting. What hooked me, though, was a discussion at lunchtime with some of my fellow students and one of the philosophy faculty. We were debating one of the perennial questions about the relationship between religion and morality: are right actions right because God commands them, or does God command them because they are right? The philosopher encouraged me to read Plato’s Euthyphro, in which (famously) a version of that very question arises. As soon as I read it, I fell in love with philosophy. I went to Vanderbilt as an undergraduate intending to major in philosophy as preparation for law school, but I quickly realized that I enjoyed philosophy too much to do anything else.
3:16: You've just produced a new translation of Augustine's Confessions. So why do you think we needed another translation, what’s distinctive about yours and were there any surprises for you?
TW: My initial motivation for doing a translation of the Confessions, besides the fact that Hackett asked me to, was that I thought we needed a philosophically sensitive translation. The Confessions is more than a philosophical work, but it’s no less than that, and a lot of the philosophical content gets blurred (to put it very, very nicely) by translators who aren’t sensitive to that dimension of the work. The arguments about being and goodness in Book 7, for example, are razor-sharp in the Latin, but philosophically flaccid in a lot of English translations. There’s an extended analysis of voluntas, will, in Book 8, but you’d never know it if you’re reading it in the widely used translation that renders voluntas variously as “urge,” “whim,” “impulse,” “inclination,” and “volition.” This is not to say that one Latin word should always have the same English translation – voluntas, for example, is sometimes “meaning,” as in what someone wants to say – but a close conceptual analysis is really not the time to pull out the thesaurus.
I still think philosophical sensitivity by itself is a perfectly good justification for another translation of the Confessions, but over the course of the long gestation of the project another aim became just as central to my work, and arguably more distinctive of the translation as a whole. (Serious illness over the last three years had something to do with how slow the project was in coming to fruition. I won’t say anything about that except to note that my condition made me unusually sensitive to Augustine’s frequent use of metaphors of gastrointestinal distress, and I’m pretty sure mine is the only topical index to the Confessions that has an entry for “vomit.”) Augustine’s language in the Confessions is shot through with Scripture. But by and large Augustine doesn’t “quote” Scripture; he simply speaks the language of Scripture as his own language. The Scriptural bits shouldn’t be set off with italics or quotation marks, as if they somehow required to be spoken in your special BBC Evensong Voice. Augustine’s own language and the language of Scripture should interpenetrate seamlessly, so that the whole thing reads as one unified voice. What that means in practice is that Augustine should sound like Scripture, and Scripture in a somewhat elevated but not archaic register. So my Augustine sounds rather like the Revised Standard Version and the Psalter of the 1979 Book of Common Prayer.
One of the surprises for me was the realization of how relentlessly Jesus-y the work is (pervasively Christocentric, as I put it when I’m on my best behavior). I’ve been teaching the Confessions for years, but it was only when working through it in the slow, meticulous way that translation requires that I came to see that aspect of the Confessions clearly. I suspect that shows in the translation. (I’m pretty Jesus-y myself. Despite the fears of my teachers at the aptly named Evangelical Christian School that going off to study philosophy – and at a secular school, no less – would destroy my faith, it didn’t. I’m an Episcopal priest of the most tediously orthodox sort.) Certainly there are a lot of footnotes that bring out the Christocentric character of passages that might otherwise escape notice; I suppose I have as much Jesus in my footnotes as Chadwick has Plotinus in his. Another surprise was how much I came to love Book 13, which presents an extended allegorical reading of the creation story. Nobody pays any attention to Book 13; I could have filled those pages with cookie recipes and no one but the copyeditor would have noticed. In fact, Peter Constantine’s 2018 translation just quits in the middle of a paragraph with fifteen pages of Book 13 left to go, and I’d be willing to bet I’m the only person who has noticed it. But I’ve come to love Book 13, every breathless leap and kaleidoscopic shift in it.
3:16: You’re an expert in medieval philosophy and Anselm and Duns Scotus loom large in your work. So let’s start with the eleventh century’s top Christian philosopher St Anselm of Canterbury. One thing you show is that Anselm hasn’t been well served by exegesis and therefore many of the ways he is supposedly refuted are beside the point because they are refuting a position he didn’t actually hold. Can you say a little about this, why you think he was so unfortunates in this respect and perhaps give an example of where objections are not well taken? And given your approach to studying history of philosophy via Anselm, what is the current state of historiographical method in Anselm scholarship and what lessons can be learned from your work in this area about method?
TW: Anselm has had a couple of misfortunes, one in his own day and one in ours. In his own day he was the target of a brilliant and arresting counterexample to an argument that he never meant to give – more on that in a bit. In our day he has the misfortune of being read as the quintessential “classical theist.” Now it’s true that Anselm holds a lot of the views that get lumped together as a package under the heading of “classical theism.” He holds that God is absolutely perfect, unchanging, all-powerful, all-knowing, and perfectly good; God depends on nothing outside himself to be what he is (that’s the doctrine of divine aseity, from which follows the doctrine of divine simplicity, which states that God’s goodness, power, etc., are not features God has but are the very same thing as God – for if God’s power, or whatever, were in any way distinct from God, God would depend on something other than himself to be what he is). But the problem with treating Anselm as a classical theist is that you can end up ascribing to him the views classical theists affirm, or even the views you think classical theists ought to affirm, when in fact Anselm said nothing about the issue either way, or even said the opposite. It’s a surprisingly frequent thing in the literature.
As your question suggests, there’s an important point here about method in the history of philosophy. It’s the apparently trivial but really quite important observation that philosophers say what they say, and don’t say what they don’t say. Suppose a philosopher says A, B, and C. Suppose further that A, B, and C together entail D. What can you conclude about the philosopher’s views on D? The answer is Absolutely nothing. Perhaps the philosopher never thought or wrote about D one way or the other. Perhaps the philosopher did think about D but didn’t see the entailment and affirmed not-D. (Also, perhaps you’re wrong in thinking that A, B, and C entail D.) For example, maybe, given some of his other views, Anselm should have held that God knows what a rational creature chooses because the creature chooses it. But in fact he considers that question in only one place, and when he does, he says that of course it can’t be the case that God’s knowledge depends on creatures. (That would, rather obviously, violate divine aseity.) There’s some interesting historical and philosophical work to be done in figuring out how Anselm might have thought all the pieces of his view fit together, but the historical work of figuring out what his view was on this particular question is easy. He said what he said, and he didn’t say what he didn’t say.
3:16: Why do you reject the consensus reading of this argument? You examine Gaunilo’s Lost Island response on behalf of the fool to establish that the consensus view of what Anselm was doing here is wrong don’t you? Can you tell us about this.
TW: Philosophers love a good counterexample, and Gaunilo’s Lost Island argument is one of the most memorable and ingenious counterexamples ever. As Gaunilo read Anselm, Anselm had depended crucially on the premise that something is greater if it exists in reality than if it exists only in the mind; that than which a greater cannot be thought must therefore exist in reality, for if it existed only in the mind, it would not in fact be that than which a greater cannot be thought – an obvious contradiction. Gaunilo parodies that argument by using the same style of reasoning, and the same key premise, to derive a silly conclusion. Imagine, he says, an island so wonderful, so glorious, that no greater island can be conceived. We’ll call it the Lost Island. Now by your reasoning, Anselm, the Lost Island must exist in reality, because if it didn’t, we could conceive of an island greater than the Lost Island; after all, it is greater to exist in reality than to exist only in the mind. But of course there’s no Lost Island. And (this is me now, not Gaunilo) there is nothing special about islands; you could use the same form of argument to reach a host of ridiculous conclusions: that the Greatest Conceivable Cockroach exists in reality; that poor Linus Van Pelt has too little faith, not too much, because there is not only a Great Pumpkin, but a Pumpkin Than Which No Greater Can Be Thought; and on and on.
As I say, a memorable and ingenious counterexample. The only problem is that if we’re to believe Anselm, it’s a counterexample to an argument he never gave, employing a premise he never affirmed. Now maybe we shouldn’t believe Anselm. Nicholas Wolterstorff has a wonderful paper arguing that Anselm knew perfectly well that he had no adequate response to Gaunilo and instead blusters, dissembles, and talks around the issue. But I find that hard to square with the fact that Anselm himself wanted his reply to Gaunilo to circulate with his original text (the Proslogion) and Gaunilo’s “Reply on Behalf of the Fool.” To write a transparently non-responsive response to a devastating objection and then make sure that your bluster circulated with the unanswered objection would be quite a bizarre piece of performance art. And if you look at Anselm’s response with the idea that he, at least, thought it was a successful response, and that it clarified the original argument in important ways, you can very quickly see that Anselm and Gaunilo are just talking past each other. It’s not merely that Gaunilo hasn’t understood the argument Anselm thought he was making, but that Anselm doesn’t even quite cotton on to how Gaunilo was reading the original argument.
In his reply Anselm makes it clear that he thinks his original argument was this:
(1) If that than which a greater cannot be thought can be thought, it exists.
(2) That than which a greater cannot be thought can be thought.
Therefore, (3) that than which a greater cannot be thought exists.
(The language here is obviously a bit difficult. If it helps, one doesn’t lose much by reading the argument like this: If God is conceivable, God exists; God is conceivable; therefore, God exists.)
He sees Gaunilo as having objected to both (1) and (2), and he offers defenses of both. But (he asks) where did you get this idea that I defended (1) by invoking some premise about how it’s greater to exist in reality than only in the mind? I never said that. What I said is that there are characteristics of that than which a greater cannot be thought that cannot belong to something that exists only in the mind and not also in reality.
Anselm gives nine different versions of this defense of (1). Some of them are more successful than others. Consider this one:
'Something that exists without a beginning is greater than something that has a beginning of existence. Suppose that that than which a greater cannot be thought can exist, but doesn’t. That’s an incoherent assumption, because if it can exist but doesn’t, the only way it could exist would be to begin to exist; but if it began to exist, it wouldn’t be that than which a greater cannot be thought, because what has a beginning of existence is not as great as something that exists without a beginning.'
Not good. It’s a bit like arguing that I don’t have a possible but non-existent older brother because my mother is past her child-bearing years. But here’s one that’s pretty promising:
'Something that exists necessarily is greater than something that exists only contingently. Suppose that that than which a greater cannot be thought can exist, but doesn’t. That’s an incoherent assumption, because if whatever it is you’re thinking about did exist, it would exist only contingently; that is, it would be the kind of thing that could fail to exist (after all, you’re thinking right now that it does fail to exist). So that than which a greater cannot be thought is not the kind of thing that can fail to exist. If it can exist, it must exist.'
According to Anselm, this is the kind of argument he was giving in the Proslogion. (In fact, the fifth of the nine versions is quite clearly intended to be a restatement of the original Proslogion argument.) So why do people read the argument in the way Gaunilo did? One reason is that Gaunilo’s counterexample is so arresting that we naturally read Anselm’s argument in light of this exceedingly clever objection. But then why did Gaunilo read it – misread it – in that way in the first place?
The answer is that in the Proslogion Anselm tried a stylistic experiment that failed and, indeed, backfired. He tried to write a philosophical work in the style of his prayers. He went for stylistic grace at the expense of clarity, and at crucial moments in the Proslogion argument his fondness for elegant antitheses got the better of him. In his reply to Gaunilo he resumes his usual precision of expression, and we can see clearly what the argument of the Proslogion was supposed to be. Anselm would go on writing for another thirty years, but he would never again try to write philosophy in the style of his prayers.
3:16: It’s your view that Anselm wasn’t just out to prove God’s existence but also to tell us something about his attributes. You say that his ontological argument works as a sort of divine-attribute-generating machine. Can you explain how so, and what attributes did he think flowed from the concept of that which nothing greater can be thought, and are they consistent when some seem, on the face of it, contradictory, such as his justice and his mercy?
Anselm tells us in the Prologue to the Proslogion that when he looked over his previous work on the existence and nature of God, the Monologion, he noticed that it was “constructed out of a chaining together of many arguments”: one argument that there is a highest good, another that there is a highest being, yet another that this being has supreme power, and on and on. He became obsessed with the idea of finding a single argument that would do everything, and that is what the argument of the Proslogion is meant to be. God is a being so great that a being greater than God is inconceivable, so we know that God is whatever it is greater or better to be than not to be. It is greater to be unlimited in power than limited in power, greater to be unlimited in knowledge than limited in knowledge, greater to be independent of everything else than to be dependent on anything else, greater to have one’s existence complete in one all-encompassing eternal now than to have an existence in time that is constantly slipping away into nothingness, and on and on.
Now precisely because this form of argument spits out so many divine attributes (they’re not really many, as the doctrine of divine simplicity tells us, because each divine attribute just is God, but you know what I mean) you get apparent conflicts between them, as well as conceptual puzzles in figuring out how to characterize particular attributes. Omnipotence is a famously difficult one. Can God make a stone so heavy that he can’t lift it? If you say yes, there’s something God can’t do: he can’t lift the stone. If you say no, there’s something God can’t do: he can’t make the stone. So either way, there’s something God can’t do. Although Anselm doesn’t consider the paradox of the stone (as it’s called) specifically, he does have an answer. That God is omnipotent doesn’t mean that any sentence that starts out “God can” comes out true. God can’t lie, because he is perfectly good; but lying is a manifestation of weakness, not of power, so God’s “inability” is actually a consequence of his power, not a limitation on his power. And the “power” to make something that escapes one’s own control would likewise be, not a power at all, but a weakness; and there is no weakness in God.
In addition to the paradox of the stone, there is the lesser-known paradox of the burrito, from The Simpsons: can Jesus Christ microwave a burrito so hot that he himself cannot eat it? My philosophy of religion students usually see quite quickly that there’s no real paradox here: according to orthodox Christology, Jesus is fully human, so of course the answer is yes.
It is greater to be just than not to be just, so of course God is just. But what about mercy? Is it greater to be merciful than not to be merciful? Anselm argues that it is, and that in fact mercy is a manifestation of justice: not God’s justice to his creatures (obviously), but God’s justice to himself. It is just for God to be good not only in punishing the wicked but also in sparing them; he is, so to speak, doing the right thing by himself in being so good that he is good even to the wicked. What we can’t hope to know, Anselm says, is why God mercifully spares this wicked person and justly condemns that wicked person. The ontological argument gets you quite a lot, but it does have its limits.
Another limit (I’m not sure anyone ever talks about this) is that it doesn’t give you the Trinity. Anselm does affirm that that than which a greater cannot be thought is a Trinity, but he doesn’t suppose for a moment that you can somehow show that it is greater to be a Trinity than not to be a Trinity. In this way theProslogion argument falls short of the Monologion, in which Anselm actually argued for the Trinity on (purportedly) purely rational grounds. This is tricky – it’s not as if Anselm thought that Aristotle would have come up with the Trinity if only he hadn’t been such a slacker – but it’s at least clear that he’s willing to go much further than Augustine, who is emphatic that reason can’t establish the claim that God is a Trinity; that doctrine comes by revelation, and reason can only try to make some sense of it and show that it’s not incoherent. But whatever exactly we’re to make of Anselm’s arguments in the Monologion for the doctrine of the Trinity, it’s clear that he doesn’t think you can get there just by reflecting on the idea that God is that than which a greater cannot be thought.
3:16: What is Anselm’s position on freewill, which he defines as ‘the power to preserve rectitude of will for its own sake’? I guess to do so you need to tell us what he meant by ‘rectitude’ and explain what he meant by freedom here. Why do you say Anselm is not an early version of existentialism, a kind of medieval Sartre, whereby free choice affords creatures the possibility of creating their own characters?
TW: Rectitude in general is a matter of something’s being what it ought to be. A statement has rectitude when it’s true, because statements ought to be true. A hard substance has rectitude when it scratches a soft substance, because that’s what hard substances ought to do. A will has rectitude when it wills what it ought to will. For example, I ought to will to fill out my tax return honestly. But suppose I will that, but I do so because I’m afraid of what the IRS will do to me if they catch me cheating. What I will is correct, but why I will it is not; and every instance of willing, Anselm says, has both a what and a why. For me to be just, both the what and the why have to be right. That’s what “for its own sake” is doing in the definition of free choice. I am not just if I fill out my tax forms honestly out of fear of the IRS, or give money to a charity so that people will think I’m a great guy. I am just only if I do the right thing because it’s the right thing. Justice is rectitude of will preserved for its own sake, for the sake of rectitude itself: doing what’s right because it’s right. Why “preserved”? Because rectitude is initially a gift of God. We can’t acquire it by our own efforts, though we can easily throw it away by our own choice. (We can’t make ourselves just any more than we can make ourselves exist.) Now Anselm worries that if God gives us only a desire for justice, we won’t be able to choose what’s right because it’s right; we will choose what’s right because we’re programmed to do so – that’s the only option God has given us – and so we won’t be praiseworthy for choosing what’s right any more than a diamond is praiseworthy for scratching a soft surface. So God gives us two fundamental inclinations, an inclination to choose what’s right and an inclination to choose what we think will make us happy – and a will that isn’t determined by either of those inclinations, so that if we choose justice, it’s because we want to, not because we have to.
Yes, the way Anselm poses the issue might sound a bit archaic, and even more so when you realize he’s talking specifically about angels rather than human beings. (You can’t talk to a medievalist in philosophy for very long without hearing something about angels, though, for the record, that business about “How many angels can dance on the head of a pin?” doesn’t derive from the Middle Ages, but from benighted early modern folks who didn’t realize that angels functioned for the medievals in much the same way that thought experiments function for philosophers nowadays: as heuristic devices for bringing out the essentials of a philosophical problem.) But it’s easy to see that the problem itself is still a live one. Anselm is worried that if there’s nothing to our choices over and above the way God created us and the inclinations he gave us, we won’t really be agents at all, but inert conduits for divine agency. We might equally well worry that if there’s nothing to our choices over and above our heredity and upbringing and various deterministic physical processes, we won’t really be agents at all, but inert conduits through which other influences, ultimately beyond our control, do their work.
In any event, for Anselm there is a lot that’s just given. We don’t make ourselves the kinds of beings we are, and we don’t give ourselves our inclinations. God gives us an inclination for happiness (and, given the kinds of creatures we are, not just anything can make us happy, or even appear to make us happy), and God gives us an inclination for justice (which we can throw away on our own but can’t recover on our own). Anselm also seems to be pretty pessimistic about the prospects for developing virtue in the way that Aristotelians think about virtue, as firmly rooted dispositions of character, acquired by repeated acts, that allow us not only to act well but to do so reliably, readily, and with pleasure. The language of character, habituation, and virtue is almost entirely absent from Anselm. Our possession of justice is always precarious: one act of disobedience and justice is gone. Any increase in justice, whether in intensity (clinging more tenaciously to the good) or in extent (doing what’s right in some new domain, as when someone who is truthful also becomes chaste), will be a result of divine gift and not of our own efforts, just like the acquisition of justice in the first place. There’s just no room in Anselm for self-creation even of an Aristotelian kind, let alone of the existentialist variety.
3:16: You argue that Anselm’s interest in art of painting led him to articulate belief in terms of beauty and awe. Can you say something about Anselm on sacred beauty?
TW: Whether Anselm was particularly interested in painting in general, I couldn’t say, but he did apparently get rather worked up about paintings in which Jesus was portrayed as unattractive. In what I like to call the “Jesus was hot” passage in Cur Deus Homo 1.1, Anselm insists not merely that Jesus should not be portrayed as ugly, but that he was in fact exceedingly good-looking: “beautiful in appearance beyond the children of men,” as he quotes from the Psalms. Truth is beautiful, and incarnate Truth is embodied beauty. Theological truth shares in this beauty: it evokes awe, wonder, delight. But (pressed by his interlocutor, the delightfully named Boso) Anselm acknowledges that the beauty of theological truth works only on those who already acknowledge it as truth. Suppose you can tell the Christian story in a way that is extraordinarily beautiful: does that give me any reason to believe it? Not really. I don’t believe because I am moved; I am moved because I believe.
As a musician myself – I’m a decent pianist and a respectable singer (you’d be glad to have me in your choir, but you wouldn’t pay me) – I understand the skittishness about the role of beauty that Anselm derives in part from his general Platonist sensibilities and in part from his great debt to Augustine. Augustine worries about his enjoyment of church music: do I actually love Jesus, or am I just really, really into his soundtrack?
3:16: Duns Scotus came some time after Aquinas and is a natural theologian isn’t he? Can you say what a natural theologian is and how far he agreed with Aquinas and where he parts company regarding knowledge of God and revelation, in particular with his arguments for univocal predication against any doctrine of analogy?
TW: Natural theology is the effort to establish the existence of God and to understand the nature of God using natural reason alone – that is, using premises that are in the public domain, so to speak, rather than drawing premises from any purported supernatural revelation. Medieval Christian thinkers were generally quite favorably disposed to natural theology. The best of the pre-Christian philosophers had clearly reached some knowledge of God, they thought, and there is Scriptural warrant as well for thinking that one can derive knowledge of God by inference from the created world: “Ever since the creation of the world [God’s] invisible nature, namely, his eternal power and deity, has been clearly perceived in the things that have been made” (Romans 1:20, RSV – every natural theologian’s favorite Bible verse). Confidence in the power of reason, unassisted by revelation, to establish truths about God begins to wane a bit in the thirteenth century. Aquinas thinks reason can go quite far; a generation later, Scotus remains fairly confident but issues some caveats and qualifications around the edges of the enterprise; a generation after Scotus, Ockham is much less optimistic about the power of reason to establish anything much that’s religiously interesting about God.
On the issues of what reason can prove about God, Aquinas and Scotus are largely in agreement. Where they part company is in how they understand the way in which our language applies to God. For Aquinas, words that apply to both God and creatures can’t be used univocally (with exactly the same meaning), because creatures are effects that fall short of their cause. Creaturely goodness is not God’s goodness; creaturely being is not God’s being. But neither are such words used equivocally (with totally different meanings, as in “bank” for the sloping ground next to a river and “bank” for the financial institution), because creatures do resemble God. It’s not a linguistic accident that we say both “Socrates is wise” and “God is wise”: the wisdom of Socrates is a partial, limited, fragmentary imitation of the wisdom that (as Aquinas puts it) “pre-exists” in God as a total, infinite, unitary perfection.
So Aquinas says that words that apply to both God and creatures are used analogically, that is, with different but related meanings. I illustrate this for my medieval philosophy students by bringing in a picture of one of my nieces and saying, “This is my niece.” I use the expression “my niece” analogically of the photograph and of the girl. I clearly don’t mean exactly the same thing – my sister did not give birth to a photograph – but the two uses of the expression are related in obvious ways. The photograph resembles my niece; the expression “my niece” applies primarily to her, only secondarily to the picture. In the same way, creatures resemble God; expressions that designate perfections apply primarily to God, only secondarily to creatures, precisely in virtue of that relation of resemblance.
The doctrine of analogy is often presented in what I think is a dangerously garbled way by people who say that all our language about God is metaphorical. It should be the business of every Thomist not to let such nonsense be uttered without immediate rebuke. “A mighty fortress is our God” is metaphorical – God is not in any literal sense a building – but “God is good” is perfectly literal. Predicates that imply limitation or imperfection, such as “fortress,” “lion,” “Morning Star,” can apply only metaphorically to God; predicates that don’t imply limitation or imperfection, such as “good,” “wise,” “powerful,” apply literally, though analogically.
Now Scotus thinks this can’t be right. Our words must apply univocally – with exactly the same meaning – to God and creatures. Why? Because the only concepts we have are concepts that derive from our experience of creatures; if we can’t apply those concepts to God, we have no concepts at all that we can apply to God. And that means we can say nothing about God. We could perhaps say what God is not, but as Scotus says (in one of those wonderful moments that remind you that he was, after all, a Franciscan), “We do not have supreme love for negations.”
One of Scotus’s arguments for the doctrine of univocity, as it is called, is drawn from Anselm. Since God is a being than which no greater can be thought, we can – we must – attribute to God whatever it is in every respect better to be than not to be. Scotus calls such attributes “unqualified perfections.” An unqualified perfection is any perfection that does not imply limitation. So Scotus claims that unqualified perfections can be predicated of God. But he takes this a step further than Anselm. He says that they have to be predicated univocally of God; otherwise the whole business of unqualified perfections won’t make any sense.
Here’s the argument. If you’re going to use Anselm’s test, you first have to come up with your concept—say, of good. Then you check out the concept and see whether it is in every respect better to be good than not-good. You realize that it is, and so you predicate “good” of God. That test obviously won’t work unless it’s the same concept that you’re applying in both cases.
You can see this more clearly if you consider the two possibilities. You might say that the concept of the unqualified perfection only applies to creatures, and the concept we apply to God has to be something different; or you might try it the other way around and say that the concept of the unqualified perfection only applies to God, and the concept we apply to creatures has to be something different.
Take the first possibility. If you come up with the idea of an unqualified perfection from creatures and don’t apply the same concept to God, you’re saying that you can come up with something that is in every respect better to be than not to be, but that doesn’t apply to God. Such a view would destroy the idea that God is the greatest and most perfect being.
So then you might try the second possibility: it really applies only to God. Scotus points out that that can’t be right either, because then the perfection you apply to creatures won’t be the unqualified perfection any more, and so the creature wouldn’t be better off for having this pseudo-perfection. But the whole way in which you came up with the idea of the unqualified perfection in the first place was by considering perfections in creatures—in other words, by considering what features made creatures better in every respect. So this possibility gets the test backwards: it says that we have to start with knowing what features God has and thereby determine what is an unqualified perfection, but in fact we first figure out what the unqualified perfections are and thereby know what features God has.
What’s at stake in all this? For Scotus – and I am very much Team Scotus here – it’s the very intelligibility of theological language. Anything short of univocity means that in making utterances about God we are actually like the grown-ups in the Charlie Brown TV specials, making noises but conveying no meaning. Obviously if you’re not a theist, you might well find such a thesis plausible; but for Scotus, it’s a disastrous prospect.
3:16: You’re impressed by Scotus’s natural theological argument for the existence of God. Can you sketch it for us and say why you think it ‘one of the most outstanding contributions ever made to natural theology’?
TW: As I’ve said, the project of natural theology is to provide arguments for theological truths using non-theological premises – no fair relying on the contents of any purported revelation. (Once you bring in revelation, you’re doing revealed [or “dogmatic”] theology, which is not philosophy anymore.) So the premises have to be acceptable to everybody, at least in principle. Now who “everybody” is obviously depends on all kinds of historical contingencies. Scotus’s “everybody” is people who know their Aristotle inside-out and accept his basic stock of concepts, distinctions, and so forth. Obviously, then, to run Scotus’s argument in a contemporary context would require a lot of translation, if it’s even possible at all; we are not Scotus’s “everybody.” But I can at least state a version that gives the flavor.
In outline, Scotus first argues for the existence of a being who is first in efficient causality (“efficient cause” is Aristotelian jargon for what normal speakers of English just call “cause”), then for a being that is first in final causality (that is, a being that is the ultimate goal of all activity), and then for a being that is first in what he calls “preeminence” (in other words, a maximally excellent being). He then argues that a being that is first in any one of these ways will also be first in the other two. Scotus then argues that any being that enjoys this “triple primacy” will also be endowed with intellect and will, and then that any such being is infinite. Finally, he argues that there is only one such being.
That outline, of course, says only what Scotus argues for, not how he argues for it. I’ll look here only at the argument for a first efficient cause, which is quite elaborate, though its basic structure will be familiar to anyone who knows Aquinas’s Five Ways (the first two, in particular). We can think about two kinds of series of causes. In one kind of series, the later members of the series depend on the earlier members for their causal activity; in the other kind, they don’t. As an example of the first kind of series, consider the motion of my golf club. My golf club moves only because my arms move, and my arms move only because my shoulders move. If the movers higher up the causal chain stop moving, the movers lower down the chain stop moving as well. Call this an “essentially ordered series.” As an example of the second kind of series, consider my grandfather, who caused my father, who caused me. My grandfather doesn’t have to keep causing my father in order for my father to cause me. (And thank goodness for that, because my grandfather died when my father was five.) Call this an “accidentally ordered series.” An essentially ordered series can’t be infinite – the argument for this is quite complicated, but the upshot is that it makes no sense to posit an infinite series of dependent causes with nothing they ultimately depend on. You have to have something that has the causal power perfectly, intrinsically, and independently in order to make sense of things having causal power imperfectly, derivatively, and dependently.
So, fine, if there’s an essentially ordered series, you have to have a first agent. But why not just think we have an accidentally ordered series? My dad was caused by his dad, who was caused by his dad, who was caused by his dad, and so forth; at each step, you have a perfectly good explanation, so it seems as though an infinite regress is perfectly intelligible for an accidentally ordered series. And indeed Aristotle thought – as did Thomas Aquinas – that there was nothing unintelligible about an infinite accidentally ordered series. Scotus agrees. But, he says, any accidentally ordered series will depend on some essentially ordered series. For within the accidentally ordered series, each cause depends on a certain form in order to exercise its causal activity; and that form is not itself a member of the series. It is a higher-order cause, and so we’re moving up a notch in an essentially ordered series, which (as we now know) must “twist up to a first,” as James F. Ross and Todd Bates memorably put it.
3:16: How does his belief that we can know the truth without divine help give him resources to attack both scepticism and illuminationism?
TW: In Christian theology there’s a heresy called “Pelagianism,” which says that we can act rightly, or even attain salvation, by the exercise of our own natural powers, without any need for the divine help that goes (in that context) by the name “grace.” There was a tendency in Christian philosophy to apply a kind of anti-Pelagian reasoning not only to the moral domain but also to the epistemological domain, though the motivation was somewhat different. In the moral domain, the problem was supposed to be that our desires had been disordered and our free will compromised by the Fall. In the epistemological domain, the problem was not with any damage to our natures, but rather with our natures themselves. Our finite and fickle minds are just the wrong sorts of things to grasp with certainty the infinite and ever-fixed truth; we need divine help (which in this context is called “illumination”) not to restore our damaged moral nature but to make good the deficiencies that would limit our cognitive powers whether our nature were damaged or no.
Scotus thinks that if our natural powers were as deficient as the illuminationists say they are, we couldn’t attain certainty even with divine help. Henry of Ghent (whom Scotus seems to have read primarily in order to get annoyed and thereby inspired to write blistering corrections) had argued, among other things, that the soul is mutable, and so it needs something immutable to save it from wavering and uncertainty. That immutable something is an “uncreated exemplar” – an idea in the divine mind – that we cannot know by our natural powers, but only by divine illumination. Scotus objects that the act of understanding provided by divine illumination is still mutable, as everything in the human soul is mutable; so if the mutability of the soul is an obstacle to certainty, divine illumination leaves us no more certain than before. Henry’s arguments, far from showing how we can attain certainty, actually lead to a pervasive skepticism.
And skepticism is just false. There are all sorts of things we can be certain of: self-evident principles, the validity of proper syllogistic inference, the truth of conclusions derived from self-evident principles by means of proper syllogistic inference, many of our own internal acts, and certain propositions about present sense experience. All these things we can know with certainty, simply by employing our natural intellectual powers.
3:16: How do his ideas about natural law impact on morality and God? Presumably he would say that we obey God’s law because they are necessary rather than that God willed them, but then again not all moral laws are on the face of it necessary, which seems to suggest that some of God’s morality are arbitrary. Is this part of a voluntarist position he adopts, or is it more complicated?
TW: Well, people desperately want it to be more complicated than that, and there has been, and continues to be, no shortage of attempts to mitigate the apparent arbitrariness in Scotus’s account. But in my view all those attempts founder on Scotus’s clear, repeated, and unambiguous insistence that most of the moral law is contingent, that God freely determines the truth values of all contingent truths, and that there is no explanation for why God wills as he does with respect to contingent truths. Scotus says what he says, and he doesn’t say what he doesn’t say. He says that nothing about God’s rationality, God’s justice, or human nature explains why God wills as he does with respect to the moral law; he doesn’t say that God’s aesthetic sensibilities, or anything else, for that matter, limits or constrains God’s moral legislation. I get why people don’t like this view – it’s pretty bonkers – but Scotus seems to have been quite comfortable with it; whenever he expresses it, he does so in the most unqualified way available to him. He has plenty of opportunities to rein in his voluntarism, but he never takes them.
But that’s answering the last question first. Here’s the view. Some moral truths are necessary; others are contingent. “God is to be loved” is necessary, and God himself cannot make it false; given the very nature of God, he ought to be loved and not hated. (Ockham will later say that God could command us to hate God, in which case it would be right for us to hate God. By thinking in terms of the truth value of moral propositions rather than in terms of commands, Scotus avoids a number of problems that Ockham has to confront. But I’ll spare you my ritual invocation of Ralph Cudworth [1617-1688] on the problems of divine command theory and just note that for Scotus “God is to be hated” is necessarily false, and divine power extends only to what is possible.) But beyond “God is to be loved,” and the correlative negative truths like “No irreverence is to be done to God,” that’s it for necessary moral truths. Moral propositions like “Adultery is not to be committed” and “Parents are to be honored” – the moral truths that underwrite the parts of the Ten Commandments that don’t have to do directly with God – are contingent. If they’re true, they’re true because God made them true. And God needn’t have made them true. There’s nothing about God (his love, his justice, his artistry, anything) that requires him to will one way or another, or to will anything at all, concerning those contingent truths. If God had made it true that one ought always to walk around trees counterclockwise, then it would be the case that one ought always to walk around trees counterclockwise; and then it would be my joy and my delight to walk around trees counterclockwise, not because there’s anything intrinsically (or, for that matter, instrumentally) valuable about walking around trees counterclockwise, but because doing so would be a path that God had prescribed for us “if you wish to enter into life” – there I quote from Scotus, who in turn is quoting from Matthew 19.
3:16: How does his view of freewill differ from Aquinas’s? Is it because he thinks that there is a fundamental problem with ethics that aims at happiness? Does morality for Scotus have a role quite independent of human flourishing, and how un-mediaeval is this?
TW: Here you bring me into some of the trickiest and most highly contested bits of the debate, and I will try to say things that I think are true without getting hopelessly into the weeds. One way to get a handle on the differences between Aquinas and Scotus on freedom is to ask why it is, on their accounts, that we are free – that is, roughly, why we at least sometimes have alternative possibilities available to us – and other animals are not. Aquinas says it’s because of our cognitive complexity. We have intellect, which allows us to grasp universals, so that we can desire not just this or that particular good thing – this slice of cheesecake, this act of sexual intercourse – but more general goods, such as health, temperance, justice, integrity, and the like; and we have reason, by which we can evaluate particular goods as instances of those more general goods and as constituents of, or means toward, the most general good of all, which is happiness. Except for happiness itself, any particular good can be evaluated in various ways: contemplating taking a day off work and heading for the beach, I can regard that course of action as a pleasant diversion from the tedium of thinking about medieval moral psychology, even as a salutary rest needed to renew my energy for carrying out my philosophical vocation; I can also regard it as a shameful abandonment of responsibility or as a frivolous expense. If I form the first sort of judgment (and stick to it), I’ll head for the beach; if I form the second sort, I’ll head to the office. There’s no one way Ihave to think about taking a day off, and so there’s no one course of action I have to take. I’m free to do one thing or the other, and that freedom derives ultimately from the complexity of my thinking.
For Scotus, by contrast, freedom belongs to the will in its own right; freedom does not derive from the intellect. Regardless of what judgment I come to, regardless of how I weigh the pros and cons or subsume a proposed course of action under this general description or that one, I can choose to go to the beach, and I can choose to go the office, and there is nothing beyond the will’s own choice in the matter that explains why I do what I do. One doesn’t trace the will’s choice back to the intellect’s judgment, because the will could choose differently even if the intellect’s judgment were the same. The will has what Scotus calls “indeterminacy of superabundant sufficiency.” (The last time I used that expression at a conference, people actually laughed – and these were medievalists! – but that’s what Scotus calls it, and I’m so far gone in Scotus-land that it seems to me a perfectly sensible expression.) Scotus’s point is that usually, as Aristotelians, we recognize that when something is indeterminate, it stands in need of something else to determine it, as matter, which is indeterminate, requires determination by form. That’s because matter has insufficient actuality to determine itself. The will, by contrast, has so much actuality that it can determine itself, and in more than one way; it requires no determination from without. There has to be such a thing as indeterminacy of superabundant sufficiency, or else we couldn’t make sense of God’s power to do everything that is possible. God determines himself to do as he does, and our will likewise (though obviously on a lesser scale) determines itself to do as it does.
Now suppose, as some of Scotus’s predecessors had supposed, that whatever we choose, we choose for the sake of happiness. (Call that view “eudaimonism in moral psychology.”) Then the will would await the intellect’s judgment as to what, here and now, is conducive to happiness, and it could chose only that; the will’s choice would be determined by the activity of the intellect, rather than by its own superabundant sufficiency. Moreover, since (so Scotus thinks) the intellect works deterministically – in a given set of circumstances, with the existing inputs from our cognitive environment and our intellect’s habitual ways of working, there is only one way our deliberation can actually go – there would be no indeterminacy anywhere in the system, and so no freedom. Contrary to Aquinas, the greater complexity of our cognition as compared to the other animals wouldn’t make room for freedom; it would just make the deterministic process of choice a good deal fancier in our case than in theirs.
So that is his reason for rejecting eudaimonism in moral psychology. But there’s also eudaimonism in normative ethics: the view that the content of morality is specified by the character of happiness and the acts and dispositions that are necessary to attain happiness. Scotus rejects that eudaimonism too. One reason is that he just doesn’t think moral requirements are that tightly connected with happiness. Can we be happy if we commit adultery, fail to honor our father and mother, or take other people’s stuff? Sure. You simply can’t derive moral requirements from any good other than the ultimate good, God, and as we saw above, not much follows even from the goodness of God.
Yet happiness remains an important concept for Scotus. Though not entirely bound to will happiness, the will does generally will happiness, and no acquired disposition in the will can ever be stronger than its natural inclination toward happiness. And happiness is, in a perfectly conventional and recognizable sense, our ultimate end: perfect love of God for God’s own sake. I don’t think anyone has really sorted all the different eudaimonisms in Scotus, all the different roles that happiness (variously conceived) plays in his ethics, his moral psychology, and even his account of the human person. That’s one of the tasks I’ve set myself in the book I’m working on.
3:16: And finally, are there five books you could recommend to the readers here at 3:AM that will take us further into your philosophical world?
TW:
As unseemly as this is, I have to start with Augustine’s Confessions, because it has had such a profound influence on me and shaped the way I think about philosophy and theology; I come back to it again and again. But to move very quickly away from what cannot help but look like self-promotion (a vice Augustine recognized and hated in himself: see Confessions 10.36.59-39.64), I would suggest John M. Rist’s tremendously helpful book, Augustine: Ancient Thought Baptized (Cambridge University Press, 1994). I know of no better overview of Augustine’s thought.
On the question of free will, Peter van Inwagen’s An Essay on Free Will (Oxford University Press, 1986) has shaped the way I (and many others) think about the matter. To see how the Scholastics dealt with these questions and for understanding the historical developments that led to views like Scotus’s, Bonnie Kent’s Virtues of the Will: The Transformation of Ethics in the Late Thirteenth Century (Catholic University of America Press, 1995) is indispensable.
There’s been a strand in recent theology known as “Radical Orthodoxy” that casts Scotus, and in particular the doctrine of univocity, as an unmitigated disaster. With vastly greater charity and patience than I would ever be able to muster, Daniel P. Horan, O.F.M., pushes back against that anti-Scotus narrative in Postmodernity and Univocity: A Critical Account of Radical Orthodoxy and John Duns Scotus (Fortress Press, 2014) and exposes Radical Orthodoxy’s pervasive misrepresentations of Scotus.
In something like the same vein, it’s also worth mentioning that Paul DeHart offers a corrective to Radical Orthodoxy’s idiosyncratic reading of Thomas Aquinas in Aquinas and Radical Orthodoxy: A Critical Inquiry (Routledge, 2012).
ABOUT THE INTERVIEWER
Richard Marshall is biding his time.
Buy his new book here or his first book here to keep him biding!