Can God Be Free
December 15, 2005 — 12:03

Author: Trent Dougherty  Category: Existence of God  Comments: 26

I just found out this week that we’ll be reading Rowe’s Can God Be Free? in Ed Wierenga‘s seminar next semester at Rochester. I haven’t read the book, but I’ve seen some references to the argument, and I’m puzzled none of those references mentioned what seemed an obvious solution: satisficing.
The essence of the problem is that in some sense "God must do the best thing" is true, but there’s no best world.  [Technically, there’s another horn of the dilemma: if there is a best world then God is not free not to create it. A. I don’t think there’s a best world ("couldn’t there be just one more dancing girl"); B. If there is, then I think Clarke’s moral necessity response is a good one.]
I suppose the structure of the argument must be something like this:

  1. If God exists, then he creates the best.
  2. If the best world is created, then there is a best world.
  3. There is not a best world.
  4. Therefore the best world is not created. 2,3 MT
  5. Therefore God does not exist. 1,4, MT

But it’s common in rationality theory to allow satisficing. There’s an old story about the gods having a bottle of wine that’s very good but only gets better with age. Question: When do they open it? Answer: It’s stupid not to ever drink it, so open it whenever. It seems to me that all that is required of God is that he satisfice. More below the fold.


According to an electronic search, the words "satisfice" and "stisfaction" do not appear in the book, but I gleaned this from a review by Tim O’Connor.

If an omniscient being creates a world when there is a better world that it could have created, then it is possible that there exists a being morally better than it. (112)

From what I can tell the defense of this thesis is based on the idea that if one being has higher standards than another then that being is better. Of course that’s false, but the real issue it seems to me is that most theists are going to say that it’s not a matter of "personal standards" but that there is an objective fact of the matter with regard to some world whether it is sufficiently good to license creation (I’ve tried to word this to allow for vagueness).
Now admittedly, I haven’t read the book, so there might be much more to it (I rush to add that I have the utmost respect for Rowe as a scholar and philosopher of religion due to his work on evil). But I think it’s often good to start with what seems to oneself to be the intuitive thrust of the argument. I look forward to reading the book, but perhaps your comments will help me prepare for that by commenting on the issues above. (And again, to be explicit, I’m not claiming to present Rowe’s argument, but rather my impression of it based on a few casual sources. However, I find those impressions worth considering, and if it turns out that my impression is not that far off, then I’ll be disappointed with the book.)

Comments:
  • For whatever it’s worth, I have a review of Rowe’s book forthcoming in Faith and Philosophy, so I’m pretty familiar with the argument you’re alluding to. Rowe denies (1),
    1. If God exists, then he creates the best.
    He denies (1). Since (by hypothesis) there is no best world, requiring God to create (or, actualize, if you like) the best is asking the impossible. It therefore cannot be a moral requirement. Not even a perfect being can do the impossible. To avoid this problem he instead assumes (1′),
    1′. If an essentially perfectly good being creates a world, then there is no better world He might have created instead.
    Then to (2′),
    2′. For any world that an essentially perfectly good being creates there is a better world that he might have created instead.
    This leads Rowe to the conclusion that,
    3′. An essentially perfectly good being does not create a world.
    Rowe then suggests that you couple (3′) with a premise suggested by H. Kretzmann (Dan Howard-Snyder makes a similar suggestion),
    4. An essentially perfectly good being must create a world.
    5. If an essentially perfectly good being exists then he both creates a world and does not create a world. !@#
    6.:. There exist no essentially perfectly good beings.
    The satificing objection does not strike me as an obvious solution. It is not at all unreasonable to hold that, other things equal, if being A performs morally better actions then being B, then B is less than essentially morally perfect. There is a being in the vicinity that is morally better than B. But there is obviously a lot to talk about here.
    The solution (I claim) is to deny that there is any contradiction forthcoming. True, an essentially perfectly good being is morally prohibited from actualizing any world in the infinite sequence,
    a. O~A0 & O~A1 & . . .& O~An
    True, he is also morally required to actualize some world or other in the sequence,
    b. O(A0 v A1 v . . . v An)
    In the context of infinitely improving worlds, it is impossible to generate a contradiction from these two claims. Further, (a) and (b) do come out true, but other formulations do not. For instance (a’) and (b’) are both false.
    a’. O(~A0 & ~A1 & . . . & ~An)
    b’. OA0 v OA1 v . . .v OAn
    Rowe fails to notice that the theorems necessary to generate a contradiction are invalid in contexts of infinitely improving worlds. What you have instead is a conistently describable situation in which a perfect being is facing a moral dilemma. Again for whatever it is worth, I present just this solution (obviously in much more detail) in Jon Kvanvig’s upcoming conference in philosophy of religion. If you’re interested, here’s a link to the paper, http://philofreligion.missouri.edu/papers2005.html

    December 15, 2005 — 16:24
  • jon kvanvig

    Hi Trent, here’s a version of the buridan’s ass paradox. Instead of two equally attractive bales, let there be a sequence of bales such that each successive one is better by the ass’s own lights to the previous one. Let the sequence be infinite, and let the ass be susceptible to dying of starvation for always aiming at the better one. The ass dies.
    Perfect analogue, I think, of an infinite sequence of better worlds, so long as it is better to create than not.

    December 15, 2005 — 18:30
  • Matthew Mullins

    Trent,
    This looks a lot like Rowe’s a priori argument for atheism on the no prime worlds hypothesis. Klaas Kraay has an article on the argument in the most recent issue of Faith and Philosophy. Kraay surveys four recent responses to the argument that he doesn’t think are successful, but notes that the argument rests on a couple of undefended premises that are conflated in the O’Connor quote. You can find the paper on Kraay’s personal site.

    December 15, 2005 — 23:44
  • David Hunter

    I’d be cautious of translating the staisficing concept as used in talk of rationality into moral talk there are pretty good reasons to doubt the acceptability of satisficing as a moral theory (See Tim Mulgan’s delightful paper “How Satisficers get away with murder”
    International Journal of Philosophical Studies, volume 9, 2001, pp. 41-46.) especially when applied to an omnibenevolent being. (IMHO if you are going to apply some form of consequentialism to God maximising utilitarianism seems to be the way to go.)
    In any case it depends on whether when you are applying what it in what I refer to a hard or soft variety of satisficing. The distinction is that a hard satisficer claims that one you reach the satisfactory level any more utility generate is worth nothing. A soft satisficer says its just no longer obligatory to generate. Obviously you need satificing to be of this first sort, but that is the least plausible version of satisficing.

    December 16, 2005 — 7:47
  • The wine example is John Pollock’s (see ‘How Do You Maximize Expectation Value?’ Nous Vol. 17, No. 3 (1982), sec. II on ‘EverBetter Wine’). Pollock first suggests a fascinating solution using a mixed-strategy that has an infinitely valuable payoff, but he concludes it is not clear what we are rationally obligated to do. But there is again no model of this problem in which a perfectly rational agent must fulfill all of his rational obligations: i.e., it is invalid that [](OA -> A) and so the principle cannot govern perfect beings in that context. We have instead a rational dilemma (cf. also Michael Slote, Beyond Optimizing, Harvard (1989) chp. 5, for a similar rational dilemma; I don’t have a citation, but I understand that Slote’s dilemma was first proposed by Tim Williamson).

    December 16, 2005 — 9:08
  • Mike, you write

    "It is not at all unreasonable to hold that, other things equal, if being A performs morally better actions then being B, then B is less than essentially morally perfect. There is a being in the vicinity that is morally better than B."

    I suppose it’s not unreasonable, but A. I don’t see that creating a better world is necessarily a better action, and B. if it were I don’t see any reason at this point to think the stated consequence follows. It just sounds like a confusion to me. If someone can make that sound better, fine, but it does not strike me as the least bit compelling. Also, it doesn’t seem like the problem is set up right anyway since A and B here have to be God Himself in different possible situations.
    I find your proposal very interesting, but I worry that it trades on an accident of provability rather than addressing the the heart of the matter. If I really thought there was anything to the premises, I don’t think I’d be terribly encouraged by (my current understanding of) your solution.  I’ll take a look at the paper.  Maybe one way to state my worry is that even if we can’t syntactically generate the contradiction, from the semantic level there seems to be an obvious tension.  Before I dive into the paper, does this objection make sense to you? If not, I’ll try to flesh it out, but if so, I’d like to hear what you have to say.

    December 17, 2005 — 21:26
  • Hey Jon, I really like the adaptation of Buridan’s ass. So it seems to me like the rational (and moral) thing for the ass to have done (assuming that the ass’s life contributed to someone’s well-being) would be to satisfice. Were you endorsing that or what? It seemed to me like a really good way to illustrate my original intuition.

    December 17, 2005 — 21:29
  • David,
    Thanks for the distinction–of which I was unaware.  However, I don’t see that I have to be committed to any very specific moral framework here. From what (little) I know of moral theory many theories are compatible with some contexts being essentially consequentialist.  This seems like one of those times, especially since if God does not create no other beings share in his goodness. The old Saint Joseph’s Baltimore Catechism’s third item is "Why did God make us?" "God made us to show forth His goodness and to share with us His everlasting happiness in heaven.
    For example, suppose I’ve got a fund which draws more interest every day. I can only cash in the fund all-or-nothing, no partial withdrawals.  I want to give the money to the poor, but every day I think "Well, tomorrow I’ll have more money to give to the poor if I just wait."  It seems to me the rational and moral (I’m not even committed to those being different notions ulitmately) act is to satisfice.  I think that leaves a lot wide open with respect to a more general moral theory.

    December 17, 2005 — 21:38
  • _I suppose it’s not unreasonable, but A. I don’t see that creating a better world is necessarily a better action, and B. if it were I don’t see any reason at this point to think the stated consequence follows. It just sounds like a confusion to me_.
    It would be nice to know where the “confusion” is. What’s confused, where? It’s s simple idea. Other things equal, if you harm someone you are thereby a worse person than if you benefit someone. Of course we can bring in a thousand moral points about intentions, motives, ability, responsiblity and so on. But these are not relevant here. This is why we have other things equal.
    I have no clue why you claim that actualizing a better world is not a better act.

    December 18, 2005 — 6:31
  • _I find your proposal very interesting, but I worry that it trades on an accident of provability rather than addressing the the heart of the matter. If I really thought there was anything to the premises, I don’t think I’d be terribly encouraged by your solution_.
    Most of this I don’t understand. I have no idea what “the heart of the matter” is supposed to be. I haven’t a clue what “trades on an accident of provability” is supposed to mean. I have no idea what the concluding conditional is supposed to convey.
    Any clear talk about this problem has to include a model for infinitely many improving worlds. Most of the alleged implications of the problem are based on a poor grasp of the relevant models (or,frankly, no grasp at all). The impressionistic appraisal of what those premises entail is obviously open to assessment. It is no “accident” that in models for infinitely improving worlds we cannot derive the alleged contradiction. Impressions are simply wrong on that score. It is no accident either that we can derive a consistent moral dilemma. Impressions are right that there is some moral tension here, and we can say just where it is. It is no accident that “perfection principles” are invalid in the relevant models. The impressions of some (Reichenbach comes to mind) were fairly accurate about that, but not so many others.
    Our understanding and appreciation of the Ontological Argument, the Cosmological Argument, and the Evidential Argument from Evil have benefitted immensely from the use logical modelling. That’s no accident either. Lots of confusions and mistakes about these arguments have happily been put to rest. The same benefits are forthcoming for the Argument from Improvability.

    December 18, 2005 — 8:41
  • Mike, my syntax/semantics comment was meant to suggest that we might be able to see a problem here without being able to formally derive a contradiction due to technicalities with infinite cases.  I’ll look at the mechanics of your paper though to see how you formalize the issue.
    But it seems to me that you switch the problem when you write

    "if you harm someone you are thereby a worse person than if you benefit someone"

    for your original comment was

    "if being A performs morally better actions then being B, then B is less than essentially morally perfect"

    and to perform a less good action is not necessarily to harm in any very substantive sense.  Obviously in the context of discussions of distributive justice and "harm principles" we sometimes talk like that, but even that is not implied by the original quote, for talk of satisficing makes no sense apart from a sufficiently good act, like in my Charity Case.  So I still don’t see any warrant for the conditional.  To get it, you’d need something like

    AP  ∀s(s is essentially morally perfect → ∀α((α is an act of s) → ~(∃s’∃β((βis an act of s’) & (β > α))))).

    I also don’t think that it follows that

    "There is a being in the vicinity that is morally better than B".

    You could get that if it were true that

    AB ∀s,s'(s performs a better act than s’ → s is a better being than s’)

    But why think that?

    December 18, 2005 — 19:48
  • Trent, you say,
    AP ∀s(s is essentially morally perfect → ∀α((α is an act of s) → ~(∃s’∃β((βis an act of s’) & (β > α))))).
    Yes, (AP) pretty much covers what defenders of this argument assert. The essentially morally perfect being is one that is “unsurpassable”. That is, there is no *possible* being that is better. I suppose I see no particular problem for (AP). If s and s’ are otherwise exactly alike, then what they decide to do does affect our assessment of them as persons. But I agree that the assessment of persons is a very complex matter, and there are no obvious and unique ways to do it. This is one reason why I don’t fuss this point: it looks like as good an answer to how we assess people as any other answer we might get. So we are not going to get much movement against the argument on this score.
    There are people who do fuss this point: William Hasker does and so do Tom Morris and Dan Howard-Snyder. But, as I said, they don’t get a lot of leverage on it. I don’t press (AB) for similar reasons. Sure you can make points on (AB) as well, but it doesn’t go very far at all (see the same authors above).
    About the models for the argument, they are not particularly controversial; or rather not any more than any modal argument is. Only a few assumptions are needed to generate the model, so it is hard to see a particular point of controversy. Still, maybe someone will try to reformulate the argument in a way that cannot be modelled in the way I suggest.

    December 19, 2005 — 14:01
  • It occurs to me that there’s another style of response to this sort of problem. I didn’t notice it in the comments, but I just skimmed through. Maybe someone mentioned it, and I missed it.
    This solution involves a notion of incommensurable levels of value, e.g. Mill’s statement that it’s better to be a human suffering than a pig immensely satisfied. You could make one pig better and better off, and it would indeed make for a better world, but adding one more dissatisfied person makes the world so much better that in comparison changing the pig is like a drop in the bucket. The Stoics say something similar about the value of the virtues in comparison with the value of what they call the indifferents. Some of the Roman Stoics, at least, say the indifferents as not being of no intrinsic value, but the value is so insignificant that we can’t really compare it to what’s of true value.
    What would be crucial for this solution to work is that there be something that could be of such value that if it’s present all other things are truly better if they serve that end even if there are intrinsic goods and bads on the lower level that would, apart from other considerations, make the world better or worse. But making something intrinsically worse on the lower level might make the world better in terms of the higher consideration. I think this is exactly what Leibniz had in mind when he said that this is the best possible world. Though Malebranche was critical of that expression, I think his view amounts to the same thing. I see similar sorts of claims in Augustine and Aquinas, with some hesitation from Aquinas sometimes to endorse the expression Leibniz would later use for it (because he thinks the world could always be made better in some sense but not in an important enough sense).
    The trick here would then be that any increase in the ways that really would in themselves make the world better would have to make it worse in some larger way. It’s not a matter of pain and pleasure outweighing each other but a matter of one kind of good being truly more valuable in an incommensurable way than any intrinsic good or bad on the lower level. If this could be worked out so that the vagueness issues only come up at the lower levels (but could appear on any lower level than the highest level), I think the classical view can be maintained. I have no argument that this is consistent, but I would maintain that it seems possible. I don’t think there’s a good argument against it.

    January 10, 2006 — 8:31
  • I guess I should also say that Leibniz didn’t think he had proved empirically that this is the best of all possible worlds. He just thought that he could show that it had to be given what else he thought about God. That’s what I would want to say about this as well.

    January 10, 2006 — 8:33
  • According to Norman Kretzmann–pretty fair Aquinas scholar–Aquinas held that there were infinitely many improving worlds. Nevertheless Aquinas thought (Kretzmann too) that God could actualize less-than-the-best world since actualizing the best world is a logical impossiblity. God could not be required to do that.
    But that aside, do you mean by the “classical view” the position that at least the following is true: if God exists, there must be a best possible world?

    January 10, 2006 — 13:21
  • By ‘classical view’, I simply meant the view I was attributing to Augustine, Aquinas, Malebranche, and Leibniz. It’s not merely that if God exists then there’s a best possible world. It’s the thesis that this world is the best possible world in some sense. As you pointed out, Aquinas denies this in one sense, and I already mentioned that Malebranche denies in it terms of this way of describing it, but I think all four would admit to thinking of this as the best possible world in some sense.
    Aquinas does have this passage that talks about how each world could be a little better and so on. That must be what Kretzmann is getting at. There’s another place where Aquinas asserts a view more like what I’ve been saying here. I think the way to put the two together is to admit that in some sense all those other possible worlds are better. In terms of the lower level of value, they are better. But the whole of this world taken all together produces some goods that are on a level incomparable with the lower level of goods, the ones we can easily count. He treats the doctrine of God’s providence as requiring that when all things are considered there’s nothing God could have done that would overall have been better. So I think Kretzmann is guilty of selective use of evidence, as is typical of many Aquinas scholars in other areas (e.g. free will).

    January 12, 2006 — 7:31
  • If it takes such a baroque theory of value to justify the conclusion there is a best possible world, then not many will be persuaded to that conclusion. It is just much more intuitive and plausible that worlds are infinitely improvable. Further, it presents no interesting objection to the existence of a perfect being.
    I guess, too, I’d hesitate before attributing a frankly sophomoric regard for evidence to a scholar and philosopher of N. Kretzmann’s stature. His Cambridge University Press volumes on Aquinas are regarded as first-rate by scholars who know Aquinas very well. Eleonore Stump, to give one example.

    January 12, 2006 — 13:19
  • It’s not a charge of sophomoric use of evidence. It’s what Anthony Kenny says virtually all Aquinas scholars do almost all the time. His body of work is so huge that it’s easy to ignore part of it. This is particularly exaggerated, according to Kenny, with the issue of the will (and in particular whether we classify Aquinas as what we now call a libertarian or what we now call a compatibilist).
    As it happens, those who fall into each camp are primarily using one set of texts, and those who fall into the other are primarily using a different set of texts. When he speaks of providence in the SCG, it sounds awfully compatibilist. When he speaks of the will in ST, it sounds awfully libertarian. When he speaks of providence in the SCG, you get the sense that this is the best world there could be. When he speaks of omnipotence in the ST, you see that he thinks that in one sense things could always be better.
    Pointing this out is not a slam on Kretzmann’s scholarship. It’s a disagreement that the view he finds in Aquinas is an absolute view. It must be placed in tension with a view that seems opposite, and ideally we should try to see if both views can be simultaneously held.
    I have the Kretzmann books you are referring to. They are good for what they are, which is a selective study of two books of one work of Aquinas, with little attention paid to other parts of his work. I’ve found the first volume in particular very helpful for working out the details of some of his different arguments for the existence of God.
    What I said about Kretzmann’s statement, by way of reminder, is that it is right as far as it goes. There are, however, other passages in Aquinas that seem to require that God cannot do better than he did in terms of how the world was made. The view I’m explaining tries to fit those two seemingly contradictory statements together. I don’t care enough about whether I’m right about Aquinas to find the passage I’m thinking of. It’s one that I’ve used in my teaching, but I’d have to look through pages of notes probably, and I have other things to do. My point was more that this view makes sense of a lot of very important philosophers’ views.
    I’m not sure why you call this a baroque theory of value. It’s actually fairly common. Mill is probably its best defender in his famous pig passage. Kant holds a view that takes this form when he distinguishes between moral value and other things that he considers not to have any moral value but still have value. Plato and the Stoics hold something like this about virtue as opposed to whatever else someone might think would count as valuable in some other sense. It’s fairly common to think one set of factors is of ultimate value, and other considerations can have value of a certain sort but not enough to change the value of things overall when the higher-level factor is in play.

    January 12, 2006 — 18:25
  • “It’s what Anthony Kenny says virtually all Aquinas scholars do almost all the time. His body of work is so huge that it’s easy to ignore part of it.”
    I never said that Kretzmann did not ignore part of Aquinas work. What I deny is the Kretzmann ignored a part of Aquinas’s work that is relevant to the question of the ordering of worlds. Ignoring relevant evidence–especially for an historian of this caliber and on a topic he is well-known for–is an utterly sophomoric error. The excuse that there’s so much material in Aquinas amounts to the plea, “but, gosh, you didn’t actually expect me to be sure I reviewed all of the relevant material before publishing to the entire academic community my conclusion on what Aquinas’s ultimate position was, did you?”
    Maybe lots of mediocre commentators have done this (frankly, these must be very mediocre to have made such an embarrassing and totally unprofessional mistake) but this is certainly no excuse Kretzmann would make. I have no idea who Kenny is referring to, but it is certainly not to the best people in the field.
    The claim that Kretzmann’s view is true “as far as it goes” is just damning with feint praise. ‘2 + 3 = 4’ is also true as far as it goes. All you forgot to do was add one more. Otherwise, it’s true.

    January 12, 2006 — 19:56
  • Kenny was talking about how the best Aquinas scholars are divided in their general approach and in the texts they tend to favor on certain issues. When Aquinas is either inconsistent or dealing with different aspects of an issue in different places, it’s not surprising that scholars will tend to favor one over the other if they happen to think that one of the two places is more definitive of his overall thought than the other. There’s nothing sophomoric about that.
    This is from his “Stump’s Aquinas” in Philosophical Quarterly” vol.54, issue 216, pp.457-462. It’s 2004, I believe July. Here is a quote:

    [He lists books by Aidan Nichols, Eleonore Stump, Robert Pasnau, and John Finnis.] The striking thing about these four books is that the Aquinas who emerges from each of them is so different from the Aquinas in any of the others… [He says this is ironic in the case of Stump and Pasnau, because both were Kretzmann students.] … The possibility of very divergent interpretations is inherent in the nature of Aquinas’ Nachlass. The Saint’s output was vast, well over eight million words, and most of us nowadays are able to read Latin only at about the same pace he wrote it. Any modern study of his work, therefore, is bound to concentrate only on a small portion of the surviving corpus. Even if one concentrates, as scholars commonly do, on one or other of the great Summae, the interpretation of any portion of these works will depend in part on which of many passages in other works one chooses to cast light on the text under study. Especially now that the whole corpus is searchable by computer, there is great scope for selectivity here.


    He goes on to talk about translation problems, where many scholars assume Aquinas is using words where there are English cognates in the way people today use them in philosophy, which distorts Aquinas. He says some scholars try to make Aquinas fit with contemporary Catholic doctrine. Others try to make him more like contemporary philosophers or scientists. Stump, for instance, seeks to make Aquinas more like Plato than Aristotle, while others do the opposite.
    Most of Kenny’s review is very harsh on Stump. He doesn’t recognize her Aquinas. But you’ve listed her as a first-rate scholar. If that description is accurate, then he’s certainly including the first-rate scholars in his list. I don’t know Nichols (which says nothing, really), but Finnis and Pasnau are first-rate from everything I’ve heard. Kenny concludes that “Aquinas was an intellectual giant, and those of us who try to interpret him to a twenty-first century audience are like Lilliputians trying to tie him down with our conceptual netting. It is not to be wondered at if we all have different perspectives on this gigantic figure.” So he includes himself too, and he’s unquestionable on the top tier of Aquinas scholars.
    There’s no sense in which 2+3=4. There is a sense in which Aquinas believes this world to be less good than it could have been.

    January 12, 2006 — 20:36
  • Sorry. That blockquote should end after the first paragraph, and this isn’t my post so I can’t fix it.
    Here is Aquinas on the perfection of the world from SCG 3.71:

    Perfect goodness could not be in creation if there were not found an order of goodness among creatures, some being better than others: or else all possible grades of goodness would not be filled up; nor would any creature be like God in having pre-eminence over another. The perfection of the universe requires both grades of goodness. But it is the care of a ruler to uphold perfection in the subjects of his government, not to make it less. Therefore it is no part of divine providence wholly to exclude from creation the capability of falling away from good. But upon this capability evil ensues: for what is capable of falling away, sometimes does fall away; and the mere lack of good is evil

    He talks as if the creation is perfect in some sense because it has every grade of good and evil. That suggests something like the view I’ve been explaining. From the same section:

    The good of the whole takes precedence of the good of the part. It belongs then to a prudent ruler to neglect some defect of goodness in the part for the increase of goodness in the whole, as an architect buries the foundation under the earth for the strengthening of the whole house. But if evil were removed from certain portions of the universe, much perfection would be lost to the universe, the beauty of which consists in the orderly blending of things good and evil

    This could be interpreted as a simple outweighing of good with the good so much more because of the evil, but I think higher-order goods makes more sense of it.
    But my whole point was that this obviously can’t mean that the world is perfect in any sense, or there would be no way to describe it as being able to be improved in various aspects (which he would then say take away from the good of the whole in some sense). So what Kretzmann said is right, but Aquinas says more elsewhere from a different perspective. Here is another passage from SCG 3.94:

    The difference between the two is this, that in the pre-arrangement of order the providence is more perfect, the further the arrangement can be extended even to the least details: there would be not many parts of prudence in him who was competent only to arrange generalities: but in the carrying of the order out into effect the providence of the ruler is marked by greater dignity and completeness the more general it is, and the more numerous the subordinate functionaries through whom he fulfils his design, for the very marshalling of those functionaries makes a great part of the foreseen arrangement. Divine providence, therefore, being absolutely perfect, arranges all things by the eternal forethought of its wisdom, down to the smallest details, no matter how trifling they appear.

    Providence is absolutely perfect. There’s no way God could have done worse in creating, and I think he would also mean there’s no way God could have done better. The more you press into the details, the more you see perfection in God’s design. He’s speaking as if the divine plan is as perfect as it can be. That seems to run contrary to the omnipotence passages that I’m sure Kretzmann is talking about. Here is another from SCG 3.97:

    God by His providence directs all things to the end of the divine goodness, not that anything accrues as an addition to His goodness by the things that He makes, but His aim is the impression of the likeness of His goodness so far as possible on creation. But inasmuch as every created substance must fall short of the perfection of the divine goodness, it was needful to have diversity in things for the more perfect communication of the divine goodness, that what cannot perfectly be represented by one created exemplar, might be represented by divers such exemplars in divers ways in a more perfect manner.

    It sure sounds like the universe is as good as it could possibly be.
    I wrote all this before going back to the omnipotence passage in ST. Now I think Kretzmann is just wrong. In that passages he says that any individual thing could not be better than it is, but of any individual thing there could always be a better one. Kretzmann probably extends this to the world. I’d have to get my book out, and it’s in the baby’s room right now so I can’t, but that’s the only thing I can think of that he might do. That would misunderstand what a thing is for Aquinas, though. The world isn’t a thing but a collection of things, a heap. It’s not a substance. So it’s no surprise that in the reply to the third objection in the omnipotence section he says:

    The universe, the present creation being supposed, cannot be better, on account of the most beautiful order given to things by God; in which the good of the universe consists. For if any one thing were bettered, the proportion of order would be destroyed; as if one string were stretched more than it ought to be, the melody of the harp would be destroyed. Yet God could make other things, or add something to the present creation; and then there would be another and a better universe.

    This is exactly what I’ve been saying Aquinas’ preferred view was. I’ve been trying to fit what Kretzmann says into that because I remembered him saying what I just described about individuals, but I think Kretzmann must be reading a modern notion of substances into Aquinas unless there’s some other passage I’m missing, and he’s certainly ignoring this reply to the third objection, which flatly contradicts Kretzmann’s reading of Aquinas. So I’m more convinced even than I was before that my reading of Aquinas is correct.

    January 12, 2006 — 20:57
  • This is what you’ve been saying?
    The universe, the present creation being supposed, cannot be better, on account of the most beautiful order given to things by God; in which the good of the universe consists. For if any one thing were bettered, the proportion of order would be destroyed; as if one string were stretched more than it ought to be, the melody of the harp would be destroyed. Yet God could make other things, or add something to the present creation; and then there would be another and a better universe.
    What he says here is that THIS particular universe we are living in could not be better in the sense that anything you might add to this universe that would make it better would make it a different universe.
    If THAT is what you mean by saying that this world could not be better, it makes no difference to the point I’m defending. I agree entirely with the metaphysics. Were something added to this world it would be a different possible world. No question. Were an improvement made, it would be a different possible world. But that is just what is meant in claiming that the actual world could have been much, much better. That term ‘the actual world’ in that sentence is not intended as a rigid designator. As a rigid designator it is true that the actual world could not be better. Used in an attributive sense, it is false: there are many other worlds that are much better and that might have been actual.

    January 13, 2006 — 6:39
  • I guess I was pretty tired when I wrote that last part. I don’t remember noticing that last sentence at all. So I return to my original thesis. Aquinas thinks that in one sense you could always have a better universe. It would be better in the sense that it has all the things in this universe plus something else that would count as more good. But I’m not convinced that his description of that as a better universe is absolute, given that it would destroy the perfect balance that we have in this universe. If you didn’t have that, then on a more fundamental level the universe with the more good in it would not be better after all. That’s what I think the first part of the ST quote says, and it’s what I see much more carefully drawn out in the SCG. I can see now how Kretzmann wants to put these things together by minimizing those other statements, but I think the overwhelming sense of those other statements gives reason to minimize the last sentence of the ST quote.

    January 13, 2006 — 7:29
  • This quote is the evidence you adduce for the claim that Aquinas scholars ignore/overlook evidence relevant to their attributions to Aquinas
    “[He lists books by Aidan Nichols, Eleonore Stump, Robert Pasnau, and John Finnis.] The striking thing about these four books is that the Aquinas who emerges from each of them is so different from the Aquinas in any of the others… [He says this is ironic in the case of Stump and Pasnau, because both were Kretzmann students.] … ”
    Where on earth is the irony? It’s nothing short of bizarre that Kenny (whose word you seem to take as nearly final) would assume that two scholars who studied with the same person should not have very divergent views. There is not enough space or time to list the counterexamples to this bizarre assumption.
    “The possibility of very divergent interpretations is inherent in the nature of Aquinas’ Nachlass.”
    Obviously. No one has ever denied it.
    “The Saint’s output was vast, well over eight million words, and most of us nowadays are able to read Latin only at about the same pace he wrote it.”
    Most of us? Most of who?? It is nothing short of ridiculous to claim that Kretzmann (and lots of other very talented medieval philosophy scholars) read Latin so incredibly slowly that, poor dolts, they just didn’t have time to finish their research.
    “Any modern study of his work, therefore, is bound to concentrate only on a small portion of the surviving corpus.”
    I’ve no doubt this is true; but it is painfully misleading. The fact that most scholars concentrate on a small portion of the corpus does not entail that any scholar interested in, say, the number of actualizable worlds (or any other particular topic), will ignore anything RELEVANT to their particular conclusions. Further, much of Aquinas’s endless corpus has nothing to do with any interesting and current philosophical worry. Much of it is pure theology. So, obviously, it will be ignored by most philosophers. So to conclude from the fact that the corpus is large and in Latin that philosophers will ignore what is RELEVANT to their conclusions is just a very poor argument.
    “Even if one concentrates, as scholars commonly do, on one or other of the great Summae, the interpretation of any portion of these works will depend in part on which of many passages in other works one chooses to cast light on the text under study. . .”
    This is at best a truism and at worst a reckless accusation. Am I to take this as accusing those other “scholars” (naturally, scholars no where near as thoughtful, careful, responsible, smart, intuitive and sensitive as Kenny) of narrowly focusing on those parts of one or the other summa that serve their interests and ignoring the rest? That assertion is just reckless and self-serving. It is also painful to read.
    “Especially now that the whole corpus is searchable by computer, there is great scope for selectivity here.”
    And this is presumably to reinforce the suggestion above that those other so-called scholars (i.e. viz., those presumably who are not identical with Kenny) have the great opportunity to present their deceptive research based on selective passages gleaned from web searches of Aquinas’s corpus.
    If this is the evidence you have that Normas Kretzmann (without question, one of the most highly regarded Aquinas scholars) is an ill-informed or poor or biased or deceptive Aquinas scholar, then it’s not much.

    January 15, 2006 — 21:06
  • Mike, I’m not continuing this conversation.

    January 15, 2006 — 21:54
  • Yes, I understand. I just don’t appreciate a superb philosopher getting such short shrift.

    January 16, 2006 — 5:00