P IO O H HL S P Y ELG OO Y

MARQUETTE

U N I V E R S I T YQ U A R T E R L Y

Volume 8 N d e r 3 Spring 1994

PHILOSOPHY & THEOLOGY Volume 8, Number 3 Spring 1994

Table of Contents ................................................................. 197

The Silence of Descartes John Conley S.J. .................................................................... 199 The Dialectic of Human Freedom: Schelling on Love and Evil Frank Schalow ....................................................................... 2 13 Sub Quadam Specie Aeternitatir: Signt)5cation et V a h r de Cette Expression Chez Spinozd Jaquet Chantal ....................................................................... 23 1 Amelm, Gaunilo, and Lost Iskand Keith Burgess-Jackson............................................................ 243 Faith as Tnrst and Beliefas Intellecfwl Credulity: A Response to William Sweet Hendrick Hart ....................................................................... 25 1 Faith, Beliefand Religiour Truth:A Rejoinder to Hart William Sweet ........................................................................ 257 Editor i Page Philip J. Rossi, S.J. ................................................................. 267

ANSELM, GAUUNO, LOST AND ISLAND

Keith Burgess-Jackson
University of Texas at Arlington
Abstract The received view is that Gaunilo's attempted refutation of AnseIm's ontological argument fails. But those who believe this do not agree as to why it fails. The aim ofthis essay is to show that whether the attempted refutation succeeds depends crucially on how one formulates the so-called greatmaking principle on which Anselm's argument rests. This principle has largely been ignored by contemporary philosophers, who have chosen to focus on other aspects of the argument. I sketch two analyses of metaphysical greatness and suggest that on one of them, which Anselm may have held, his argument avoids Gaunilo's criticism.

Did Gaunilo rehte Anselm's ontological argument by showing that analogous reasoning entails the existence of objects, such as Lost Island, that we know do not exist? Most contemporary philosophers of religion believe that Gaunilo did not refute the argument (exceptions to this generalization are Grim 1982 and Gregory 1984), but oddly there is no consensus as to why. The most commonly expressed view-which, incidentally, Anselm shared-is that God and Lost Island are disanalogousentities, that they differ in such a way that only the argument for God goes through. In his third reply to Gaunilo, Anselm wrote:
I can confidently say that if anyone discovers for me something existing either in fact or at least in thought, other than "that than which a greater cannot be conceived," and is able to apply the logic of my argument to it, I shall find that "Lost Island" for him and shall give it to him as something which he will never lose again (quoted in Hick & McGill 1967, 23).

Exactlywhy God and Lost Island are disanalogousis never made clear. Some commentators, including the early Anselmian Bonaventure, suggest that an island, unlike God, is a created thing, and hence

244

Burgess-Jackson:Anselm, Gaunilo, and Lost Island

ontologically imperfect (see Hick & McGill 1967, 24n12). Some maintain that islands are contingently existing objects while God is not (Hartshorne 1965, 19; Hick 1973, 17; Davis 1975,445). Some island, unlike the concept of a claim that the concept of a ~erfect perfect being, is incoherent (Wainwright 1988,35). Still others insist that whereas the nonexistence of Lost Island is conceivable, the nonexistence of God is inconceivable (Allen 1967, 273; Hick 1967, 539; Mackie 1982, 54). I want to challenge the received view that Gaunilo's refutation fails. I do this not by showing that the refutation succeed, as others (for example, Grim 1982 and Gregory 1984) have tried to do, but by showing that whether it succeeds depends crucially on how one characterizes the so-called greatmaking principle on which the argument rests. Those who believe that Gaunilo's refutation fails have not addressed this topic. My aim, in short, is to clarify the issue, not resolve it, although I do venture a thought here and there about what is at stake in the various resolutions of the issue and about what Anselm may have had in mind when he formulated his famous argument. Consider the following reconstruction of Anselm's argument (for an alternative reconstruction, see Back 1983, 189):
1. For all objects x, if x exists in the understanding alone and it is possible that x exists in reality, then x is an inferior thing of its kind. 2. God (the greatest conceivableobject) exists in the understanding alone. 3. It is possible that God exists in reality. Therefore, 4. God is an inferior thing of its kind (from 1, 2, and 3). Therefore, 5. The greatest conceivable object is an inferior thing of its kind (from 4 and the definition of "God").

Anselm resolves the evident absurdity by rejecting premise 2, from which he infers that God exists in reality as well as in the understanding. This, of course, was the aim of his argument. Now consider Gaunilo's refuting analogy. By replacing "God" with "Lost Island" (his name for the greatest conceivable island), Gaunilo deduces

Philosophy & Theology

5a. The greatest conceivable island is an inferior thing of its kind.

245

Logically speaking, 5a is no less absurd than the original (5), so it follows, by parity of reasoning, that Lost Island exists in reality. It would appear that Gaunilo has succeeded in his refutation, for nobody, includingAnselm, will accept this proposition. (Here I adopt the ad hominem version of refutation by logical analogy. According to this version, S refutes T's argument A if and only if S constructs a second argument, B, such that (1) T admits that Aand B have the same form, (2) T accepts the premises of B, and (3) T rejects the conclusion of B. Hence, T avoid the alleged refutation by either (1) denying that A and B have the same form, (2) rejecting one or more of B's premises, or (3) accepting B's conclusion.) Have we correctly stated the greatmaking principle? Suppose we change it to
la. For all objects x, if x exists in the understanding alone and it is possible that x exists in reality, then x is an inferior thing.

This premise, which omits the final three words of the original (I), generates the following conclusion:
5b. The greatest conceivable island is an inferior thing.

But this proposition, unlike 5a, is not absurd; indeed, it may be true! So Anselm is no longer committed to rejecting a premise of the argument, and hence is not committed to accepting the real existence of Lost Island. Gaunilo's refutation by logical analogy fails. (That is not to say that no other refutation or criticism succeeds; I say nothing, for example, about criticisms made by Kant). Anselm's reply to Gaunilo, as every student of the subject knows, was brusque and cryptic. As we saw, he claimed only that there is a relevant difference between God and Lost Island. Was this just rudeness toward an unexpected critic, or did Anselm believe that Gaunilo was being disingenuous?Anselm's brusqueness makes sense (I do not say "is justified") if he understood the greatmaking principle as formulated in la, for l a quite clearly does not generate an absurdity of the sort Gaunilo needed (and thought he had deduced). Perhaps

246

Burgess-Jackson: Ansefm, Gaunifo, and Lost Island

Gaunilo (mis?)understood Anselm's greatmaking principle as formulated in 1. If so, then whether Gaunilo refuted Anselm's argument does not depend, as is often claimed (by, e.g., Rowe 1993, 35), on whether the revised premise 3 ("It is possible that Lost Island exists in reality") is true-or, in keeping with the ad hominem interpretation It of the alleged refutation, whether Anselm accepts that depends, rather, on the nature and acceptability of the greatmaking principle. Since this essay is not meant to be exegetical, I make no claim about what Anselm actually had in mind. What I would like to do is clarify the difference between the two greatmaking principles I have distinguished and sketch the ontologies in which these principles are or might be embedded (for discussion ofAnselm's ontology and how it is related to those of Plato and Augustine, see Brecher 1974 and Back 1983). Let us assume, with Anselm, that existence in reality is a greatmaking property-that, other things being equal, real existence augments the greatness of a thing. There are two ways we might cash this out, one relativistic and the other absolutistic. The relativistic analysis maintains that, relative to a particular class of objects, the objects in that class that have real existence are greater, ontologically speaking, than those that lack real existence. Real horses are greater than imaginary horses; real coins are greater than imaginary coins; real unicorns are greater than imaginary unicorns. This is not to say that real coins are greater than imaginary unicorns, because that claim makes no sense on a relativistic analysis. Judgments of comparative greatness must be made within, not between, classes of objects. Of course, this does not preclude the relativist from saying that one type of object-say, a human being-is ontologically superior to another-say, a horse. This comparison is on the basis of something other than real existence (rationality, for example). Within the class of horses, those that are real are greater than those that are (merely) imaginary; but horses, qua horses, are inferior to human beings qua human beings in that they lack rationality. What we have is an ontological hierarchy that allows for both intra- and interclass comparisons. The absolutistic analysis of greatness, as the name implies, does not make greatness relative to a class. To say that X is greater than Y,

Philosophy & Theology

247

in this way of understanding, does not presuppose that X and Y are members of the same class of objects. It says that whatever classes X happens to be a member of, it is greater than Y, whatever classes Y happens to be a member of. To say that real existence is a greatmaking property is just to say that anything that really exists is ontologically superior to anything that does not exist. Any existing thing is superior to any nonexisting thing. An existing rock or worm is superior to a nonexisting (merely imaginary) human being. Again we need to be cautious. The absolutistic analysis does not prevent comparisons from being made within each class. Within the class of really existing objects, for example, God might be ranked highest, followed by human beings, horses, and islands. The same superiority ranking might obtain within the class of imaginary objects. The fundamental divide, in this case, is between real and nonreal (that is, merely imaginary) existence. O n each side of that divide further gradations are or might be imposed. Here are sample hierarchies: Absolutistic analysis of greatness Existing God Existing human Existing horse Existing island Nonexisting God Nonexisting human Nonexisting horse Nonexisting island Relativistic analysis of greatness Existing God Nonexisting God Existing human Nonexisting human Existing horse Nonexisting horse Existing island Nonexisting island

It may be (and here, admittedly, I speculate) that Anselm had an absolutistic conception of greatness but that Gaunilo thought he (Anselm) had a relativistic conception. In other words, Gaunilo may have assumed that comparisons on the basis ofgreatness must be made

248

Burgess-Jackson:Amelm, Gaunilo, and Lost Island

within classes, which would indeed make his Lost Island argument go through, for a real island, in Anselmian terms, is superior to a merely imaginary island. But Anselm could respond that a real island is still ontologically inferior to other things-to a real God, to a real human, to a real horse, and so on (see Brecher 1974, 104; Devine 1975,257; Davis 1975, 438; and Downey 1986, 42 for a similar conclusion, although reached on other grounds). The virtue of this approach as compared to others is that it makes sense of Anselm's quick dismissal of Gaunilo's alleged refutation. Anselm may have been puzzled that Gaunilo could not see what he saw, namely, that his argument works only with the greatest conceivable thing, not the greatest conceivable human, horse, or island-and there is, logically, only one such thing (on this point see Brecher 1974, 100). For the reasons I have given, the modern-dayAnselmian cannot embrace the version ofthe greatmaking principle expressed by premise 1. But then, Anselm may not have wanted to. The version expressed by premise 1a, which presupposes an absolutistic analysis of greatness, allows Anselm's argument to go through but cuts off analogous arguments such as that ofGaunilo. Whether the underlying ontology is acceptable and whether this version of the greatmaking principle can withstand criticisms such as those lodged by Kant are matters best left to others, or for another occasion. (I thank Steven Shepherd for making me think hard(er) about Gaunilo's alleged refutation, which I originally thought successful, then came to believe was a failure. I now think that whether the refutation succeeds depends on how one construes the greatmaking principle. I also thank the editor and readers of Philosophy & Theology for helpful comments on an early version of the essay.) Works Cited Allen, R. E. 1967. Gaunilo. In The Encyclopedia ofPhilosophy 3:273. Edited by Paul Edwards. New York: Macmillan Publishing Company. Back, Allan. 1983. Anselm on Perfect Islands. Franciscan Studies 43:188-204. Brecher, R. 1974. 'Greatness' in Anselm's Ontological Argument. The Philosophical Quarterly 24:97- 105. Davis, Stephen T. 1975. Anselm and Gaunilo on the 'Lost Island'. The SouthernJournal ofPhilosophy 13:435-48.

Philosop/y 6 Theology

249

Devine, Philip E. 1975. The Perfect Island, the Devil, and Existent Unicorns. American Philosophical Quarterly 12: 25 5-60. Downey, James Patrick. 1986. A Primordial Reply to Modern Gaunilos. Religious Studies 22:4 1-49. Gregory, Donald R. 1984. On Behalf of the Second-Rate Philosopher: A Defense of the Gaunilo Strategy Against the Ontological Argument. Histoty ofPhilosophy Quarterly 1:49-60. Grim, Patrick. 1982. In Behalf of 'In Behalf of the Fool'. International journalfor Philosophy ofReligion 13:33-42. Hartshorne, Charles. 1965. Anselm i Discovety: A Re-&mination of the OntologicdlProoffor Godi Existence. LaSalle, IL: Open Court Publishing Company. Hick, John. 1967. Ontological Argument for the Existence of God. In The Encyclopedia ofPhilosophy 5:538-42. Edited by Paul Edwards. New York: Macmillan Publishing Company. , and McGill, Arthur C., eds. 1967. The Many-Faced Argument: Recent Studies on the OntologicalArgument for the Existence of God. New York: The Macmillan Company. . 1973. Philosophy of Religion. 2nd ed. Englewood Cliffs, NJ: Prentice-Hall. Mackie, J. L. 1982. The Miracle of Theism: Argumentsfor and Against the Existence of God. Oxford: Clarendon Press. Rowe, William L. 1993. Philosopb of Religion: An Introduction. 2nd ed. Belmont, CA: Wadsworth Publishing Company. Wainwright, William J. 1988. Philosophy of Religion. Belmont, CA: Wadsworth Publishing Company.