God and the Island

By ALEX HATHAWAY

Upon purchasing a tin of tobacco from the market, Bertrand Russell began his routine trek back to the campus of Cambridge University. Suddenly, as if struck by Zeus’ bolt, he threw his hands into the air and exclaimed, “Great Scott, the ontological argument is sound!” (Pojman 2). This epiphany-like experience has not been uncommon among philosophers of both the classical and modern eras. Beginning with its original formulation by St. Anselm, the ontological argument for the existence of God has confounded philosophers for over nine centuries, and it continues to be a subject of profound debate.

The following are several integral reasons for its sustained salience in the field of philosophy. First, it is an analytic argument justified strictly by reason, rather than experience (a priori), making the preponderance of “arm-chair” philosophers content. Secondly, it is a deductive argument, meaning that if the truths of the premises are confirmed, the truth of the conclusion follows necessarily, making its potential power staggering. Thirdly, even if one is steadfast in one’s particular theistic disposition, the argument maintains its significance, in that it presents interesting philosophical questions unrelated to belief in God.

For these reasons, and undoubtedly many more, the ontological argument for the existence of God clearly warrants the ensuing analysis, a minute branch in the apparently timeless dialogue. In particular, the following few pages will endeavor to explain in detail the ontological argument’s two premises and resulting conclusion, while the latter few pages will scrutinize the first noteworthy challenge to the argument; Gaunilo’s “Perfect Island” essay.

The initial construction of the ontological argument is put forth in St. Anselm’s Proslogion, its foundation established here:

So even the fool must admit that something that which nothing greater can be thought exists at least in his understanding… And surely that than which a greater cannot be thought cannot exist only in the understanding. For if it exists only in the understanding it can be thought to exist in reality as well, which is greater… Therefore, there is no doubt that something than which a greater cannot be thought of exists both in the understanding and in reality (Pojman 4).

St. Anselm goes on to produce a Reductio ad absurdum version, but the former will be advantageous in its simplicity. The following is a conventional argument structure deduced from the simplistic version:

  1. Something than which nothing greater can be thought exists at least in the understanding.
  2. It is greater to exist in reality than to solely exist in the understanding.
  3. Thus, something which a greater cannot be thought of exists both in the understanding and reality.

Upon viewing this argument for the first time, one might believe premise (1) to be ambiguous and the meaning somewhat opaque. In order to clarify, it is generally accepted that the “Something than which nothing greater can be thought” is a definitional equivalent to God. This assumes that God has all maximally and objectively great properties such as omnipotence, omniscience, omnibenevolence, and omnipresence. These specific “omni” properties will be examined in detail subsequent to the further presentation and explanation of the argument.

A newcomer to the ontological argument also may initially be suspicious of premise (2). In order to elucidate the tyro, one may think of a situation in which there is a choice between two nearly identical diamonds. Diamond A has the following properties: flawlessness, iridescence, incalculable monetary worth, and existence in reality (entails existence in the understanding), while Diamond B has these properties: flawlessness, iridescence, incalculable monetary worth, and existence in the understanding. Thus, the sole distinction between these two diamonds is that Diamond A exists both in reality and in the understanding, while Diamond B exists only in the understanding. What follows is unquestionably the selection of Diamond A. However, the newcomer may not be satisfied with this example, for one wonders if the former result applies to a situation in which the choice is between two parking tickets. Ticket A has the following properties: given for unjust reasons, requires sixty dollar payment, and exists in reality, while Ticket B has these properties: given for unjust reasons, requires sixty dollar payment, and exists in the understanding. What follows, though, is unquestionably the choice of Ticket B, which might indicate that existence in reality is actually not greater than existence in the understanding. It seems, though, that this predicament can be remedied by just slightly altering premise (2)

2. It is greater to exist in reality than to solely exist in the understanding for objectively good things.

Now Ticket B can safely be chosen by the tyro without fear of undermining the argument.

While this elucidation might have aided the newcomer’s comprehension of premise (2), it has also brought forth from the shadows the complex problem (originally raised by Immanuel Kant) of whether existence of any sort can correctly be called a property. For instance, the meaningfulness of a sentence such as, “God is existent.”, is very questionable, in that the existence of God must be presupposed for the sentence to be coherent. One cannot declare that the sentence is either true or false; it is an empty predicate which does not have meaning. Perhaps then, existence cannot actually be used in the way Anselm uses it in the ontological argument. But, being such an immense difficulty and one which requires a thorough examination, this question cannot properly be analyzed in an essay of this length. Thus, existence as both a property and a predicate will be assumed through the duration of the essay.

Now that the newcomer has become fairly erudite concerning the ontological argument, the earliest major refutation of it must be examined. Immediately after the publishing of St. Anselm’s Proslogion, a monk named Gaunilo set forth what has been known as the “Perfect Island” essay (although it is not clear that he ever uses the word “perfect”). In this essay, Gaunilo attempts to show the absurdity of the ontological argument by overloading its possible conclusions. He asserts that one could “prove” the existence of any object “more excellent” than all other objects of its type, through this argument. This would overload and trivialize the argument by demonstrating that various absurd conclusions could be deduced soundly from it. Speaking sardonically of St. Anselm, Gaunilo states (subsequent to a description of a most excellent island):

But if this person went on to draw a conclusion, and say, “You cannot any longer doubt that this island, more excellent than all others on earth, truly exists in reality. For you do not doubt that this island exists in your understanding, and since it is more excellent to exist not merely in the understanding but also in reality this island must also exist in reality” (Pojman 5).

The following is a more conventional structure, mirroring the ontological argument’s construction, taken from the above passage:

  1. An island more excellent than any other islands exists at least in the understanding.
  2. It is more excellent (greater) to exist in reality than to solely exist in the understanding.
  3. The island more excellent than any other islands exists both in the understanding and reality.

It seems as if the newcomer’s questioning of the ambiguity of premise (1) in the ontological argument, applies equally to premise (1) of Gaunilo’s “Most Excellent Island” (MEI) argument. Thus, a discussion of what grouping of properties might constitute a MEI is now needed.

One inherent problem is that most, if not all, of the properties of a MEI are merely subjective. For instance, one person’s conception of a MEI might contain many coconut trees (because he likes the taste of coconut), while another’s conception might contain many banana trees and no coconut trees (because he likes the taste of bananas and is allergic to coconut). Again, a Nubian’s conception of a MEI might have a temperature property of 92 degrees Fahrenheit while an Inuit’s conception of a MEI might have a temperature property of 42 degrees Fahrenheit. Therefore, it seems impossible to conceive of an objective or universally accepted notion of a MEI.

The case is different, however, in regards to St. Anselm’s “greatest possible being”, which can be defined as a being that possesses all objectively good properties to their maximal extent. One clear example of this is knowledge (a type of objective good) which is obtained by God to its maximal extent through the property of Omniscience (will later be defined in detail). Other objective goods, though, are not so visibly expressed by the “omni” properties. But once probed deeply enough, the “omni” properties are actually found to covertly contain these goods. For instance, an objective good such as justness is obtained by God through the property of omnibenevolence, which presupposes justness to its maximal extent. Therefore, while Gaunilo’s MEI is subject to people’s relative notions of goodness, St. Anselm’s “greatest possible being” can be conceived objectively and through some degree of universal consent.

Laying aside this aforementioned impediment to Gaunilo’s conclusion in order to form yet another more concrete impediment, one may assume that there are, in fact, certain objective properties of a MEI. Maximal amounts of landmass, of foliage, and of sand may be considered objectively excellent properties of an island. However, Alvin Plantinga contends that “maximal” (greatest possible) properties do not make coherent sense in this circumstance. “The qualities that make for greatness in islands – number of palm trees, amount and quality of coconuts, for example – most of these qualities have no intrinsic maximums. So the idea of a greatest possible island is an inconsistent or incoherent idea” (Plantinga 91). He compares these properties to “a natural number than which it’s not possible that there be greater”, which he declares is an unmistakably incoherent concept. Thus, Plantinga concludes that even a conception of a MEI is not possible, which in turn mortally wounds Gaunilo’s attempted refutation of the ontological argument.

“But the wound has been inflicted on more than one victim”, the now astute student of the ontological argument contends, “St. Anselm’s argument undoubtedly has the same demise in regards to properties without intrinsic maximums”. To analyze this claim further, a list of God’s (greatest possible being’s) theoretical properties is necessary.

The four foundational properties that are generally agreed to by philosophers and theologians alike are the following: omnipotence, omniscience, omnibenvolence, and omnipresence. In defining omnipotence; the ability (or power) to either perform or not perform any possible action. In defining omniscience; having the knowledge of whether any possible proposition is either true or false. In defining omnibenevolence; for every action that is performed, it is performed through a purely good will. In defining omnipresence; being in every possible spatial relation to every actual object.

In order to demonstrate that these “omni” properties do have intrinsic maximums, one must imagine a possible world in which there are only a few possible actions. If a being inside this possible world has the ability to either perform or not perform any of the given actions, he would be omnipotent in relation to that world. Now in transferring this case to the much larger-scale actual world; for all the given possible actions, it is coherent to conceive of a being which could perform or not perform any of those actions. This being would be omnipotent.

As for omniscience, Plantinga asserts “And certainly knowledge, for example, does have an intrinsic maximum: If for every proposition p, a being B knows whether or not p is true, then B has a degree of knowledge that is utterly unsurpassable”(Plantinga 91). It is possible, therefore, that in the actual world there are (n) number of (p) propositions, and for every (p) a being knows whether (p) is either true false. This being would be omniscient.

In the case of omnibenvolence, one could surely conceive of a being which, for every action that is performed by the being (surely a finite number), that action is performed out of a wholly good will. Lastly, in regards to omnipresence, it is admitted that this property does not have as clear of an intrinsic maximum as do the other “omni” properties; however, it may be coherent to think of a being which is present at every relation in space (possibly a finite number) to every actual object. All of this is merely intended to show the weakness in premise (1) of Gaunilo’s argument while maintaining the viability of premise (1) in St. Anselm’s argument.

Hopefully, the student of the ontological argument will now have serious qualms as to the force of Gaunilo’s essay. Specifically, in that the properties which are obtained by a MEI are both subjective and lacking in intrinsic maximums (rendering the mere conception of a MEI impossible), as opposed to the properties obtained by Anselm’s “greatest possible being” which are both objective and intrinsically maximal (allowing for the conception of it to be both possible and coherent). But one final objection to Gaunilo’s argument seems valuable for putting at ease the student who may still sense the strength of the MEI.

As to their nature, there is a great divide between the concept of “greatest possible being” and the concept of a MEI. This is because the MEI is by definition a type of an island, which is an object inherently lacking in certain great-making properties. This creates a problem for Gaunilo’s argument while lending credence to Anselm’s argument because of the aforementioned addition to premise (2) of Anselm’s original argument, which states that, “It is greater to exist in reality than to solely exist in the understanding for objectively good things”. For it is more likely that an object with no lack of great-making qualities (God) is objectively good than it is for an object with inherent deficiencies to be objectively good. And an island, by its very nature, does not have: consciousness, a will, an ability to perform actions etc.

“But” interrupts the student, “perhaps the objective MEI does have the properties of consciousness, a will, an ability to perform actions etc.”. Very well, but as soon as a MEI is given consciousness, a will, ability to perform actions etc. it begins to look very similar to our conception of God. And at this point Gaunilo’s argument becomes mute, and the ontological argument returns unscathed by the MEI, meanwhile recommencing its effort in attempting to prove God’s existence.

The broad aim of this essay has been to examine the ontological argument’s intricacies and thereby strengthen its appearance, which may have seemed feeble at the outset. In refuting Gaunilo’s argument, various apparent deficiencies were found evident in the ontological argument, but through examination they were revealed to be mostly unsubstantiated. While this is true, there are still numerous other objections to St. Anselm’s ontological argument, and accordingly its dissection will in all probability continue for yet another nine centuries.

Works Cited

Alston, William P. “The Ontological Argument Revisited”. The Philosophical Review, Vol. 69, No. 4. (Oct., 1960), pp. 452-474

Oppy, Graham. “Ontological Arguments.” Philosophical Database. 8 Feb. 1996. Stanford Encyclopedia of Philosophy. 22 Feb. 2008 <http://plato.stanford.edu/entries/ontological-arguments/#TaxOntArg>.

Plantinga, Alvin. God, Freedom, and Evil. Grand Rapids, MI: William B. Eerdmans Publishing Company, 1974.

Pojman, Louis P., and Michael Rea. Philosophy and Religion. Ed. Worth Hawes. 5th ed. Belmont, CA: Thomson Wadsworth, 2003.


Alex Hathaway (’10) is a Philosophy major at University of Colorado.

Homepage art courtesy of beremski.

One thought on “God and the Island

  1. I understand that it is beyond the scope of this paper to deal with Kant’s objection as to whether existence is a property, but I find that there are some problems that could have been further addressed with regard to the “four foundational properties,” specifically omnipotence.

    Some could argue omnipotence is itself an incoherent concept. The concept of omnipotence needs to be further examined because if the claim is that God can do anything that means he can do the logically impossible, even contradictions. That mean God can both exist and not exist, make an option be traveling both up and down simultaneously, and have 2+2=34. If any grammatically correct sentence that “God can X” is the proposed definition of omnipotence, (including “God can do the logically impossible”) then one runs into the absurdity of there being nothing that is impossible…

    If one argues that God can do the logically possible the definition is still incoherent because what determines what is logically possible? If X is omnipotent because X can do whatever is logically possible for X to do, then it logically follows that a rock is omnipotent because the rock does whatever is logically possible for it do.

    One common dilemma raised against omnipotence is whether “Can God create a stone that he/she cannot lift.” If God cannot create this stone, he is not omnipotent; if God can lift the stone than he is not omnipotent either because there is something he cannot do, namely, create a stone he cannot lift.

    You’re discussion with omnipotence…
    “If a being inside this possible world has the ability to either perform or not perform any of the given actions, he would be omnipotent in relation to that world. Now in transferring this case to the much larger-scale actual world; for all the given possible actions, it is coherent to conceive of a being which could perform or not perform any of those actions. This being would be omnipotent.”

    could be improved in better dealing with what constitutes a “possible action”

Leave a Reply

Your email address will not be published. Required fields are marked *