Sunday, August 13, 2006

Paper: Reference and Religion

Reference and Religion

Introduction

The occurrence of the word ‘God’[1] in a sentence provides ample material for philosophical reflection. Anselm thought that he could prove God’s existence from a consideration of the meaning of the word ‘God’. Others have taken the meaning of ‘God’ to include contradictory properties, leading to a reductio argument against His existence. God tells us not to take His name in vain. But how are we to make sense of the way in which our use of the name ‘God’ refers to God? Is it because the name ‘God’ means something like, “the creator of the universe” or “the omnibenevolent,” and God is the creator of the universe or the omnibenevolent? If this is the case, which of these does the name really mean? Is it the case rather that the name directly refers to God without carrying any of these meanings? How, then, do we come to acquire the use of such a name and in what way does it refer to God? By situating these questions within the framework of 20th Century philosophy of language, we will find ourselves in a better position to critically discuss and answer them. In the course of this work we will come to formulate a coherent picture of the way in which the names of God directly refer to Him. This picture will then be subjected to rigorous scrutiny in an attempt to give it a clearer exposition. While formulating this picture, we will find it necessary to re-appraise some of the work we have considered. It will be necessary to challenge and amend certain aspects of the theories under consideration. It will be shown that while particularly relevant to our consideration of ‘God’, these problems must be considered by anyone working on questions regarding names.

We will begin by discussing one attempt to work out these questions from within the tradition of analytic philosophy. William Alston gives a theory of reference to God along the lines of Saul Kripke’s causal picture. While he does much to both formulate and address the topic and its implications, it will be shown that Alston’s account is ultimately unclear and insufficient. This will give us reason to demand a clearer and better developed formulation of what a causal picture of reference to God might look like.

Next we will come to subject this revised picture to scrutiny. We will consider three points of conflict relevant to any such formulation. The first problem to be dealt with will be our inability to allow that nothing like sense or descriptive connotation can be relevant in the case of some names, including the name ‘God’. We will consider one attempt to deal with this worry, but will find it misguided. Through a reformulation of the Kripkean picture, namely one that takes into account something like Millian connotation while maintaining our Kripkean intuitions about direct reference, we will find ourselves in a better position to accommodate this issue within our account. Following this, we will encounter the problem of vacuous names. Inasmuch as many have claimed that God does not exist, we will demand that our picture be able to make sense of names that do not refer to existing bearers. We will consider a few attempted solutions to this problem before finally settling on one that takes into account something like Meinongian objects. Finally we will come to consider our topic under the aspect of the causal chain and its continuity. We will come to entertain doubts as to the conditions necessary for guaranteeing continuity of reference along Kripkean causal chains. While it will be shown that we lack the resources to give an account of precisely how a chain might be continuous, this will not in and of itself render the causal picture entirely impotent. This and the previous two issues must be considered if we are to give a worthwhile formulation of a causal picture of reference to God. However, they must not be taken as refutations.

We will conclude our investigation by considering what kind of issues such a picture of reference to God allows us to coherently discuss. It will be shown that within our picture we are able to make better sense of debates as to the existence of God, of questions regarding religious pluralism, and of questions regarding the abilities of persons of differing theological sophistication to successfully refer to the same being.

Exposition of Alston’s account

In his essay, "Referring to God,"[2] William Alston undertakes the task of generating a theory of direct reference to God. Alston’s theory is proposed in opposition to a descriptivist view of reference to God. Such a theory would claim that we refer to God only through the use of certain uniquely identifying definite descriptions (e.g. “creator of heaven and earth,” “righteous judge,” “merciful redeemer,” etc), or, that the name ‘God’ is shorthand for some number of these.[3] According to this descriptivist theory, some person ‘S’ “refers to X only if the description S is employing to pick out a referent is true of X” (Alston 1988, 113). If the relevant description does not hold true of X, S has failed to refer to X. Since logically incompatible descriptions cannot hold true of the same thing, it is not possible that two speakers who utter logically incompatible definite descriptions with the intention to refer can be talking about the same thing. If two definite descriptions are used in an attempt to pick out some object, the only one that will actually succeed in referring to that object will be the one that is actually true of it. In the case of religious descriptivism, therefore, someone who refers to a triune God could not refer to the same thing as someone who referred to a God that is wholly one. Additionally, the anthropomorphite and the negative theologian must necessarily refer to different things, if either refers at all.[4] If there is not actually some X of which these definite descriptions are true, then none of these seeming referrers has actually succeeded in referring to anything or making a meaningful statement.[5]

Alston sees such a view as problematic. His first criticism is that it seems difficult to determine just what definite description (or set of definite descriptions) would be sufficient to ensure reference to God. Alston says:

…one typically has a number of putatively identifying descriptions among which there is no obvious choice of a primus inter pares. There could be a practice of reference in which the psychological processes of speakers are so structured that exactly one putatively identifying description plays a central role in the process, and in such a way that this central description stands out in consciousness. But our referential practice is not of this sort.[6]

He is not entirely clear about what he means by saying that one has access to “a number” of descriptions. He could either be referring to the multitude of theological descriptions within religion in general[7] or within a particular religion.[8] If the worry were that there are so many different religions, each with their own concept of the divine, that one is unable to determine which of these actually picked out God, we would be facing a different question. In this case the problem would not be which description of God should be used to pick Him out, but which putatively identifying descriptive tradition is the correct one—which tradition’s description(s) actually pick out the divine. This would be a question relevant to questions of religious pluralism and the comparative study of religions. The latter understanding of Alston’s worry seems more relevant to the subject of his paper. We may presume that Alston is maintaining that within a particular religion we do not know exactly which of the descriptions is crucial over and above the others—which one is the identifying definite description. This worry alone should not be sufficient to make us wary of giving a descriptivist account of reference to God. For, just because we do not know what description(s) the name ‘God’ is shorthand for, that does not mean that ‘God’ is not necessarily shorthand for some description. It may simply be that we lack knowledge of the relevant description. Now, what about Anselm’s “that than which nothing greater can be thought”? Would not most theists agree that this description picks out God? Even if God is “that than which nothing greater can be thought,” the more important issue with respect to our discussion is: How is it that the name ‘God’ refers to God? Is it because ‘God’ is really just shorthand for “that than which nothing greater can be thought,” or some other description?

The thrust of Alston’s first criticism of this account is that while we have access to a number of descriptions believed to apply to God, we are in no way able to determine which of these is the most important or fundamental. That is, we are unsure which description serves to determine reference. Furthermore, Alston’s will later maintain that despite our uncertainty, we do succeed in referring to God. Alston believes that this occurs even if the specific descriptive attributions we make of Him do not hold true. Even when they do hold true of God, Alston maintains that it is not in virtue of this that we refer to God. Even if the name ‘God’ were in fact shorthand for some description, Alston maintains that it would still not be in virtue of this that we refer to Him. Alston’s intuition is that there is something more basic at stake in determining the reference of ‘God’. Building on this intuition, he endeavors to provide an alternative account of reference in which the details of our descriptive beliefs and utterances play more of a secondary role.

Alston argues that a Kripkean account of reference to God is in a strong sense “more fundamental than descriptivist reference” (Alston 1988, 113). By this he means that while definite descriptions may be used to refer to God, these only succeed in virtue of a more primary mode of reference, namely, direct Kripkean reference. On this account, it is only after establishing reference to God through Kripkean means that we can refer to Him through the use of descriptions. Alston’s reasons for this are threefold. First, he maintains that “almost all uniquely identifying predicates themselves contain one or more singular referring expressions…. We can rarely identify something by purely qualitative predicates” (Alston 1988, 120). That is, Alston maintains that inasmuch as these uniquely identifying definite descriptions will frequently require the occurrence of some other referring expression, such as a name, reference by means of definite descriptions presupposes some prior ability to refer without them. Unless there were an infinite regress in which, for example, ‘Alexander the Great’ meant “the student of Aristotle,” ‘Aristotle’ meant “the student of Plato,” ‘Plato’ meant “the student of Socrates,” ‘Socrates’ meant “the husband of Xanthippe,” etc, it would be necessary for there to be some point at which this kind of descriptivism would ultimately rely on some other kind of reference, namely the Kripkean kind.

In Naming and Necessity,[9] Saul Kripke presents what he terms a “picture” of causo-historical reference. According to Kripke’s account, it is not in virtue of our possession of or a name’s being shorthand for some uniquely identifying definite description(s) that we succeed in referring to an object with a name. For, we may consider cases in which we hold limited or false beliefs about a thing yet nevertheless succeed in referring to it with a name. Furthermore, we can conceive of possible worlds in which these putatively identifying definite descriptions do not hold true of our intended referent. Nevertheless, we are capable of succeeding in referring to it with some name. According to Kripke, this is because we stand in some direct causal relationship to the thing to which we refer. This is the result of that object having been “baptized” in the past with the rigidly designating proper name that we presently use. A rigidly designating proper name is a name that refers to the same being in all possible worlds. If this being exists in all possible worlds, the name is termed “strongly rigid.” This name has been passed down to us through a chain of people that used the name to refer to the actual thing that was originally given that name. So long as in our present use of the name we intend to be referring to the thing that stands at the beginning of this causal chain, our reference succeeds. Direct reference is secured regardless of the things that we have come to believe about the referent.

Alston takes this picture and applies it to reference to God. According to Alston's view, such a manner of reference begins with an experience of God.[10] During some initial experience of God, a prophet succeeds in fixing the reference of the proper name ‘God’[11] for himself by baptizing[12] the object of his experience, namely God, with the name. Following this experience, the manner of referring to God with the name ‘God’ is passed down to some community of worshippers, according to a Kripkean causal chain. The community then successfully talks about God, regardless of whether or not the individual members have their own veridical experiences of Him. If subsequent members of the tradition do in fact have experiences of God, then the name becomes multiply grounded. Either way, the referent of ‘God’ becomes fixed in such a way that it becomes possible to refer directly to Him without the need of any definite descriptions. This occurs not because the things that one says about Him are true (although they might be), but because of the referrer’s historical dependence on the original baptizer. Since this community is subsequently able to refer to God through the use of a directly referential proper name, it is no longer essential that the multitude have access to some uniquely identifying definite description(s). Even if they have come to have false or contradictory beliefs about God, they are still able to refer to Him with the name ‘God’. It becomes possible that both the most literal and the most figurative talk about God, though perhaps false, nevertheless succeed in being about God. That is, propositions containing false descriptive attributions made of God, rather than being about nothing and therefore meaningless, become meaningful though false, inasmuch as they are actually wrong about God.

After first outlining this general account of the way in which we come to inherit the names of God, Alston presents two examples that he takes to demonstrate the superiority of a causal theory of reference to God. He says: “These cases are divided into (a) those in which nothing uniquely satisfies the description(s), but S nevertheless succeeds in referring to X, and (b) those cases in which the description(s) are uniquely true of Y, but S is referring to X” (Alston 1988, 116). We will consider these in the opposite order that they are given. In Alston’s second situation, the majority of the descriptions attributed to God by a given community are true not of Him but of something else. In this case, Alston argues that it is still possible that the causal chain along which the name ‘God’ has been inherited ultimately leads back to God Himself (Alston 1988, 122). Although the members of this community are grossly mistaken as to the true nature of God, inasmuch as their causal chain originated with Him, they do in fact succeed in referring to Him through their use of the name ‘God’.

It is my understanding that this example is given as analogous to Kripke’s example of “the man who proved the incompleteness of arithmetic.” In this example, Kripke imagines that “Gödel was not in fact the author of this theorem. A man named ‘Schmidt’…actually did the work in question” (Kripke 1980, 84). Kripke looks at the situation thusly: If names are merely disguised definite descriptions and reference is determined by the satisfaction of this (or these) definite description(s), then since ‘Gödel’ is synonymous with ‘the man who proved the incompleteness of arithmetic,’ whenever we use the name ‘Gödel’ we actually refer to Schmidt. Kripke’s intuition here is that in this case “we simply are not” referring to Schmidt (Kripke 1980, 84). Rather, Kripke will maintain that it is not in virtue of Gödel’s having or not having proved the incompleteness of arithmetic that we refer to him. The name ‘Gödel’ is not shorthand for any description. There is a more fundamental causal relationship between our reference to Gödel with the name ‘Gödel’ and an original reference to him with the name ‘Gödel’. Even though in some possible world he might not have proven the incompleteness of arithmetic, he is still Gödel and ‘Gödel’ still refers to him.

However, there is a problem with taking this case and working out its analogue with respect to God. There is a fundamental difference between calling someone ‘Gödel’ and calling something ‘God’. We can conceive of someone being Gödel without having proved the incompleteness of arithmetic. If it turned out that Gödel was not the person who proved the incompleteness of arithmetic, we might be disappointed or confused, but we would still allow him to be Gödel. While our intuitions with the case of Gödel might induce us to maintain that we have still referred to Gödel even though he did not prove the incompleteness of arithmetic, it is unlikely that our intuitions hold in the same way here. For, it seems as though we would be unlikely to refer to something with the name ‘God’ if we came to find out that it did not satisfy at least some important definite descriptions. Alston anticipates this issue. He says that if someone “came to realize the true nature of the situation they will stop using ‘God’ to refer to the being in question…but that doesn’t alter the point that when they are making a referential use of the name, it is the mechanisms of direct reference that are determining the referent” (Alston 1988,125). However, despite this caveat, Alston fails to appreciate the force of this problem. We ought not to say as Alston does that the thing standing at the beginning of ‘God’’s causal chain could actually be God without satisfying at least certain of these descriptions. If it turns out that the being we refer to within ‘God’’s causal chain is actually not “the merciful redeemer,” “the righteous judge,” etc, but that something else is, in what sense would we want to say that the thing we call ‘God’ is in fact God? If the community of worshippers came to find out that the thing that they have been referring to as ‘God’ is actually just some weak entity that holds little sway over their salvation and bears no relationship to their creation, would they not abandon this being in favor of the thing that actually satisfies these descriptions (if such a thing actually exists) or stop using the name ‘God’ altogether if it were shown that nothing could satisfy at least some of these descriptions? It seems as though the community of worshippers would indeed have succeeded in referring to something as ‘God’, namely whatever stood at the beginning of the causal chain, but they would quickly cease to worship that thing or refer to it as ‘God’ once they found out that it fails to satisfy their descriptions. That is, they would hold that the thing they had been referring to as ‘God’ turned out not to be God at all. We will return to this point below.

First let us consider Alston’s other example. He imagines a situation in which some impostor ("the devil, one's internalized father figure, or whatever") "represents himself as the true God, creator of heaven and earth, righteous judge, merciful redeemer, and so on" (Alston 1988, 121). This being is revealed to some person. The impostor succeeds in convincing the person that it is actually God who has been revealed and that he is e.g. “the creator of heaven and earth.” Alston’s intuition is that although most people would take the description “the creator of heaven and earth” to apply to God, if the dupe were to refer to the impostor using this definite description then he would nevertheless refer to the impostor rather than the thing that actually satisfies the definite description, namely, God. This is meant to oppose a descriptivist theory where the dupe would actually refer to God as a result of that name being an abbreviation of “the x such that x is the creator of heaven and earth.” "Moreover," Alston argues, "if a community grows up on the basis of these revelations and epiphanies, and the practice develops in that community of using 'God' to refer to the focus of the worship of the community, we will have a Satan worshipping community in which the members use 'God' for Satan" (Alston 1988, 121). Alston posits the Divine Impostor Example in order to make sense of how a direct reference theory might be able to come to terms with allegations that the referent of ‘God’ might not satisfy certain descriptive conditions. However, this example is rather unclear and misleading for two reasons.

First, Alston fails to stipulate whether or not this case is the first experience in which some supposed prophet has referred to something as ‘God’. If it is the first then we are led only to confusion. For, what sense can we make of how something would represent itself as God without the referent of ‘God’ already having been established? There would already need to be some established practice of referring to something as ‘God’ before different descriptive qualities could become associated with Him. Without the prior association of such descriptive attributes as “the righteous judge” and “the merciful redeemer” with some particular being, it makes little sense to say that Satan or one’s internalized father figure would be representing itself as God by means of these attributes. If the meaning or reference of ‘God’ has not already been established then there is little sense to be made of a case in which something that represents itself as possessing these attributes could thereby mislead someone into thinking he has experienced God. Rather, the dupe would hold that the thing he experienced was “the creator of heaven and earth” even though he would appear to be wrong from an objective point of view. However, he would not necessarily think that that thing ought to be called ‘God’, even if it tells him that that is its name, unless he already associated these descriptive conditions with that name. If this is not the first case of something being called ‘God’, then perhaps the example can get off the ground. Perhaps the impostor would perform some kind of magic like the wizard of Oz.

Even if this is not the first case of someone using the name ‘God’ in tandem with some set of supposedly uniquely identifying definite descriptions, there is a methodological problem with Alston’s argument. This methodological issue is that throughout the essay Alston has been relying on two different versions of direct reference without clearly delineating the difference between them. While proclaiming an implementation of Kripkean reference, Alston is implicitly making use of Donnellan’s distinction between referential and attributive uses of definite descriptions.[13] According to Donnellan, when we use a definite description referentially we use it to refer directly to some thing that we have in mind, whether or not that thing actually satisfies the description. An example of this is the case of our referring to someone with the definite description, “the man in the corner drinking a martini,” regardless of whether or not there happens to be a martini or water in his glass. On the other hand, when we make an attributive use of a definite description, we use it to refer to whatever it is that satisfies the definite description. If we take the dupe to be using the definite description “the creator of heaven and earth” referentially, then he has actually referred to Satan even though Satan does not satisfy this description. In this case, we might get Alston’s Satan-worshipping community that uses the names and descriptions commonly held of God to refer to Satan. However, if we take the dupe to be using the definite description “the creator of heaven and earth” in the attributive sense, and grant that this truly applies to God, then he would actually be referring to God even though he has only had experience of Satan. However, Alston does not stipulate which of these pictures concerning definite descriptions applies to the dupe. Moreover, Alston vacillates between a picture of direct reference concerned with names as distinct from definite descriptions and one concerned with definite descriptions that are used to refer.

Peter Byrne[14] seems to be aware of this tension in Alston’s paper. However, Byrne is too charitable to Alston in fleshing out this distinction for him. Rather than noting Alston’s vacillation between theories of direct reference, Byrne assumes that implicit within Alston’s picture is the assumption that the definite descriptions mentioned above are being used in the referential sense. That is, we must assume that the dupe is referring to the impostor through referential uses of definite descriptions. Even though these descriptions are actually true of God, the dupe mistakenly uses them to refer to Satan. Byrne believes that the reason why Alston’s account is confusing is that he is only making use of a Kripkean causal theory.[15] However, it is not the case that Alston is working within a merely Kripkean framework. He refers in passing to Donnellan’s work on the referential-attributive distinction (Alston 1988, 125). We must require of Alston that he stipulate which of these is the case in the Divine Impostor Example and why. Without stipulating the way in which these descriptions are being used or how a use of a definite description is functionally equivalent to a directly referring name, Alston’s account is incomplete and unclear.

Alston fails to articulate clearly the kind of causal theory at stake. He goes back and forth between an explicitly Kripkean account of names and an account of definite descriptions that is implicitly influenced by Donnellan. These are entirely distinct theories of direct reference. While the two theories share the intuition that we may refer to something without access to a uniquely satisfying definite description, the way in which this gets cashed out differs. Donnellan accounts for the way in which the descriptivist theory fails to take account of cases in which we refer to something through the use of a definite description despite the fact that the thing referred to fails to satisfy the particular description. Donnellan’s account is pragmatic, inasmuch as it is concerned with how we might on some occasions actually use definite descriptions to refer to something other than the being of which they are true. Kripke, on the other hand, gives an account of how it is that referring to something through the use of a rigidly designating proper name allows us to use a name to successfully refer to something without the name being shorthand for some uniquely identifying definite description. Donnellan’s theory applies to our ability to refer to something directly through the use of a description while Kripke’s picture applies to our ability to refer to something directly through the use of a name, without that name being shorthand for any uniquely identifying definite description(s). If we take a Kripkean picture then we could have a Satan-worshipping community that uses the name ‘God’ to refer to Satan. In this case there would have occurred either a shift in reference (if the name ‘God’ had previously been used to refer to God) or an original baptism of Satan with the name ‘God’. However, if we make use of Donnellan’s referential-attributive distinction, we could have a Satan-worshipping community that directly refers to Satan with the definite description “the creator of heaven and Earth”. However, while making use of both, Alston fails to show what the relationship is within his theory between Kripke’s account of proper names and Donnellan’s account of definite descriptions. By failing to articulate clearly what kind of a picture of direct reference to God he intends to give, Alston’s considerations are ultimately unclear and insufficient to establish anything resembling a theory of reference to God. While this does not in and of itself render vain any attempt to apply contemporary philosophy of language to the case of religious language, it does give us reason to demand a clearer exposition of any such attempt.

Let us therefore give a clearer statement of a causal picture of divine reference. By divine reference we mean referring to God. Rather than give an account of referential definite descriptions, for the purposes of this essay we will hereafter be considering a picture that applies strictly to names. According to such a picture:

S refers to God with the name ‘G’ (where ‘G’ is a name used to refer to God—e.g. ‘God’, ‘Adonai’, ‘Ha-Shem’, etc) in virtue of S’s place within a causal history of referring to God with the name ‘G’. It is not in virtue of some descriptive condition(s) that S refers to God with the name ‘G’.

Fundamental Problems: A. Being God, B. Vacuous name, C. Continuity of chain.

It will now be necessary to subject this picture to scrutiny. In the course of this scrutiny we will encounter three problems. One problem with any such account of a causal picture of divine reference stems from our misgivings about worshiping something that does not satisfy the conditions necessary for being God. This problem will lead naturally into our second problem, which deals with our uncertainty as to God’s existence. We may also raise a third problem with giving a causal account of reference to God. This is that continuity of reference, while problematic in all Kripkean accounts, is much more difficult to grant in divine reference.

Problems with any such account

A. Being God?

The first problem that we must consider is that of the conditions necessary for the being at the beginning of our causal chain to be God without satisfying some definite description(s). It will be shown that the issue of being God is not as fundamental to our giving a picture of referring to God as we might have thought. It need not be the case that the being we call ‘God’ is God. First let us consider one attempted solution to the problem of our requiring that the thing we refer to with the name ‘God’ actually be God.

In the opening of his essay “The Name of God,”[16] Jerome Gellman presents a brief picture of reference for the name ‘God’ that bears similarity to Alston’s. According to Gellman, users of the name ‘God’ intend to be referring to some being that was originally named ‘God’ and experienced by the prophets.[17] The name ‘God’ is then passed down to others “by means of the referential chain” (Gellman 1995, 536). In order to deal with our worry that the being at the beginning of this causal chain might not be God, Gellman endeavors to provide rules of naming. Gellman attempts to go beyond Alston by raising the issue of the criteria necessary for the naming of something with the name ‘God’ to be valid. It must be noted that Gellman’s rules apply to naming something ‘God’ and not referring to something with the name ‘God’. He gives conditions for baptism rather than conditions for reference. This can be seen as a way of dealing with our objection to Alston’s position that the thing we refer to with the name ‘God’ is God even though it does not satisfy some description(s). If the thing that we stand in a causal relationship to with respect to the name ‘God’ was improperly named, that is, is not God, then Gellman believes that no legitimate act of naming ever actually took place and we are able to withdraw the term from our language.

Gellman posits that talk about God takes place within a “God naming-game,” in which a certain set of requirements must be met in order for there to have been a legitimate naming of something as ‘God’. He argues that while these descriptive conditions do not determine the reference of ‘God’, they must nevertheless be considered in naming something ‘God’. If we attempt to name something ‘God’ that does not fit at least a certain, unspecified amount of the conditions, we fail to make any naming at all. According to Gellman, for something to be named ‘God’ within the Western religious tradition, it must be:

(1) The most perfect actual being,

(2) who is very high on the scale of perfection.

(3) whose perfection is vastly greater than that of the second most perfect actual being, and

(4) upon whom all other things in some important way depend.[18]

However, he never says why these are the conditions of the “God naming-game.” He merely takes them as self-evident. The things referred to as ‘God’ in Alston’s examples do not satisfy at least a certain, unspecified number of these conditions. Therefore, if we grant Gellman’s account of the “God-naming Game,” and the thing we call ‘God’ does not satisfy these conditions, then the entirety of discourse using the name ‘God’ has actually been meaningless, since we failed to name anything with the name in the first place. Since the thing originally thought to have been baptized ‘God’ did not fit the above conditions, contra Alston, no act of naming has taken place and it is not possible that the thing could nevertheless have been God.

Next, Gellman considers an example similar to Alston’s, in which it turns out that what was thought to be an experience of God at the time of the original baptism was in fact an experience of a projected father image. This example is similar to Gellman’s example of the “animal naming-game.” In the “animal naming-game,” Gellman imagines that some people have been playing a game in which they give names to animals. If it turns out that something that was thought to have been an animal (and was therefore named according to the rules of the game) instead turns out to have been a pile of rags, Gellman argues that we should conclude that no legitimate act of naming ever took place. With respect to God, if the thing that was supposedly named ‘God’ turns out to have been a father-image, then Gellman believes we have been shown that “God did not exist” (Gellman 1995, 541). Gellman concludes this example by giving his reasoning: “A father image lies outside the limits of the God naming-game no less than a pile of rags lies outside the limits of the animal naming-game” (Gellman 1995, 541).

There are two problems with this account. First, Gellman wants to argue that if it turns out that we had merely been referring to a projected father image through our use of the name ‘God’ then we ought to conclude that God does not exist. However, within Gellman’s rules, would we not merely conclude that the thing that we called ‘God’ was not actually God? Although we made a mistake in our playing of the “God naming-game,” we have not received evidence either way regarding God’s existence. That something wrongly called ‘God’ is not God does not necessarily entail that nothing is God. The fact that some initial user of the name ‘God’ fails to successfully name God with that name does not mean that God does not exist. God could exist whether or not He was ever named or encountered. In Gellman’s example of the thing named ‘Kuzu’ that turns out to be a pile of rags rather than an animal, we do not conclude upon finding out the truth of the matter that animals do not exist. Just because we failed to name an animal with the name ‘Kuzu’, that does not mean that there are not any animals. We merely conclude that we did not on this occasion name an animal with the name ‘Kuzu’. Similarly, from the fact that we failed to name God we ought not to conclude that God does not exist. Rather we ought to conclude merely that on this occasion we did not name God with the name ‘God’. Now, we can conclude this without yet knowing or being committed to God’s existence. All that this should tell us is that the thing we named ‘God’ is not God.

Second, Gellman’s appeal to rules governing naming is misguided. In positing his “God naming-game,” Gellman sets forth a set of conditions governing the naming of something. He then proposes that we ought to see whether or not these conditions hold true of the being referred to by our forefathers.[19] If it does fit the conditions, then it is God and has properly been named ‘God’. If it does not fit the conditions then no act of naming has taken place. However, it is evident that most people do not begin to conceive of God as whatever it is that satisfies some set of conditions (many never at all). Furthermore, scripture shows us that persons have been acquainted with God as ‘God’, ‘Allah’, ‘YHVH’, etc (their translational equivalents), and then refer to Him as such. We do not have evidence of any explicit act of naming ever having taken place. Additionally, Gellman proposes rules governing the naming of things as God’ but does not account for rules governing uses of the name ‘God’ to refer. This seems wrong. There does not seem to be anything logically precluding our naming something ‘God’ even if it does not satisfy these supposed conditions. I used to have a fish that I named ‘God’. It is merely custom (a custom that Gellman seeks to appeal to and reinforce) that keeps us from naming certain kinds of things with certain names. However, custom is not the same thing as logical impossibility. In “On Referring,”[20] P. F. Strawson rightfully remarks: “At present our choice of names is partly arbitrary, partly dependent on legal and social observances” (Strawson 1950, 355). Furthermore, it is evident that an act of naming has taken place. In the case of the pile of rags that has been mistakenly baptized ‘Kuzu’, it is still in virtue of our having named it ‘Kuzu’ that we are able to refer to it. Even if it turns out that the being at the beginning of our causal chain does not satisfy Gellman’s “God-naming” conditions, that alone does not mean that we have really failed to name it.

Through a consideration of Gellman’s rules governing the naming of something as ‘God’, we have seen that although we might desire to provide restrictions as to what type of thing could stand at the beginning of the name ‘God’’s causal chain, we do not seem to be justified in doing so through laws governing naming. Gellman’s rules are problematic not only inasmuch as he misinterprets their consequences, but also insofar as rules of naming simply do not apply in fixing reference. Even if we consider Kripke’s stipulation that it need not be necessary that some explicit act of naming take place[21], we are in need not of rules governing the naming of some being with the name ‘God’, but of rules governing reference to God with the name ‘God’.

Still, we have not allayed our concern that the conditions associated with the name ‘God’ are different than, for example, the name ‘Gödel’. For, the name ‘God’ seems to be rather different from some other proper names. We have come to agree with Mill that a city may be named ‘Dartmouth’ even though it does not lie at the mouth of the River Dart. The name ‘Dartmouth’ can be divorced of its descriptive content. However, we are unsure as to whether something should be named ‘God’ without satisfying at least some definite descriptions concerning its nature. As Richard Gale remarks in his On the Nature and Existence of God[22]:

A second disanalogy between ‘God’ and the sort of ordinary proper names to which the historical-cum-indexical-reference theory applies is that whereas it is not an analytic truth or true by definition that the referent of an ordinary proper name satisfy some description, this does not appear to be so for ‘God’. At any time at which ‘God’ is used, there will be some descriptive sense that it has by definition.[23]

We have seen that Gellman’s naming-rules are insufficient to cope with this issue.

Is there a better way of looking at this problem? I think there is, but it will require a re-appraisal of Kripke’s work on names. The problem that we have encountered is that while we have relinquished a theory of reference in which we refer to something because of the fact that it satisfies some descriptive conditions, we have nevertheless found a problem with considering the name ‘God’ as purely denotational. Inasmuch as Kripke’s picture disavows the role of any kind of Fregean sense, we have found ourselves ill-equipped to make sense of what it means for the thing we call ‘God’ to be God.

Mill is aware of this tension regarding the connotative content of certain names that are nevertheless strictly speaking denotational. In his A System of Logic[24], Mill writes:

But there is another kind of names, which, although they are individual names, that is, predicable only of one object, are really connotative…It may be significant of some attribute, or some union of attributes, which, being possessed by no object but one, determines the name exclusively to that individual…‘God,’ when used by a monotheist, is another. [25]

Now, what does it mean to say that a denotational name is “really connotational”? Does it mean that the name has a meaning? On Mill’s account, yes. Does this meaning determine reference? On Mill’s account, yes. However, I stand behind our intuition that this is not the case. The connotation of a name does not give the conditions necessary for something being so named. Nor does something named with a connotative name necessarily need to satisfy the connotation. Rather, the connotation of a name gives us secondary information about its bearer in the form of either the word’s literal meaning or something else that has been attached to it. This does not mean that the bearer must satisfy this information. However, a connotative name will bring along this secondary information. Strictly speaking, it is possible for any name to have a connotation attached to it. Furthermore, these connotations can change over the course of time. For example, the name ‘Mohammed’ now has a connotation attached to it (in addition to its meaning “the praised one” in Arabic) that is dependent upon the prophet Mohammed having had that name. However, prior to his having become an important historical figure, this particular connotation did not adhere to the name ‘Mohammed’. There may have been some other connotation prior to this, but it is important to note that connotations are not fixed once and for all. Furthermore, connotative names can cease to require that their bearer satisfy the relevant connotation. Eventually this connotation might disappear for all but the most pernicious linguists. For example, the name ‘Cicero’ is derived from the Latin word ‘cicer’, which means chickpea.[26] Even though Cicero inherited his name from an ancestor whose nose resembled a chickpea, it was not necessary for him to satisfy this connotation in order to bear the name ‘Cicero’. [27] Now, keeping Kripke’s objections to descriptivism in mind, we might maintain that while the connotation can tell us about something that has the name, it is not necessarily the case that the referent in fact satisfies the connotation. That is, connotation gives something like the meaning of a name or some secondary information but neither determines reference nor must be satisfied by its referent. Despite the fact that the name carries along with itself some connotative content, reference is still secured otherwise than by descriptive or connotative satisfaction.[28] This is not to say that the bearer of a name cannot satisfy the name’s connotation. It is of course just as possible for a person to be named ‘Cooper’ who is a barrel-maker as it is for a person to be named ‘Cooper’ who is not a barrel-maker. In the case of the one who does satisfy the connotation, it is simply important to note that it is not in virtue of this that reference to him with the name ‘Cooper’ is secured.

Here we will see the first and most important contribution of applying a Kripkean conception of reference to the case of God. That is, a Kripkean picture allows us to refer to something despite the fact that it may not actually satisfy some descriptive association. Reference is not fixed by the satisfaction of some description. Now, this does not in and of itself rule out that the name does have some connotative concept attached to it. All that the Kripkean picture claims is that it is not in virtue of some descriptive concept that we refer to something. Regarding names, while it is the case that some of them do in fact have a connotative concept attached to them, we nevertheless use them to refer to their referent even if we are ignorant of the specifics of this concept or the referent does not satisfy the connotation. That is to say, despite the fact that in the actual world the name ‘Dartmouth’ has some descriptive concept attached to it, e.g. “being the town located at the mouth of the River Dart,” etc, it is not in virtue of this that we refer to Dartmouth. Rather, we refer to Dartmouth because we have inherited the practice of referring directly to it with the name ‘Dartmouth’. Thus we may still refer to the city in counterfactual situations, not because of the qualitative things it would have satisfied in those situations, but because of our relationship to the city itself. That is, we do not refer to Dartmouth in the possible world in which it was named ‘Dartcity’ because of its having been named ‘Dartcity’ in that world. Rather, we refer to Dartmouth because of the relationship between ourselves and our actual Dartmouth. It is our Dartmouth that has been named ‘Dartcity’, not some city named ‘Dartcity’ that happens to be our Dartmouth. Furthermore, we can make sense of how it is that something else, for example, a college, can be named ‘Dartmouth’. Even though the fact that the university has this name is dependent upon the city (and its Earl) having had that name, it still has the same name. Our revised picture of reference now states:

S refers to x with the name ‘N’ in virtue of S’s place within a causal history of referring to x with the name ‘N’. Even though ‘N’ may have a connotation, it is not in virtue of this or some descriptive conditions that S refers to x with the name ‘N’. Furthermore, it is not necessary that x is N. Additionally, there may be some other entity y that is also named ‘N’.

Now, while it does not seem necessary for Dartmouth to satisfy some descriptive condition(s) in order for us to refer to it with the name ‘Dartmouth’, are we in the same situation with respect to God? Does God need to satisfy some descriptive condition(s) or not in order for us to refer to Him? Yes and no. Yes, certain things need to be true of the being we call ‘God’ in order for it to be God.[29] No, the thing we call ‘God’ does not need to satisfy any descriptive conditions in order for us to refer to it with the name ‘God’. That is, in order for God to satisfy the connotation of ‘God’, He needs to satisfy certain conditions. However, on our revised account, in order for us to refer to something as ‘God’, it does not. Even if the being that we call ‘God’ does satisfy the name’s connotation, it is not in virtue of this that we refer to it. It would not be the fact that the being we call ‘God’ satisfies some set of descriptive conditions that we would succeed in referring to it with the name ‘God’. Rather, it must be in virtue of the causal history of the name ‘God’ that we refer to Him.

Furthermore, it might turn out, as in Alston’s example, that the thing we call ‘God’ really does not satisfy the connotation of ‘God’. Now, if we were to find this out (I’m not sure how we would), we would probably stop calling that thing ‘God’. But this does not mean that we had not referred to it or could not continue referring to it with the name ‘God’. We might still refer to it as ‘God’, but we would not necessarily mean that the thing we call ‘God’ is God. Perhaps we would begin to call it ‘Pseudo-God’, or something like that. Now, as we have noted above, this would not mean that God does not exist. All we would know is that the thing we have called ‘God’ is not God. Let us then reformulate our picture of divine reference:

S refers to x with the name ‘G’ (where ‘G’ is a name used to refer to God—e.g. ‘God’, ‘Adonai’, ‘Ha-Shem’, etc) in virtue of S’s place within a causal history of referring to x with the name ‘G’. Even though ‘G’ may have a connotation, it is not in virtue of this or some descriptive conditions that S refers to x with the name ‘G’. Furthermore, it is not necessary that x is God. Additionally, there may be some other entity y that is also named ‘G’.

There seems to be an important sense in which we should retain our Millian intuitions that in the case of a name like ‘Dartmouth’ it is not necessary that the thing referred to satisfy the conditions of being at the mouth of the River Dart. However, at the same time, some names have connotations that require us to take something like sense or meaning into mind. However, this does not mean that the sense or connotation determines the reference. We can still directly refer to something with a connotative name even if the thing does not satisfy the connotation. Furthermore, we do not necessarily refer to the thing that actually satisfies whatever conditions are contained in the name’s connotation. We refer to the thing that we refer to.

B. Vacuous name?

“All names are names of something, real or imaginary.”

-J. S. Mill (System of Logic Book I, Chapter I: §3)

Now, even if we allow that ‘God’ is the kind of name to which Kripkean intuitions may apply, even though it does carry with it some associated connotative concept, we still have two more doubts to address. Our second worry is that our picture of reference cannot account for reference to nonexisting things. Many have claimed that God does not exist or is in some sense dead. Now, if the latter is true then we are not in that bad of a position in terms of giving a Kripkean account of reference to God. If it is merely the case that God has died then that means that He was once alive and therefore we can give an account of reference to God much in the same way that we would with respect to the late Mr. Aristotle. Additionally if it is merely the case that God is not alive then we can still refer to Him much in the same way that we would refer to an inanimate object. However, if, as the former says, God does not exist, never has existed, and could not possibly exist, then we are faced with a more difficult problem. If this is the case then it does not seem to make any sense to provide a Kripkean account of reference to God. For, on this account a name rigidly designates the thing that it refers to. However, something that does not exist cannot have a name attached to it. Therefore, a name cannot designate something that does not exist. This problem is especially pertinent to any attempt at formulating a picture of divine reference. Richard Gale gives voice to this worry in his On the Nature and Existence of God. Gale says: “Another fundamental inadequacy in Alston’s causal theory of reference for ‘God’ is that it cannot account for how two persons who use ‘God’ can be coreferrers even if God does not exist” (Gale 1991, 301). Gale believes that we should be able to account for reference to God without depending on His necessary existence. The problem as Gale states it is that within a Kripkean framework it seems impossible to say that two people could succeed in using the same proper name to refer to the same nonexisting thing. While the problems of vacuous reference can be expanded to include the difficulty that two people cannot successfully refer to the same nonexisting thing according to orthodox Kripkeanism, the problem is more fundamental. It is impossible not only for two people to be co-referrers, but for one person to refer at all within a Kripkean picture of reference if the thing seemingly referred to does not exist. This problem is germane to any purported reference to historical figures. We can never be certain or prove that anybody from the past actually existed. It does not follow from the fact that we purportedly refer to historical figures that they actually did exist, no matter how strong the evidence. It is as impossible to prove that Homer existed as it is to prove that anyone existed. No amount of historical reference is sufficient. For, it is just as possible that our predecessors only seemed to refer. If we are unsure of someone or something’s existence, we should be unsure of whether or not we are making a genuine reference within a Kripkean picture.

Now, this is a problem within Kripke’s picture as a whole. It is not only relevant to the case of reference to God. For, there seems to be an important sense in which names that we know lack existing bearers do in fact refer. It is not merely the case that we seem to refer. But if a name serves to contribute its referent and these are cases in which there is not an existing referent, how do vacuous names refer? That is, what do they contribute to propositions containing them? Bertrand Russell raises this issue in “The Philosophy of Logical Atomism.”[30] He writes:

It is obviously a perfectly significant statement, whether true or false, to say that Romulus existed. If Romulus himself entered into our statement, it would be plain that the statement that he did not exist would be nonsense, because you cannot have a constituent of a proposition which is nothing at all.[31]

Theories of direct reference claim that names function to contribute their referent to statements in which they occur. If there is not some existing referent to be contributed, statements containing vacuous names must be missing a vital component and therefore be meaningless or mistaken. Russell does a good job of raising this problem but fails to adequately solve it. He goes on to claim:

[T]he name ‘Romulus’ is not really a name but a sort of truncated description. It stands for a person who did such-and-such things, who killed Remus, and founded Rome, and so on…. If it were really a name, the question of existence could not arise, because a name has got to name something or it is not a name, and if there is no such person as Romulus there cannot be a name for that person who is not there.[32]

He argues that since it seems as though we are forced into a logical problem if we take names to contribute their referents to statements in which they occur, names that purportedly name non-existing things are not really names at all. Rather, they, like all proper names, are merely shorthand for definite descriptions. The question of how we can make seemingly meaningful propositions concerning the existence of things using what appear to be proper names pervades Russell’s work. In “Descriptions,” [33] Russell maintains: “And so, when we ask whether Homer existed, we are using the word “Homer” as an abbreviated description: We may replace it by (say) “the author of the Iliad and the Odyssey.” The same considerations apply to almost all uses of what look like proper names” (Russell 1919, 332). All that a positive or negative existential statement says on Russell’s account is that there is or is not some x of which such-and-such. However, we must not be led into this kind of a solution for vacuous names since we are not willing to accept it for ordinary names. Our Kripkean intuitions regarding counterfactuals must still apply. An author might have chosen for his character to have done q instead of p. Nevertheless it would still have been the same character. Even if in some other version of the story some other character would have happened to do the things that our character does in our version, the two characters are distinct and their names refer directly to them without being shorthand for definite descriptions. For example, Sherlock Holmes is still Sherlock Holmes and ‘Sherlock Holmes’ still refers to him whether or not Conan Doyle decides to make him do one thing or another. Neither ordinary nor vacuous names are “truncated descriptions”. How, then, could it be that statements containing vacuous names are nevertheless meaningful and seemingly about some specific individual? Although Kripke’s picture cannot account for vacuous names, it does seem that they refer. This is evidenced by the fact that in both common and sophisticated discourse, persons seem to successfully discuss non-existing things and things the existence of which we are uncertain. Examples of such reference include the discussion of characters occurring in fiction and things that are logically impossible. Further sources of confusion include real people who appear in fiction[34], characters appearing in multiple works of fiction (either by the same author or by different authors)[35], fictional characters created by fictional characters[36], fictional characters whose defining feature is their non-existence[37], etc. As in these cases there is either not an actual existing thing to which a name has been attached, or an actually existing thing that becomes the subject of fiction, it seems impossible or at least confusing for there to be a causal chain leading from the fiction to the referrer. According to the causal picture, someone’s use of a vacuous or fictional name should be meaningless when in fact it is not. The bare causal picture is unable to countenance the problem of vacuous and fictional reference

Enquiry into possible solutions

We must therefore see whether or not there is room to solve this problem within the context of a Kripkean picture of reference. Following our consideration of various ways in which the problem of fictional and vacuous reference has been dealt with, we will see whether or not any of these are sound for any fictional term. First, we will consider pretense-theory solutions. Second, we will consider solutions that make use of “blocks”. Third, we will consider a solution that makes use of object theory. We will eventually come to favor a variant of the third kind of solution. Finally, we will consider whether we ought to apply such an account to the problem of referring to God given our ignorance of His existence.

Pretense Theory? No.

Some philosophers have attempted to account for vacuous and fictional reference through so-called “pretense theory.” These philosophers maintain that in cases of vacuous reference, we only seem to refer in virtue of some pretense or make-believe. That is, we merely pretend that there is an x named ‘N’ of which p and q. For our purposes, then, such a solution would maintain that we merely pretend that there is something named ‘God’ that is, e.g. “the creator of heaven and Earth”. One of the most active proponents of this position is Kendall Walton. He describes his picture thusly in “Existence as Metaphor?”[38]:

The central idea of the make-believe approach is, of course, that what seem to be commitments, by speakers or theories, to nonexistent entities are to be understood in a spirit of pretense or make-believe. They are, in one sense, not meant seriously and not to be taken seriously. When, having exposed our parents’ lies about Christmas, we go on to “talk about” Santa Claus anyway, and when we “talk about” Sherlock Holmes, the Fountain of Youth, Vulcan, and ether, while accepting that there aren’t really any such things, that they don’t exist, we are recognizing and utilizing fictions to the effect that they do exist, or better, fictions to the effect that in using names and referring expressions like “Santa Claus,” “Sherlock Holmes,” “the Fountain of Youth,” “Vulcan,” and “ether,” we successfully pick out (existing) objects.[39]

Walton’s account does at first seem reasonable. After all, when we open a book, we surely enter a world of make-believe or at least willfully suspend disbelief. We know that a fictional character does not exist, but for the sake of the story, we merely pretend to believe that he does. However, in cases where we are not sure whether a work is fiction or non-fiction, pretense theory fails. For, it would be wrong to suppose that a person who does not know whether or not some character actually exists outside the novel is nevertheless merely pretending that that character exists (whether they realize it or not) just in case there is no real-world referent. Here it seems rather that the person who is ignorant of some character’s actual existence is merely taking himself to be reading about the actual character, whether or not the character also exists in reality. Upon eventually coming to learn that the character does not actually exist, this person might pretend that the character does exist. However, we must note the fact that the thing about which our reader now pretends to believe existence is the same character whose existence he was previously unsure of. Moreover, there are cases in which one would refuse to claim that they were merely pretending to refer to some non-existing entity. For example, if it were to be the case that God does not exist, then it would be ludicrous to suppose that every religious person is merely pretending to believe that there exists some being, namely God, about which they make statements, to whom they pray, etc. As Edward Zalta notes near the end of his “Pretense Theory and Abstract Object Theory,”[40] “Ponce de Leon was not pretend-searching for the fountain of youth. We do not pretend fear when we wake up screaming in the middle of the night, having dreamed about a monster” (Zalta 2000, 143). Likewise, the believer is not pretend-worshipping. The religious person does not pretend that something named ‘God’ exists. This is a fundamental problem with “make-believe” and “pretense” theories of fictional and vacuous reference.[41]

Blocks? No.

In his article “Gale on Reference and Religious Experience,”[42] Andrew V. Jeffrey attempts to give a different kind of solution to the objection that ‘God’ may be a vacuous name. Jeffrey attempts to deal with Gale’s objection to the possibility of giving a picture of direct reference to God despite the possibility of God’s non-existence through an appeal to Keith Donnellan’s work on vacuous names.[43] Jeffrey says:

Donnellan also suggests that two distinct negative existence statements…can express the same proposition if the blocks in the histories of the terms are related in the right way…If Donnellan is right, then we can conclude that when two or more uses of a term end in a common “block,” or in blocks sharing the right kind of historical connection, the statements in which these uses appear preserve intentional identity, i.e., are “about the same thing,” despite the statements’ lack of a real world referent.[44]

Jeffrey’s solution to the problem of applying the causal picture of reference to the case of God, if God does not exist, is that if it happens to be the case that my use of the name ‘God’ ends in the same block as, say, Moses’ use of the name ‘God’ (or its translational equivalent), then my use of the name is successful inasmuch as it bears a causal relationship with some historically prior reference.[45]

Let us therefore consider Donnellan’s solution to the problem of vacuous names. In “Speaking of Nothing,” Donnellan has been giving a way of analyzing the truth-conditions of negative existential propositions through non-descriptivist means. He is attempting to give a solution to the problem of how a negative existence statement can be true within a “historical”[46] theory of reference without there existing a real thing referred to. Donnellan writes:

I will suggest a rule, using the notion of a block, that purports to give the truth conditions for negative existence statements containing a name. This rule, however, does not provide an analysis of such statements; it does not tell us what such statements mean or what proposition they express. This means that in this case we are divorcing truth conditions from meaning.[47]

The rule does not tell us what these statements mean. It merely tells us what be the case in order for the statement to be true or false. The rule is:

(R) If N is a proper name that has been used in predicative statements with the intention to refer to some individual, then “N does not exist” is true if and only if the history of those uses ends in a block.[48]

Now, keep in mind that Donnellan defines a block as: ”When the historical explanation of the use of a name (with the intention to refer) ends…with events that preclude any referent being identified” (Donnellan 1974, 23). So, (R) states that given some proper name ‘N’, “N does not exist” is true if and only if it is impossible to identify the referent. Donnellan is concerned at this point about the conditions necessary for negative existence statements being true given that there is not actually some existing thing of which the statement is true. He has not yet given an account of how it is that different vacuous names seem to refer to different non-existing things. Donnellan goes on to amend his rule in order to make it capable of differentiating between different empty names and their corresponding blocks. He says:

Certain uses of the name ‘Aristotle’ in predicative statements will have similar histories, histories that will distinguish them from other uses of the name. Each use of the name will, of course, have its own historical explanation, but these may, at a certain point, join up. […] It is possible that the histories may join at what I have called a block. Another possibility, however, is that although different uses of the name end in different blocks, these blocks are themselves historically connected.[49]

According to Donnellan, if two uses of the name ‘Aristotle’ (in tokens of the statement, “Aristotle does not exist”) join up in the same (or “historically connected”) situation(s) in which it is impossible to determine the referent, then these two uses have the same truth value. If two empty names are historically connected, negative existential statements containing them will express the same thing. Donnellan gives an example of a case in which this would happen. His example concerns the sameness of the propositions, “Santa Claus does not exist” and “Père Noël n’existe pas.” What, then, is the purpose of a name the use of which will always end in a block? That is, what does the name do? It seems as though according to this theory the name should point us towards the historical connection that an omniscient observer would need to look at in order to get confirmation that some negative existence statement is either true or false and whether it expresses the same thing as some other statement. But what is this historical connection, and what distinguishes it from other chains? How is it that two blocks join up or are historically related?

This is a classic example of putting the cart before the horse. It is not the case that “Sherlock Holmes does not exist” is true because my use of the name ‘Sherlock Holmes’ somehow links up with Arthur Conan Doyle’s original use. It would be true that Sherlock Holmes does not exist whether or not Conan Doyle ever actually wrote any of his stories or named their protagonist ‘Sherlock Holmes’. This is the case simply because Sherlock Holmes, unlike Conan Doyle, really does not exist. It has nothing to do with some historical use of the name.

Object Theory? Yes.

“Let us grant to those who work in any special field of investigation the freedom to use any form of expression which seems useful to them; the work in the field will sooner or later lead to the elimination of those forms which have no useful function. Let us be cautious in making assertions and critical in examining them, but tolerant in permitting linguistic forms.”
-Rudolf Carnap
[50]

An alternative hypothesis about vacuous and fictional names may be given along the lines of object theory. Such a theory would be able to account for why it is that it predicative statements or negative existence statements really seem to be about something, despite the fact that the thing that they are about does not really exist. One such solution to this problem may be informed by a reading of Terence Parsons’ work on semantics and intentionality in his series of articles and his book, Nonexistent Objects.[51] In these works, Parsons argues for a Meinongian account of intentional objects. Intentional objects are simply the objects about which we make statements and judgments, have thoughts, etc. According to such a theory, intentional objects do in some important sense exist. However, they do not exist in our usual sense of what we mean when we say of something that it exists. This account comes out of the Austrian philosopher Alexius Meinong’s theory of objects, in which some objects have our typical notion of existence and others do not, yet all objects do in some sense exist as what Parsons terms “existents”. One motivation behind Meinong’s theory is that all of our judgments presuppose there being something that is judged. That is, in order to make a judgment about something, there must be something that is the subject of the judgment. For example, in “The Theory of Objects,”[52] Meinong writes: “In order to know that there is no round square, I must make a judgment about the round square” (Meinong 1904, 83). There must be some thing that lacks the property of existence in order for us to make sense out of negative existential statements according to this Meinongian account.[53] This thing is what we will call an existent. In making use of Parsons’ and Meinong’s work on objects and semantics, we will come to formulate a picture of reference that contains under its scope reference to the same kind of Meinongian objects in cases of both ordinary and vacuous proper names.[54] That is, our account will apply to both existing and non-existing existents. Internal to our semantics we will be relying on something similar to a “’Fido’-Fido” view of proper names.[55]

Here it must first be noted that Parsons and Meinong will be considering their objects as constituted by a set of properties. Parsons’ account of names in “A Prolegomenon to Meinongian Semantics,” makes use of sets of properties as determining the reference of names of objects such as Agatha Christie and God. In Nonexistent Objects, Parsons says that within his language, proper names will be placed “before one-place predicates, just like definite descriptions. So we write ‘Pegasus flys’ just as PF” (Parsons 1980, 120). He continues to say that “a sentence of the form is true or false depending on whether or not the object that A refers to has α or not, and that if A fails to refer, then such a sentence is automatically false. This makes names behave just like definite descriptions” (Parsons 1980, 120). However, Parsons allows that names can refer to objects that do not exist. For example, he maintains, “’Pegasus’ refers to a certain object, to the winged horse of Greek mythology” (Parsons 1980, 121). While Parsons shares our intuitions about our ability to refer to non-existing objects with proper names, he disagrees with our rejection of descriptivism. Rather, Parsons denies the validity of the causal picture of names. He claims that it rests on a mistake about the role played by the referent itself in the causal chain. He writes:

The mistake is that the referent of the name must itself be a causal agent in the chain. I don’t think that that is right even in the case of certain existing things. For example, the novel The Wind in the Willows has a certain name (namely, “The Wind in the Willows”); but if we trace back our present use of that name causally, we don’t come to the novel, but rather to a copy of the novel. The novel itself is not a physical object, and doesn’t enter into causal relations. But coming to a copy of the novel is good enough; we need one more link in the chain, but it’s not a causal one; rather, it consists of something like exemplification or tokening. I think that reference to Pegasus or Sherlock Holmes is like this. We trace our use of the name causally to the novel, or to a telling of the myth, but then, instead of encountering what Donnellan calls a “block,” which is something like a break in the chain, we make one more non-causal step to its referent.[56]

Parsons maintains that the referent of a proper name does not itself enter into the causal chain governing its reference. However, the example that he gives to prove this point is misleading. There is a difference between the name of a novel and the name of an ordinary object. It is unclear what exactly the name of a novel itself names. Is it an idea in an author’s head? Is it the original manuscript? Is it the copyright? This is an interesting question but it is not one that should distract us from our inquiry. Rather than claim that our use of the name of a novel originates from some copy of a novel, we must maintain that the name of the novel originates from the novel itself. After all, we are able to successfully refer to a novel regardless of whether or not we have ever encountered it in the form of an actual copy. To the best of my knowledge I have never encountered a copy of The Wind in the Willows. Now, if I were to, I know that the name of the novel would be on the title page of the copy. Now, where does this title originate? Surely it does not originate with the copy. If it did, how would a printer ever know what title to print on what copies of novels? The name of the novel must originate with the novel itself, whatever that may be. Now, if we were to name some particular copy of a novel, then that copy of the novel would itself enter into our chain for its name insofar as it would stand at its beginning. This would be similar to how an individual person enters into the causal chain relevant to himself and his name.

It will therefore be necessary to add to Parsons’ work in order to generate a picture of reference to objects, both existing and non-existing, that coincides with our Kripkean intuitions about names. If we intend to give a Kripkean account of reference to objects, we will need to provide further argumentation as to how it is that we refer directly to objects without reference being determined by the satisfaction of some (set of) description(s). Now, the bare fact that something is in some sense constituted by properties does not in and of itself mean that reference to it must be descriptivist. For, even though in this world something does happen to have some properties, it is not in virtue of these that we refer to it. That is, although it is the case that in our world Richard Nixon was the 37th President of the United States, it is not in virtue of the name ‘Richard Nixon’ being an abbreviation for “the 37th President of the United States” that we refer to him with the name ‘Richard Nixon’. Now, there are certain properties, such as “the 37th President of the United States,” “being named ‘Richard Nixon’,” etc, which only contingently apply to Nixon. Following Timothy Sprigge’s distinction, Kripke calls these properties “external.” However, there are also more fundamental properties of Nixon that cannot change across possible worlds. Among these are, for example, the properties of “being human”[57] and “being a child of Francis and Hannah Nixon.”[58] Again following Sprigge’s terminology, Kripke terms these properties “internal.” Now, according to our account it should still not be in virtue of his having these internal properties that we refer to Nixon. Of names, Parsons writes:

Names are to name objects, which means that we require that g(a) Î q for each name a. For example we might let ‘a’ name the object which is the correlate of Agatha Christie:

g(a) = {p: Agatha Christie has p}

Then ‘a’ would play the role in this language that the name ‘Agatha Christie’ plays in English. Similarly we might have:

g(b) = {p: p is ascribed to God by such and such a religion}

Then ‘b’ might play the role of ‘God’.[59]

On Parsons’ account ‘p’ denotes a property, ‘a’ denotes an object, and ‘q’ denotes the set of all objects. To say “that we require that g(a) Î q for each name a” means that for every name, the thing named must be contained within the set of all objects. “Being an object” is interpreted on the widest possible scope, allowing for all intentional objects to fall under its domain.

It is my aim to sketch a way of reconciling the problem of fictional reference by an alteration of the causal picture of reference that takes into account a reappraisal of Alexius Meinong’s work in spite of the criticisms levied against it by Bertrand Russell. Such a picture will be shown to make better sense of reference to both existing and non-existing objects, with special attention paid to the existent Abrahamic God to which both the Abrahamic theist and the atheist refer. A Kripkean picture would require that we say that S refers to some object x with the name ‘N’, not because x satisfies the descriptive conditions associated with the name ‘N’, but because of the fact that some prior person, say, R, referred to x with the name ‘N’. This picture must be extended backward until we get to the original baptizer of x, if there is one. So, we may say that R referred to x with the name ‘N’ because some prior person, Q, referred to x with the name ‘N’, with this ultimately leading back to some person, A, referring to or baptizing x with the name ‘N’. That is, ‘N’ rigidly designates x. Furthermore, it is necessary to note that the same Meinongian object may have multiple names attached to it. For example, we can conceive of the fact that the Meinongian object corresponding to Clark Kent has two names attached to it, namely ‘Superman’ and ‘Clark Kent’. These both directly refer to the same existent object that does not exist, even if they carry with themselves different connotations.

According to such a picture, it is to the thing to which we have an intentional stance that we refer to when we use a proper name. That is, it is not the actual existing, embodied, living object that is referred to directly, but the metaphysically distinct existent that bears the property of either existence or non-existence. Now, this is not to say that we refer to our idea of the thing or some sort of concept. We still directly refer to the thing itself. It just so happens that the thing-itself is not merely the “actual existing, embodied, living, concrete object.” There is a corresponding Meinongian object that is either instantiated as an existing thing or not. This shows that the object theorist’s claim is in a strong sense prior to the pretense-theorist’s position. For, our person must now be pretending that the same intentional object (about whose existence he was once unsure) exists even though he knows that it does not. The claim is also prior to a solution like Donnellan’s. This is because our child who has come to learn that Santa Claus does not exist really is talking about the same thing that he previously thought existed. “Sherlock Holmes does not exist” is true simply because Sherlock Holmes really does not exist. “Santa Claus does not exist” and “Père Noël n’existe pas” express the same proposition because they both deny existence of the same being, namely Santa Claus.

In order to engage with Parsons’ work, it will be necessary to adopt his logical terminology. Parsons retains the existential quantifier, but re-interprets its ontological and metaphysical scope. In his system, the existential quantifier may be reformulated as saying something like, “There is some object x such that…,” rather than saying something like, “There exists some object x such that….” In order to designate what we would commonly mean by existence, Parsons makes use of the predicate “E!,” which, interestingly, is borrowed from Russell and Whitehead’s Principia Mathematica. It may be the case that some object does not have this kind of existence. However, we may still place such an object within the existential quantifier’s scope. Therefore we may have existing existents (i.e. things that really do exist or have existed: $x(E!x); e.g. me) and non-existing existents (i.e. things that really do not exist and never have: $x(~E!x); e.g. Sherlock Holmes). There is an important distinction here between existents that actually do not exist but are logically possible (contingently non-existing existents: $x(◊E!x & ~E!x); e.g. unicorns) and existents that actually do not exist and are logically impossible (necessarily non-existing existents: $x(~E!x); e.g. the round square).

It must be noted that there are intentional objects whose existence we are unsure of, yet to which we nevertheless refer. For example, while lacking evidence as to whether or not some actual man Homer existed, we may nevertheless succeed in referring to “him.” That is, as a result of some causal relationship that I bear to some existent intentional object that may or may not have actually existed and may or may not have been the author of The Iliad and The Odyssey, statements of ours in which the proper name ‘Homer’ occur do succeed in referring. If we were to find out conclusively that Homer either actually existed or did not, we would still be referring to the same object Homer that we do now. Moreover, we have inherited this practice of talking about Homer via some causal chain. I may say something like, “Homer wrote The Iliad,” without ‘Homer’ being shorthand for “the author of The Iliad.” This picture also provides a sensible interpretation of the way in which we refer to things that we are reasonably sure never existed. When we refer to Sherlock Holmes we refer to the Meinongian Sherlock Holmes object that most people would be willing to admit does not exist. It is not our idea or concept of Sherlock Holmes, but Sherlock Holmes himself (even though he does not exist). Some people, it must be noted, do believe that actually non-existing existents do exist. For example, many children believe that Santa Claus[60] does in fact exist. Now, eventually these children might come to learn that Santa Claus does not exist. When they then make a statement such as “Santa Claus does not exist” they refer to the same object Santa Claus about which they might previously have said, “Santa Claus does exist”. It is for this reason that a community of users may successfully refer to objects that do not exist or whose existence is unclear. For example a referential community may refer to Atlantis even though some believe the island to be purely fictional or mythological. These non-believers still refer to the same thing as those who believe it once did.

Now, it is important to note that even in these cases, there is a real causal history at work in our use of empty names. Empty names are acquired in precisely the same way as any other. They are inherited from our predecessors and we pass them on to our descendants. Now, empty names may come into our language in different ways than ordinary proper names. Inasmuch as there is no concrete object to be baptized or ostended, an empty name’s origin will perhaps be less easily pinned down. Some of these might come into being as directly referring proper names through the act of literary creation. Others might come into being through hallucination or games of make-believe. However, the details of the origin of an empty name are just as irrelevant to our currently using it to refer as are the details of the origin of an ordinary proper name. Once introduced, the practice of referring with both ordinary proper names and empty proper names is transmitted along Kripkean causal chains.

The two most common kinds of objections to such an account have been given by Russell and Quine. The first kind of objection, which makes use of little argument, maintains that the Meinongian account simply betrays a sense for reality or results in an overpopulated universe which is aesthetically displeasing. Russell attacks Meinong with respect to his sense of reality. Quine attacks him on the basis of his universe of objects being displeasing. Objections of the second kind appeal to the incomprehensibility or incoherence of Meinong’s account.

Let us first examine objections of the first kind. In “Descriptions,” Russell writes: “In such theories, it seems to me, there is a failure of that feeling for reality which ought to be preserved even in the most abstract studies” (Russell 1919, 324). Similarly, in “The Philosophy of Logical Atomism,” Russell writes:

Meinong maintains that there is such an object as the round square only it does not exist, and it does not even subsist, but nevertheless there is such an object, and when you say ‘The round square is a fiction’, he takes it that there is an object ‘the round square’ and there is a predicate ‘fiction’. No one with a sense of reality would so analyse that proposition…. I cannot bring myself to suppose it. I cannot believe that they are there in the sense that facts are there.[61]

In these quotes Russell has levied no more argument against Meinong than that he seems to have lost touch with reality. Russell’s Meinong is a pitiful example of a philosopher who has “gone off the deep end.” However, what Russell fails to understand throughout his later work is that Meinong’s work is at least no worse off with respect to its grasp on reality than his own. It makes just as little sense to say that names are shorthand for definite descriptions as it does to say that there are things that exist and there are things that do not exist. On our revised Meinongian view, however, we have been able to explain the way in which vacuous names, being just like ordinary names except for their not referring to concrete objects, succeed in directly referring to their bearers. We have been able to make sense of how it is that vacuous names are passed down along Kripkean causal chains which preserve reference. All of this has been accomplished without contradicting our intuitions about the nature of names. We have not been forced to transform names into disguised definite descriptions. We have allowed names to remain names whether or not they name existing or non-existing objects. Furthermore, we have allowed for the existence of nothing other than that which exists. Our Meinongian objects themselves do not exist. They merely are. Only those Meinongian objects that are related to existing existents find their way into the actual world, and even here it is not the objects but the existing things that exist in the world.

In “On What There Is,”[62] Quine provides a caricature of a philosopher named ‘Wyman’, who, like Meinong, maintains the existence of possible but not actualized objects. Quine says: “Wyman’s overpopulated universe is in many ways unlovely. It offends the aesthetic sense of us who have a taste for desert landscapes, but this is not the worst of it. Wyman’s slum of possibles is a breeding ground for disorderly elements” (Quine 1961, 4). Quine goes on to ask questions about how many possible fat men there are in a certain doorway. He also wonders how we ought to define identity between possible objects. However, rather than attempting to give serious consideration as to how these issues might be accounted for, Quine merely concludes that “…we’d do better simply to clear Wyman’s slum and be done with it” (Quine 1961, 4). However, our revised Meinongian universe is no more overpopulated than our own. It is possible for us to consider some possible fat man, and to wonder whether or not he and his fat friends are in our doorway, but this does not in any sense imply that there actually are any Meinongian fat men in the doorway. Our universe still only consists of those things that exist, despite the fact that we may also consider those things that do not exist. We have merely provided a framework within which to raise these issues.

Our framework, in accordance with the spirit of Carnap’s principle of tolerance, is one in which we have introduced a “system of new ways of speaking, subject to new rules” (Carnap 1956, 206). There is a difference between asking and answering questions within our framework and asking and answering questions external to our framework. We are concerned not with questions of the latter kind, for it is uncertain that they may ever be coherently answered. Rather, we are concerned with questions that may be posed and answered within our framework. This is not to maintain that every kind of entity and manner of speaking within our framework should be taken to exist or make sense external to it. To say that there are unicorns or fictional characters is not to imply that they actually exist as concrete objects in the world. Some fictional characters also exist in reality, but no truly non-existing existents actually exist. Our universe is just as clean and aesthetically pleasing as any other. No amount of metaphysical speculation will change that.

Objections of the second kind often appeal to either a violation of the law of contradiction or a refusal to differentiate between different meanings of “exists.” In “On Denoting,”[63] Russell raises an objection as to what he perceives as Meinong’s allowance of objects that violate the law of contradiction. He says: “This is in itself a difficult view; but the chief objection is that such objects, admittedly, are apt to infringe the law of contradiction. It is contended, for example, that the existent present King of France exists, and also does not exist; that the round square is round, and also not round, etc. But this is intolerable…” (Russell 1956, 45). This quotation both raises Russell’s concern about the law of contradiction and shows his inability to comprehend two different meanings of “exists.” First, Russell is worried that objects such as the round square are both round and square. This violates the law of contradiction since nothing can be both round and square inasmuch as the two are logically incompatible. However, what Russell fails to note is that it is precisely this logical impossibility that renders the round square a necessarily non-existing object on our account. While it is impossible for the round square to exist, it may still be allowed within the scope of our theory of objects. Now, as to Russell’s worry that “the existent present King of France exists, and also does not exist,” we must maintain that the only sense in which the present King of France exists is in our reformulation of things that fall under the reformulated existential quantifier (i.e. distinguishing “There are…” from “There exist…”). We have not maintained that objects like the Present King of France both exist and do not exist. Rather, we have maintained that objects such as this are existents but do not exist in the same way that other existent objects do exist. Russell’s inability to make sense of our two notions of “exists” leads to the inapplicability of his objections. He has not presented any serious problem for the theory of objects.

At this point we may reformulate our revised picture as stating:

S refers to x with the name ‘N’ in virtue of S’s place within a causal history of referring to x with the name ‘N’. Even though ‘N’ may have a connotation, it is not in virtue of this or some descriptive conditions that S refers to x with the name ‘N’. Furthermore, it is not necessarily the case that x is N. Nor is it necessary that either x or N exist. Additionally, there may be some other entity y that is also named ‘N’.

In the case of reference to God, the picture becomes:

S refers to x with the name ‘G’ (where ‘G’ is a name used to refer to God—e.g. ‘God’, ‘Adonai’, ‘Ha-Shem’, etc) in virtue of S’s place within a causal history of referring to x with the name ‘G’. Even though ‘G’ may have a connotation, it is not in virtue of this or some descriptive conditions that S refers to x with the name ‘G’. Furthermore, it is not necessarily the case that x is G. Nor is it necessary that either x or G exist. Additionally, there may be some other entity y that is also named ‘G’.

Here we begin to see the lessened role played by a speaker’s beliefs and the actual facts about a referent’s ontological status in the course of determining reference. A person who believes that something exists may refer to the same thing that someone else thinks does not.

As in the case of Santa Claus, someone who believes that he exists refers to the same thing, i.e. Santa Claus, as someone who believes that he does not. This will allow us to see how it is possible that both the Abrahamic theist and atheist succeed in referring to the same thing. While the theist believes of some x (namely God) that it exists and the atheist believes of this x that it does not exist, it is nevertheless possible and likely that they are talking about the same thing.[64] Such a realization seems to make a great deal of sense. Arguments as to the existence or non-existence of God would seem more futile than they already are if it were maintained that the atheist and the theist were necessarily referring to different things with their uses of God’s proper names. On our account, however, we can make sense of how some atheists might argue that God is necessarily non-existing (e.g., because of incompatibility of attributes, brute fact, etc) while some theists will argue that God is necessarily existing. However, the case of God is one in which it is difficult, some would say impossible (I leave the question open), to discover whether or not God is an existing existent or a non-existing existent. That there is some existent object does not mean that that object exists. For, that there is a chain between some object and some name does not entail that the object itself exists. In the case of God, that there is some Meinongian object God to which the name ‘God’ refers does not mean that that object exists.[65] Nevertheless, it is possible that we have inherited names that situate us within a direct causo-historical chain of reference to this existent.

C. Continuity of reference?

The previous two problems that we have discussed, namely the connotation of certain names like ‘God’ and the problem of vacuous reference, have been coped with through amendments to Kripke’s picture of reference. It has been shown that ‘God’ is not the only name that should give us reason to re-appraise Kripke’s work. Furthermore, we have not amended the picture in order to account for reference to God. Rather, we have seen that the picture needs to be amended with respect to all names. Now, we have not yet shown exactly how we may give a causal picture of reference to God. We have merely shown how a causal picture should not work, and then amended it in order to show how it could work. However, we must stick to the issue at hand: Can we give a causal picture of reference to God? In order to complete our inquiry, we will need to consider a third problem. This final issue concerns the question of continuity of reference. We must ask ourselves: How exactly is it that the causal relationship between me and those prior to me ensures that I have referred to some object with a name and that we have referred to the same object? While I may stand in some historical connection with prior users of some name, what is it about this connection that renders my current use of the name successful and in accord with those before me?

Knowledge of the history of our own acquisition of the name? No.

One way that we might consider making sense of the role played by the causal chain in our picture would be to say that it relies on our being able to historically trace back uses of the name to the object that was originally baptized. In the case of reference to God, we may be led to start by looking at scriptural genealogy as providing evidence of the continuous transmission of reference to God from Adam downward. It might be thought that if we can show how our own genealogy links up with these then we can be assured that reference has been preserved. We have a history of attempts to establish precisely this kind of traditional authority. For example, in the beginning of Ibn Daud’s Book of Tradition, he writes:

The purpose of this Book of Tradition is to provide students with the evidence that all the teachings of our rabbis of blessed memory, namely, the sages of the Mishna and the Talmud, have been transmitted: each great sage and righteous man having received them from a great sage and righteous man, each head of an academy and his school having received them from a great sage and righteous man, each head of an academy and his school having received them from the head of an academy and his school, as far back as the men of the Great Assembly, who received them from the prophets, of blessed memory all.[66]

Although Ibn Daud is talking about the continuity of the contents of the oral rabbinic teachings, we might be tempted to believe that this same inherited tradition would be sufficient for guaranteeing continuity of reference. For, in teaching one’s students, it seems as though along with preserving the content of the teachings one would preserve the subject of which these teachings discuss. Another attempt at establishing the reliability of the Jewish faith through an appeal to the continuity of tradition is given by Judah Ha-Levi in his Kuzari. In the Kuzari, the King of the Khazars exhorts his Jewish interlocutor to “illustrate [for me] the manner of tradition, because this (manner) must prove its trustworthiness” (Kuzari III.64). Ha-Levi presents a picture of the uninterrupted tradition of Judaism, from the prophets onward. Aviezer Tucker discusses this work in relation to Kripke’s causal theory of reference in his article, “Kripke and Fixing the References of ‘God’.”[67] In the course of his criticism of Ha-Levi’s argument, which he rightly sees as being based on an intuition resembling a Kripkean conception about names, Tucker notes that the traditions of Judaism have not persisted uninterrupted. There have been times when Passover and Rosh Hashanah were not celebrated, etc. From this, Tucker concludes that we cannot appeal to historical tradition in proving God’s existence. This seems right, however, Tucker might have made a stronger claim. Rather than basing his criticism on the fact that the chain has not been uninterrupted, Tucker might have noted that even if there were a continuous chain of reference leading from the prophets to us, this would not be sufficient to guarantee that the chain has managed to preserve reference. Even if there had never been times in which the Jewish holidays were not celebrated, we would not be able to base the success of our current reference to God on the success of past or original reference to God. No matter how extensive or conclusive our genealogies, they will be insufficient to guarantee continuity of reference. We cannot prove continuity of reference from continuous succession. Knowing that we have inherited a practice of referring is not sufficient to establish continuity of reference. For, the fact that a name has been used by many people by no means establishes that it has been used to designate or actually designated the same thing in each circumstance. Even the most “authentic” disciple could make a mistake. History and genealogy cannot guarantee successful or continuous reference.

Speaker’s intentions or Helm’s intuition? No.

Kripke anticipates our misgivings with respect to the conditions necessary for continuity of reference, but fails to give a sufficient treatment of the issue. He repeatedly shies away from this problem, staunchly maintaining that he has only given a picture. In order to guard against such divergence in reference, Kripke stipulates: “When the name is ‘passed from link to link’, the receiver of the name must, I think, intend when he hears it to use it with the same reference as the man from whom he heard it” (Kripke 1980, 97). The problem with Kripke’s picture of proper names is that names can be borrowed or mistakenly applied to things other than their original referents. That is, while I may inherit some name from a person prior to me in its causal history, I may then either take that name and consciously use it to name something else or I may take that name and accidentally use it to refer to something else with the intention of referring to the inherited referent. That is, I may inherit a name and then use it to refer to something other than that object to which it rigidly referred prior to my inheriting it. This involves a new use of a name but not a new name simpliciter. Additionally, we must consider that following my original aberrant use of the name, I may continue referring to something else with the name while still believing myself to be referring to the original referent. Others may subsequently inherit this practice of mistaken reference from me. A tangential causal chain may then arise which leads back to a second rigid designation for the same name. This new causal chain may eventually become dominant. All of this can occur without our becoming aware of the shift in reference. Kripke’s stipulation that the inheritor of a name must intend to refer to the same thing as the person before them in the causal chain is therefore insufficient to guarantee continuity of reference.

One example of this problem can be seen in Kripke’s own example of two men who observe someone raking leaves. They “think they recognize him as Jones” (Kripke 1980, 25). Following this recognition, one man says, “What is Jones doing?” The other man replies, “Raking the leaves.” Now, it turns out that the man raking leaves is actually Smith. Now, Kripke rightly points out that in this case “they are referring to Smith, even though they both use ‘Jones’ as a name of Jones” (Kripke 1980, 25). Now, strictly speaking, this is true. The two men have succeeded in talking about Smith even though they called him ‘Jones’. If they walked over to the man raking the leaves in order to help him out rather than just talking about him, they might then come to realize that they had been calling Smith by Jones’ name.

However, what about if this is the case: Our two men again see Smith raking the leaves but think that it is Jones. They still refer to Smith with ‘Jones’. However, a fourth person comes along. He’s new to town. He sees our two men standing there and asks them who the man raking the leaves is. They say, “Oh, that’s Jones.” Our fourth man then goes up to Smith and talks to him, never asking him his name. The fourth man then goes home, builds a temple, and begins worshipping the man he calls ‘Jones’. For whatever reason, this new religion takes off, becoming a competitor among world religions. Now, in this case, ‘Jones’ really does refer to Smith. A causal chain for the name ‘Jones’ then becomes established as the religion begins proselytizing, even though it really leads back to Smith. All the converts intend to refer to the same being as our fourth man, who intends to refer to the same being as the original two men. Our community is a Smith-worshipping community that calls Smith ‘Jones’. Now, what would have to occur in order for them to realize that it was Smith they were worshipping the whole time? Near the end of his book, Eternal God[68], Paul Helm addresses the question of the plausibility of a causal account of divine reference. He says:

Continuity of reference can be secured, on the one hand, by observing the continuity of whatever factors served, in the first place, to fix the reference. These factors are contingently connected to the substance or phenomenon in question. They could fail to be present, or they could be present under different conditions, in different possible worlds. In the case of God the reference-fixing factors are not fairly constant and repeatable factors, such as the awareness of certain colours under certain circumstances, or they need not be. For reference to God is fixed by reference to his actions and these reference-fixing factors might be expected to be intermittent, and certainly not repeatable in accordance with a set of laws of nature, or capable of being arranged by human beings. While this is another significant difference between reference-fixing in the case of heat, and in the case of God, it is not an insurmountable obstacle to a theory of theological reference. For there is no reason why, provided that the idea of divine activity or speech is allowed in the first place, it should not be possible to decide whether or not two or more actions or utterances on subsequent occasions are to be attributed to the same individual, God, or to some other factor or factors.[69]

Helm is right to point out the importance of providing an account of the conditions necessary to guarantee continuity of reference to God. However, an appeal to continuity of observable phenomena will not be sufficient to guarantee continuity of reference to God or any other thing.

First, given a conception of God as not being a physical entity, it seems difficult to say what exactly would need to be observed. It’s not as though we could recognize God on the basis of hair color, costume, height, facial structure, etc. What about continuity of the content of religious experience? Could we observe something continuous in religious experiences that would be sufficient for our granting that these have all been of the same thing, namely the being named ‘God’? Without getting into too detailed of a discussion of religious experience, we may merely note first that religious experiences tend to be characterized as ineffable. Inasmuch as the subject of these experiences will often claim that they are unable to describe what they experienced, it seems difficult to say what common characteristics of religious experience, other than ineffability, would be sufficient to determine whether or not these share a common object. Furthermore experiences closely resembling religious experiences can be caused by things other than God. For example, someone may have an experience that they take to be religious as a result of anesthesia or drug use. These considerations should make us wary of appealing to phenomenal similarity of experience in attempting to establish sameness of the experienced.

Second, even if our conception of God did allow for some kind of necessary physical property, an observation of this in something referred to with the name ‘God’ would not be sufficient to grant continuity of reference. Even in cases where observable phenomena remain fixed, it is difficult to rely on continuity of observed phenomena for preserving continuity of reference. Let us consider another version of the Smith-Jones problem to see why this is the case. So maybe Smith comes over to see what all the fuss is about. As soon as he walks into the temple the fourth man recognizes him and throws himself at Smith’s feet and says, “Praise be to Jones”. Smith will say, “Hey, you’ve got the wrong guy. I’m Smith, not Jones.” Here it seems possible that everyone will realize the mistake. They might then continue worshipping Smith. Either they will call stop calling Smith ‘Jones’ and begin calling him ‘Smith’ or they will retain the name ‘Jones’ and keep using it as a name for Smith. Either way they know who they are referring to. We know that our fourth man and his brethren have just been calling Smith ‘Jones’ and it seems in principle possible that everyone would come to realize their mistake.

However, is this as easily accomplished in the case of God? What if Smith never comes over, or he only comes over after he’s had plastic surgery. Would the community then ever come to realize their mistake? Say Smith comes over after having drastic plastic surgery or a sex change. He (or she) would not be recognized and would come to think that these people were worshipping our original Jones. Now suppose Smith has a twin brother, Schmith. Schmith knows what’s going on and decides he’d like to be worshipped. He comes over to the temple and everyone bows down to him and praises him, etc. Now, Smith (our physically altered Smith) might cry out, “That’s not Jones, that’s Schmith!” The worshippers might say, “Of course it’s Jones, look at him!” Smith might insist that it’s not Jones. The worshippers might then kill Smith (in this case Smith’s murderer might really be insane), saying that he has blasphemed against Jones. Now Schmith becomes the referent of ‘Jones’.

We have come to see the problems with maintaining causal reference. In cases where there have been divergent uses, we have seen that experience should lead us to realize and correct our mistakes. When we have the actual person available to point at, talk to, etc, it seems easier to establish continuity of reference. However, we have seen that it is also possible that similarity of experience does not necessarily guarantee continuity of reference. Schmith has been able to deceive everyone except Smith. Smith is dead and ‘Jones’ refers to Schmith, even though it was previously used to refer to Smith (and let’s not forget Jones). Even if later believers have experience of something that they take to be God, how are we to know that what they experienced really was God and not something else? If they are misled or duped, how can we even be sure that they have been duped in the same way as those previous to them? Other than scriptural authority, and our belief that we are referring to God, what reason do we have to maintain that all of these experiences were of the same thing or that that thing was not an impostor?[70] That is, what do we have other than belief to think that we are referring to God, or anything for that matter? Let us then restate our picture:

S refers to x with the name ‘N’ in virtue of S’s place within a causal history of referring to x with the name ‘N’. However, the causal history itself does not determine reference. S does not necessarily refer to the same x as those before them in the causal chain. Even though ‘N’ may have a connotation, it is not in virtue of this or some descriptive conditions that S refers to x with the name ‘N’. Furthermore, it is not necessary that x is N. Nor is it necessary that N or x exist. Additionally, there may be some other entity y that is also named ‘N’.

Let us now give one final formulation of our picture of divine reference:

S refers to x with the name ‘G’ (where ‘G’ is a name used to refer to God—e.g. ‘God’, ‘Adonai’, ‘Ha-Shem’, etc) in virtue of S’s place within a causal history of referring to x with the name ‘G’. However, the causal history itself does not determine reference. S does not necessarily refer to the same x as those before them in the causal chain. Even though ‘G’ may have a connotation, it is not in virtue of this or some descriptive conditions that S refers to x with the name ‘G’. Furthermore, it is not necessary that x is G. Nor is it necessary that G or x exist. Additionally, there may be some other entity y that is also named ‘G’.

While all of these facts deserve being pointed out, it must also be noted that none of these are sufficient to make continuous chains of reference to God or any other object impossible. Furthermore, Tucker is right to point us towards tradition within religion as a source for preservation of reference. Even though it has been shown that continuity of tradition is insufficient to guarantee continuity of reference, such things as tradition and holidays do serve some purpose in the preservation of reference within a faith. Performing the same rituals, saying the same prayers, and celebrating the same holidays as our predecessors certainly does serve at least some purpose in facilitating the possibility of continuous reference. Without tradition it would certainly be much more difficult or even impossible for us to base our reference to God on a historical or causal picture.

Implications and Conclusion

What, then, are the ultimate implications of looking at religion from within our Kripkean-Meinongian framework? First, Alston is right in pointing out that people generally come to learn about some thing called ‘God’ through their being initiated into a form of religious life. He says: "First, it is obvious that we, at least most of us, acquire our religion, including our practices of talking to and about God, from a community. We did not think it all up ourselves; nor were most of us privileged with special revelations from God" (Alston 1988, 119). That is to say, the name ‘God’ is passed down to us from past members of whatever religious tradition we have been born into or entered. Within the context of our religion, we inherit "the sub-practice of referring to God, of referring to the object of worship our predecessors in the community had been referring to" (Alston 1988, 119). As our predecessors and those before them ultimately gain their manner of speaking about God from the initial baptizer at the beginning of the causal chain, Alston's account seems to make sense. We see a relation between scriptural reference and our own present-day reference. Gale sees this point as one of the successes of a causal picture of divine reference. He says: “It is this reliance on an ongoing linguistic community that will have a fruitful application to the case of ‘God’, in which the linguistic community is replaced by a religious community” (Gale 1991, 6). Furthermore, it is important to note that a similar historical dependence holds for both the theist and the atheist. Neither conceives of or refers to God with the name ‘God’ (whether he takes God to be an existing or non-existing object) on his own. We are not like Ibn Tufayl’s character Hayy Ibn Yaqzan[71], who, alone on a desert island, develops a systematic theology through pure reflection, without even the knowledge of language.

While this relationship does exist, it is not sufficient to ensure that we actually succeed in referring to God. For, while we may have inherited some practice of using the name ‘God’, the fact that some historical people used it to refer to God does not mean that we do so as well. There could easily have occurred shifts of reference between their use of the term and ours. The name could have been taken and then re-applied to something else. If this turns out to be the case and we come to inherit a practice that though purporting to link directly back to the prophets actually refers to something else as the result of a shift in reference, we must conclude that we do not refer to God but to something else as ‘God’. It is not clear what conditions would have to be satisfied to ensure that our current use of the term ‘God’ refers to the same thing as the prophets’ uses of the term. This is the reason why we, like Kripke, are only able to give a picture of reference.

A second significance of an Alstonian account of divine reference is that it establishes the possibility that uses of the names of God are co-referential across Abrahamic traditions. Inasmuch as Judaism, Christianity, and Islam all recognize the Old Testament prophets and draw influence from their testimonies, it seems as though we might be able to establish that each of these three religions is in fact referring to the same thing with their different divine names. It would not be necessary that the descriptive notions associated with these names bear any role in determining their reference. Divorced of their descriptive or qualitative differences, it becomes possible that all three of these traditions refer to the same being along Kripkean lines. It is at least in principle possible that members of all three Abrahamic traditions succeed in referring to the same thing, i.e. God, despite the different specific qualitative attributions they make of Him. This contrasts with a descriptivist account of divine reference, whereby Christians, Muslims, and Jews necessarily refer to different gods in virtue of the differing propositional content of the descriptions employed. Near the end of his article, Alston notes:

[Second,] the prospects for taking radically different religious traditions to all be referring to and worshipping the same God are greatly increased. If one’s referent in religious discourse is determined by what one takes God to be like, then we, the Hindus, and the ancient Greeks and Romans cannot be credited with worshipping the same being. But if reference is determined rather by the real contacts from which a referential practice stems, then there may indeed be a common referent, in case these traditions, including their referential traditions, all stemmed from experiential contacts with the one God.[72]

However, it would be untoward for us to assume some idea that every religion refers to the same thing merely as a result of some neat causal chain leading back in time. While direct reference theology accounts for the possibility that different faiths can be worshipping the same thing, it should not claim that this necessarily occurs. Claiming a shared historical connection with earlier referrers is not sufficient to establish that continuity of reference has been maintained. Neither our intending to refer to the same being as someone prior to us in a causal chain nor our ability to present referential genealogies are sufficient to guarantee successful or continuous reference. As the practice of using these names to refer to the God of the prophets has been passed down through the community of worshippers, it is possible that any person’s use of these will have the same referent. Whether this does in fact occur can only be realized through an investigation of how reference is maintained. Insofar as it has been established as possible that members of the three Abrahamic religions (and different persons of different sects and theological sophistication therewithin) could be referring to the same thing (whether or not it is an existing thing), there is room for both modest ecumenism and intense exclusivism. For, one may be inclined to either allow different persons to make different descriptive attributions of the same thing to which he too refers or one may be inclined to vehemently reject the tenets of other faiths and take issue with the fact that these conflicting faiths attribute descriptive properties to one’s own God. Similarly there will be the problem of conflicts between atheists and theists who nevertheless refer to the same thing.

While it seems prima facie more likely that members of the Abrahamic traditions could be talking about the same thing, it is also possible that members of other religious traditions could be talking about the same God. This would be a case similar to the example of two entirely independent cultures coming up with different names for the same mountain.[73] It is unclear what would have to be the case for us to realize that e.g. the being we call ‘Adonai’ is the same being as the one that historical Norse or Germanic peoples called ‘Odin’. However, we have no reason to exclude this possibility altogether.

A third success of an Alstonian account of divine reference is that it establishes the fact that the sophisticated theologian and the common believer can nevertheless refer to the same thing. Inasmuch as both the sophisticated theologian and the common believer could have inherited their practice of referring to God from the same source, or from different sources that eventually link up with the same being, it is possible for them to refer to the same thing through their use of the names of God. Again, however, whether or not two actual people (one a sophisticated theologian, the other a common believer) refer to the same thing and whether that thing is in fact God is a different matter altogether.

Let us now look at the negative implications of the Kripkean-Meinongian picture of divine reference. First, it has not been shown that this existent actually exists. While we may stand in a causal relationship with some existent, and this existent may even be the same as that referred to by the prophets, the mere fact that someone prior to us referred to it is not sufficient to ensure that the existent actually exists. This is the problem with historical and ontological proofs of God’s existence.

Secondly, while it has been shown that it is possible that we succeed in referring to something as ‘God’ through our causal acquisition of this practice, it has not been shown that the existent we refer to as ‘God’ is the same existent that those before us referred to as ‘God’. There could have occurred divergences in reference between the first reference and our own. While we may use the same names as the prophets and the church fathers, that does not necessarily mean that we refer to the same thing as them. Similarly, while people of other faiths or other degrees of theological sophistication may use the same names as us, that does not necessarily mean that we are talking about the same thing. While the names could have a common origin, there may have occurred a shift in reference at some moment along the chain. This divergence may have occurred at a point of syncresis or appropriation, or even at some less obvious point of history. We cannot be sure if and when these shifts might have occurred. For this reason members of one faith may claim that members of another have gone astray. For example, a Jewish person might claim that even though he and a Moslem both take themselves to be referring to the God of Abraham, only the causal chain leading to himself and his brethren actually successfully leads back to God whereas that of the Moslem relies on some divergence. A Protestant might say this about a Catholic, etc.

Finally, given these two implications, it is impossible to say at this point that even if we take for granted the fact that the prophets did in fact experience and refer to an existing existent we succeed in referring to it as well. Even if the prophets experienced an existing God that they then referred to, it is not certain that this means that we do as well. It may be the case that although the prophets experienced the existing God we do not succeed in referring to Him despite our use of the same divine names. A parallel example will illustrate this point nicely. Given that the name ‘Santa Claus,’ as a variation of ‘Saint Nicholas,” did at some point in time refer to some existing saint, this is not enough to prove that our current use of the name refers to an existing existent. A child might think Santa Claus exists and, if he were told about the life of Saint Nicholas, might even believe that the same saint is his current benefactor. The child might therefore take himself to be referring to the same thing as the original users of the name ‘Santa Claus’. That the child thinks this does not mean that Santa Claus exists. We would be confused if it did. For, we take it for granted that Santa Claus does not exist. Even if the child believes that he is referring to the same thing as those who were in direct contact with the historical saint, that does mean that he actually succeeds. There has occurred a shift in reference somewhere along the chain that the child is ignorant of. We see that while some people used the name ‘Santa Claus’ to refer to an existing existent (the historical St. Nicholas) and our possession of the name ‘Santa Claus’ (in addition to its connotation) is in fact dependent on this, we nevertheless do not refer to an existing existent. It is possible that while the prophets used the names of God to refer to an existing existent, we do not, even though we might think we do. For, we might refer to something different altogether.

From all of this we can see that while our consideration of certain contributions to the philosophy of language and metaphysics can in fact help us to make better sense of what exactly might be going on when we talk about God, this is insufficient to prove a) that God exists, b) that we refer to God with the name ‘God’, c) that we refer to the same thing as the prophets, d) that people of every Abrahamic faith refer to the same thing, and e) that the sophisticated theologian and the common believer refer to the same thing. All of these things are possible, and we can now give an account of how they would be the case, but we are in no position to know for sure. The causal picture can only give us a better framework within which these issues can be coherently discussed.

Bibliography:

1. Alston, William. “Referring to God,” Philosophy of Religion 24 (1988), pp. 113-28.

2. Alston, William. Perceiving God: The Epistemology of Religious Experience. Cornell University Press. Ithaca, 1991.

3. Bochenski, Joseph M.. The Logic of Religion. New York University. New York, 1965.

4. Byrne, Peter. “Response,” Referring to God. ed. Paul Helm. St Martin's Press, 2000.

  1. Carnap, Rudolf. “Empiricism, Semantics, and Ontology.” Printed in Meaning and Necessity. University of Chicago. Chicago, 1956.

6. Crane, Tim. “Intentional Objects,” Ratio 14, 2001, 336-349.

7. Donnellan, Keith S. “Reference and Definite Descriptions,” The Philosophical Review 77 (1966), pp. 281-304.

8. Donnellan, Keith S. “Speaking of Nothing,” The Philosophical Review, Vol. 83, No. 1. (Jan., 1974), 3-31.

9. Frege, Gottlob. “Letter to Jourdain, Jan. 1914.” The Frege Reader, ed. Michael Beaney. Blackwell. Oxford, 1997.

10. Friend, Stacie. “Real People in Unreal Contexts.” Empty Names, Fiction, and the Puzzles of Non-Existence. Ed. Anthony Everette and Thomas Hofweber. CSLI Publications. Stanford, 2000.

11. Richard Gale. On the Nature and Existence of God. Cambridge University Press. Cambridge, 1991.

12. Gellman, Jerome. “The Name of God,” Nous Vol. 29, No. 4 (December 1995), pp. 536-43.

13. Gellman, Jerome. “The Metaphilosophy of Religious Language,” Nous Vol. 11, No. 2 (May 1977), pp. 151-61.

14. Ha-Levi, Judah. Kitab Al-Khazari. Trans. Hartwig Hirschfeld. Bernard G. Richards Co. New York, 1927.

15. Helm, Paul. Eternal God. Clarendon. Oxford, 1988.

  1. Ibn Daud. The Book of Tradition. Trans. by Gerson D. Cohen. Routledge. London, 1967.

  1. Ibn Tufayl. Hayy Bin Qazan. Contained in Medieval Islamic Philosophical Writings. Ed. Muhammad Ali Khalidi. Cambridge. Cambridge, 2005.

18. Jeffrey, Andrew V. “Gale on Reference and Religious Experience,” Faith and Philosophy Vol. 13, No. 1 (January 1996), pp. 91-112.

19. Kripke, Saul. Naming and Necessity. Harvard University Press. Cambridge, 1980.

20. Meinong, Alexius. “The Theory of Objects,” Untersuchungen zur Gegenstandstheorie und Psychologie, 1904. Reprinted and translated in Roderick M. Chisolm, Realism and the Background of Phenomenology. The Free Press. Glencoe, 1960.

21. Mill, John Stuart. A System of Logic. Longman’s, Green and Co. London, 1906.

22. Miller, Richard B. “The Reference of ‘God’” Faith and Philosophy Vol. 3, No. 1 (January 1986), pp. 3-15.

23. Parsons, Terence. Nonexistent Objects. Yale University. New Haven, 1980.

24. Parsons, Terence. “A Prolegomenon to Meinongian Semantics,” The Journal of Philosophy. Vol. 71, No. 16 (Sep., 1974), pp. 561-580.

25. Parsons, Terence. “The Methodology of Nonexistence,” The Journal of Philosophy. Vol. 76, No. 11, (Nov., 1979), pp. 649-662.

26. Parsons, Terence. “A Meinongian Analysis of Fictional Objects,” Grazer Philosophische Studien 1 (1975), pp. 73-86.

27. Quine, W. V. O. “On What There Is.” From A Logical Point of View: Logico-Philosophical Essays. 2nd Edition. Harper Torchbooks. New York, 1963.

28. Russell, Bertrand. “The Philosophy of Logical Atomism.” Logic and Knowledge. Ed. Robert C. Marsh. George Allen and Unwin. London, 1956.

29. Russell, Bertrand. “Descriptions.” First appeared in Introduction to Mathematical Philosophy. George Allen and Unwin. London, 1919. Reprinted in Readings in the Philosophy of Language. Ed. Peter Ludlow. MIT Press. Cambridge, 1997.

30. Russell, Bertrand. “On Denoting.” First appeared in Mind 14. 1905. 479-93. Reprinted in Logic and Knowledge. Ed. Robert C. Marsh. George Allen and Unwin. London, 1956.

31. Ryle, Gilbert, "Meaning and Necessity," Philosophy, 24 (1949), 69-76.

32. Sprigge, Timothy. “Internal and External Properties,” Mind New Series, Vol. 71, No. 282 (Apr., 1962), pp. 197-212.

33. Strawson, P. F.. “On Referring,” Mind 59.235 (July 1950), pp. 320-344.

34. Tucker, Aviezer. “Kripke and Fixing the References of ‘God’,” International Studies in Philosophy Volume XXXIV, 4. 2002.

35. Walton, Kendall. “Existence as Metaphor?” Empty Names, Fiction and the Puzzles of Non-Existence. ed. Anthony Everette and Thomas Hofweber. CSLI Publications. Stanford, 2000.

36. Zalta, Edward N. “The Road Between Pretense Theory and Abstract Object Theory.” Empty Names, Fiction, and the Puzzles of Non-Existence. ed. Anthony Everette and Thomas Hofweber. CSLI Publications. Stanford, 2000.

37. Ziff, Paul. “About God.” Contained in Religious Experience and Truth. Ed. Sidney Hook. New York University. New York, 1961.



[1] For the purposes of this paper, we will be treating the word ‘God’ as a name rather than a concept or title. When a name appears in single quotes (e.g. ‘God’, ‘Smith’, ‘Holmes’, etc) we will be discussing the name itself. When a name appears without single quotes, we will be discussing the thing taken to be designated by that name (e.g. God, Smith, Holmes, etc).

[2] William Alston. “Referring to God,” Philosophy of Religion 24 (1988), pp. 113-28.

[3] A position such as this would take Russell and Searle’s work on names as its starting point.

[4] The variety of linguistic stances toward the divine has been used as evidence for strong atheism. As Cicero notes: “The manifest disagreement amongst the most learned on this subject creates doubts in those who imagine they have some certain knowledge of the subject” (Cicero On the Nature of the Gods, I.VI).

[5] For an example of this kind of an account of religious language, see Paul Ziff’s essay, “About

God.” A more rigorous version has been given by Joseph M. Bochenski in his The Logic of

Religion. See especially pp. 53-68, which culminate in Bochenski’s statement:

“[T]he term “God,” as used by the bulk of today’s believers, is a description. It is an abbreviation

for a substitution in the formula

(x){jx}

where j is substituted by the product of the predicates attributed to God by the creeds

concerned” (Bochenski, 68).

[6] Alston 1988, 114.

[7] e.g. “Most glorious of immortals, Zeus”; The Most-Ancient, (and) ever O Mazda, (as) the Youngest”; “God, supreme in loyalty” (Cleanthes “Hymn to Zeus”; Yasna XXXI, 8; Rumi Masnavi I: 2200).

[8] e.g. the many epithets of Zeus, the 101 names of Ahura Mazda, the 99 names of Allah.

[9] Saul Kripke. Naming and Necessity. Harvard University Press. Cambridge, 1980.

[10] For an understanding of what Alston takes to count as an experience of God, see his Perceiving God: The Epistemology of Religious Experience. Cornell University Press. Ithaca, 1991.

[11] Of course none of the Old Testament prophets would have used the English word ‘God’ to refer to the being they experienced. However, instead of using the original Hebraic words, we will use the English translations and transliterations of the names of God. We will follow the same standard for other names, e.g. using ‘Socrates’ rather than ‘Σωκράτης’.

[12] This is inconsistent with scripture. We are never presented with any act of naming God prior to acts of referring to Him as ‘God’ in either the Pentateuch or the Qur’an. In the Pentateuch Adam names all the creatures he encounters but never names God. The first reported reference to God is actually made by the serpent in conversation with Eve. In Chapter 2 (“The Cow”) of the Qur’an, Adam names nothing but is rather taught the names of things by God.

[13] See Keith Donnellan. “Reference and Definite Descriptions,” The Philosophical Review 77 (1966), pp. 281-304.

[14] Peter Byrne. “Response,” Referring to God. ed. Paul Helm. St Martin's Press, 2000.

[15] “Alston operates on a very narrow base, using only one source of causal notions of reference (Kripke). Yet causal theories of reference have been developed by a number of writers subsequent to Kripke…” (Byrne 2000, 63).

[16] Jerome Gellman. “The Name of God,” Nous Vol. 29, No. 4 (December 1995), pp. 536-43.

[17] With respect to his claim about the baptism of God as “God,” Gellman’s reading of scripture is equally suspect as Alston’s. Gellman too claims that the being referred to as ‘God’ was named at the beginning of that name’s causal chain, in someone’s experience of Him (Gellman 1995, 536).

[18] Gellman notes that these conditions are borrowed from Robert Nozick.

[19] In laying out the conditions for naming God, Gellman either departs from or contradicts his view in “The Metaphilosophy of Religious Language.” In this earlier work, Gellman asserts that a theory of religious language is adequate if it accounts for the way religious language is actually used (Gellman 1977, 151). “The Meta-Philosophy of Religious Language,” Nous Vol. 11, No. 2 (May 1977), pp. 151-61.

[20] P. F. Strawson. “On Referring,” Mind 59.235 (July 1950), pp. 320-344.

[21] Kripke 1980, 162.

[22] Richard Gale. On the Nature and Existence of God. Cambridge University Press. Cambridge, 1991.

[23] Gale 1991, 7.

[24] John Stuart Mill. A System of Logic. Longman’s, Green and Co. London, 1906.

[25] Mill 1906, 20-1.

[26] Plutarch, Life of Cicero.

[27] For another example of names that no longer require their bearers to satisfy their connotation, we may consider patronymic names which no longer require that their bearers stand in the relationship of “being the son of x”. For example, someone can have the surname ‘Milošević’ without being the son of somebody named ‘Miloš’.

[28] On my account, names that do have some connotative quality, e.g. ‘Davidson’ or ‘Cooper’ can directly refer to multiple people, even if neither satisfies the description called to mind. That is, two people can be named ‘Davidson’ even if neither is the son of someone named ‘David’ or two people can be named ‘Cooper’ even though neither is a barrel maker. It is not the case that Cooper = Cooper is analytically true, since we can have more than one Cooper without having more than one ‘Cooper’-name. It could be the case that Cooper = Cooper and Cooper ≠ Cooper are both true. It is therefore informative to learn that Cooper = Cooper and that Cooper = Henry (if ‘Henry’ is one of Cooper’s nicknames).

[29] This is a metaphysical issue that we will not be able to concern ourselves with here.

[30] Bertrand Russell. “The Philosophy of Logical Atomism.” Logic and Knowledge. Ed. Robert C. Marsh. George Allen and Unwin. London, 1956.

[31] Russell 1956, 242.

[32] Russell 1956, 243.

[33] Bertrand Russell. “Descriptions.” First appeared in Introduction to Mathematical Philosophy. George Allen and Unwin. London, 1919. Reprinted in Readings in the Philosophy of Language. Ed. Peter Ludlow. MIT Press. Cambridge, 1997.

[34] E.g., Napoleon’s appearance in Tolstoy’s War and Peace.

[35] An example of the first kind would be Sherlock Holmes in multiple works by Conan Doyle. An example of the second kind would be Sherlock Holmes in Billy Wilder’s film, “The Private Life of Sherlock Holmes.”

[36] A clever example of this kind is the fictional Icarus created and then lost by another character, the writer Hubert in Raymond Queneau’s The Flight of Icarus.

[37] E.g., Agilulf, the protagonist of Italo Calvino’s The Nonexistent Knight.

[38] Kendall Walton. “Existence as Metaphor?” Empty Names, Fiction and the Puzzles of Non-Existence. ed. Anthony Everette and Thomas Hofweber. CSLI Publications. Stanford, 2000.

[39] Walton 2000, 70-1.

[40] Edward N. Zalta. “The Road Between Pretense Theory and Abstract Object Theory.” Empty Names, Fiction, and the Puzzles of Non-Existence. ed. Anthony Everette and Thomas Hofweber. CSLI Publications. Stanford, 2000.

[41] For a similar objection see also Stacie Friend’s “Real People in Unreal Contexts.” Empty Names, Fiction, and the Puzzles of Non-Existence. Ed. Anthony Everette and Thomas Hofweber. CSLI Publications. Stanford, 2000.

[42] Andrew V. Jeffrey. “Gale on Reference and Religious Experience,” Faith and Philosophy Vol. 13, No. 1 (January 1996), pp. 91-112.

[43] Keith S. Donnellan. “Speaking of Nothing,” The Philosophical Review, Vol. 83, No. 1. (Jan., 1974), 3-31.

[44] Jeffrey 1996, 101.

[45] Chronologically speaking, Richard Miller’s article “The Reference of ‘God’” is the first known attempt to apply Kripke’s work in the philosophy of language to the case of religious language. He says that his interest in such an application began with his “transforming the question “Does God exist?” into “Does ‘God’ refer?”” (Miller 1986, 3). Miller also works out the problem of referring to God even though ‘God’ may be a vacuous name by recourse to “blocks”.

[46] Donnellan’s term for what we’ve been calling “causal”.

[47] Donnellan 1974, 25.

[48] Donnellan 1974, 25.

[49] Donnellan 1974, 26.

[50] Rudolf Carnap. “Empiricism, Semantics, and Ontology.” Printed in Meaning and Necessity. University of Chicago. Chicago, 1956.

[51] See: Nonexistent Objects, Yale University. New Haven, 1980; “A Prolegomenon to Meinongian Semantics,” The Journal of Philosophy. Vol. 71, No. 16 (Sep., 1974), pp. 561-580; “The Methodology of Nonexistence,” The Journal of Philosophy. Vol. 76, No. 11, (Nov., 1979), pp. 649-662; and “A Meinongian Analysis of Fictional Objects,” Grazer Philosophische Studien 1 (1975), pp. 73-86.

[52] Alexius Meinong. “The Theory of Objects,” Untersuchungen zur Gegenstandstheorie und

Psychologie, 1904. Reprinted and translated in Roderick M. Chisolm, Realism and the Background of Phenomenology. The Free Press. Glencoe, 1960.

[53] Meinong’s theory of objects is in some ways similar to Aristotle’s theory of the objects of knowledge that must in some sense be in order for us to know that they do not exist. Aristotle’s account includes things like goat-stags and chimaeras. See Prior Analytics 1.38. Aristotle will later stipulate that strictly speaking objects of knowledge such as the goat-stag cannot really be known. See Posterior Analytics II.7. This concept of objects of knowledge will later be discussed by figures such as Averroes, Aquinas, Duns Scotus, Ockham, Francisco Suarez, and others.

[54] By “ordinary proper names” I mean proper names of concrete objects and things that we would all agree actually exist, i.e. proper names of what will become existing existents. By “vacuous proper names” I mean proper names of non-existing objects, i.e. proper names of what will become non-existing existents.

[55] This is the view, criticized by Gilbert Ryle, that for every name and proposition there is some existing entity that is the bearer of that name or proposition. Our view will claim that at least for every use of a name there is some existent (but not necessarily existing) entity that is the bearer of that name. See Gilbert Ryle, "Meaning and Necessity," Philosophy, 24 (1949), 69-76.

[56] Parsons 1980, 121.

[57] Consider Timothy Sprigge’s discussion of whether or not “the particular we call the Queen” could possibly have had the property of not being human. He says: “The internalist suggests that we cannot imagine that particular we call the Queen having the property of at no stage in her existence being human. If the antiinternalist admits this, admits that it is logically inconceivable that the Queen should have had the property of, say, always being a swan, then he admits that she has at least one internal property. If on the other hand he says that it is only a contingent fact that the Queen has ever been human, he says what it is hard to accept. Can we really consider it as conceivable that she should never have been human?” (Sprigge 1962, 203).

[58] Recall Kripke’s consideration of the fact that from an individual born of certain parents in one world cannot be born of others in a different world. That is, the sperm and egg out of which an individual is formed can only come from two specific people. (Kripke 1980, 110-13).

[59] Parsons 1974, 568.

[60] An adult version of this example may be given through a consideration of the fictional poet Ossian.

[61] Russell 1956, 223.

[62] W. V. O. Quine. “On What There Is.” From A Logical Point of View: Logico-Philosophical Essays. 2nd Edition. Harper Torchbooks. New York, 1963.

[63] Bertrand Russell. “On Denoting.” First appeared in Mind 14. 1905. 479-93. Reprinted in Logic and Knowledge. Ed. Robert C. Marsh. George Allen and Unwin. London, 1956.

[64] Tim Crane gives an example of such a case in his “Intentional Objects.” He says: “Consider a debate between an atheist and a theist, and suppose for the sake of argument that the atheist is right: God does not exist. (Let’s suppose that the debate they are having is a straightforward one over the existence of the Christian God as traditionally conceived: the all-powerful creator of the universe who loves us as a father loves his children etc.) If the atheist is right, then the theist has been talking about (and thinking about) something which does not exist. Yet the theist’s words made sense, it seemed that he was able to put these thoughts about God into words. His thoughts are thoughts about something that does not exist.” Tim Crane. “Intentional objects,” Ratio 14, 2001, 336-349.

[65] As Parsons notes in a discussion of the applicability of his work to the ontological argument: “According to others (e.g. Kant), existence isn’t a property because it is trivial (it adds nothing) to say of something that it exists. This view would appear to identify ‘there is something such that…’ with ‘there exists something such that…’. If this identification were correct, and if we could establish that there is a god, then we would automatically have established that a god exists. This would actually vindicate part of the ontological argument. Unfortunately, no historical version of the argument successfully establishes that there is a god. The theory of objects does assert this, but it can’t be combined with the view in question to show that a god exists, for it rejects the triviality of existence statements” (Parsons 1980, 216).

[66] Ibn Daud, 3.

[67] Aviezer Tucker. “Kripke and Fixing the References of ‘God’,” International Studies in Philosophy Volume XXXIV, 4. 2002.

[68] Paul Helm. Eternal God. Clarendon. Oxford, 1988.

[69] Helm 1988, 212-3.

[70] E.g. In Genesis 26: 24 God says to Isaac: “I am the God of Abraham thy father.” In Genesis 28: 13 God says to Jacob: “I am the LORD God of Abraham thy father, and the God of Isaac.” In Exodus 3: 6 God says to Moses: “I am the God of thy father, the God of Abraham, the God of Isaac, and the God of Jacob.” These quotes can be read as instances of God informing successive prophets that the thing they are experiencing is the same thing that they have inherited a process of referring to from their ancestors. We have nothing other than the authority of scripture to make us believe that this is actually the case. An Alstonian impostor could have represented itself to any one of the prophets without our knowing. This impostor could have made claims such as those cited above.

[71] Ibn Tufayl. Hayy Bin Qazan. Contained in Medieval Islamic Philosophical Writings. Ed. Muhammad Ali Khalidi. Cambridge. Cambridge, 2005.

[72] Alston 1988, 126.

[73] See Gottlob Frege’s “Letter to Jourdain, Jan. 1914.” The Frege Reader, ed. Michael Beaney. Blackwell. Oxford, 1997.

2 Comments:

Anonymous Sadjed Tayebi said...

Hi Dear Sam,
I read your paper "Reference and Religion", and I found it so intresting and precise and I think I will send you some comment as soon as possible. I also want to refer to it in a discussion about Language of Religion in our Department. Could I possibly ask you to tell me whether you have published it and where? In addition for citing your name I think I should know your full name. Please help me.
Thanks a lot

1:33 AM  
Anonymous Sadjed Tayebi said...

Hi again,
I forgot to write my email address: s_tayyebi@yahoo.com
I appreciate your help

12:34 PM  

Post a Comment

<< Home