Kripke’s claptrap on proper names — 3

by pieterseuren

Reference is a question of inference. It involves (a) the speaker’s intention to focus on a specific world entity, the reference object, and (b) the listener’s correct reconstruction and taking over of that intention. It works because we are conspecifics, betting on the similarity of our cognitive processes. Reference is not part of the semantics of a language but results from the use of the language in question. Any language must be such that reference is possible, in conjunction with the necessary cognitive structures and processes. All this sounds a bit trite, yet in theoretical semantics, especially as practised by philosophers and/or logicians, this trvial truth has still not been discovered.

Reference, as thus defined, allows for a certain amount of play. Suppose someone tells you a story about an imaginary professor Nereus, of the Camp Klanx Institute in a town called Jegnenim in a country called Donhall, and in his story he says that this man produces drivel. Has he or has he not referred to the really existing professor Seuren of the Max Planck Institute in Nijmegen, Holland—that is, to me? Well, yes and no. A keen listener or, rather, reader will soon see through the anagrams and infer the storyteller’s malicious intent, keying the story to me. Can I accuse the man of telling lies about me (supposing, of course, that I do not produce drivel)? Not really, because he can always coyly fall back on the formula that all likeness to existing persons is purely accidental. Kripke’s analysis does not allow for such reference play: the name Seuren refers to me (in the present context) and that’s it. What the name Nereus refers to remains unclear in his analysis: his theory of proper names does not allow for the naming of nonexistent entities and thus blocks any possibility of referring to them.

This brings us to one further grave objection to Kripke’s analysis: it does not allow for reference to nonexistent entities. Yet when using natural language we, humans, refer to nonexistent entities and give them names all the time. We can do so because we, wittingly or unwittingly, invent nonexistent entities, for whatever reason. They may occur in our dreams or we may believe in them, or hypothesize that they exist, we may plan to realize them in the actual world, we may want to scare others with them, or build a career on them, and so on. Having invented them, we naturally quantify over them, and can even formulate true propositions about them. Thus, when I say The Greeks had nine Muses, what I say is true. Likewise it is true to say Apollo was worshipped on the island of Delos, but false to say Poseidon was worshipped on the island of Delos, because it is historical fact that Apollo was, and Poseidon was not, worshipped on Delos in ancient times—even though neither the Muses nor Apollo nor Poseidon have ever existed.

In contrast to traditional European philosophy, most of Anglo-Saxon metaphysics adamantly excludes the notion of nonexistent, often called ‘possible’, entities. This is probably correct in so far as one tries to construct a picture of what the world ‘out there’ is really like ‘an und für sich’, in Kant’s terms. Nonexistent entities are, by definition, excluded from such a ‘de re’ metaphysics. But these philosophers forget (a) that all humans naturally and involontarily, and throughout their lives, construct ideas of what the world is or could be like and see if those ideas match what can reasonably be taken to be the case, (b) that, while doing so, they hypostatize their thought-up virtual entities, that is, they bestow virtual reality on them, so that they can think and talk about them, and (c) that, therefore, it must be possible, in the use of natural language, to refer to nonexistent entities. I discuss this problem extensively in Chapter 2 of my Language in Cognition of 2009, and the answer I give there is that there is no other solution than to accept that human cognition makes an ontological distinction between actually existing and virtual entities, whereby often the question arises of whether the latter do or do not actually exist—that is, are instantiated in the actual world. Even if one rejects the reality of virtual entities in a (necessarily speculative) ‘de re’ ontology, one is forced to accept that the human construal of the world, which is what we have to do with when we study natural language, gives rise to virtual, nonexistent entities. Indeed, no such construal would be possible without them.

Quine wrote a famous and enormously influential tirade against ‘possible entities’ in the first few pages of his From a Logical Point of View of 1953, a tirade that owes its fame and influence mainly to its thundering rhetoric. Yet it fizzles out when one realizes that nonexistent or ‘possible’ entities only have a place in the construal of reality as performed by humans (according to an innate programme, to be sure), not, of course, in reality itself as it may be assumed to be regardless of any human construal. Quine, and many others with him, failed to appreciate this distinction. He applied the identity criteria taken to be valid for actually existing entities to virtual entities, which made it easy to reduce the latter to absurdity. This would have been relatively harmless had he not applied his ‘de re’ metaphysics to natural language, just as formal semanticists apply their ‘de re’ logic to natural language, forgetting that language is grafted on cognition, not on ‘de re’ metaphysics or logic. This basic error reduces his entire tirade, like so many other tirades of his, to hot air.

The question of reference to nonexistent entities stands out with particular clarity when proper names are taken into account, and even more so in the context of Kripke’s analysis of them. One cannot ‘baptize’—not even in Kripke’s sense of the word—entities that do not exist. Yet one can worship them, fear them, look for them, believe in them, think, dream, or talk about them, one can plan or design them (as when I design a house yet to be built), one can postulate them hypothetically—and one give them names. The question is: what is it that does not exist but that I worship, believe in, dream about, plan or assign a name to? My answer is, in principle, that when reference is made to nonexistent entities, then the reference objects are nonexistent or virtual entities and when reference is made to existing entities, then those are the reference objects (see Chapter 5 in my Language in Cognition, OUP 2009, for more detailed discussion). This applies also to identity statements of the kind Descartes is Cartesius or Zeus is Jupiter, the former involving reference to really existing and the latter to virtual entities. The identity predicate be says that the mental representations of the reference object(s) of its two terms are to be merged into one representation giving rise to one virtual entity, so that any instantiation in the actual world will result in one actual entity. And exactly the same applies to identity statements containing DRPs, such as The man over there is the psychiatrist who lives on Main Street.

Language reflects mental structures and processes, not the world as it is taken to be by philosophers engaging in ‘de re’ metaphysics. Language has predicates with their argument terms (subject, object). Predicates presuppositionally mark their arguments as being either extensional or intensional: an extensional argument requires actual existence of the reference object for possible truth; an intensional argument does not and thus allows for the virtual reality of the reference object. A predicate like hit is extensional with regard to both its subject and object arguments (If John hit Bill is true, then both John and Bill must have actual existence), but predicates like worship, dream about, or call by the name “Z” are, though extensional with regard to their subject term, intensional with regard to their object term (if John worships Hanuman is true, then John must actually exist, but Hanuman need not). If this is so, then, again, the ‘baptism’ theory must be given up and replaced with the much less ambitious analysis according to which a proper name Z stands for ‘the x such that x is called “Z” ’, because one can give names to nonexistent objects but one cannot baptize them, other than in a metaphorical sense. … to be continued