of demonstrative identification, arguing that even if an individual cannot itself be demonstratively identified, “It may be identified by a description which relates it uniquely to another particular which can be demonstratively identified.”6 Much of this book is an examination—and rejection—of that claim. In this chapter, I shall address the problem of how a story can introduce characters indefinitely, for example as “a” young man, or “a” woman with a flow of blood, yet go on to refer to these characters using anaphoric chains of co-referring definite or singular terms.
How can a definite term co-refer with an indefinite term? A definite term can refer back 7 to an indefinite one.
There was a young man. The man was wearing a linen cloth.
But what about that initial description or name? What term does it refer back to, and what individual does it refer to, simpliciter? In the previous chapter, I provisionally defined co-referring terms as those whose meaning requires that, if they have a referent at all, they have a common referent. So it seems that the terms “a young man” and “the man” have a common reference, assuming the young man existed at all. But can an indefinite term like “a young man” refer?
Transportability
I shall show that singular terms in a story-relative context, even proper names, do not have a meaning that is transportable.8 By a transportable meaning, I mean one that can belong to any token, regardless of context or order. Consider
There was a young man. He was wearing a linen cloth.
The two sentences imply that some young man was wearing a linen cloth. Can any other term in the narrative have the same meaning as the pronoun “he”? Certainly, if the term occurs later in the natural order of reading. It will co-refer in some sense, and is anaphorically connected with the antecedent indefinite description “a young man.” But no term that comes before the antecedent can have that meaning, for the whole purpose of an indefinite term is its indifference to what comes before: “a young man” means any young man, not necessarily a specific man already mentioned. That is:
He was wearing a linen cloth. . . . There was a young man.
does not imply that some young man was wearing a linen cloth. Perhaps “he” was an old, or a middle-aged man, or just some other man. But the inference depends on the meaning of the pronoun, any term with that meaning will validate the inference, yet no term prior in the order of reading can have such a meaning. Pronouns cannot refer forward to an antecedent term in the way that they refer back: the back reference is a part of their meaning, therefore the meaning is not transportable as I have defined it.
It may be objected that in English, as well as Latin and other languages, a pronoun can anticipate a postcedent that occurs later in the text. For example: “In his Enquiry Concerning Human Understanding, Hume raises doubts about our knowledge of necessary connection,” where “his” anticipates “Hume.” Or in Latin: Is qui bene se exprimit hoc scripsit, where “is” anticipates “qui.” In reply, in such cataphoric co-reference, the semantics of the definite cataphoric term is in suspension, rather like a musical suspension, until the postcedent is identified, at which point the co-reference is understood, and the meaning of the cataphor is clear. Thus, the meaning of the cataphor is not transportable to a position before the postcedent, even if the term itself is transportable.9
Such non-transportability is obvious in the case of the pronoun, but the same is true of proper names. The sentences
There was a young man called “Mark.” Mark was wearing a linen cloth.
together, and in that order, also imply that some young man was wearing a linen cloth. The proper name has some semantic connection to the indefinite antecedent “a man called ‘Mark’” that licenses the inference. But this is no longer valid if the name occurs before the antecedent. For example
Mark was wearing a linen cloth. . . . There was a young man called “Mark.”
Tokens of the same proper name may have different meanings, and the indefinite “a man called ‘Mark’” could be introducing us to another person with the same name. For example, Acts 12:12:
He went to the house of Mary the mother of John, also called Mark
Is this the same as Mark the evangelist? This idea is suggested by 1 Peter 5:13, where Peter says “She who is in Babylon, chosen together with you, sends you her greetings, and so does my son Mark,” but this is not a logical inference, but a probable or possible one. Proper names are essentially ambiguous, unless their meaning is resolved by context. In the case given earlier, the context is the indefinite antecedent, which can only occur before in the sequence of reading, never afterwards. In a story-relative context, not even proper names have a transportable meaning.
It could be objected that the singular anaphor term could be replaced by an indefinite noun phrase that included the content of the antecedent.10 Then we could translate the two sentences
(1) A young man was wearing a linen cloth. He ran away.
as something like
(2) A young man was wearing a linen cloth. A young man who was wearing a linen cloth ran away.
The definite noun phrase “he” is replaced by the indefinite noun phrase “a young man who was wearing a linen cloth.” This yields the required inference: the content of the conclusion is identical to the content of the second premiss. But this is problematic for a number of reasons. First, we have to suppose that every singular term in the anaphoric chain means the same as an indefinite term containing everything that was previously asserted in the chain. So the term “Moses” in “Moses was a hundred and twenty years old when he died” is equivalent to “a man who was born into the tribe of Levi, who was abandoned by his mother, taken in by Pharaoh’s daughter, killed an Egyptian etc.” This is wholly implausible. The term “Moses” does not mean that, and the sentence itself does not contain that information.
Second, what if there is another such man?
(3) A young man was wearing a linen cloth. Another young man who was wearing a linen cloth ran away.
The second sentence of (3) includes the content of the second sentence of (2), which is supposed to signify that the man, that is, he, is the same man, but the term “another” says that the man is different. It would be like saying “another the same man”!
The same line of reasoning shows that the meaning of unique definite descriptions cannot be captured by a unique indefinite description. Consider:
(4) A uniquely omnipotent being is creating. That same being is loving.
We want to describe the content of the second sentence in way that captures how the being that is loving is the same as the omnipotent being that is creating. No analysis based on Russell’s theory of descriptions will do the trick.11 Suppose we analyze the second sentence as follows.
(5) A uniquely omnipotent being who is creating, is loving.
But we could equally say
(6) Another uniquely omnipotent being who is creating, is loving.
which includes the content of (5), so (5) cannot possibly mean that the being is the same one as mentioned in the first sentence of (4). Of course, (6) contradicts
