Objection four: Donnellan

1. Donnellan’s Distinction

2. Referential vs descriptive

a. “Use the red from yesterday” can, in context taken to function as picking out from among available colors one that isn’t even red but I succeed in referring by using it. I’m telling you what to get, not describing it as Russell would want. Russell, of course, wants properties of sentences but Donnellan goes far beyond. Referential uses can occur because of failures in presupposition, failures of commonly held context rules and so on. So speaker reference and speaker meaning seem to part company with the regimented point of view.

3. Donnellan argues that definite descriptions can perform two different jobs, and that we cannot discover which is in question without attention to the particular context of use.

4. Consider one of his examples. We find the gruesomely mutilated dead body of Smith. One can say ‘Smith’s murderer is insane’ meaning roughly ‘whoever murdered Smith is insane’. In such a case there may not even be a murderer; the speaker may have no idea who it is if there is one. Donnellan calls this an attributive use of the definite description ‘Smith’s murderer’. Roughly, what Russell says about definite descriptions fits this use. It is a matter of denoting: providing a description that actually fits a particular individual.

5. Imagine now that Jones is on trial for Smith’s murder. We observe his weird behavior. Assuming his guilt, I say ‘Smith’s murderer is insane’ meaning roughly ‘Jones, that man behaving oddly, is insane’. In such a case, there is a particular individual I intend to pick out, but my way of picking him out may be mistaken: Jones may be innocent. Donnellan claims, however, that in this referential use of the definite description I have still managed to refer to Jones, even if my phrase does not denote him. Roughly, what Russell says about genuinely proper names fits this use.

6. Donnellan allows that usually there is some sort of presupposition or implication that the F exists, whichever kind of use is in question. But if that assumption is false, in the attributive case there is no further question — since Smith was not murdered we cannot say anything about his murderer. But in the referential use, the falsity of our assumption does not undermine our claim: its truth or falsehood depends on whether Jones is insane, not on whether he was actually the murderer. The referential use can work without the hearer or even the speaker actually believing the assumption involved. A further difference that Donnellan discerns is that in the attributive case our assumption is simply general: someone or other is the murderer; whereas in the referential case it is particular: this individual is the murderer (and so, something that can only be extracted from the context of use).

7. Lycan introduces a possible way around Donnellan’s distinction, by appeal to a different and more general distinction: what a linguistic expression means and what a speaker means in using it. I use ‘Smith’s murderer’, which means in context that Jones is Smith’s murderer, to pick out Jones, whatever he may or may not have done. So I use a literally false sentence to say something possibly true. But Lycan suggests that this strategy will not always work and that we should preserve some version of Donnellan’s distinction.

8. Lycan considers the case “I know that’s right because I heard it from the town doctor.” From the emphasis, it seems that you mean it’s right because it was the doctor as opposed to an idiot. But, if you say, “I know that’s right because I heard it from the town doctor,” meaning that the doctor knows what’s up. One case is attributive, the other referential. Lycan wonders if the speaker and the semantic referent exhaust the types of referent. Mckay and Kripke push the issue wondering if anything is ever left over after we identify these two.

Objection five: Geach

9. Anaphoric expressions inherit meaning from another expression, its antecedent that occurs earlier. Thus,

a. The man who lived around the corner was eccentric. He used to snack on turtle heads.

i. Anaphoric because of where ‘he’ picks up its referent.

10. Russell can handle such well behaved anaphora but what about:

a. A rabbit appeared in our yard after dinner. It seemed unconcerned.

b. I shot an elephant in my pajamas. How it got there, I’ll never know.

i. In the first case, “a rabbit seemed unconcerned” misses the point as a paraphrase because the sentence is picking out a rabbit in the yard, not just any silly rabbit.

11. Consider

a. A rabbit appeared in our yard after dinner The rabbit seemed unconcerned.

i. Russell is forced to analyze this as if it had a built in uniqueness quantifier. But the sentence doesn’t strongly commit one to saying that only one rabbit appeared. Perhaps contextual consideration rule this possibility out but Russell cannot mobilize them within his theory.