 Russell on descriptions. We'll look at Russell on descriptions today and on Monday. On Monday, we'll see that, though this is a very, one might even say, technical subject, it's a little bit subtle, a little bit official. I'd say you can't quite see what the big issues are. You'll see on Monday that these questions about names and descriptions that Russell is talking about are really right next to some of the most basic questions in philosophy. And in fact, the whole time in this class, we'll be keeping it relatively technical. But lots of the big questions about what conception of truth we have, what our language, what kind of conception of reality we have, what the relation between sense data and the world of concrete physical objects, these questions are always right next to the topics we're discussing. The topics we're discussing, it seems to me, are in some ways more basic than these questions. But they certainly have implications for them. So today's bombshell is that descriptions aren't referring terms. I can hear the subtle reaction, but that kind of gives the big reveal away. But the way that we've been working so far is that there are all these singular terms, terms identifying particular objects. And this is the kind of naturalist that they seem to be working with, that they include. The terms that specify particular objects, that you're then going to go on to say something about, that would include descriptions like the tall man who called earlier, the point of intersection of A and B, as well as names like Sally or California, or demonstrators like this and that. Yes, so we've got this kind of rag bag of different singular terms. And Frege's idea seems to have been, and Serl's idea was certainly, you can take the class of proper names, Sally, California, and so on, and explain how they work in terms of definite descriptions. Frege and Serl writers, though, you could have a language that didn't have any proper names. It only had definite descriptions. And when you introduce proper names, they're justizing the talk that you're doing in terms of definite descriptions. So in Frege's account, it's that the name is abbreviating a definite description. And Serl's account is the name is a peg on which to hang a bag of definite descriptions. So the basic picture of Frege is you get the sign, which will be the singular term. Then you get the sense, which is specified by description, or as in Serl, a cluster of descriptions. And that's how you get to the reference of the sign. The reference of the sign is wherever matches the description. Russell's point is that none of this can be right. There's got to be a basic class of names that gets tied up to objects, but not by way of descriptions. Descriptions themselves are not referring towns. So you're going to have a class of names for which something like this is the right picture. You get the sign and you get the object. And if a sense is something like a description, there is no sense mediating the connection between the sign and the object. The hook between the link between the sign and the object has to be being set up without a appeal to descriptions. So I hope to make that completely convincing by the end of this hour. That's the task. So what is a name? Here's Russell on a name. A name is a simple symbol whose meaning is an individual or particular. So you specify the meaning of the name by saying what it stands for. That's the thing I was doing last time with line and circle and smudge and all that. That these are terms that you specify their meaning by saying which thing they stand for, which individual or particular. There are questions I agree like is love a particular, but we're kind of bracketing for the moment. There's a question as to how far the notion of a concrete object goes. We will touch on that on Monday, actually. But if you ask what's a name, well, it's a name. If you specify its meaning by saying what it stands for, and it doesn't have any parts that are themselves symbols. So a name like Sally has got parts all right. There's the s and the a and the l and the l and the y. But those are not themselves symbols. So Scott is a simple symbol because though it has parts, namely separate letters, those parts are not symbols. That's all right. That's the simplicity of a name. I mean, of course, there are words in there like caught is in there, but that's kind of an accident. Yep, letters are standing for sounds. Well, you could think of letters are standing for sounds. I think that's right, but there's something more basic about the spoken language within the written language. That's what you're saying. You don't think of the written sounds as standing for letters. So you might say, well, you only really get what the letters are when you get that they have translations into sound. But the thing is the notion of standing for is different in these two cases. As you go back to semantic value, the idea is that the term is making a contribution to the truth or falsity of a sentence containing it. And what the name Scott stands for, that makes a difference to the truth or falsity of a sentence containing it. That's all right. But the thing, what verbal noise the sign stands for in that sense, that has nothing to do in any direct way with the truth or falsity, with what the truth or falsity, the sentences, the symbol isn't making a contribution to the truth or falsity of the sentence by standing for a particular sound. You see what I mean? So there's kind of an ambiguity in stand for. One has to do with semantic value. The other one doesn't. It's important, but it's something else. OK? So so much for Scott, a simple symbol. On the other hand, take the author of Waverly. That, as you can see, has all kinds of junk in it. It has all kinds of bits and pieces. So that's not a simple symbol because the author and Waverly, these are separate words that are themselves symbols. So if it is separate words in it that are themselves symbols, then it's not a simple name. So in particular, if you look at the author of Waverly, then what's in there? I mean, what are those bits and pieces in there? That's the basic question about descriptions. What are those bits and pieces in there? And what are they doing in there? What are you guys all doing in there? What is happening? They're all in this kind of, there are seats over here. There are all these kind of old friends jumbled together in this playpen, the author of Waverly. But what are they up to in there? What is going on? What are they doing in there? So just to start breaking it down, the puzzling thing is, although at the start, you think, well, this looks like a singular term, well, when you get some focus in this thing, you say, who are these guys anyway? One of them is it looks like a general term. So how come there's a general term in there as this is a singular term? What's it doing in there? Right, I mean, that's like, it's tall. If you ask what its semantic value is, it's a map from people onto truth values. That's all right. But every name of a person, you can put it in there and you get a different truth value. And really, there's two bits in there. It's really a relation. Blank is an author of something and the name Waverly. So whatever's going on, this is certainly not a simple symbol. But there's a further question. How are these bits and pieces being put together to generate a meaningful composite? The Russell says, we have then two things to compare a name, which is a simple symbol, directly designating an individual, which is its meaning. So that would be something like line or square, the way I was using them last time. And having this meaning in its own right. So if you introduce a name like line or square and you say that's I hereby dub this mark line, then that connect between the name and the object was set up independently of any connects being set up between any other symbols and an object. So that name has that meaning in its own right independently of the meanings of all other words. But it's different with the author of Waverly. Once you fixed the meanings of this, author, of, and Waverly, you had thereby fixed the meaning of the whole thing. So the meaning of that whole composite depends on the meaning of other words. So a description consists of several words whose meanings are already fixed and from which result whatever is to be taken as the meaning of the description. So this should all seem plain as day. And in a way kind of laboriously spelling out the obvious, it should seem immediately compelling as I explain it. If I'm explaining it correctly. The question that we have is that we're going to have is, how does this go? What are you doing when you fix the meanings of those words and then you get the meaning of the description as a result? How does that work? What are the mechanics of this? But that's just first on the basic contrast between names and descriptions. So it should already be seeming a bit shaky that descriptions actually are singular terms. Something different is going on with names. Sorry. Something different is going on with names than with descriptions. They're working in different ways. Is that plain as day? OK. So what is Russell's theory? Russell's theory is about how the components of a description work together to generate the meaning of the whole description. So take the author of Waverly was Scotch. You take this fine novel Waverly and you remark about it, the author of Waverly was Scotch. So let's look at that bit, the author of Waverly. The thing is, this looks just like you were saying Scot to Scotch. This looks like you've got a singular term here and a general term being applied to it. It looks like Sally is tall. The point is that appearance is really misleading. So in that thing, in this sentence, we can distinguish two general concepts. There's the wrote Waverly part and there's the was Scotch part. Two general terms. OK, class, what is the semantic value of this one? Wrote Waverly. It is a map from objects to truth values. So let me give an example. Suppose you take a function, like squared, squared. Now you put in a number here, like 2. And what do you get out? 4, very good. Remember cooking. So I put in another number here. What do I get out? Great. So what this is, what this is, what this is, if you just take square, the square function, this is a map from numbers to numbers. Right? That's all right. So I could similarly take a function that was a map from words onto numbers. If I put in a word, then what comes out is the number of letters in the word. You could have a function like that. So I write in Seattle and I get out. Yes, that's what I make it. OK, that's all right. So this is a map from letters to numbers. Yep. So the idea is that you get something like this. This is a map from people to truth values. You put in Robert Birkenau there, wrote Waverly. And what truth value do you get out? What's Waverly? What's Waverly? OK. I'm glad this came up. Waverly is, I think, a 19th century novel written by the great Scots novelist Sir Walter Scott. OK. Robert Birkenau, despite his many qualities as a Renaissance man, not being a distinguished 19th century Scots novelist is not the man. Right. So does that clarify that? OK. Maybe we should have Waverly on the reading list. OK. Aha. Right. Still false. Very good. OK. That was worth it, right? OK. So if you put Robert Birkenau as your input here, you get his output false. If you put in Muhammad Ali in here, what do you get his output? False, right? You put in practically anyone here, you get it coming out false. Yes? But it's still a function, right? It's still a map from people onto truth values. That's all right? Yeah. OK. So this, too, is a map from people onto truth values. Yep. That's it, semantic value. Maps people onto truth values. They're kind of like functions. Uh-huh. Where was it? OK. So you've got these two general terms in there with those kind of semantic values. They're mapping people onto truth or falsity. Just one last exercise. If you put in Scott here, what do you get his output? Very good. Right. OK. That was the right answer. OK. So let's just focus on the author of Waverly part. If you say the, oh, yeah, yeah. OK, I've been glossing that. Yeah, it really is people. You specify which person you're putting in by putting a name in. Yeah. So to get a whole sentence, you've got to put a name in. But whether the sentence is true or false depends on who that name refers to. That's right. In this kind of case, it's a square. It's really a map from numbers to numbers. But the only way I can tell you which numbers I'm talking about is to write one end. You see what I mean? So what you've, literally what you put in the slot is a bit of language. But the semantic value of this bit of language, the general term, is the map from objects to truth values and which object you're mapping onto a truth value depends on which object the name stands for. Yeah. Yeah. You could write a man wrote Waverly and you'd still get an output. That's right. That's very good. In a way, that's the key point today. So let me just, let's just pause a bit in that example because that is a very good point. A man is not the name of any particular man. Yes? Yes. OK. There might be something called a man. But that's not strictly to the point. That's an accident. In the usual meaning of the term in English, a man is not the name of a particular man. Yeah. OK. So what is going on? Because a man wrote Waverly can be true. Yeah. And not only, not only because somebody might be called a man, but because it just might be true that a man wrote Waverly. Now a man is what Russell calls an indefinite description. Yeah. And he says, look, that's really colloquial. If you say a man came to see me today, you might say that's a very indefinite description. That's fine, right? You're not, well, there's something indefinite about it. You're not saying which man. Yeah. OK. But that can be true. A man wrote Waverly. So if the semantic value of Waverly, of wrote Waverly, is a map from objects to truth values, just take it for granted that what I just said is true, that the semantic value of this is a map from objects to truth value, how can it be true that a man wrote Waverly? Because you've not specified an object. How can it even be meaningful? Because you've not specified an object that can be mapped onto a truth value. It's not easy, is it? How's it going on? OK. Suppose you take this function. This function takes you from numbers. Suppose this function takes you from numbers on to a, well, this is going to be too complicated. Let me give you the answer here, and then I might give you the analogy with numbers. I think what's going on with a man is it says something like, there is some name such that, if you put that name into the slot here, the whole thing would have truth as the output. So what I said, what a man says is there is a name n such that n wrote Waverly maps onto truth. That's right. Yes, right. Thank you. Of a man. Yeah. There is a name of a man n such that, if you put n into that slot, then you'll get something true. So it's still true that the general term is only mapping objects onto truth values. But what you've done is you've said something like, and that should be is, right? You see what I mean? It's saying something like how many. So once you've got this kind of term mapping objects onto truth values, you could have inputs to this. They don't actually specify which person, but they just say how many. They say there's at least one man, such that if you put the name in there, then you get something true out. Or you might say, well, if the human race had jointly written Waverly, then you could put any name in there and get something true out. So if you took some general term like perishes, presumably you could put anything in there and you get something true out. You see what I mean? So in that bleak mood, you can say everything perishes. And what you're saying is, this is a map from objects onto truth values, all right? But what you're just saying is, put any name in there and you get something true out. So you can have a how many term going there. I wasn't thinking of the human race as itself for a fairing term. All I meant was, if we all had written Waverly jointly, then each individual name you put in there the thing would be coming out as true. So I just meant, the example was, Armand wrote Waverly, but it could be everyone, or it could be at least two. What about at least three people wrote Waverly? How would that work? Anyone? How would you explain what that meant in these terms? Well, okay, you don't need me here. The only thing I'd change about that is, it's not quite M, N and P wrote Waverly, it's M wrote Waverly and then wrote Waverly and P wrote Waverly, yeah. But that's completely correct, what you say, okay? Are you perfectly happy with the last comment? Put your hand up if you're not perfectly happy with the last comment, yeah. Right, that is the one weak spot for this approach. Maybe there are things that don't have names, yeah? In that case, you'd have to either say, let us, we just would say there is a potential name, yeah, because you could give it a name. Maybe not everything has a name, but everyone could have a name. Here we have this poor little orphan child, yeah? But you could give it a name, yeah. Oh, yeah, sorry, I'm losing, yeah, that's right. The reason I'm putting it this way is to make very clear the connection between what happens when you explain the meaning of the predicate here in terms of a map from names under reference onto truth values, given that you've got the framework set up in terms of names, what is the connection between that and this kind of phrase, I'm on, yeah, and that's why I'm doing the shuffle with name. To get the, and I think you can, I think you can work with your kind of problem that you were raising by talking about potential names. I think you can do that. But you're right, there are further issues here. Yeah, sorry, it's going one, two. That's right, that's right, it leaves it open, yeah. So you'd want it to be at least true that you could have M wrote Weberly and N wrote Weberly and whatever the other one was, P wrote Weberly, yeah, yeah. You'd want those three truths and for those three not to be identical to each other. And you just leave it open whether it was anybody else who'd had a go-to. You just keep silent about that one. Does that make sense? Yeah, yeah, yes, yeah, because it was joint authorship. Okay, we should shift to, yeah, we should maybe shift from wrote Weberly to was an author of Weberly or was at least a contributing author of Weberly, something like that, yeah. Oh, that was your question too, yeah, yep. Is there not? Walter, Walter, exactly. Please. I mean, it seems like we're stuck. You've got to be using the name in a particular sense, that's right, but we are just taking that for granted at the moment, yeah. Whatever, we're not, we're setting the names to one side and assuming that's been solved for the moment, yeah. And now we're saying, what's going on with those descriptions, yeah. I'm taking the connection between the name and the object for granted. We're just asking what's going on with the descriptions. But you know you're completely right, that has to come up in the end. It's just that being unable to think of talking parallel, we have to do one thing at a time, yeah. Okay, so what is it saying, so far, this has been about Aman, yeah. So you can think of that as a warm-up exercise for the author of Weberly, so the really hard thing is the meaning of the, people sometimes say to me, in theory of meaning, you must study such interesting and wonderful words. The whole history of the culture is in the language and no, you know, we do the, ah. Yeah, oh well, the question is what is the saying? Oh well, the question is what is the saying? And if we're going to have the same approach of saying it's kind of like how many, if the author of Weberly with Scotch is true, then how many people wrote Weberly? Could Weberly have been written, if the author of Weberly with Scotch is true, then could Weberly have been written by a machine so that there was no author? No, it makes no sense, right? There had to be at least one person who was an author of Weberly and kind of been entirely generated by a machine. So that's good, and you've got to say, at least one person hits that spot, at least one person has it true of them that they wrote Weberly, yeah? Is that enough? Suppose Weberly was jointly authored. Suppose Weberly was written by a commune of the People's Republic of Scotland. Could it, would it then be true to say the author of Weberly with Scotch? No, I mean, if you say the, it's got to be just one, right? I mean, if there's some great disaster somewhere in the world and I say, well, the rich country should help, meaning America, you might reasonably say, what do you mean the rich country should help? There are plenty of rich countries. You see what I mean? The, or if I raise a question to the class and I say, well, what does a smart student think? You might reasonably be indignant at that. You see what I mean? It doesn't just mean there's at least one. It's saying there's exactly one, yeah? So you've got to hear, you can say the author of Weberly with Scotch, you've got at least one person wrote Weberly, but you also got, and most one person hits that spot. There's a most one person where if you put a name for them into that slot, then you get something crew out. You can't put two or more people into that slot and get something crew out, okay? I promise you the payoff for all, this is really quite dramatic, but it won't come till Monday. Okay, so what we got so far is at least one person wrote Weberly and at most one person wrote Weberly. That's all right. But the whole sentence here is saying something a bit more than that, because you get the Scotch part in. So what I was saying is there's something, whatever hits the Weberly spot, hits the Scotch spot, Scotch spot. Yeah? So anything that hits the wrote Weberly spot, hits the Scotch spot. That's all right, too? Okay, so that's Russell's theory of descriptions. What this is saying, the author of Weberly with Scotch, is at least one person wrote Weberly and most one person wrote Weberly and anyone who wrote Weberly with Scotch. That's all right, that's pretty plausible. Now, note the implication of this. Is there a referring term in here in the top sentence? Where is it? I mean, apart from Weberly. Apart from Weberly. Is there a referring term in the top sentence? No, there is no referring term there. Is there a referring term in the second sentence? No. Is there a referring term in the third sentence? There is nothing there but those how many terms? There are no referring terms here at all. I mean, except Weberly. Scotch is not a referring term. Why Scotch is a predicate? There are many Scotch people. Not, we are legion. You see what I mean? So, we thought descriptions were referring terms. It looks like with a description, you're singling out something to go on to talk about it. That's not what is happening. Descriptions are not referring terms. So, let's just go over this. The question is, do descriptions refer? And you see why this is important? Because for, well, one reason it's important is for Frege and Searle, the whole idea is that descriptions are the basic referring terms. And all other references explained in terms of them. But if this kind of point is right, then Frege and Searle are just completely, really completely mistaken. You can't have that kind of approach of explaining reference in terms of descriptions. And they look like referring terms. And here's why they look like referring terms. Because it looks like, when you get a statement like this, the author of Weberly with Scotch, what you're doing is you're referring to an object. And whether this is true or false will depend on how things are with that object, yeah? That's why it looks like reference is a map from object onto truth values. Then this looks like it's presenting a specific object and to be mapped onto a truth value, yeah? So it looks just like reference. It looks like you're just singling at a particular object and then applying a general term to it. But that forgets that these descriptions are complex. It treats them as if they're simple terms. So the meaning of a description can't be just that it tags some object down here that you then go onto. There's more going on than just tagging an object. Just because it's complex is obvious there's more going on than just tagging an object, yeah? It does. I mean, well, it doesn't this way. If you look at, ah, man, right? That's an indefinite description. Then nobody would have any impulse to say that's the referring term. Well, that's not true, actually. Some philosophers have said it's the referring term. But nobody's sensible would say that's the referring term. You see what I mean? That picks out an indefinite man, kind of shifty forts of character. That's interesting. I think he's probably talking about both in the cluster theory, but the point is that we can go over this, but in CERL's picture there's got to be a level of reference possible that is purely descriptive. And that would be the use of definite descriptions. But we can go over this in a moment, actually. The reason this is looking like it must be proper reference is that it seems to be just presenting an object over here to be mapped onto a truth value. Where the complexity of the symbol means that can't be the right picture, the meaning of the symbol isn't a matter of just tagging an object. Russell's saying, when you've got a proper name, that just tags an object. That presents an object to the general term to be mapped onto a truth value. And if you ask, well, what is it for Sally to refer to a particular person? Well, what's going on is the general term is always mapping Sally onto a truth value, depending on how things are with that person. I hope at this point that you're familiar with this kind of ring bottle. Yep, yes, right, yes. Well, CERL's picture requires that you should have a level of descriptive discourse that's functioning without yet having names in play. Yeah, for both, that's how you can explain names and terms of their relation to the descriptions. So if you have purely general terms, you only have is it, F terms, yeah? Then you don't yet have any names. But how can you make sense of a language with just the purely general terms, but no names? Let me go on just a second. Suppose you take the explaining the name Sally. Once you explain the name Sally to someone and you say, this is Sally, well, the key thing you've got there is you now know how all these endlessly many things you could say about Sally are to be evaluated. You have to look at this particular person and see how things are with them. So if you say Sally weighs 150 pounds, Sally is six feet tall. You've always got to look at that single object to find out whether any of them is true. So that's what it means to have a name which is a simple symbol. But there is a sense of which is simple, but it's simple, but it also depends on the general terms. You only understand what's going on with the reference of a name when you understand what the implications are or is used in the context of sentences with general terms in them. But that's with the author of Waverly with Scotch. You could have explained, well, let's see, it's really an optical illusion that you just get a tag here because what you've got is this mapping from at least one into the slot and most one into the slot and anyone who wrote that or anyone who hits this slot, hits that slot. You could easily have made this at least two, at most three, at least one, and at most two. So you could vary what's going on here in lots of ways. And when you varied it in any of those ways, it would be immediately obvious to you you didn't have a referring term here. That's all right. I mean, if I said, amen, that just says at least one and misses out the second sentence. If I said at least one and at most two, nobody would be inclined to say that's the referring term. You see what I mean? Because that would allow, there were two people, you're not singling them out. This is identifying something, this is specifying the meaning of a sentence at a much more abstract level than just saying which object. You've got a really powerful abstract way here of saying how many things meet one predicate and about anything that meets one predicate meets the other. And it's only a coincidence that there's that particular tuning of this powerful apparatus that could be used for saying at least a million, at most two million. You've got that specific case where this powerful apparatus happens to be tuned in on one object. And that looks like reference, but it's not really, it's an optical illusion to say that this is reference because it just happens to be the case where there's only what it's saying, there's exactly one. You wouldn't have thought it was reference if it had said there's exactly two or exactly three. I'm not sure if I've explained that very well, but okay. If I have, if the test for whether I've explained this very well is you should now have no impulse whatever to say that the author of Waverly with Scotch involves a referring term. That's not referring to an individual and saying something about it. You could have a name that doesn't refer, you could try to have a name that doesn't refer, that's right. So that was my example last time, a smudge. But that's not what's going on here. The point is not that this is a term that tries to refer but might fail. This isn't even trying to refer. This is just saying, this is a how many term rather than the referring term. So when I was explaining Aman, there's something a little bit fancy about, I mean, I don't mean that pejoratively, but they're a little bit fancy terms like Aman because you've got the ground floor plain English where you say Sally is tall. The cat is on the mat. And you, that cat is on that mat and you name particular objects and to say something about their properties and the relations between them. That's the ground floor. And then one step up, you say, there's a cat on a mat without specifying which one. And that depends on your knowledge of the ground floor level of language where you specify the particular cat and the particular mat. You could have one person in mind but you wouldn't thereby be referring to them. I mean, if I say a man came to see me this morning, I might really have one person in mind but I still didn't refer to them. So the point is, a description is a phrase like Aman. A definite description is a phrase like Aman. It belongs to that higher level of language than the ground floor language at which you say who you're talking about. You actually specify which one you have in mind. At this level, you're not specifying which one you have in mind. You're characterizing it more abstractly. You're saying there's someone with these general characteristics and anyone with these general characteristics as a further general characteristic. One, two, three, yeah. That's right. That's exactly it, yeah. That's what I'm saying is the optical illusion. It seems like that, right? But suppose you had at least two people wrote, at least one and at most two people wrote weatherly. Yeah. Right, that's the illusion, right? That's what I'm calling the illusion, yeah. Because really when you, if you put it in terms of semantic value, yeah, can I put it in terms of, okay. So if I got my semantic value for a simple meme is just the object, right? That's the concrete object, right? And then my semantic value for a general term is a map from objects, objects get mapped to crucial falsity, yeah? Then something like a man is much fancier kind of semantic value. It's semantic value is you got your whole universe that you're talking about and you're saying, you say a man came to see me, it's something like they're always individual objects and it's a map from that domain. So a man maps from this domain onto predicates and objects and truth values, yeah? That's right, that's right. And it's just a happy accident that we're in the case in which that's exactly one, yeah. So given that domain, something like a man is mapping from the general term onto truth values is a much fancier kind of semantic value than just you get the name and you get the object, yeah? Yeah, in a sense, yes. It's saying there's exactly one and in that sense it's more specific but it's a very same kind of abstract schematic description rather than specifying which object. That's just the way, we'll have a suggestion to think of it, yeah. Okay, so it's really an illusion that this is referring to the author of Waverly. Okay, I realize that this is a little bit technical. We'll see on Monday that this is really large ramifications at this point.