 Good morning. I'll have to end a few minutes early today. I hope you'll bear with me about that. So today we're looking at Crip Keys naming a necessity again on Monday. We'll again be looking at naming a necessity and going back over the question is it true as Frege and Serral thought that descriptions fix the references of ordinary proper names. Crip Keys got a quite different take on this question. I'm a quite different picture of how the connect between the sign and the object gets set up. It's probably the view of reference that's the most common today. It's taken for granted I think by most people working on language today. So I'll set that out in the second section and then assuming time permits we'll move on to look at questions about informative identities once more. There's no class to be complete without us looking at informative identities right? Okay so last time we were looking at that. What fixes the reference of the name Gödel relative to different possible worlds? And what I was saying was if you take the name of someone famous like Gödel there are two questions. One is how is this reference fixed in the actual world? How does the name get hooked up to an object in the actual world? And then the other question is when you talk about what could have happened to Gödel, what would have happened to Gödel under various counterfactual circumstances is it still a description that's fixing the reference? And last time I was arguing no it's always the same object that the name is referring to in different possible worlds even though the descriptions that the thing is satisfying might be quite different. How about that? That's all right so far. Any questions on that? Okay Plain is dead? Okay. So you get the cluster of descriptions associated with Gödel, the serral style generation of the dossier and then you get the idea that that is so that's perfectly well reasonably well defined. This is well defined as anything in this area tends to be. So then the question is is that what's fixing the reference? Now Sel admits that there can be, indeed he says it's an advantage of his account that it acknowledges types of indeterminacy about which descriptions are in the cluster. You know if only one or two people put descriptions into the cluster does that get in? If some descriptions get in that conflict with other descriptions do those get in? What kind of weight do you give to particular descriptions in the cluster and how many descriptions in the cluster does an object have to match to be the reference of the term? So there's kind of indeterminacy about all these things. So that gives the view a lot of flexibility. It's difficult to get a knock down objection to the view when it has that kind of flexibility in it. But Kripke's argument is nothing like that is right. That whole picture is wrong from start to finish and it can't be fixed by juggling with the pieces here. As I say there's room for flexibility but Kripke is going to argue none of this is right. Ordinary proper names don't have their references fixed by descriptions at all. And one key example he gives you really have to read the whole thing partly because it consolidates everything we've done so far. That's kind of a sound effect. Here comes the causal theory of reference. Okay here it comes. And here's a key example. Kripke gave a very influential example. Goudal, Kurt Goudal, very famous as discovering the incompleteness of arithmetic. Proving that for any finitely stateable set of axioms for arithmetic there are going to be truths of arithmetic that can be recognized to be true but not provable in that system. Okay so we all know who Goudal is. He's a person who discovered the incompleteness of arithmetic. So can we say that that's what fixes the reference of the term? Well and Kripke outlines this picture in which Goudal who was of course an unassuming and immensely distinguished Princeton mathematician. Goudal and Kripke's picture may actually have been a somewhat desperate character. Here he is. And you can see what Kripke means. These are his accomplices in I guess early mid-century Austria and suppose that someone else, not Goudal, was the brilliant and hard-working mathematician who carried out the proof of the incompleteness of arithmetic. Goudal never did any of that stuff. Goudal was simply the brains of the outfit that stole this other guy's discovery. You can imagine that scenario. As Kripke says in the lecture I hope Professor Goudal is not present. So Schmidt let us call him the guy who did prove the incompleteness of arithmetic. Here's Kripke's scenario. Suppose Goudal was not in fact the author of this theorem. A man named Schmidt whose body was found in Vienna under mysterious circumstances many years ago actually did the work in question. His friend Goudal somehow got hold of the manuscript and it was thereafter attributed to Goudal. So this is the scenario. Poor old Schmidt lying there somewhere on the rain swept street. It's still true that even if Goudal did not discover the incompleteness of arithmetic the name still refers to the person who stole the credit. So for any set of descriptions you could get this phenomenon. You could get this with Aristotle. Maybe back in Athens Aristotle actually stole the credit for all those works. It makes perfect sense to say perhaps Aristotle was a fraud. Perhaps Goudal was an imposter. He passed himself off as having done this proof. That makes perfect sense. The story is not a contradiction. When I describe the scenario to you I am not describing a scenario in which it turns out that Goudal was not this man. Goudal was actually Schmidt. If you see what I mean. Was that a bit too fast? If you assume that the name Goudal has its reference fixed by the discoverer of the incompleteness of arithmetic then in this scenario what you're discovering is that Schmidt was Goudal. But that's not right. What you're discovering is that Goudal was a fraud. Was that too fast? Okay so if I've explained that correctly that should be completely obvious. But if that's right then it really is over for the description theory of the way names get their references fixed. You can always find out that the set of descriptions you associate with a name they do not actually apply to that person. Those descriptions got associated with that person as a result of some error. Kripke goes on to point out that it's also true and it's almost trivial really that point we've made many times before in this class that if you take the descriptions most people associate with a name they may be wildly off the mark. Maybe most people do think that Einstein insofar as you've heard of him at all was the inventor of the atomic bomb. That does not show that Einstein was not the guy who showed that E equals MC squared. That just shows that the reference is not being fixed by those descriptions that are popularly associated with the name. One other example people say there are people who say well after all those shape plays commonly attributed to Shakespeare they were not written by Shakespeare. Shakespeare stole the credit for them from Bacon. But when people say this they are saying Shakespeare was a fraud. They are not saying Shakespeare was Bacon. As you can see these are two quite different people. One of them wears a hat. You see what I mean here we have two people one is Shakespeare one is Bacon. If Shakespeare stole the credit for Bacon's plays so that the descriptions we commonly associate with the name Bacon. So that the descriptions we commonly associate with the name Shakespeare are all in fact true of Bacon. That does not show that Shakespeare and Bacon are one in the same person. They are two different people. It's just that we associated all the wrong descriptions with the name Shakespeare. Therefore the reference of the name Shakespeare is not being fixed by the set of descriptions we associate with the term. The reference of the name Einstein is not being associated with the totality descriptions we associate with the term. The reference of the name Gödel is not being fixed by the totality descriptions we associate with the term. Here ends the first lesson. Yes that's right. I think that is Charles idea. You could always cast off some of the descriptions but the whole boat could still keep a float. Something like that. Yeah but the thing about this Gödel example is that Gödel is quite well known in academic life. Are you quite well known anyway. But that's the stuff that everybody knows about him. I mean there isn't anything left. You see what I mean there isn't anything left to fix the reference. Once you've thrown out those descriptions these are the central things we know about Gödel. On any kind of picture where you have some kind of weighted vote being given to a whole bunch of descriptions anyone would say these mathematical discoveries they're the most important things. Or if you take Shakespeare you know very little is known about Shakespeare's private life. The vast majority of what's known about Shakespeare has to do with him writing those plays. So the thing about these examples is there's nothing left when you subtract those descriptions. It's fair enough for some to say I can juggle with it a bit at the fringes. But this is the center. Yes you could find a new description that's right. But the thing is that is coming after the fact. What's happening is that you've got the name. You've got the object. You've got the connect between the name and the person. You are discovering that all the descriptions you associated with that name are not true of the person. And now you say well what is true of this guy. Well he was a fraud and so on. And so you now revise what descriptions you associate with the name. That is what would happen. But the point is that that's assuming that there's a connect between the name and the person that has been set up independently of the descriptions. And that connect is what you're using to drive which descriptions you associate with the name. Yeah. Yeah. So the description that actually applies to Bacon when I say Shakespeare, am I talking about Shakespeare or am I talking about Bacon? Right. Well on a description theory you'd be talking about Bacon. But the claim is that's the wrong answer. You're still talking about Shakespeare. I mean Shakespeare is Shakespeare. Oh you see what I mean. That's the argument anyhow. That's very important I think. Yeah. I mean if you say Shakespeare. I will actually give some examples of that kind. But if you said look Shakespeare was born just two months ago. Yeah. He's only two months old. He didn't write any of those plays. Yeah. Then is it really the same person? Yeah. That is a point for the description theory. I agree. And I agree to that Kripke's formulations leave it wide open that that possibility could be correct. That you could still be referring to Shakespeare even though you were so wildly wrong about the descriptions. And that doesn't seem right. Yeah. The positive thing that Kripke does get though is if you say I mean how should I put this. If you say well I've got to get there are boundaries right. That's the way of putting the point. There are boundaries. So it can't be kind of lived all that long after the time we think he lived at. He kind of lived all that long before. If you said well Shakespeare was born on the stone edge. They couldn't even write then. I mean he couldn't have written those. But yeah. That's the same thing. So you see there are some kind of boundaries and you know he can't actually have been a frog. I mean the animal. Yeah. Then you say well those boundaries around which descriptions could be true. And Kripke doesn't really have any place for those boundaries around the descriptions that can be true. But it seems right that there are boundaries. But just saying it's got to be something within those boundaries isn't going to be enough to fix a reference. Yeah. That's the key thing that Kripke still has. Yeah. So I think your point is extremely important. But it doesn't bring back the description theory. Yeah. It says there was something right about the description theory. Yeah. One two. But we should move on because I will try not to talk for so long. But we should move on after this because I will have to stop a little bit carry on. Yes. Yes. Yes. Yes. Well I mean even with family members I mean this is more like a kind of spooky movie. But even with family members can't you imagine this kind of thing. I mean your best friend says do you remember that walk we took when we were children by this by the cliffs at the seaside so long ago. And you say yes that's one of my most precious memories of you. That's right at the heart of the cluster of descriptions I associate with your name. And he says nah that was my evil twin. You never knew about my evil twin. You see what I mean. Or that woman says I'm not your mother. You see what I mean. I don't want to make your spine tingle too much but these things could happen. Yeah. You could make those discoveries. Yeah. You still be talking about that same person despite all these revelations. Yeah. That's the idea anyway. Yeah. Last one. Is it is not quick. Okay. Let's talk about it later. Yeah. Okay. So Kripke's positive picture. Well one way to think about what Kripke's positive picture is is to think about photographs. So I have here a simple goose. I think that is a goose isn't it. Yeah. It's a bird. Let's just call it a bird. Right. I have here a picture of a simple bird. Now there are tons of birds quite like that. Yeah. So let us ask the question for photographs. You can ask the question that's parallel to the question about reference. Which thing is this a photograph of. So there's a question of often is for photographs as of the question of reference for signs. Yeah. So you can think well one theory you could have that would be like the description theory for photographs would be that a photograph is a photograph of the thing that is most like it. So in order to find which bird this is a photograph of what you'd have to do is go among the geese and discover the one that looks most like that. So that's what you might call a match theory of who the photograph is of. But of course that is not actually how we think of photographs. I mean it could turn out that around this goose this has taken a civic center in San Rafael where there are literally just thousands of geese all very similar. And let us suppose that among all those geese there is one that is just a bit lighter than this one. But because of some mistake I mean some error in the photo processing this coloration here is a bit lighter than the coloration of the bird that was actually in front of the camera when the picture was taken. Was that too fast? No it was too slow. So let us suppose there is a goose there round about that matches the photo better than the one that was in front of the camera when the picture was taken. That is not sure that this is really a picture of that other goose that I never ever pointed the camera at. Do you see what I mean? If you ask what makes a photograph a photograph of one thing rather than another what matters for it being a photograph of one thing rather than another is that that is the thing light bouncing off which entered the camera lens and affected the camera sensor resulting in the production of this photograph. So it's the thing that was causally involved in the production of the photograph. Some goose that matches this photograph very well but was often the other side of the planet that's not the goose that this photograph is of. Photographs are of the things that were causally involved in the generation in the production of the photograph. That's how it can happen that you get a photograph of yourself and you say I don't look like that at all. I look much better than that you think. That's not a photograph of me. I mean if the match theory were right you could actually argue no that's a photograph of someone else because it doesn't match me at all. You see what I mean? But if you were actually sitting in front of the camera when the shutter went then you are the person that photograph is a photograph of and it doesn't matter whether you really look like that or not. So I think that is similar to Kripke's picture of how it works for names that Kripke's picture is what goes on in the fixing of the reference of a name is you have some kind of initial dubbing I would call this child Saul and then you have a chain of communication in which gossip news information about that person circulate in a community. You got a chain of communication in which the name is passed from speaker to speaker. So what happens with you is that there's some kind of initial dubbing of you and then as time goes on you radiate information about yourself into the community. Your name gets passed from person to person and there is a cluster of descriptions connected to you but what matters is not who best matches the cluster of descriptions what matters is how that cluster of descriptions was generated in the first place. So you might think people get it all wrong about me I'm not the way they think I am no one knows the real me but still you are the referent because you are the person that was causally implicated in the generation of all those descriptions. So what matters is how the cluster of descriptions is generated not what best matches the cluster of descriptions. That's like the case with photographs it's not what best matches the photograph it's how it was generated how the photograph was made. So you get two stages here you get the initial baptism in which you maybe point to the thing or you say by description I'll call my oldest son Saul you get that initial description in which you say low Saul or whatever and then you get the second stage that chain of communication and there is a question what's meant by chain of communication here and after all if you say what he's got in mind here is something like this you got the talk of he's talking about the physicist Feynman suppose the name is spread from link to link as if by a chain and a certain passage of communication reaching ultimately to the man himself does reach the speaker the speaker is then referring to Feynman there's been that chain of talk from the initial dubbing through to you or I right now and then we're referring to whoever it is that was at the other end of that chain even though you or I can't identify them uniquely a chain of communication going back to Feynman himself has been established by virtue of his membership in a community which passed the name on from link to link is not that sitting alone in your study you said I suppose out there somewhere or other there is a great physicist I shall call this guy Feynman that's not what happened that can happen but it's a very unusual case usually what happens is there's just that actual chain of talk that you pick up on there's stuff going out there in the in the community that you pick up on and using the name and you are referring to whoever's at the end of the chain that you picked up on think of the cluster so you can think of the cluster of descriptions circulating in the community around Feynman as a kind of communal photograph of Feynman so if you take cells dossier of descriptions cells cluster of descriptions think of that as a kind of communally generated photograph it might be a it might be a good or an imperfect photograph of Feynman but that cluster of descriptions is like the content of the photograph and what you're talking about when you say Feynman is not whoever best matches the photograph what you're talking about is whoever was causally involved in generating that cluster of descriptions so that's where the cluster of descriptions the idea that description six reference turns out to be completely wrong not just wrong in some point of detail but completely wrong yes that's right yeah I mean this is this is the tragedy of life right that people can spread rumors about you they're still talking about you yeah that's what's so terrible about it the stories I could tell you but you see what I mean and when people sue because somebody's spreading scandal about them you can't defend yourself by saying well the description didn't fit so naturally I was talking about someone else see settle you see what I mean it doesn't work like that yeah uh one yeah yeah yes right absolutely true I mean the problem of informative identities does not go away it just gets worse on this picture and I'll bring this out towards the end I mean that's really why I started out with it it may seem like a kind of well surely you can handle that kind of question right at the start but actually the depth of the problem just gets more and more apparent the more you think about these questions so I'll come on to that at the end but you're completely right unicorns yes right right the yeah I mean you're aware that there are no unicorns yeah I mean that's the way you you put it by saying and the unicorns started all off there was no unicorn right right right good okay right I'm sorry okay but you're right the name unicorn seems to have some kind of meaning even though there's no causal chain going back to anything yeah how did that happen yeah that's a question yeah that I mean the two arguments for the description theory were one informativeness of identities and the other meaning without reference yeah so this second question meaning without reference is also a hard question and I don't actually want to suggest that it just goes away the question how a name like unicorn kind of meaning even though it doesn't refer because there is no just as you say there is no causal chain going back to you we've got an account of meaning here that doesn't doesn't on the face of it explain that yeah so I don't want to try and close off what you're raising there with a sound bite there isn't a way of closing that off with a sound bite that is still a hard problem yeah yeah right yeah yeah okay someone's gonna capital P in capital right right right yeah okay these are two different cases right with it with the pink hat example um well I assume the girl exists it's not just a kind of joint hallucination right you yeah that is another case but in that case you write it's not her agency it's not that she collaborated in this but still light was bouncing off her hat and hitting you retina you see what I mean she was causally involved even though she wasn't deliberately bringing it about so causation here is not the same thing as intentionally making it happen yeah um but the the other case of yeah the artist who imagines something or yeah someone with an imaginary friend right we've all got them they've all got imaginary friends right and you give them a name right so so what is going on there um well I don't want to just shut up the question so let me say something brief about this in the unicorn case and one thing you might say is well in an imaginary kind of case what you're doing is after all you're imagining that there is an object there you're imagining that you are causally connected to something out there um so you're imagining that the name has a meaning in the case of unicorn or in the case of uh the imaginary friend the thing the name doesn't really have a meaning but you've just invented a kind of vivid game in the context of which you're imagining that there are all the trappings there the object the causal chain in virtue of which the name would have a meaning there's a pretty radical view but it says in talking about unicorn or talking about imaginary friends always involves some element of pretence or make-believe and it's only in that context a pretence or make-believe that you can have the idea of meaning or reference at all in these cases it doesn't really mean anything we will discuss this further okay okay um but they are both these are both just the questions to be raising the informativeness and the non-existence okay um I see now that I think of it you guys are actually anticipating everything I was going to say uh so let me um let me shape up to the example that was given earlier of uh Shakespeare being a French peasant or whatever it was not every sort of this is Kripke not every sort of causal chain reaching from me to a certain man will do for me to make a reference and it says you can hear a name and decide to use it to refer to some unrelated object and after all you get your new puppy you say what do I call a puppy and you say Wittgenstein what a great name for a puppy um you can do that and that's causally connected to your knowledge of the philosopher right that doesn't mean that when you say Wittgenstein needs feeding now or whatever that doesn't mean um isn't isn't Wittgenstein adorable um that doesn't mean you're talking about the philosopher even though there is a causal chain going back from your use of the name to the philosopher yeah um so what's the right kind of causal chain some kind of causal chain of the right ones and some are not um and Kripke addresses this by saying when the name is passed from link to link is it as you go on in the chain the receiver of the name must I think intend when he learns it to use it with the same reference as a man from whom he heard it so this is like um that would block the Wittgenstein case because when you use Wittgenstein as a name for your dog you don't intend to use it with the same reference yeah but let me give you another couple of examples this is an example that was actually suggested to me by someone in this class a little no this when I was teaching this class a couple of years ago um suppose you have a family this is kind of a heart-rending story but suppose you have a family with a beloved dog dear old lopured spot yeah greatly loved by the children greatly loved by everyone and one day the family is out and the father comes home alone and what should he find the poor old spot dead as a doornail and he is concerned about the tragedy he's concerned about how upset everyone should be so he um takes a shovel and um disposes of spot um and very rapidly finds another dog that looks exactly like lopured spot and gets the new dog in place before the rest of the family return so over the years what happens here is everyone talks about spot all the way through no one else realizes there's been that unnoticed substitution so when this is not like the I'm going to call my dog vittgenstein case this is a case in which everyone intends to use the name with the same reference all the way through now just think about what happens in this scenario and at first when the children come back and say spot dear old spot if they knew the truth the right answer would be that's not spot that's an imposter right yes I mean you know that would be kind of a stubborn reaction but that would be true yes okay but now suppose that um uh suppose that the new spot lives with the family for another 15 years they have many happy memories of all their rumps together those evenings by the fireside um all that stuff at the end of the 15 years when people say spot which dog are they referring to the new dog or the old dog I mean at that point it seems absurd to say well spots an imposter what they're talking about is is a new spot even though they intended to use the name with the same reference all the way along the reference has shifted so even though there's a causal connect back to the original dog and even though they intended to use the name with the same reference all the way through by kripke's lights the name should be referring all the time to the original dog but it's not here's um suppose you just came upon this suppose that you work on ancient philosophy I mean guys I mean my impression is it with the pre-socratics you very you don't get the pre-socratics you don't have books by them right you don't have even have articles you just have these fragments that are found right so suppose you think a lot about pre-socratic philosophy and then you come upon this passage there are no unknowns there are things that we know that we know there are no unknowns that is to say there are things that we now know we don't know but they're also unknown unknowns there are things we do not know we don't know and you think that's that's wonderful um that's probably about 3000 bc um this yeah and and you think I what a discovery and you see the author's name is rumsfeld and you think okay this is a pre-socratic philosopher all right um and fits right in and you speculate about his influence on Aristotle um you you wonder whether it's which community he was really central in right that's what you do with a fragment like that right so um you speculate that this was a pre-socratic epistemologist um laying down some of the foundational thoughts for um uh succeeding centuries of epistemology um now I mean this to be like the Shakespeare example earlier that if that's what you think rumsfeld is right if you think rumsfeld lived um thousands of years ago at the dawn of uh philosophy are you referring to the politician so is this the guy you're referring to if you say I think he was a pre-socratic philosopher um this is the same point really I mean you were ahead of me that um there's that box of limitations on how far wrong you can be in the description you are you are if you pick up the name here you are intending to use the name rumsfeld with the same reference as everybody else you know it's not like you're just calling your dog rumsfeld you mean to be talking about the same thing as everyone else but you just made up such a radical mistake that it seems crazy to say you have very unsound views about um uh a late 20th century politician you see what I mean it seems crazy to say that that's who you're referring to so this is the same point that there are limitations in how far wrong you can go in your use of descriptions and that this oops sorry got the wrong way this two-stage thing of um kripke's passing the name from link to link with the intention to preserve reference that's not enough there's some other kind of conditions on what the right causal chain is okay let me um finally I just quickly want to go through um uh a quick point of contact between um russell and kripke remember russell's got this picture there's that basic class of names that is getting hooked up to the objects but not by means of descriptions and is connected to your sense data or whatever and um you say well uh uh that read doesn't sound right that gives us this picture on which each of us is fundamentally alone trapped behind our sense data unable to think and talk ultimately about anything else couldn't you keep a picture like russell's where you have a basic class of names and they refer but they refer to concrete objects to things in the public world to things in the shared world on the face of it that's just what kripke is doing kripke's names like when kripke's talking about names is just like russell in that he's saying these are not defined in terms of descriptions I mean you see I got a basic class of names here and the connect to the world this was driving the rest of language well the causal chain is what is hooking up the names to the world and driving the rest of language that seems like a better clearer answer than russell's as to what the basic connect is between names in the world is not us associating descriptions it's the stuff out there causing the use of the symbols by us it's really not a matter of convention or anything like that this I mean if you think of um what a photograph is of that's not a matter of us laying down conventions about what we take a photograph to be of who a photograph is of is just an objective fact about the way the world works or what the causal connections are in the world it's nothing to do with um conventions we lay down and similarly for his basic class of names and what fixes reference is just this objective fact about the way the world causally operates so you could think well russell's acquaintance you should really just russell should really just have interpreted that in terms of a causal connection because it's through a causal connection to the world that you get knowledge of what's going on around you that basic impacting on you by the external objects that's what makes um thought and talk possible the things that are impacting on you um it's not a description theory make it seem as though what goes on in your head is kind of autonomous from what goes on in the world you just dream up all these conceptions as to what's out there and then the question is does the world shape up does the world match up but this picture turns out in its head and says what's going on in your mind is driven by what's out there oh i should have warned you if you're not if you're of a nervous disposition you might want to shut your eyes okay so okay here we have here we have one um unfortunate individual here we have another individual not so unfortunate are they the same is this person that person well suppose i told you that's one in the same person is that an informative identity i mean not at the same age of course is that an informative identity yes that was highly informative right um and what it's saying is not whoever meets this description meets that description what it's saying is whoever was at the end of the causal chain generating this photo was at the end of the causal chain generating that photo yep that's a comments informative because these are two different photos two different things causally generating them yeah that's all right yeah that's how that's a comments informative but now consider these identities if your nerves are up to it um now consider the identity of this person with this person is that informative aha why not because i agree that this photo matches this photo the descriptive content of this photo is the same as the descriptive content of that photo but that doesn't guarantee that they're one in the same suppose they are identical twins one could have been at the causal chain at the end of the causal chain generating one photo the other a moment later at the end of the causal chain generating another photo i mean we're so sensitive to humans that you might think well it's very unlikely they'd be just so similar but it's not impossible and if i did it with geese it would clearly be possible yeah so this this identity isn't um uninformative because it's not a priori or analytic that what was causally involved in generating this image was causally involved in generating that image and similarly here i mean that after all might be two different people on different occasions they just posed exactly the same way so neither of these identities are uninformative but then once you think well that's to be a causal theorist in general thinks about identity how could you sorry thinks about reference how could you ever get an uninformative identity on a causal theory of reference on that point you know i i have i have to go okay thank you