The charlatan and dilettante non-philosopher Saul Kripke on ”Naming and Necessity”
By: Demokrates Sorensson, a few terms into Theoretical Philosophy, 2007.
During the last few days some thoughts have emerged, which I fear may be too important to be lost. For the sake of not losing them, I therefore take the opportunity to write them down, and turn them in to you.
In his much read three lectures collectively called “Naming and Necessity”, Saul Kripke attacks the theory, maintained by Frege and Russell, that names refer by means of definite descriptions. Kripke starts out by making a reference to Keith Donnellan, who was the one who first tried to refute the theory. Donnellan had shown, he claims, that you can make a reference to some object while using a definite description that isn’t satisfied by that object.
“It is a point, made by Donnellan, that under certain circumstances a particular speaker may use a definite description to refer, not to the proper referent, in the sense that I’ve just defined it, of that description, but to something else which he thinks is the proper referent of the description, but which in fact isn’t. So you may say, 'The man over there with the champagne in his glass is happy,' though he actually only has water in his glass. Now, even though there is no champagne in his glass, and there may be another man in the room who does have champagne in his glass, the speaker intended to refer, or maybe, in some sense of ‘refer’ did refer, to the man he thought had the champagne in his glass.” [All italics inside the citations of this paper are there in the original, unless otherwise stated.]
I can confirm that this really is a correct paraphrase of a part of Donnellan’s argument. It is based on a simple misconception. In the above situation, you are said to “intend” to refer to a certain person, the person you think are drinking champagne. Even when you discover that he wasn’t drinking champagne, you still think you meant to refer to him. Well, actually you did, so long as you had a description in your mind, satisfied by the person in question, even without him really drinking champagne. This person is what you “intend” to refer to. And a satisfied definite description is what we usually have when we point people out, for ourselves aswell as for others. When we point something or someone out, simultaneously ascribing some property to it, we think that the necessary and sufficient conditions of it being what we ascribed the property to is simply that it is there, whatever it may be, and that it is really pointed out by us. The former is most often satisfied, because hallucinations and the like are rare. The latter is true as a matter of tautology. If it is pointed out by us, it is pointed out by us. In the case of the champagne-drinker, it is natural to implictly assume that what is referred to by the statement “The man over there with the champagne in his glass is happy” is whatever object that is present on the spot pointed out, and which is being pointed out. In “pointing-out”-situations, this is simply how our human “theory of mind” wants to have it.
Next, Kripke cites Frege. The citation illustrates Frege’s well known idea, that a name may have many senses and yet the same reference. As in the example of the citation, the reference of “Aristotle” may for some be determined by the meaning“Plato’s disciple and the teacher of Alexander the Great”, for others by “the stagyrite teacher of Alexander the Great”. Kripke adds that even for a single person, it might be difficult to decide on a meaning for the name “Aristotle”. He then gets weird: “In fact, he may know many things about him; but any particular thing that he knows he may feel clearly expresses a contingent property of the object”. So what? Why should a name’s reference always be determined by necessary properties? And necessary in what sense? If I say that “Aristotle” was “Plato’s disciple and the teacher of Alexander the Great”, the name does refer, for there was such a person, and only one. But was it necessary that such a person came to be? I think not, not in a metaphysical sense at least. At least either way can't be proven, because metaphysics is not science. But perhaps Kripke with the opposite of “contingent” means necessary in another way, namely substantial. A substantial property is a property without which an object would cease to be such an object as the property is considered substantial of. If this is what he means, I would agree with him that sometimes it is difficult to decide on whether some property is substantial for something or not. But that’s entirely a matter of taste, a matter of will. If I want to call “Aristotle” any single thing alone satisfying the definition “Plato’s disciple and the teacher of Alexander the Great”, and nothing else being required, then for sure, the property satisfying this definition is substantial for “Aristotle”. And I can decide any property I like, and ascribe to any name I fancy.
Actually, Kripke confuses these two kinds on necessity, that is metaphysical necessity, and logical necessity (if I define a thing in a way, it is a logical necessity that if a thing is in that way, it is the thing I mean by the definition). This is clear from his saying that “if 'Aristotle' meant the man who taught Alexander the Great, then saying “Aristotle was teacher of Alexander the Great”, would be a mere tautology. But surely it isn’t; it expresses the fact that Aristotle taught Alexander the Great, something we could discover to be false.” Well, we couldn’t discover that Aristotle did not teach Alexander the Great, if Aristotle is defined as whoever did that. What we could discover is that no one person taught Alexander the Great, and that therefore, the definition is not satified, or in other words: it is true for all things that they are not and have never been this "Aristotle". But then there is no thing of which it could be false that it alone taught Alexander the Great, and what cannot be false, can hardly be discovered to be false. And even if we said that being the man who taught Alexander the Great it is not a necessary condition to be Aristotle, i.e. a substantial property, it would be, given that we knew that the single person satisfying whatever other characteristics we defined Aristotle as having, also in fact was the man who did teach Alexander the Great, impossible to discover that he was not.
Citing Ludwig Wittgenstein’s Philosophical Investigations, Kripke moves on to the description cluster theory of reference inspired by Wittgenstein and propounded by John Searle. According to the cluster theory, the reference of a name is not given by a single description, but by some number of a set of descriptions. In the citation, Wittgenstein wonders how many descriptions of Moses found in the Bible he is willing to substitute for the name Moses. This is because if a name is just an abbreviation for a number of descriptions, the descriptions should be possible to put in the place of the name, and have any utterance of them say the same thing. The idea seems to be that many things may be associated by a name, and out of these things, some number of them does determine the reference of the name. This is a very blurry idea. If a name is used, it’s used in a sense translateable into a definitine description, otherwise it’s not a name. It is a fact that many definite descriptions exist that describe the same object, that many different senses can have the same reference. But it is not the case that when something is referred to, we may use more than one description. For sure, we may assume as the sense of the name many different but referentially consistent senses previously associatied with the name, but when assumed together as the sense of the name, then these many senses are really become just one, namely the sense of the name they’re assumed as the sense of. To say that it is sufficient in order to be the referent of a name to be something satisfying only part of the sense, that is one or several but not all of the senses assumed as the sense, is nonsense. Or wrong at least. For in the remaining part of the sense, there may be something ruling out the thing in question as a referent of the name.
Another misconception comes into the picture with the statement by Kripke that “if Moses is not synonymous with any description, then even if its reference is in some sense determined by a description, statements containing the name cannot in general be analyzed by replacing name by a description, though they may be materially equivalent to statements containing a description”. If a name is not synonymous with any description, you can’t describe it. But if you can’t describe the object of your thoughts, how can you possibly refer to it? It may be noted that a description need not be put in a natural language. It might aswell be a mental picture or idea, which are composed out of the symbols of the individual mind. Even if some such picture or idea would not be describable in the symbols of a natural language, it is at least possible to substitute it as is for the name of it’s object. While I doubt that there are actually names of such things, things not describable in ordinary words, if indeed there are, Kripke is still mistaken. For if there is no description of “Moses”, verbal or otherwise, then there is nothing named by “Moses”. It couldn’t be in any other way, since to be able to refer to something, or even intend to refer to something, we must think of a thing to refer to, but thinking of a thing implies being able to describe it, because fundamentally, the very thought of the thing is it’s description, either this thought is verbal or not.
The main part of Kripke’s failure is found in the second lecture, in his step by step attempt at refutation of a what he proposes to be the upgraded Russelian description theory of reference propagated by Searle. According to Kripke, the theory has the following main theses (and I cite):
“(1.) To every name or designating expression “X”, there corresponds a cluster of properties, namely the family of those properties φ such that A believes that “φX”.
(2) One of the properties, or some conjointly, are believed by A to pick out some individual uniquely.
(3) If most, or a weighted most, of the φ, of one unique object y, then y is the referent of “X”.
(4) If the vote yields no unique object, “X” does not refer.
(5) The statement, “If X exists, then X has most of the φ’s” is known a priori by the speaker.
(6) The statement “If X exists, then X has most of the φ’s” expresses a necessary truth (in the idiolect of the speaker).
(C) For any successful theory, the account must not be circular. The properties which are used in the vote must not themselves involve the notion of reference in such a way that it is ultimately impossible to eliminate.”
Having first examined the last thesis, and found it sensible, Kripke takes a look at thesis (6). He now reiterates what, concerning the purported problem of determining the reference of “Aristotle”, was related above. If only one property out of a cluster of properties is “given any weight at all” (it’s unclear what is meant by “weight” here, but nevermind for the moment) to pick out the referent, “for example, Aristotle was the philosopher who taught Alexander the Great – then certain things will seem to turn out to be necessary truths which are not necessary truths – in this case, for example, that Aristotle taught Alexander the Great.” But as was shown above, if we define Aristotle simply as the only person who was a philosopher and who was teacher of Alexander the Great, then it follows by logical necessity that if Aristotle did exist, he was just that person. It must be so, for being Aristotle in that case is nothing other than satifying the definition in question. To be is to be the value of a variable.
But Kripke just keeps going. “This is not. It just is not [sic!], in any intuitive sense of necessity, a necessary truth that Aristotle had the properties commonly attributed to Aristotle.” It’s no small wonder that a man so famous for his thinking on modal concepts can be so utterly out of focus on this. If the common description of Aristotle is satisfied, then yes, it is logically necessary that Aristotle, in the sense of this description, did have the properties included in the description. What eludes Kripke is that what is meant by this is not that it is necessarily true that there existed such a person as the Aristotle we all know, but that given that there existed such a person, it is a necessarily true that he was Aristotle.
Catastrophically flawed and all, the above “argument from Aristotle” is nevertheless and as a matter of fact the only objection Kripke has got against thesis (6). Still, he thinks that given “these remarks, this means we must cross off thesis (6) as incorrect.” He certainly is bold, this Kripke.
Strange enough, Kripke has no problem of understanding the same conditional cited by thesis (6), when cited by thesis (5). He concludes that if he has named a planetary body seen in a certain location and time “Hesperus”, that is, determined the denotation of “Hesperus” by a definite description, it is a priori true that if there is any “Hesperus” at all, it is that thing seen there and then. But what does this mean other than that it is necessarily true? Does Kripke mean that something can be known a priori, without being necessarily true? Such a view is desperately in need of defence, which Kripke doesn’t deliver.
Kripke’s next outburst is that the truth of theses (3) and (4) is an empirical “accident”, which “the speaker hardly knows a priori”. What about this? The fact is that (3) and (4) are definitions of two necessary conditions for being a name. They are true by matter of will: if you accept these definitions as defining necessary conditions for being a name (as most people do), then it is true that what you mean by being a name, is among other things to saitisfy these definitions. This is totally, utterly and completely clear and lucid a priori.
Thesis (1) is a definition, writes Kripke. I agree. It is a definition of one of the necessary conditions for being a name, according to one definition of what it is to be a name. Pity only that Kripke does not see this as regards to the rest of the definition. (I will return to the theses’ not really being theses later.)
We now move on to thesis (2). Here, Kripke uses two examples to try a refutation. The first concerns Cicero. Kripke claims that when most people think of Cicero, they think just of a “famous Roman orator, without any pretension to think either that there are only one famous Roman orator or that one must know something else about Cicero to have a referent for the name.” If this was really the case, most people would simply be wrong to think that Cicero has a referent, or else they must either have an unusual definition of what it is to be a name, or use Cicero in the manner not of a name, but as the abbreviation of an indefinite description or as a predicate. But what most people who use the name Cicero to stand for a famous Roman orator do mean by the name is actually not what Kripke says. I think you would agree with me that most people who use the name Cicero in this sense also ascribes to the name the property of being the famous Roman orator so named by experts in the relevant fields of science.
Kripkes next example concerns Richard Feynman, a famous physicist. Kripke claims that common people, when they talk of “Feynman”, do refer to Richard Feynman, even though they lay nothing more in the sense of the name than being a famous physicist. Thus, this example is just a repeat of the above one, and quite as miserably flawed. On the basis of these two examples, Kripke erroneously concludes that thesis (2) is refuted.
Going on to thesis (3), Kripke becomes really annoying. The example is a hypothetical situation where “Gödel” is defined as whoever discovered the incompleteness theorem of arithmetics. In this hypothetical situation, "Gödel" wasn’t in fact the one who discovered that theorem – that was done by Schmidt. Still, Kripke insists, ordinary people would not be referring to Schmidt when talking of “Gödel”. This, he concludes, means that thesis (3) is incorrect: one unique thing satisfies a definition of “Gödel”, and yet a person who thinks that “Gödel” is whatever is such a thing would still not think that this particular thing is “Gödel”. Do we need a genius to realize what’s wrong with this argument? Obviously, Kripke is using two different definitions of “Gödel”. The first definition says that "Gödel" is whoever discovered the theorem in question. Since Schmidt did this, he is this “Gödel”. But the other definition is the commonplace one, saying namely that “Gödel” is the man born in 1906, an Austrian-American mathematician and logician, answering to the full name Kurt Gödel, having a certain apperance, and whatever else is known by the experts. Sincerely, I can’t believe that this wasn’t apparent to Kripke himself. Either this is the result of some temporary madness on his part, or else he must’ve been consciously trying to mislead.
He continues with equally galling nonsense. He claims that since many people believe Columbus to have been the first European to set foot in the Western hemisphere, and that he was the first to realize that the Earth is round, these people, using the name “Columbus”, really refer to in the former case some Norseman, and in the latter case some ancient Greek. “But they don’t”. Well, mr Kripke, if they really do define Columbus in this way, and they consider Columbus to be a name such as defined by your seven "theses", I agree, for in that case, since the definite description of the name’s referent would be satisfied by two different persons, the name wouldn’t refer. Had on the other hand the same person satisfied the two criteria of the definite description, the name would have referred; and given that those criteria were the only ones contained in the description, people using the name Columbus in this sense could not refer to any other person. Since your thinking is so substandard on this point, we need not be disturbed that the person most people understand as Columbus is the historical person called Columbus by historians and well known in popular culture, and that for most people, this historical person would still be Columbus even if he wasn’t the first to discover America or that the Earth is round.
Thesis (4) is dealt with in the same manner as thesis (3). Kripke claims that even if we defined the name “Gödel” as referring to the discoverer of the incompleteness theorem, we would still think that “Gödel” exists even if he didn’t in fact discover the theorem. The error is just the same as above, namely a confusion of the commonplace definition of "Gödel", and an hypothetical one, à la Kripke. It is a sad fact for Kripke that definitions work the way they do, but unfortunately there is nothing to be done about it.
At this stage, Kripke thinks that he has refuted "theses" (2), (3) and (4), and so, since he thinks the truth of thesis (5) dependent on them, he disbelieves this too. Recalling his musings on "Gödel" and “Schmidt”, he adds: “I think that my belief about Gödel is in fact correct and that the “Schmidt” story is just a fantasy. But the belief hardly constitutes a priori knowledge...” What is this supposed to mean? What we have here is just the same confusion as before. Kripke has two definitions of “Gödel”, one his own, the other the hypothetical one. The reason why he can come to the conclusion that a thing satifying the hypothetical definition is still not the “Gödel” which this definition defines, is that he operates with the hidden assumption that in order to be “Gödel” it is, believe it or not, necessary that a thing satifies the coommonplace, different definition, in order for any other definition of “Gödel” to be able to be satisfied by it. The fact being that the “Schmidt” of the example wouldn’t do this trick, he doesn’t think that “Schmidt” would satisfy any definition of “Gödel” at all. This is as if someone would define “pure water” as Coca-Cola, but then say that Coca-Cola is still not “pure water”, because it is not pure water, according to the ordinary definition of water. Normally, this kind of mental operation is not performed. In connection to thesis (5), Kripke thinks that if this "thesis" was true, his belief that "Gödel" (Gödel) and not “Schmidt” is “Gödel” would be a priori true. But we can all easily see that that is not by any sound reason what the "thesis" says. It says simply that given that there is such a thing that it satisfies whatever definition of “Gödel” we choose, it is “Gödel”, and that this can be grasped a priori. If I define “an idiot” as one not being able to realize such a clear and distinct truth, then surely if Kripke satifies this definition, it is necessary that he be “an idiot”.
As we are now done with Kripke’s famous "arguments" against the Wittgenstein-inspired theory of reference of Searle, or at least what Kripke claims to be this theory, on now to his own alternative. Instead of a definite description giving a name its reference, Kripke argues that this is done by a “causal chain”, or “chain of communication”, which he seems to use synonymously. That several people manage to use a name and refer to the same object is not because they let the name stand for definite descriptions which have the same reference, Kripke maintains, but because the name, originally given to name some object, has travelled from mouth to mouth in a “causal chain”.
First, the main objection. Even if Kripke would be right that “causal chains” are what makes several persons use a name with the same reference, that wouldn’t be interesting for philosophy, but for linguistics. What we are looking for is not what causes a name to have reference, or the same reference for several individuals, but what, given that the name does refer to something, is the suffient and necessary logical conditions for it so doing. It is as if we had a triangle drawn on a blackboard, and to the question “what makes this triangle be a triangle”, got the answer that “that it was caused to become triangle” by the person drawing it. We want to know what it is to be a triangle, not what made that particular triangle! I may now state what I alluded to earlier: the “theses” examined by Kripke are not really “theses” at all, but definitions, definitions of a set of conditions which together are the necessary and sufficient conditions for being a name, that is, the definition of what it is to be a name.
The second objection concerns the “causal chain”. What really is this chain? And how does it get started? As for the latter question, Kripke actually allows that the reference may be given by a definite description, or by ostensive definition. These two, however, I will readily subsume under the category of definite description. As for the former question, once the name has been given, the name spreads by “various sorts of talk” and “from link to link as if by a chain”. More precisely what this means Kripke doesn’t explain. But, perhaps we can think for ourselves. When one has named something, one has agreed with oneself to let that name stand for whatever was intended as the reference of the name. In order to know what one intend to name, one must be able to understand one’s own intention. This is done either in words, or in mental symbols. So the name was put on something thought to exist and be such that it fits the description given by interpreting one’s intention. If such an object really did exist, the name did refer to it. Now, Kripke talks about a “causal chain” spreading a name among people, keeping the reference of the name intact. What do we require for that to be the case? Let’s assume that I have named something, for example a stainless steel rat hiding under my sofa. I call it “Pag-rått-sky”. Nevermind this silly name, just follow the example. Next, I talk to my neighbour and tell her about this rat. But being afraid for what she might think of it, I doesn’t say that it’s a stainless steel rat, or that it is intelligent and does a lot of my homework. I don’t even say that it’s in my room. The only thing she hears from me is that I once saw a rat and named it “Pag-rått-sky”. According to Kripke, when she then start using the name, she will refer to the rat under my sofa. But how could she possibly do that? “Pag-rått-sky” could be just about any rat there is, since I said almost nothing of where or when I saw it. In what way does she use name, to have it refer to the same thing as I? Well, the obvious – and only – solution is that she does as a matter of fact get this reference by means of a definite description. If she uses the name without being able to say what rat she is talking about, she isn’t referring to the same rat as I did by it, and, possibly, to nothing at all. If she does refer to a specific rat, she must be able to tell us, or at least herself, that that is what she does. But then, she will be employing a definite description. In case she wants to refer to the same rat as me, then because of the lack of information on it, she has only one choice. She will have to say that for her, “Pag-rått-sky” will stand for any rat such that it alone satisfies the description by which I have given “Pag-rått-sky” it’s reference.
I could go on about further errors, misconceptions and suspected sofisms I have found in Kripke’s text, but my time being limited, I would prefer to end here. If in any case someone comes up with a defence for Kripke, I will be sure to be interested, but alas, I dare not hope for it.
(c) Demokrates Sörensson; sprid gärna men ta aldrig betalt.