By Rick Grush
I.
1.
[All page references to the Collected Papers volume.]
Kripke argues against the prevalent description theory of names, and proposes what E calls the Causal Theory.
Evans begins by distinguishing two questions concerning proper names.
The first question has to do with the 'speaker's denotation'. I take it that this is the entity which the speaker's utterance on a particular occasion denotes. The second has to do with a name's denotation simpliciter. To see the contrast, consider the following example:
I am overhearing a conversation about a guitarist who used to transcribe songs for Frank Zappa, who replaced Yngwei Malmsteen in the band 'Alcatrazz', and who did two albums with David Lee Roth after Roth emasculated Van Halen by departing from the group. Because of a misunderstood sentence in this conversation, I take it that the name of this guitarist is 'Dan Folgleberg'. Now, if I then go on to say that I think Dan Folgleberg is the ass-kickinest guitarist on the planet, what *I say* is that Seve Vai is the ass-kickinest guitarist on the plantet. My us of the name 'Dan Folgleberg' denoted, on this occasion, on my lips, Steve Vai. But the name 'Dan Folgleberg, itself, denotes another individual who certainly does not fit that description at all.
>>[See Note 1 at bottom of these notes]
These questions are different. To paraphrase Evans: for a speaker to denote x it is not sufficient for him to simply utter something which is x's name.
There are thus two description theories of names not distinguished by Kripke, or even the description theorists. The description theory of speaker's reference holds that a speaker denotes x with the name 'NN' just in case x uniquely satisfies most of the descriptions S associates with the name 'NN'. Reference will be determined by fit between the speaker's beliefs about the bearer of 'NN' and objects in the world.
The description theory of name denotation says that in order for N to denote x, there must be a description, or set of descriptions, cullable from the speakers in a group who assume they are using the name to refer to the same individual, and this description or set of descriptions must be true of x.
E claims that Kripke's attack is leveled at the first theory only, and passes the second by. Kripke argues:
a) a subject can denote x by using his name NN, even if he has no description of x which he can fashion, and even when the only description S has associated with NN is false of x (the Godel example: can refer to Godel, even if the only description is 'proved the incompleteness theorem', and even if Godel did in fact not prove this). (thus the descriptive knowledge is not necessary)
b) a speaker S can fail to denote x by using a name NN when S's description is true of x (S does not denote Schmidt in the example, hence the descriptive conditions are not sufficient).
The descriptive conditions for speaker reference are obviosly not sufficient. If so, then I would be saying something true when I said that Dan Fogelberg is the ass-kickinest guitarist on the planet. This is clearly a reductio.
But what is important is the weaker thesis that some descriptive content is necessary for a speaker to denote something. This position is, Evans claims, the fusion of two thoughts:
i) in order to say something by uttering a sentence, one must have certain intentions: one must aim at something with the use of the name.
>>[Note: in Chapter 11 of VR, this condition is miantained. But what is intended is that one participate in a practice.]
ii) in order to intend something or believe something about x, one must possess a description uniquely true of it.
Regarding (i) Evans claims that it has whatever force it does because of a bad inference: the wind and parrots don't say anything because they don't intend to say anything, therefore, in order to say that p one must intend to say that p. E claims that the appropriate conclusion should be: in order to say that p one should intend to say something. If we stuck to the stronger conclusion, then much of our actual discourse would be on a par with the wind and parrots, because we, in many cases, are unable to intend to say that p, because we don't know enough to do so.
Regarding (ii), it means that in order for S to believe that a is F, then the following must hold:
There is a property phi (this is the descriptive content) such that
1. S believes that the phi is F
2. a is uniquely phi.
Accordingly, someone who contemplates a thought which is 'empty' is in essentially the same sort of state -- we can ths still atribute a content to empty thoughts (this is the motivation for going via russellain descriptions).
Evans says that his objections to (ii) are essentially Wittgenstienian. It may be context, and not description, which relates a thinker to an object.
>>[See Note 2 at bottom of these notes]
2.
p. 6:
"The Causal Theory as stated by Kripke goes something like this. A speaker, using a name 'NN' on a particular occasion will denote some item x if there is a causal chain of reference-preserving links leading back from his use on that occasion ultimately to that item x itself being involved in a name-acquiring transaction such as an explicit dubbing or the more gradual process whereby nicknames stick. I mention the notion of a reference-preserving link to incorporate a condition that Kripke lays down; a speaker S's transmission of a name 'NN' to a speaker S' constitutes a reference-preserving link only if S intends to be using the name with the same denotation as he from whom he in his turn learned the name."
E first examines the sufficiency of this condition. An example in favor is the pub example; S overhears talk in a pub, people using the name Louis, and saying things about this person. S can join in the conversation and say things about Louis, even if S has no accurate descriptive knowledge at all.
E claims that the Causal theory gives the right answer in this case, but does not acknowledge the role that the social context is playing here. To see the problem, imagine a much later time (after the social context has dissolved). The Causal theory claims that S will still be denoting x by using the name 'Louis', even if he says something like 'Louis was a basketball player' as the result of a confusion. Though this is not compelling, it provides some reason to doubt the sufficiency of the causal account.
E claims that it is the social context that S was in which linked his use with Louis XIII, because that context included members who would be prepared to adjudicate remarks made using the name 'NN' on the basis of facts in Louis XIII's biography (or something like that -- there is more on this example later on, where it will turn out that S is using the name deferentially for whoever the participants are using it to refer to). It is because the context provides this constraint that S can denote Louis XIII by uttering 'Louis'. [This reconstruction goes a bit beyond what Evans explicitly says.]
Another reason to be drawn to the Causal theory is that it promises to solve some instances of the problem of ambiguity: to supply a basis for taking the sense of S's utterance to be that p, when it could also be interpreted to be to the effect that q or that r. E claims that this isn't compelling, either, because two other principles (his intending to say p, and the audience finding it reasonable to conclude that he said that p), while not without problems, can also be applied to many cases which the Causal theory cannot address. Whatever theory eventually accounts for ambiguity elimination will probably not need to be supplemented by the Causal theory for just the special case of proper names.
Evans claims that the causal condition is not necesarry, either, in that one may excognitate the correct name for x and use it to denote x, even with no causal chain at all.
He closes the section by remarking that Kripke may admit all of this, but maintain that the denotation of a name *in a community* is still to be explained in terms of a causal chain of refrence preserving links (as opposed to the denotation of a use of the name by an individual not necessarily in that community).
3.
So Evans now considers this theory, the Causal theory in a community view. He says, suppose we consider a parallel theory for general terms. It involves two components: a component by which a name's denotation is established, and a component which explaines how that denotation is preserved. Such a theory could not explain how terms change their meaning, but this happens all the time without intentions to change the practice. (Evans suggests that the mechanism whereby a name gets a denotation is the same as the mechaism whereby that denotation is preserved.)
Change of meaning would be decisive against a causal theory of general terms, as change of denotation is decisive against the causal theory of names. Evans gives two exmples: the real case of the name 'Madagascar' which originally denoted a part of the African mainland, and the imaginary case of twins whose name tags were inadvertantly switched by a nurse. The explanation, in such a case, for why 'Jack' denotes x is that y was dubbed 'Jack' by the mother (or something like that).
The (unamended) causal theory is thus clearly inadequate. Somehow, the speakers' intentions must count in the determination of denotation.
Evans rehearses some final considerations, some results which an adequate theory must accomodate: The theory must allow that 'Madagascar' can change reference, but that 'Godel' does not come to denote 'Schmidt', and 'Goliath' does not come to denote the Philistine killed by David.
p. 11-12:
"For although this has never been explicitly argued I would agree that even if the the 'information' connected with the name in possession of an entire community was merely that 'Goliath was the Philistine David slew' this would still not mean that 'Goliath' referred in that community to that man, and therefore that the sentence expressed a truth. And if we simultaneously thought that the name would denote the Philistine slain by Elhannan then both the necessity and sufficiency of the conditions suggested by the Description theory of the denotation of a name are rejected. This is the case Kripke should have argued but didn't."
4.
Evans examines the two components of the description theory:
a) The denotation of a name is determined by what speakers intend to refer to by using that name.
b) The object a speaker intends to refer to by his use of a name is that which satisfies or fits the majority of descriptions which make up the cluster of information the speaker has associated with the name.
E claims that (a), while inadequate as a description of denotation fixing by individuals, might yet work for communities. (b) is the real inadequacy. It claims that the relation between the information and the name's denotation is one of fit, and Evans wants to claim that it is better to think of the appropriate relation as one of *source*. Evans makes an interesting analogy between the object denoted by a name and the object of vision. We would not claim that the object seen is the one which best fits the perceptual image, but rather is the object which is the source of the sensory information, even if it is distorted to some degree.
Evans links these two components (intended reference and information source) in the following way, p. 13:
"We must allow, then, that the denotation of a name in the community will depend in a complicated way upon what those who use the term intend to refer to, but we will so understand 'intended referent' that typically a necessary (but not sufficient) condition for x's being the intended referent of S's use of a name that x should be the source of causal origin of the body of information that S has associated with the name."
Section II.
5.
In this section Evans will sketch a theory of the denotation of proper names. He begins by making a distinction between reference and denotation. Consider the following example (which I have adapted from Donnellan): At a party I say to you 'The man in the corner drinking a martini is a spy', when in fact there is a woman, wearing a heavy rain coat, and drinking water from a martini glass, and she's not a spy. Now I can certainly *get it across* that p, or *communicate* that p (where p = the woman in the corner drinking water is a spy) by using a sentence which cannot strictly speaking by used to say that p. Also, I can refer to this person by means of the ill-fitting description, a description which does not denote that person.
In such a case, it is not the fit of the information which is at issue, but the fact that a certain object is the source of all, or the dominant amount, of the information used to make the reference. Evans then goes on to examine the notions of source and dominance.
Our beliefs about items are the result of information-gathering transactions. These can be done ourselves, or by others (and the information passed on to us). This can be by perception, or by other means (rifling through a brief case). These transactions are typically causal in nature -- that is they involve some sort of causal contact. We can thus distinguish between the item (if any) which is the causal source of the information we gather, and the item(s), if any, which the information fits.
Not all information is based on information-gathering episodes, some can
be deduced (there is a tallest man, and he is over 6 feet tall). Evans suggests
that a name introduced in this way may function much as the Description
Theory suggests (the descriptive names, like 'Julius').
Evans argues that there must be restictions which disallow the source of a legend from being the individual whom the legend is about. Furthermore, misidentification can bring it about that the source of information is not the same as the object which the information is believed to be about. Such cases point out the fact that a dossier of information may have x as an informational source for some of the information in it, even though most of the information is *of* or *about* y. Furthermore, given more (mis) information, x may come to be the dominant source of information, and become the item which dominates the cluster, so to speak.
Evans then gives an example in which he meets a man, who is shortly thereafter replaced by an identical twin. After some period of time (period variable), I meet the twin in a room, and identify him. There are three periods:
i) shortly after the switch, when the dossier is dominated by x (the first man), I have false beliefs about x ('x is in the room', for instance).
ii) a while later, neither will be dominant, and Evans claims that during this period, one will not be misidentified as the other, but there will just be confusion.
iii) after a long period, y (the twin) will dominate the dossier. Evans would then have a true belief about y, though will have misinformation about when he first met y.
>Evans makes much of this sort of example in Chapter 11 of VR.
E points out that dominance is not necessarily determined by amount -- there are other factors as well, such as interest, detail, etc.
Evans then goes into his Napoleon example. The diagram is read as follows: alpha is the person named Napoleon at birth, and his life is represented by a left-leaning slanted line. Beta is the imposter (call him Smith), whose life is represented by a right-leaning line. The heavy line represents the name 'Napoleon'. In diagram 1, alpha has the name 'Napoleon' until 1814, and then Smith takes over. In 2, the switch takes place in 1793. Thus in (1), alpha is responsible for the bulk of the exploits known as Napoleon's, except Waterloo, and in (2), it is beta who is responsible for the bulk of the exploits, including Waterloo.
In a (1) case, we would say that most peole had a false belief about who fought at Waterloo (it wasn't Napoleon after all). But in the (2) case, it would be the case that Napoleon did fight at Waterloo, and furthermore, he got him name by taking it over from someone in 1793.
Evans says that Donnellan claims that in a (1) case, someone, S, who said 'Napoleon fought skillfully at Waterloo', then beta would be the intended referent. But if he said 'Napoleon fought skillfully at Waterloo, unlike his performance in the Senate', then alpha would be the intended referent.
Evans claims that in the (1) case, though, it was simply not Napoleon who fought skillfully at Waterloo.
Page 18 has Evans' tentative definition:
" 'NN' is a name of x if there is a community C
1. in which it is common knowledge that memebers of C have in their repertoire the proceedure of using 'NN' to refer to x (with the intention of referring to x)
2. the success in reference in any paticular case being intended to rely on common knowledge between speaker and hearer that 'NN' has been used to refer to x by members of C and not upon common knowledge of the satisfaction by x of some predicate embedded in 'NN'."
This definition does not entail that 'the man we call 'NN' ' is a name, (notice that th Description Theory might, depending on some details, have this consequence) because its referential success does not depend on it being common knowledge that it has been used to refer to x.
The conditions in (1) and (2) above seem to embody what Evans will call in Chapter 11 of VR a 'Proper-Name Using Practice' (or PNUP). PNUPs are thus more stringent than Kripke's theory, because it does not allow cases where some one individual just dubbs something with a name causing it to be come common usage. Imagine a case in which a villager dubbed a girl 'Goldilocks', when there were in fact two twins present in the village. According to Kripke, the name 'G' would then refer to the girl thus dubbed, on Kripke's account, even if some of the villagers saw only the second girl. On Evans' account, there would be no reference for the name (unless the villagers happened to fall into two coherent communities, each of which had interactions with only one of the girls -- one group would then refer to one girl, and the other group the other girl, regardless of the one present at the dubbing).
Another example: scrolls are discovered which have mathermatical proofs, and the name 'Ibn Kahn' is written on the bottom. A coherent group of mathematicians uses the name 'IK' to identify the worls of the mathematician (even though 'IK' was the name of a scribe). Evans' definitions have the consequence that 'IK' becomes a name for the mathematician.
p. 19:
"Speakers within such traditions [the current-day mathematicians, e.g. -RG] use names under the misapprehension that their use is in conformity with the use of other speakers referring to the relevant item [the ancient mathematician's contemporaries and family, e.g. -RG]. The names would probably be withdrawn when that misapprehension is revealed, or start a rather different life as 'our' names for the items... One might be impressed by this, and regard it as a reason for denying that those within these traditions spoke the literal truth in using the names. It is very easy to add a codicil to the definition which would have this effect."
E claims, though, that this is not a good reason for denying that those people in the community are speaking the literal truth. But a decision on this matter is not important. What is important is that either way, the definition a) allows for change in denotation, and b) denies that the mathematicians are speaking falsely of the scribe.
In regard to (a):
There is a standard way in which people get their names (real dubbings at birth, for instance), and this conditions our use of names. I take it that Evans means here that because of this, we are deferential to the name someone is given at birth (This would account for why there would be some pull, even in the Napoleon Case 2, to say that Napoleon was not involved in Waterloo, or any of the other earlier previous military stuff. But we can leave this aside as a type, one type among others, of deference. The general theory should then give an account of how names get their reference, and this should have room for understanding deference to particular practices, particular people, or whatever, in deciding on reference relations. It would then be able to handle birth-names and such as a special case, and not as the paradigm for all reference.)
So consider nicknames as a case of a name without the baptismal baggage. Evans claims that in such cases, change of denotation, specifically because the names do not come with this additional baggage, is not as much of a problem. p. 20:
"So long as they have no reason to believe that the name has dragged any information with it, speakers will treat the revelation that the name had once been used to refer to something different with the same sort of indifference as that with which they greet the information that 'meat' once meant groceries in general."
Evans then discusses this in relation to the Napoleon example.
He then mentions briefly cases in which deference is made not to this or that event (real baptisms) in assigning reference, but cases in which there is deference to a person or group in establishing the reference ralation (elms, gold, e.g.).
Thus in the pub example, where people we discussing Louis XIII, the man S was using the name deferentially to refer to whoever the group was referring to. Even if S had an opinion about who they were talking about, he would still defer to them in case he was wrong.
Yet another example: archeaologists find a tomb, and believe (falsely) it to be of a character in the bible. In such a case, even if they learned a great deal about the man in the tomb, they would still be saying false things about the biblical character, because of the deference they would likely make to biblical scholars. [If, on the other hand, they found a name, 'Goober' inscibed on the tomb, and under a misapprehension thought that 'Goober' was the man's name, they would be using 'Goober' as a name for the man. The difference is that in this case there is no reason to be deferential to any particular usage, and the group practice takes over.]
Evans mentions that sometimes a name can be used deferentially to refer to whoever fits a description (a sort of descriptive name) like 'Jack the Ripper'.
All of this leads Evans to distinguish two reasons for withdrawing sentences using a name, p 22:
"We must thus be careful to distinguish two reasons for something that would count as 'withdrawing sentences containing the name'
(a) the item's not bearing the name 'NN' ('Ibn Kahn', 'Malachai')
(b) the item's not being NN (the biblical archeologists)."
Evans ends with an example which is worth quoting at length, p. 22-3:
"A youth A leaves a small village in the Scottish highlands to seek his fortune having acquired the nickname 'Turnip' (the reason for choosing a nickname is I hope clear). Fifty or so years later a man B comes to the village and lives as a hermit over the hill. The three or four villagers surviving from the time of the youth's departure believe falsely that this is the long-departed villager returned. Consequently they use the name 'Turnip' among themselves and it gets into wider circulation among the younger villagers who have no idea how it originated. I am assuming that the older villagers, if the facts were pointed out, would say 'It isn't Turnip after all' rather than 'It appears after all that Turnip did not come from this village.' In that case I should say that they use the name to refer to A, and in fact, denoting him, say false things about him (even by uttering 'Here is Turnip coming to get his coffee again').
"But they may die off, leaving a homogenous community using the name to refer to the man over the hill. I should say the way is clear to its becoming his name. The story is not much affected if the older villagers pass on some information whose sourse is A by saying such things as 'Turnip was quite a one for the girls', for the younger villagers' clusters would still be dominantly of the man over the hill. But it is an important feature of my account that the information that the older villagers gave the younger villagers could be so rich, coherent, and important to them that A could be the dominant source of their information, so that they too would acknowledge 'That man over the hill isn't Turnip after all.'
"A final possibility would be if they used the name deferentially towards the older villagers, for some reason, with the consequence that no matter who was dominant they denote whoever the elders denote."
6. Conclusion.
Evans here comapres and contrasts his view with the Descriptive and Causal accounts.
***************************
Note 1: Evans in VR also draws similar distinctions, but more of them. I
quote from my notes to Chapter 9:
______
p. 320:
"Clarity in the theory of reference will be served if the following
concepts are clearly distinguished: that of the referent of an expression
as used by a particular speaker on a particular occassion; that of the intended
referent of an expression; and that of the object which a speaker means."
The distinction is this:
Suppose on monday I see a man who is drunk. On wednesday, I see a different man, but mistakenly assume that it is the same person. I thus, looking at that person, say 'That woman was drunk on monday.'
In this case, the *intended referent* is the man I have a demonstrative thought about on wednesday. But even though this person is the intended referent, my *expression* had no referent, because I mistakenly uttered 'woman', and thus blundered wrt the conventions of the langauge. I thus failed to refer to anything. Had I said 'man', then the intended referent would have been the actual referent of the expression. (Recall, being the intended referent is not sufficient for being the actual referent.)
Now the *object which I mean* actually does not exist, because I have mistakenly identified the person I saw on wed. with one I saw on monday.
The fourth distinction (mentioned by Evans in footnote 16), is *the object the speaker is talking about*, and this is the object, like the cylinder, or Mr. X of the legal department, from the examples in section 5.3, p. 130-131)
[end quote from Chapter 9]
_______