« Ed Schultz Plays the Race Card | Main | A Searle-y Objection to the Causal Theory of Names »

Wednesday, April 27, 2011

Comments

Feed You can follow this conversation by subscribing to the comment feed for this post.

Your examples bring out, in my view, the problem with description theories of singular reference. Suppose that, according to religion A1, there is one god who satisfies the uniquely applying description F1. And suppose that, according to religion A2, there is one god who satisfies the uniquely applying description F2. And suppose, for sake of argument, that F1 and F2 are contraries. Nothing can be both F1 and F2. Then if religion A1 is correct, there is exactly one F1 and (from the logical properties of F1) there can be no being that satisfies F2. Conversely, if religion A2 is correct, there is exactly one F2, and there can be no being that is F1. According to a description theory, neither god can be identical with the other, if either exists.

But now suppose that both religions have the same ‘core’ holy book B, which refers to one god by name G, but which gives no uniquely defining property for that god. However, at some point in the past there was a disagreement over some heresy, and each religion now has a different supplementary holy book – B1 in the case of religion A1, and B2 in the case of religion A2. It is from these supplementary texts that the disagreement over the uniquely defining properties F1 and F2 arises. But both texts refer to god using the same name G as the core book B. Thus the intended or purported reference of G is the same for both the competing religions. The disagreement is not about the identity of the purported referents of G – for these have to be one and the same being – but rather the properties F1 and F2. Both religions agree they are worshipping one and the same god, but they disagree about whether he has the uniquely defining F1 (in the case of A1) or the uniquely defining F2 (in the case of A2).

Something like this holds for Judaism and Christianity. The purported referent of ‘YHWH’ is the same for both Christians and Jews. That is a logical conclusion from the semantics of the Old and New Testament names. But the properties are different. Christians hold that YHWH had a son, who is one and the same god as YHWH, but a different person. Jews do not hold this belief. So a description theory has the god of the Jews and the god of the Christians as being numerically different. But a reference theory (which holds that identity is determined by *intended* identity or intended referent) has them as being the same.

I know little about Islam, but I understand that the Islamic holy books are the records which Muslims believe were dictated by God to various Islamic prophets throughout the history of mankind. I.e. Moses (according to Muslims) was an Islamic prophet. Since Christians, Jews and Muslims agree that YHWH gave the laws to Moses. Thus the intended reference of ‘YHWH’ is the same for all. I.e. if YHWH exists, all three religions are referring to him, even though they disagree fundamentally, and bitterly, about his properties. E.g. Christians believe YHWH he has a son in Jesus. Muslims believe he has a prophet in Jesus, Jews disagree with both of these things, believing Jesus to have been a heretic and a blasphemer, for which reason he was executed.

I hope this comment is relevant to your post!

Thanks for the post.

A possible drawback to your tentative conclusion is that it would appear to require (say) both myself and W. L. Craig to worship different Gods (we differ on the Divine simplicity and foreknowledge, among other things), which conclusion seems intuitively implausible.

I don't know if it has a bearing on your argument, but the Koran asserts the God of Muslims and the God of the "People of the Book" to be one (29:46)

It seems to me that there is a fairly straightforward solution to this problem. Let S1, S2, and S3 be each a set of descriptions of God. Each of these sets corresponds to a set of descriptions associated with God in each of the monotheistic religions (Judaism, Christianity, and Islam).

Now, some descriptions will be in all three sets; e.g., 'the unique x which is the creator of the world'; 'the unique x which is omnipotent'; 'the unique x which is omniscient'; 'the unique x which is morally perfect'.

Other descriptions occur only in one or two of the sets, but not all three: e.g., 'the unique x which is triunite'; 'the unique x which has no son'; etc.

Let C be the largest non-empty overlapping set of descriptions such that every member of C is in each of the sets S1, S2, and S3. We are assured that in this case C is not empty. Moreover, we are also assured the C is unique. We can now take C to fix the referent, if such exists, of the term 'God' (or its translation) in each of the monotheistic religions. While it is possible that no entity satisfies all the descriptions in C, we know that if such an entity exists, then there is one and only one such entity (due to the nature of the descriptions). Call C the *referential conception* of God.

What about the other descriptions that are in one of the other sets, but not in C: i.e., are not members of the referential conception of God? Well, all descriptions that are not shared by all three monotheistic religions, and hence are not in C, can be viewed as part of the epistemic-corpus of individual religions. These may or may not be true of the only God, even if such a God exists and it satisfies all the descriptions included in the referential conception C.

Now, clearly the adherents of each religion believe all descriptions included in their respective epistemic-corpus just like they believe every descriptions included in the referential conception. Each such adherent will insist that every one of the respective descriptions in their epistemic-corpus is true of the one and only one God. But this fact need not enter into the meta-considerations discussed presently. What adherents of each religion believe about God is an internal matter to each of these religions. Hence, it cannot be taken as a criterion of reference of the term 'God'. This question is settled by considerations about how to fix the reference of the term. And these considerations are external to each religion.

Leo,

You are quite right to draw the conclusion you draw. If you conceive God as simple, but Craig does not, then you two are referring to different Gods. Trinitarians who differ on the *filioque* refer to different Gods. And so on.

If, resisting this conclusion, you say that Jews, Christians, and Muslims refer to and worship the same God, will you take the further step of saying that the foregoing plus the Brahman-worshipping Hindus all refer to the same Absolute Reality?

Will you say that the Plotinians and the Muslims refer to the same Absolute, but just differ as to how they conceive it?

Are all religions one in that they are all directed to the same Absolute?

Thanks for your comment.

This is a very interesting discussion, rather different in tone from the statements of Rev.Franklin Graham et al denying the identity of the God of Islam with the God of Christianity. It would be helpful if all discussions comparing Islam and Christianity were conducted in this manner.

Islam, Judaism, and Christianity affirm in their sacred writings the intention to worship "the God who revealed himself to Abraham." From a traditionalist perspective affirming the historicity of the Abraham story,this phrase is taken to refer to a particular agent (God)who acts-reveals at a particular time and at a particular place to a particular person. Given this characterization of their worship object / 'God' it would seem to be the case that Muslims, Christians, and Jews are committed - and should see themselves as committed - to the worship of the same Divine Reality, 'the God of Abraham,' nothwithstanding further disagreements in their theologies.

Most Christians see Jews and Christians as worshiping 'the same God,' despite their significant theological differences.
Perhaps this is intellectual confusion on their part, but consistency would seem to require that those Christians also see themselves and Muslims as worshiping 'the same God,' despite the significant theological differences. If theological differences make the 'God of Islam' a different God than the 'God of Christianity,' then consistency requires the conclusion that the 'God of Judaism' is a different God than the 'God of Christianity.'

Edward,

That is an excellent comment and (forgive me for saying it) much better than your usual comments. Your first paragraph captures exactly what I was driving at, assuming we adopt a description theory of names.

You speak of "intended or purported reference." I can agree that the intended reference of 'Yahweh' or 'God' or 'Allah' is the same for the Jew, the Christian, and the Muslim. But I spoke about successful reference. One condition on successful reference is that the item referred to exist. A second condition is that the term used in making the reference pick out that item and not some other one. It might be that the Jew, the Christian, and the Muslim intend to refer to the same being, a being that exists, but only the Christian succeeds in referring to a being that exists.

On a description theory, a subject S using a term T succeeds in referring to object O iff O satisfies the definite description that unpacks the senses of T. On a causal theory, S's use of T refers to O iff S's use of T can be traced back via a causal chain to an initial baptism.

So as I see it, the question bioils down to a question in the phil. of lang ass to whether the description theory or the causal reference thery is to be adopted.

But you speak of a "reference theory" -- very bad terminology since every theory of reference is a reference theory! -- according to which the identity of the referent is determined by what the speaker intends. But who ever held such a theory?

Note that whether or not 'God' refers to anything cannot depend on whether I intend it to refer to something. The cooperation of the world is needed for successful reference. And whether 'Obama' and 'POTUS' refer to the same being cannot depend on whether or not I intend that they should: it depends on whether the reference of the first term has the properties encapuslated in the second term. (I trust you understand what 'POTUS' and 'SCOTUS' mean in Stateside political jargon.)

Peter,

You are telling us in effect that there is something common to the Jewish, Xian, and Muslim conceptions of God, and that this common conception fixes the reference of 'God.' Yes, there is something common, as I pointed out. It would be nice, though, if you told what exactly reference-fixing is. I seem to recall that notion from Kripke.

The common conception is an abstraction from the specific conceptions, and the latter are mutually exclusive. Thus the Xian conception includes the property *triune* while the Jewish and Muslim conceptions include the complement of that property. It follows that nothing instantiates all three specific conceptions. Now if reference is determined by sense, and *triune* is included in the def. descr. that unpacks the sense of 'God' as used by the Xian, then it follows that the Jew, Muslim, and Xian cannot refer to the same being.

So I don't understand how what you said is relevant to my question which, again, is: Does the normative Christian and the normative Muslim worship numerically the same God, or numerically different Gods?

You are not just assuming that the description theory is false, are you?

I'm not sure that anybody has a logically coherent explication of the Christian notion of the trinity, so I'm not sure just how much of a role that notion can play in fixing reference. I suppose it could restrict reference to a being about whose identity there is something profoundly mysterious, but that's a pretty loose sort of designation.

Bob,

You are in the vicinity of an important point. If the doctrine of the Trinity is unintelligible, then it cannot be part of the sense of a definite description whether the description is used to fix reference, or is the vehicle of a name's reference.

Peter,

I think your talk of reference-fixing presupposes a causal theory of names. As I recall, Kripke gives the example of 'the celestial object that causes the perturbation of the orbit of Uranus' as a definite description that fixes the reference of 'Neptune.' But as you know, that is not to say that 'Neptune' refers to x iff x satisfies the definite description in question.

Bill,

1) "Now if reference is determined by sense, and *triune* is included in the def. descr. that unpacks the sense of 'God' as used by the Xian, then it follows that the Jew, Muslim, and Xian cannot refer to the same being."

I agree. Particularly if it turns out that (as Bob noted) the doctrine of the Trinity is ultimately an inconsistent doctrine. Then no expression with that sense can have a referent. Are we to say that merely because the official doctrine of each of these theistic conceptions contains inconsistencies (or falsehoods), the term 'God' contained in them fails to refer? I think that we should avoid making a determination regarding the existence of God based upon a theory of reference *alone*. And there are other phenomena to consider.

First, there may be several still unknown cases of inconsistency (implicit or explicit) embedded within the official belief systems of each of the three monotheistic conceptions. If we take all descriptions to be part of the sense, then no theistic doctrine can be said to talk about God, since the term would have no referent in any of them.

Secondly, within each of the monotheistic conceptions there are various denominations which disagree on fundamental points of doctrine. Are we to say that within the Christian conception, for instance, there are x number of different denominations each referring to a different god? Suppose there is a debate among two of these denominations on a fine doctrinal point involving God. If they do not refer to the same thing, then they cannot disagree. But they certainly appear to disagree.

Third, a similar problem arises in science (and most likely in everyday discourse). Does the term 'mass' in Newtonian physics refers to the same thing/force as used in Einstein's? If not, how can we say that Newton is wrong about mass? And so on.


It is for this reason that I suggested to take some subset of these descriptions as reference fixers. Since reference fixers are not part of the meaning (sense) of the term, two disputants may be said to talk about the same thing even if they have incompatible beliefs. Since there is enough in the three theistic conceptions that is shared, we can say that the term 'God' refers to that entity which is fixed by this common stock of descriptions. Then, we can say that where they disagree, they disagree about one and the same God fixed by their common descriptions.

2) "So I don't understand how what you said is relevant to my question which, again, is: Does the normative Christian and the normative Muslim worship numerically the same God, or numerically different Gods?"

The answer is "yes", provided, of course, they share at least some beliefs that can be used as reference fixers. And it seems to me they do. It is for this reason we can also say that they fervently disagree about other matters pertaining to the very same God.

3) " But as you know, that is not to say that 'Neptune' refers to x iff x satisfies the definite description in question."

Correct. 'if' but not 'only if':i.e., *If* x does satisfy the description 'd' for the name 't', then 't' refers to x. But this is not a necessary condition. Hence, a description is not part of the sense of a name.

4) Kripke's theory of reference does include a causal component. Of course, this in itself need not pose a problem regarding the term 'God'. A baptism may be routed through a description such as 'the creator of the world' as one of the descriptions and since all three theistic conceptions include something like this description, this would be a shared reference fixer.

5) Does the causal component create a problem regarding reference to causally inert objects such as numbers, propositions, etc., (The topic of your newer post)? Well, Kripke has a paper titled (I believe) "De-re beliefs about Numbers" in which he deals with this problem. I might have a copy (I am not sure at the moment). It is also worth noting that some followers of Kripke (e.g., Alan Berger) developed extensions of the theory which deal with a variety of other issues. I do not remember exactly Alan's theory but he has written a book called "Terms and Truth" which contains this material. I think it is relevant to some of the issues raised here.

>>That is an excellent comment and (forgive me for saying it) much better than your usual comments.

Hmm. I think all my comments are equally good, but some are more liable to confusion and misunderstanding than others. Much of the problem is terminological, plus confusion about core assumptions that need to be made explicit (e.g. about the meaning of the word 'exist', which caused endless difficulties before).

I've tried to be clearer about my terminology below, but note there is no standard and agreed meaning to terms like 'refer' and 'denote' and so on.

>>every theory of reference is a reference theory

Not every theory of individuation is a reference theory. A term individuates when it tells us which individual the sentence is about. According to a reference theory, the term tells us which individual the sentence is about without telling us anything about it. The term is a mere label, on this account. According to a description theory, a term tells us which thing the sentence is about (or aims to do so) by giving a description that one individual uniquely satisfies. The relation between the description and the individual that uniquely satisfies it is sometimes called 'denotation'.

>>I spoke about successful reference. One condition on successful reference is that the item referred to exist.

Can I challenge you on this one? It seems to me that names in fiction are merely labels (i.e. they are referential, not descriptive) yet they succeed in telling us which fictional character the sentence is about. So they are referential, in the sense defined. So perhaps there can be reference to that which does not exist (to use your preferred way of putting it, which is not my preferred way). Tolkien tells us which hobbit found the ring in a fish (Smeagol). He tells us which hobbit celebrated his birthday at Bag End (Bilbo). He told us which hobbit carried the ring into Mordor (Frodo). And so on. Yet nothing is a hobbit (or, if you like, hobbits are things which do not exist).

So, just as we can think about things that do not exist - think about a hobbit, perhaps - can we not also refer to things that do not exist? This would solve a number of your difficulties. If we can use names to refer to what does not exist, and if names have this individuating power independent of the existence of the referent, then

(1) We don't need a different theory to explain the semantics of fiction than we do to explain a historical, factual narrative. I understand the story of the garden of Eden the same, whether or not it is true. I.e. I know, because the story tells me, which of the two people in the garden picked the apple from the tree. So I know the meaning of the name 'Eve', but I can't use this knowledge to determine whether or not Eve actually existed.

(2) We can explain your puzzle about the identity of the Christian and Muslim God very easily. If the names 'Allah' and 'God' refer to the same individual entity, and they have a 'referential' rather than 'descriptive' function, then it doesn't matter what features the different religions ascribe to God. The names are simply labels. If God and Allah do not exist, then the names are labels for what does not exist.

(3) And they can do so whether or not God exists. Whether or not God exists, if the two names refer (i.e. individuate by reference, not by description) to a single entity, then the two Gods are identical. If God exists, then Allah exists, and one or both of the religions will have been wrong about the properties of God. That's all.

(4) We can explain your puzzle about naming hurricanes ('Hillary'). If we can refer to things that do not exist, we can refer to things that do not yet exist.

Of course, it may seem odd to say that there can be a relation (reference) between a name and something that does not exist. But that should be no odder than thinking about, wishing, desiring etc things that do not exist. If one is possible, why not the other?

I hope that is clear. I have tried to keep it clear by using terminology that I deplore ("reference to the non-existent"), but it will hopefully be less confusing. (Fingers crossed).

Peter writes, >>4) Kripke's theory of reference does include a causal component. Of course, this in itself need not pose a problem regarding the term 'God'. A baptism may be routed through a description such as 'the creator of the world' as one of the descriptions and since all three theistic conceptions include something like this description, this would be a shared reference fixer.<<

This is not clear to me. Consider a mundane example. I go to PetsMart to acquire a cat. I see a kitten I like and I attach the name 'Max Black' to it in a Kripkean initial baptism. I don't know what it means to say that that baptism is "routed through a description." The whole point of direct reference theories is to nail down the name-nominatum link without the use of any description. In other words, what makes 'Max Black' refer to the kitten is my slapping of that label on the poor cat. The reference is not in virtue of the sense of 'Max Black' if it has one. Suppose I name the cat 'Dog.' 'Dog' has a sense, but that sense plays no role in the securing of reference.

Suppose I have a tremendous mystical/religious experience, perhaps like the one Pascal reports. I attach some name or title to the intentional object of the experience, the name 'God' or 'Lord.' That initial baptism does not require my conceptualization of the object of the experience as "the creator of the universe."

And note that if it turns out that the being I name 'Lord' is a mere emmissary of God himself and not God himself, that does not deprive my use of 'Lord' of reference.

@Peter - what does 'fixing the reference' mean?

Suppose I say

(1) We don't know the name of the first dog born at sea, but let's call it 'Fido'.

(2) Fido was male.

In (1) we have denoted some object by a uniquely applying description: "first dog born at sea". So if I merely said "the first dog born at sea was male" I have a sentence which (if the theory of descriptions is correct) is true if such a dog existed, and if that dog is male, and is false if either (a) such a dog existed, and was not male or (b) no such dog existed. I.e. because the subject is descriptive, the negation of that proposition has two causes of truth (not being male, or not existing).

But when I 'fix the reference by description' and name that dog "Fido", and utter (2), is the proper name acting as a concealed description (for 'first dog born at sea') or not? If as a concealed description, then Bill's original argument applies. But if not, we have the puzzle of what happens if the original description fails to apply. That is, if the name in 'Fido is male' is referential, and if the sentence is false, then 'Fido is not male' is true, and implies (because of referentiality) that something (namely Fido) is not male. So Fido exists. But that seems inconsistent with the possibility that the description may not be satisfied.

In summary, a proper name whose reference is 'fixed by description' is either a concealed description or it is a logically proper name. If a description, the original problem posed by Bill remains. If not, then the proper name has existential import, which is inconsistent with the possibility that the description is empty.

Hope that makes sense.

Edward,

That makes excellent sense and you raise a good question. But I must quibble over your terminology. You shouldn't speak of names' being referential when you mean *directly* referential. If 'Fido' is a definite description in disguise, then it is not a logically proper name or Kripkean rigid designator. Now if the description which is 'telescoped' or abbreviated by 'Fido' is satisfied, then 'Fido' refers, albeit indirectly: the reference is routed through, or mediated by, the sense of the name.

Of course, you may be making a distinction between reference and satisfaction and holding that a name that is a description in disguise does not *strictu dictu* refer.

That terminological quibble aside, the question you put to Peter I would also put to him.

Suppose we introduce 'Vulcan' by means of the reference-fixing description 'the planet whose orbit is within the orbit of Mercury.' We know that the description is not satisfied. Thus we know that 'Vulcan' does not have a referent. But perhaps one could say that it nonetheless has a reference.

The underlying problem here may be how a direct reference theorist handles negative existential such as 'Vulcan does not exist.' The descriptivist won't have any trouble. But what must the direct reference dude say?

Bill, Edward,

A couple of quick points.

1) Various Related but Distinct Theories of Reference: One must distinguish between rigidity, direct-reference-theories, and causal-theories of reference. While these may be related, they are different. Some terms that are rigid may not be directly-referential, etc. Kripke has not committed officially to direct-reference. He has a causal component but it is not purely causal.

2) Fixing Reference: A description may fix the reference of a term, but not give the meaning. Kripke introduced the term 'fixing reference' in order to say that some descriptions may be used to fix the reference without giving the meaning.

A description that is used to fix the reference does so in the same way as any description picks out an object; i.e., via property instantiation or satisfaction. Except that when a description is used merely to fix the reference, but not give the meaning, then it may be false of an object in some possible world. The name used to fix the reference, however, still refers to the same object.

There may be cases where a name is introduced via a baptism and one uses a description to fix the reference of the name. There are no problems here that do not appear already with any use of a description to pick out an object.

So I am puzzled by both of your comments.

Bill >>But I must quibble over your terminology.

Indeed you might, and it was for this reason that the translators of Frege's work dropped the translation of 'reference' for Bedeutung, and used the word 'meaning' instead. I have a note here http://ocham.blogspot.com/2011/05/reference-and-bedeutung_01.html on Frege's use of the term, where it is clear that he was using it in the sense of 'direct reference'. And if we do not get the concept of 'reference' from Frege, where do we get it from?

To avoid all confusion, I try always use the term 'refer' for Frege's meaning, and 'denote' for the relation between a descriptive use of a term, and the object(if any) that satisfies it.

@Peter >>A description that is used to fix the reference does so in the same way as any description picks out an object; i.e., via property instantiation or satisfaction. Except that when a description is used merely to fix the reference, but not give the meaning, then it may be false of an object in some possible world. The name used to fix the reference, however, still refers to the same object. <<

My point still applies.

1. A name either signifies descriptively, in which case it may fail to apply to any object, or 'referentially', in which case it must apply to some object. Bill had a quibble about this, but see my note here http://ocham.blogspot.com/2011/05/reference-and-bedeutung_01.html. 'Reference' in the Fregean sense, = 'direct reference'.

2. Bill argues that there is a problem about the term 'God' if it signifies descriptively. You seem to have accepted his arguments here (correct me if I am wrong).

3. You have argued, apparently, that there is some kind of half-way house by allowing the 'fixing of the referent' using a description, then using a name referentially of the referent thus fixed.

4. I have argued that there is no such half-way house. Either the name established by referent-fixing is used descriptively or 'referentially'. If descriptively, then you must concede that Bill's argument applies. If referentially, you must explain how a name can be referential when there may be no object to which it refers. You seem to suggest that this may be possible because an empty name (i.e. that names nothing in this world) may name something in some possible world. Is that what you meant? If so, then I have other questions.

The comments to this entry are closed.

Google Search Engine

My Photo
Blog powered by Typepad
Member since 10/2008

Categories

Categories

September 2017

Sun Mon Tue Wed Thu Fri Sat
          1 2
3 4 5 6 7 8 9
10 11 12 13 14 15 16
17 18 19 20 21 22 23
24 25 26 27 28 29 30
Blog powered by Typepad