Singular Meaning

Edward Ockham of Beyond Necessity is back from his Turkish holiday and reports that, besides lazing on the beach at Bodrum, he

. . . spent some time thinking about singular concepts. Do you accept singular meaning? Either you hold that a proper name has a meaning, or not (Aquinas held that it does not, by the way). If it does, then what is it that we understand when we understand the meaning of a proper name? The scholastics held that there was a sort of equivalence between meaning and signifying ("unumquodque, sicut contingit intelligere, contingit et significare"). What I signify, when I use a term in the context of a proposition, is precisely what another person understands, when he grasps that proposition that I have expressed.

Do I accept singular meaning?  That depends on what we mean by 'meaning' and by 'singular.'  Let's see if we can iron out our terminology.

1. Without taking 'sense' and 'reference' in exactly the way Frege intended them to be taken, I would say that 'meaning' is ambiguous as between sense and reference.  Unfortunately, Edward seems to be using 'meaning' to mean 'sense.'  Of course, he is free to do that.

2. Edward also uses the word 'signify.'  I should like him to explain exactly  how he is using this word.  Is the signification of a proper name the same as what I am calling its sense? Or is the signification of a proper name its  referent? Or neither? Or both?

3.  Suppose I assertively utter a token of 'Peter is tired' in the presence of both Peter and Edward.   My assertion is intended to convey a fact about Peter to Edward.  The latter grasps (understands) the proposition I express by my assertive tokening of the sentence in question.  And of course I understand the same proposition.  What I signify -- 'express' as I would put it -- by my use of 'Peter' is what Edward understands when he grasps the proposition I express. 

4.  Now the issue seems to be this.  Is the meaning or signification or sense  I express, and that I understand,  when I say 'Peter'  a singular meaning?  More precisely: is it an irreducibly singular meaning, one that cannot be understood as logically constructed from general concepts such as man, philosopher, smoker?

5. I say No!   I don't deny that 'Peter' has a sense.  It has a sense and a referent, unlike 'Vulcan' which has a sense but no referent.  But the sense of 'Peter' is not singular but general.  So, to answer Edward's question, I do not accept singular meaning.

Corollary: the haecceity of Peter - Peterity to give it a name -- cannot be grasped.  All thinking is general: no thinking can penetrate to the very haecceity and ipseity of the thing thought about.  One cannot think about a particular except  as an instance of multiply exemplifiable concepts/properties.  This is 'on all fours' with my earlier claim that there are no singular or individual concepts.  The individual qua individual is conceptually ineffable.  So if we know singulars (individuals) at all, we do not know them by conceptualization.

If Edward disagrees with this he must tell us exactly why.  He should also tell us exactly how he is using 'proposition' since that is another potential bone of contention.  Is he a Fregean, a Russellian, or a Geachian when it comes to propositions?  Or none of those?

Comments

[arash]

Bill,

1. You say: "'meaning' is ambiguous as between sense and reference"
Could you kindly provide your definition of both of them?
2. You say: "So if we know singulars (individuals) at all, we do not know them by conceptualization"
How else do we know them?

thanks

[Bill Vallicella]

The sense of a word is what you understand if you understand the word whether or not there is anything to which the word applies. The referent of a word is the thing or things, if any, to which the word applies.

It may be that we don't know singulars as singulars. Or perhaps we do know singulars as singulars by sense perception.

[Edward Ockham]

First point: I have expanded on my comments to your earlier post http://maverickphilosopher.typepad.com/maverick_philosopher/2011/08/the-aporetics-of-singular-propositions.html , and on the comments to it by Peter Lupu and David Brightly, on my blog here http://ocham.blogspot.com/2011/08/sentence-negation-predicate-negation.html . I reaffirm my thesis that there is no ‘middle way’ between Direct Reference and the TFL position that there is no fundamental logical distinction between singular and general terms. I.e. the MPL view that there is an absolute and fundamental category distinction between singular terms – symbolised by lower-case letters - and general terms – symbolised by upper-case letters leads to Direct Reference, and the absurd and impossible result that empty singular terms have no meaning whatsoever. There is no halfway-house that allows for ‘sense’ as opposed to ‘reference’.

Accepting this point also indirectly answers Peter’s question about ‘why bother’ with an archaic (‘Ptolemaic’) system like TFL. Answer: because MPL leads to an absurd and impossible result. The category distinction between singular and general terms leads to Direct Reference, and Direct Reference is absurd and impossible. Case proved.

Moving to your questions.

>>. Is the meaning or signification or sense I express, and that I understand, when I say 'Peter' a singular meaning? More precisely: is it an irreducibly singular meaning, one that cannot be understood as logically constructed from general concepts such as man, philosopher, smoker?

I say ‘yes’. The function of a singular term is to allow me to tell you, or to signify, which individual I am talking about. A singular term such as a proper name can do this without telling you anything else (i.e. without giving you any other information such as ‘is a man’, ‘lives in London’ etc).

Your question presupposes an understanding of ‘general concept’ versus ‘singular concept’. I say a general concept is one which we can entertain under the aspect ‘predicable of more than one’. A singular concept, by contrast, is one which we can only think of or entertain under the aspect 'not predicable of more than one’. A concept signified by ‘a’ is singular when the thought expressed by ‘a is F’ and the thought expressed by ‘a is G’ are inconsistent with the thought that no F is G. The thought that Socrates is a philosopher and that Socrates is arguing is inconsistent with the thought that no philosopher is arguing. By contrast, the thought that some philosopher (Socrates) is Greek and that some philosopher (Cicero is arguing) is perfectly consistent with the thought that no Greek person is arguing.

>> I say No! I don't deny that 'Peter' has a sense. It has a sense and a referent, unlike 'Vulcan' which has a sense but no referent. But the sense of 'Peter' is not singular but general. So, to answer Edward's question, I do not accept singular meaning.

So we disagree. But your account faces a problem that we have already agreed is a real problem. For any non-singular concept or meaning G*, signified by the general term ‘G’, it is possible to use a singular term such as ‘Peter’ to deny that Peter satisfies that concept. I can say ‘Peter is not a G’. This may be false, but it is not false in virtue of the meaning of ‘Peter’ and the meaning of ‘G’. But ‘Peter is not Peter’ is false in virtue of the meaning of ‘Peter’.

>>All thinking is general: no thinking can penetrate to the very haecceity and ipseity of the thing thought about.

This is either meaningless or false. If it means that I cannot use a singular term to tell you which individual I am talking about, without telling you anything else about that individual, it is false. If it does not mean this, what does it mean?

>>One cannot think about a particular except as an instance of multiply exemplifiable concepts/properties.

Why not? Why can’t I tell you which individual I am talking about, without telling you anything else?

>> If Edward disagrees with this he must tell us exactly why.

I disagree, because it is self-evidently true that I can tell you which individual I am talking about, without telling you anything else. An example. In Bodrum last week I met an individual called ‘Graham’. I can now tell you that Graham is an accountant, and I have thus told you which person I am talking about. I.e. I am not telling you that Peter Lupu is an accountant, nor that David Brightly is, nor anyone else who frequents your blog. I am using the term ‘Graham’ in a bare, singular sense. It means nothing other than: that person. It ‘s a simple and obvious idea, and doesn’t involve any difficult technical concepts like ‘haecceity’ or ‘ipseity’ or ‘reference’ or ‘sense’ or anything like that. I can tell you what kind of individual I am talking about (an accountant, a man, a human being). Or I can just tell you which individual I am talking about, without saying anything about what kind of individual they are. This is all really really simple, and not very difficult. I think you are making it difficult because you are overloading the issue with complex theories invented by philosophers. It is not difficult at all.

>> He should also tell us exactly how he is using 'proposition' since that is another potential bone of contention. Is he a Fregean, a Russellian, or a Geachian when it comes to propositions? Or none of those?

I normally use ‘proposition’ in the scholastic sense, namely a sentence that makes an assertion. I recognise it also has another, modern sense, meaning a Platonic entity which is what is asserted by the sentence. I try not to confuse these, but a long habit of translating scholastic Latin sometimes leads to confusion.

[Bill Vallicella]

Ed,

I may have missed it, but you still haven't explained your use of 'signify.' This is one of your favorite words, so it would help if I know exactly how you are using it.

You say a proposition is a sentence that makes an assertion. But that's equivalent to saying that a proposition is a declarative sentence. That won't do. Surely there is a distinction between a decl sent and a prop even if props are not construed Platonically. The Turkish 'Deniz mavidir' makes the same assertion (to acquiesce in your way of talking without quibbling over whether people or sentences make assertions) as the English 'The sea is blue.' So the sentence cannot be identical to the proposition expressed by it. And that's just one of several arguments.

[Bill Vallicella]

Ed,

You tell me you met an individual in Bodrum named 'Graham.' And then you tell me that 'Graham' means: that person. But that is not what it means to me. It means: the male person Ed met in Bodrum who is an accountant. Recall what I quoted you as saying above:

>>What I signify, when I use a term in the context of a proposition, is precisely what another person understands, when he grasps that proposition that I have expressed.<<

Let the proposition be: 'Graham is an accountant.' What I understand by 'Graham' prior to your telling me that he is an accountant is only this: a man Ed met in Bodrum. Now obviously that is not a singular meaning! You met a lot of men in Bodrum. And the same goes for 'a man Ed met in Bodrum who is an accountant.' That too is general, not singular.

You introduced a proper name into our discourse, 'Graham,' by means of an indefinite description, 'a man I met in Bodrum.' Fine, but that does not by a long shot show that proper names have singular meaning.

And note that you had the man right in front of you so that you could use indexical expressions such as 'this man' and 'that person' to refer to him. But I don't have the man in front of me. So how can what you signify be what I understand??

And again, what do you mean by 'signify'? Point to? Indicate? Wag your index finger at?

[Bill Vallicella]

Ed sez >>I say a general concept is one which we can entertain under the aspect ‘predicable of more than one’. A singular concept, by contrast, is one which we can only think of or entertain under the aspect 'not predicable of more than one’. A concept signified by ‘a’ is singular when the thought expressed by ‘a is F’ and the thought expressed by ‘a is G’ are inconsistent with the thought that no F is G.<<

It is clear that the thought expressed by 'Socrates is a philosopher' and the thought expressed by 'Socrates is a Greek' are inconsistent with the thought expressed by 'No philosopher is a Greek.'

But it doesn't follow that 'Socrates' signifies a singular concept. Why think otherwise?

Part of the problem here is the miserable scholastic term 'signify' which you have not explained. It seems to conflate sense and reference. Obviously, 'Socrates' in all of its occurrences above refers to the same individual. But that fact does not show that there is a singular concept expressed by 'Socrates.' I am not saying that that there is no concept expressed by 'Socrates'; I am saying that there is no such singular concept.

So far your position is a muddle as far as I can see.

Tell me exactly what you mean by 'concept' and by 'signify.'

[Edward the nominalist]

I replied to your posts 4 and 5, but the reply has now disappeared. Not that it was a very good reply. Did you delete it?

You ask many questions. Let's start with this:

>>Tell me exactly what you mean by 'concept' and by 'signify.'

A concept is a mental item. A term signifies such a mental item. To signify is to use the term such that the concept occurs in the mind of the one to whom we are signifying. Signification (as I use use it) is not a word-world relation, but a world-mind one.

>>It is clear that the thought expressed by 'Socrates is a philosopher' and the thought expressed by 'Socrates is a Greek' are inconsistent with the thought expressed by 'No philosopher is a Greek.'
But it doesn't follow that 'Socrates' signifies a singular concept. Why think otherwise?
>>

It doesn't follow if you believe in direct reference. But if you don't, you allow the proper name to have a signification (you would probably call this a 'sense') even when there is no reference for the name. This 'sense' or signification cannot be general, for if so we could entertain the idea that 'a is F' and 'a is G' are consistent with 'no F is G'.

Another argument for the singularity of the 'sense' is the one I mentioned above: we can always use a proper name to deny that the referent of the name falls under some general concept. Indeed, I can now tell you that Graham was not an accountant at all. It is not part of the meaning (or 'sense') of 'Caesar' that Caesar is a man. Or I can say that Graham might not have been an accountant, or that Caesar may not have been a Roman emperor. There are many arguments to show that the sense or signification or meaning of a proper name is not general in nature. But there are many arguments to show that empty names have a sense, i.e. that the meaning or signification of a proper name is not simply its referent.

If you accept both sets of arguments, i.e. the arguments that singular terms are fundamentally different from general terms, and the arguments that empty names have significance, then you automatically buy 'singular concepts'.

[Edward the nominalist]

In fact, let me summarise where I think we agree.

1. We agree that a sentence signifies something, although you would say it has a 'sense'.

2. We probably agree that the sense of a sentence is composed from the senses of the component parts of the sentence.

3. We agree that a proper name has a sense, which it has independently of whether it has a 'referent'.

4. And so in some sense, a very trivial one, you agree with me that there is 'singular senses', namely, the sense that a proper names has.

Where do we not agree? You seem to think that the sense of a proper name is general. But then you also seem to agree with Kripke's view that proper names have different semantic properties to general terms, or disguised descriptions. And you also seem to agreed with Sommers' observation (and Ockham's) that proper names behave as though they contained an embedded universal quantifier.

[arash]

Bill,

According to your definition: "The sense of a word is what you understand if you understand the word whether or not there is anything to which the word applies"
Since I understand: "Albert Einstein" in "Albert Einstein died" even though there is no concrete and present person to which that name applies, then there must be a concept for "Albert Einstein". Do you think it's sustainable to affirm that in the that I grasp when I understand "Albert Einstein" in that sentence is a general concept?

[Bill Vallicella]

No, I didn't delete anything. You may have taken too long composing your response.

I'm glad we agree about concepts. Concepts are mental items as you say. But then you say that signification is a world-mind relation. Don't you want to say that it is a mind-mind relation? For you say, >>To signify is to use the term such that the concept occurs in the mind of the one to whom we are signifying.<< So Ed, by using 'Graham' signifies to Bill by causing the concept expressed by 'Graham' to occur in the mind of Bill. Is that what you mean by 'signify'?

[Bill Vallicella]

>>1. We agree that a sentence signifies something, although you would say it has a 'sense'.

2. We probably agree that the sense of a sentence is composed from the senses of the component parts of the sentence.

3. We agree that a proper name has a sense, which it has independently of whether it has a 'referent'.

4. And so in some sense, a very trivial one, you agree with me that there is 'singular senses', namely, the sense that a proper names has. <<

Ad 1. Yes, sentences have senses.
Ad 2. Yes, I incline to subscribe to the Compositionality of Meaning (Sense): the sense of a sentence is built up out of the senses of its sentential and subsentential parts.
Ad 3. Yes, I agree.
Ad 4. If you define the sense of a proper name as a singular sense, then I have to agree that there are singular senses. But I object to your terminology because it papers over a real difference, the difference between saying that proper names express irreducibly singular senses and saying that they express senses whose singularity is constructed from general senses.

I am open on the question whether the logical form of every declarative sentence is *Every/some S is/isn't P*. Thus I am open to the rejection of atomic propositions.

I suppose I incline to Russellian descriptivism. It is hard to make sense of negative existentials on Kripke's scheme

[Bill Vallicella]

arash,

When you hear 'Albert Einstein' the concept in your mind is general: it is perhaps the concept German-Jewish physicist who as a young man worked in a patent office and later emigrated from Germany.

[arash]

Bill,

But when I say "Albert Einstein is dead" I mean Albert Einstein and not "a German-Jewish physicist who as a young man worked in a patent office and later emigrated from Germany". Don't I?

[Bill Vallicella]

Ed wrote, >>This 'sense' or signification cannot be general, for if so we could entertain the idea that 'a is F' and 'a is G' are consistent with 'no F is G'.<<

Don't get it! Suppose the sense of 'Socrates' is 'the wisest philosopher of all time' Now that is obviously general since some other person, Schmocrates, might have satisfied the description. Now if the wisest philosopher is Greek and the wisest philosopher is snub-nosed, then this is not consistent with 'No Greek is snubnosed.'

What exactly do you mean by 'singular' and 'general'?

[Bill Vallicella]

araesh,

Which one? There are a lot of Albert Einsteins, and some are canine.

[Edward Ockham]

>>Don't get it! Suppose the sense of 'Socrates' is 'the wisest philosopher of all time' Now that is obviously general since some other person, Schmocrates, might have satisfied the description. Now if the wisest philosopher is Greek and the wisest philosopher is snub-nosed, then this is not consistent with 'No Greek is snubnosed.'<<

'The wisest philosopher of all time' is indeed singular. However we need to our account to include singular terms which are invariant under modal or temporal contexts, and those which aren't. Clearly the first syllogism below is valid, and the second isn’t.

David Cameron is a politician
David Cameron was a student
Some politician was a student

The prime minister of the UK is a man
The prime minister of the UK used to be a woman
Some man used to be a woman

[Edward Ockham]

>>What exactly do you mean by 'singular' and 'general'?

Let’s attempt this. A singular term is one which has the logical form ‘every F’, where ‘F’ signifies a concept F* and where F* is such that only one object can instantiate it, i.e. where the particular quantification ‘some F’ makes absolutely no sense. Singular terms may be verbally complex (‘the tallest girl in the class’) or simple (‘Socrates’). Note that ‘every Socrates’ is not well-formed, since the ‘every’ is already embedded in ‘Socrates’.

Thus every singular term a is such that the following inference is valid

a is F
a is G
some F is G

We will further distinguish between singular terms where such an inference is valid under temporal and modal contexts, and those where this is not the case. E.g. the inference

a is F
a was G
some F was G

is generally invalid when the singular term is descriptive (‘the tallest girl in the class’) but generally valid when the singular term is a proper name (‘Susan’). This suggests that proper names are ‘non descriptive’ or ‘irreducibly singular’. If their sense consisted of some description, then we wouldn’t be able to say, as we clearly can, that they once didn’t fit the description, or that they might not fit it.

>> I object to your terminology because it papers over a real difference, the difference between saying that proper names express irreducibly singular senses and saying that they express senses whose singularity is constructed from general senses

I think the invariance rule given immediately above answers this. Some singular terms are reducible or descriptive. Others aren’t.

[Edward Ockham]

You haven't opened comments on the Donnellan post - can I comment on it here?

Donnellan says "When the historical explanation of the use of a name (with the intention to refer) ends in this way with events that preclude any referent being identified, I will call it a "block" in the history."

This seems dangerously circular. Why do events preclude any referent being identified? Well, because there isn't a referent at all. And what do we mean by 'there isn't a referent'? Well, we mean that a referent doesn't exist. So, Donnellan explains 'Santa Clause does not exist' in terms of his 'block' definition, which can only be explained by using the word 'exist'.

[arash]

Bill,

Yes I should provide you a description. But my suspect is that the best description you could provide to identify X contains other proper names (which actually denote an existent particular) with which X stands in spatio-temporal relation. So the description can't be purely general.
E.g.: "''Albert Einstein' = 'a German-Jewish physicist who as a young man worked in a patent office and later emigrated from Germany'", but Germany is a proper name.

[Bill Vallicella]

I'll open the comments on the Donellan post so you can comment. Thanks for your interest!

[Bill Vallicella]

Ed,

You say above that the Cameron argument is valid but that the prime minister argument is not. You are right only if 'the prime minister of the UK' is being used attributively as opposed to referentially. No doubt you are familiar with this distinction introduced by Keith Donnellan back in the '60s. But if the definite description in question is being used attributively, then there is nothing singular about it.

Part of what we are disagreeing about is how 'singular' and 'general' ought to be used. It is a datum that 'the prime minister of the UK' used attributively is not always satisfied by the same person in the actual world. And though it is satisfied by Cameron now, it might not have been.

Now it seems to me that a description that is multiply satisfiable at different times and different possible worlds is properly called general.

Now IF a proper name is a def descr in disguise, as Russell held, then no proper name is, strictly speaking, singular. A Kripkean rigid designator, however, is singular.

[Edward the nominalist]

I don't agree with Donnellan's analysis, but that is another story.

[Edward the nominalist]

I sketch out an approach here http://ocham.blogspot.com/2011/08/rigid-designation.html .

[Bill Vallicella]

I'm afraid I don't understand this: >>Let’s attempt this. A singular term is one which has the logical form ‘every F’, where ‘F’ signifies a concept F* and where F* is such that only one object can instantiate it, i.e. where the particular quantification ‘some F’ makes absolutely no sense.<<

[Edward Ockham]

>>I'm afraid I don't understand this

First claim: a singular term has the logical form 'every F'. That is because, as we agreed, the argument "Socrates is running, Socrates is arguing, therefore some runner is arguing" is valid, and its validity is explained by 'Socrates' having the form of a universally quantified noun phrase.

Second (implicit) claim: we can analyse 'Socrates' into an expression of the form 'every S' where 'S' stands for a singular concept. Note this is not the claim that 'Socrates' stands for a singular concept. For 'Socrates' is a quantified expression. Rather, 'S' is what remains when we strip away the embedded quantifier. Thus 'Socrates' stands to 'S' as 'every man' stands to 'man'.

Third claim: if there is an embedded concept term 'S', only one object can instantiate it. Or rather, we want to capture the 'individuality' of the concept, the fact that it cannot be subdivided into subconcepts, in the way that the concept 'animal' can be divided into *man, *giraffe and so on. Sommers would capture this by claiming that 'some S' and 'every S' are equivalent. I prefer to capture it by saying that 'some S' simply makes no sense. The concept of Socrates has no subjective parts, in the way that that concept of animal does.

[Edward Ockham]

I looked at Donnellan's 'man with the Martini' paper again, and comment here http://ocham.blogspot.com/2011/08/referential-and-attributive-use-of.html .

I don't accept that the ambiguity in the 'Prime Minister' argument is due to the referential/attributive sense, nor do I accept that the distinction between referential and attributive is even coherent. In the sense that 'referential' means 'object dependent' (i.e. that the meaning of a singular term is reality-dependent) I hold that no terms are referential, because no terms are object dependent. In the sense that 'referential' means 'tells you who or what we are talking about, I hold that all definite descriptions are referential. It is the point and purpose of such a description to do this - that is its 'definiteness'.

Even 'the King of France does not exist' truly states which king does not exist.

[David Brightly]

Ed,

Help! A couple of queries:

1. Let's translate 'Socrates is bald' in TFL into MPL. We get 'every+ S is bald' in MPL where the quantifier 'every+' means 'every---and-there-is-at-least-one'. So the contradictory of 'Socrates is bald' translated into MPL is 'not every---and-there-is-at-least-one S is bald'. This is equivalent to 'some S is not bald or there is no S'. If 'some' in MPL means the same as 'some' in TFL, then the translation of this back into TFL is identical. But then how can 'some S' make no sense?

2. That only one object can instantiate S appears to be a condition on the extension of S. I don't at the moment see how this can relate to the possibility or otherwise of an intensional analysis of S into subconcepts. Furthermore, can't any concept be subdivided by conjoining a suitable further concept? Could you expand on this?

[Edward Ockham]

Brightly
>>This is equivalent to 'some S is not bald or there is no S'.

Yes. If 'S' is a predicate, which (if I am right) it isn't.

>> But then how can 'some S' make no sense?

I see your point.

>>That only one object can instantiate S appears to be a condition on the extension of S.

I'm still puzzling over this. Being an individual - indivisible - means somehow that you have got down to bedrock when you have reached the proper name. But if you have tried to explain the logic of proper names by the 'embedded' universal quantifier, you haven't really got down to bedrock at all.

>>Furthermore, can't any concept be subdivided by conjoining a suitable further concept? Could you expand on this?

It's not like I have a settled theory to expound. I find some of this as puzzling as you.

[Bill Vallicella]

Ed,

It seems to me that you have not adequately explained either 'signify' or 'singular.'

>>First claim: a singular term has the logical form 'every F'. That is because, as we agreed, the argument "Socrates is running, Socrates is arguing, therefore some runner is arguing" is valid, and its validity is explained by 'Socrates' having the form of a universally quantified noun phrase.<<

Bizarre! The Socrates argument is of course valid. But why should anyone think that its validity is explained by 'Socrates' having the form of univ. quant. noun phrases? 'Socrates' is not a phrase, for starters.

[Edward Ockham]

>> But why should anyone think that its validity is explained by 'Socrates' having the form of univ. quant. noun phrases? 'Socrates' is not a phrase, for starters.

Well, 'Socrates' is most certainly a noun phrase http://plato.stanford.edu/entries/names

Whether its validity is explained by its having that form is a different question. In TFL that is an economical explanation. You have a theory in which everything is a common term + a quantifier. So how do you explain proper names? Well, they must be analysable into a common term + a quantifier.

Clearly there is something wrong with that. But the alternative (direct reference = MPL) is worse.

David Brightly has raised an interesting objection above, which is potentially fatal. I was having a Pimms in the garden when I realised how fatal it was.

[Bill Vallicella]

'David Cameron' is a noun phrase, no doubt, as are most proper names. But 'Socrates,' and many other proper names are not phrases because they consist of only one word. Or will you tell me that there are one-word phrases?

[Edward Ockham]

>>Or will you tell me that there are one-word phrases?

Yes. It's some time since I did any linguistics, but I am pretty sure that (as a technical term) 'noun phrase' or 'NP' means any expression, possibly a one-word expression, that is technically a noun phrase, namely consisting of a body plus a modifier.

[Edward Ockham]

Consider here http://en.wikipedia.org/wiki/Phrase_structure_rules for example.

S -> NP + VP

Every sentence can be broken into a noun phrase plus verb phrase. If 'Socrates' were not NP, then 'Socrates runs' would not be a sentence. Note 'runs' is a VP also.

[Bill Vallicella]

This is a mere terminological matter, and I am happy to acquiesce in the notion that 'Socrates' is a one-word noun phrase.

To return to a substantive question, you think that MPL requires direct reference. Is that right? Why exactly? Correct me if I'm wrong, but for you the direct reference theory of names is the theory that the meaning of a grammatically proper name = its referent.

I think it is worth noting that one could hold that the reference of a gram. prop. name is NOT routed through (determined by) its sense, but that it does have a sense nonetheless.