Philosophy Discussion Forums | A Humans-Only Club for Open-Minded Discussion & Debate

[Terrapin Station]

Well, first, those words are not equivalents (synonyms) of "correspondence," though one or more of them can be substituted for it in certain contexts. The relevant sense of "correspond" here is correlation:

"Definition of correlate (intransitive verb):

"1a: to bear reciprocal or mutual relations : CORRESPOND
"If two things correlate, a change in one thing results in a similar or opposite change in the other thing.

"(transitive verb):

"1a: to establish a mutual or reciprocal relation between"

https://www.merriam-webster.com/dictionary/correlate

In the case of "snow," the correlation consists in the fact that that word is used to denote that "stuff" in English-speaking speech communities. It does not entail that there is any similarity, equivalence, etc., between the word and the things it denotes.

He's asking you to explain correspondence, not to define terms or to say something completely superficial and non-explanatory like "The word is used to denote the stuff . . . "--which is basically just saying "correspondence works" without explaining how.

He's asking you to explain how it works. Because in having to peg down how it works, you should come to some realizations that you're not coming to. Of course you keep avoiding this, but we can keep pointing out that you're avoiding it, and hopefully you'll eventually confront it. (Ultimately we're trying to help you.

[GE Morton]

He's asking you to explain correspondence, not to define terms or to say something completely superficial and non-explanatory like "The word is used to denote the stuff . . . "--which is basically just saying "correspondence works" without explaining how.

He's asking you to explain how it works. Because in having to peg down how it works, you should come to some realizations that you're not coming to. Of course you keep avoiding this, but we can keep pointing out that you're avoiding it, and hopefully you'll eventually confront it. (Ultimately we're trying to help you.

Egads. There is no "working" involved in correlations or correspondences. "How does it work?" is a misguided, vacuous question, and so would be any proffered answer to it. Those terms require no "explanations;" the only relevant question is how they are used, what they denote in a given speech community. A word is correlated, corresponds with a thing or set of things if, upon presentation, it evokes a distinctive set of behaviors among persons in that speech community with respect to things in that set.

There is no doubt some complex neurophysiological explanation for why those behaviors are triggered by that word, but that is a question for cognitive scientists and biologists, not philosophers.

[Terrapin Station]

Egads. There is no "working" involved in correlations or correspondences.

This is what should have the "Egads."

EVERYTHING works some way or other.

We're asking you to explain how correspondence is supposed to work.

[Terrapin Station]

Asking how something works is asking how it obtains/how it comes to be, what the exact relationships or conditions necessary for it are, how those conditions or relationships obtain, and in terms of substantial details, not just a loose, conventional manner of speaking about it that basically just names it in other words.

[GE Morton]

Asking how something works is asking how it obtains/how it comes to be, what the exact relationships or conditions necessary for it are . . .

I just answered that, to wit: "There is no doubt some complex neurophysiological explanation for why those behaviors are triggered by that word, but that is a question for cognitive scientists and biologists, not philosophers."

That question, while interesting, is irrelevant to the fact that there is a correlation/correspondence between nouns and sets of things, one readily determinable empirically, and to the meanings and utility of those words.

[Terrapin Station]

Asking how something works is asking how it obtains/how it comes to be, what the exact relationships or conditions necessary for it are . . .

I just answered that, to wit: "There is no doubt some complex neurophysiological explanation for why those behaviors are triggered by that word, but that is a question for cognitive scientists and biologists, not philosophers."

That question, while interesting, is irrelevant to the fact that there is a correlation/correspondence between nouns and sets of things, one readily determinable empirically, and to the meanings and utility of those words.

Before even starting to comment on this, it seems again like you're reverting to talking about meaning and not the correspondence theory of truth. You ignored the earlier comment about this.

[GE Morton]

Before even starting to comment on this, it seems again like you're reverting to talking about meaning and not the correspondence theory of truth. You ignored the earlier comment about this.

Oh, you're right, because we're not talking about the correspondence theory of truth. That is an inadequate theory of truth, as I agreed with Peter several posts back. We're talking about the correspondence of words, nouns, with the sets of things they denote. A word corresponds to a given set if people in a given speech community associate it with members of that set. Whether they do or not is revealed by their behaviors.

Also, I gave the theory of truth I adopt some time ago. It is the "semantic theory of truth," as outlined by Tarski, Kripke, and others several decades ago: "A proposition P is true IFF s," where s is the state of affairs asserted by P and is confirmable. If it is confirmable then P is true.

[Terrapin Station]

Oh, you're right, because we're not talking about the correspondence theory of truth. That is an inadequate theory of truth . . .

Whatever your opinion of it, we were talking about the correspondence theory of truth. That's what my comment that you responded to was about. It wasn't about meaning.

[GE Morton]

Oh, you're right, because we're not talking about the correspondence theory of truth. That is an inadequate theory of truth . . .

Whatever your opinion of it, we were talking about the correspondence theory of truth. That's what my comment that you responded to was about. It wasn't about meaning.

Ok, my mistake. I assumed you were talking about the issue under discussion in the previous several posts. I didn't suspect you'd changed the subject.

[Peter Holmes]

I gave the theory of truth I adopt some time ago. It is the "semantic theory of truth," as outlined by Tarski, Kripke, and others several decades ago: "A proposition P is true IFF s," where s is the state of affairs asserted by P and is confirmable. If it is confirmable then P is true.

But this formula - p is true iff s - is precisely the correspondence theory: 'snow is white' is true iff snow is white - which is a tautology.

And, by the way, confirmability is otiose. An unconfirmable factual assertion could be true, though there may be no reason to believe it is.

[GE Morton]

But this formula - p is true iff s - is precisely the correspondence theory: 'snow is white' is true iff snow is white - which is a tautology.

No, it isn't. That is because the theorem, "P is true IFF s" is stated in a meta-language; P is a proposition in a target language. You can't define truth-in-a-language within the language without circularity. Nor is that the correspondence theory (at, least, not the classic version of it. There are several versions).

And, by the way, confirmability is otiose. An unconfirmable factual assertion could be true, though there may be no reason to believe it is.

Of course. A proposition may be true, false, or indeterminate. Confirmability is required for assigning the truth value "true" to a proposition. If it cannot be confirmed or disconfirmed then no truth value can be assigned. Why is that "otiose"? Propositions whose truth values are indeterminate convey no information.

[Terrapin Station]

Whatever your opinion of it, we were talking about the correspondence theory of truth. That's what my comment that you responded to was about. It wasn't about meaning.

Ok, my mistake. I assumed you were talking about the issue under discussion in the previous several posts. I didn't suspect you'd changed the subject.

The comment of mine in question said:

"The correspondence in question is between the meaning an individual assigns (as the proposition) and the state(s) of affairs as perceived from an individual's spatio-temporal situatedness. There's no way to avoid something like the above in theories of truth, because we are necessarily talking about the relation of a proposition to something else."

I'm not sure how one could read that as being about something other than the correspondence theory of truth.

[Peter Holmes]

...the theorem, "P is true IFF s" is stated in a meta-language; P is a proposition in a target language. You can't define truth-in-a-language within the language without circularity. Nor is that the correspondence theory (at, least, not the classic version of it. There are several versions).

A meta-language is just another language, so it's as foundationless as all languages. And what we call truth isn't a thing that can be described in any language. All we can do is explain how we use or could use thw word 'truth' and its cognates and related words.

And the theorum states 'p is true iff...' - not 'p can be assigned the truth-value true iff... So, as I said, confirmability is otiose.

[GE Morton]

...the theorem, "P is true IFF s" is stated in a meta-language; P is a proposition in a target language. You can't define truth-in-a-language within the language without circularity. Nor is that the correspondence theory (at, least, not the classic version of it. There are several versions).

A meta-language is just another language, so it's as foundationless as all languages.

"Foundationless"? I wasn't aware that a language required a foundation to be functional or useful. What sort of foundation do you think it should have? A language is a communication tool. What sort of foundation does any tool have?

I think you have some category confusion there (applying predicates to a class of subjects that only apply to some other class).

And what we call truth isn't a thing that can be described in any language. All we can do is explain how we use or could use thw word 'truth' and its cognates and related words.

The formula I gave is a definition of "truth-in-L," with "L" being the target language. Definitions are one way --- the most common way --- we explain the uses of terms.

And the theorum states 'p is true iff...' - not 'p can be assigned the truth-value true iff... So, as I said, confirmability is otiose.

I think the word you want there is "redundant," not "otiose" (which means "lazy", or "pointless"). Yes, it is redundant. I mention it to point out that confirmation is entailed, necessitated, by the theorem.

[GE Morton]

Well, first, those words are not equivalents (synonyms) of "correspondence," though one or more of them can be substituted for it in certain contexts. The relevant sense of "correspond" here is correlation:

"Definition of correlate (intransitive verb):

"1a: to bear reciprocal or mutual relations : CORRESPOND
"If two things correlate, a change in one thing results in a similar or opposite change in the other thing.

Again, please explain how this applies to the supposed correspondence between the word 'snow' and the stuff we call snow. What 'similar or opposite change' would occur in the one or the other?

That property (reciprocal effects) doesn't apply to the correlation/correspondence between symbols and the things they symbolize, since symbols are (usually) arbitrary. They are associated only by convention within a certain universe of discourse. With some symbols the correspondence is more explicit, such as "H2O" for water. If the chemical composition of water changed, then the symbol would change accordingly.

In the case of "snow," the correlation consists in the fact that that word is used to denote that "stuff" in English-speaking speech communities. It does not entail that there is any similarity, equivalence, etc., between the word and the things it denotes.

So you agree there's no similarity, connection, harmony, equivalence - or, I assume, mutuality or reciprocity - in a word, correspondence - between the word 'snow' and the stuff we call snow. And you think 'correlation' nails it. Oh-kay.

Well, you're adding some words to what I said. I said there is no similarity or equivalence. There certainly is a connection and a "harmony" --- namely, the one I gave, i.e., that that word is used in a given speech community to denote that "stuff."

I can't believe this is so hard to get across. I just agreed that a word, a name, is not a property of a thing. What IS a property of a thing is the fact that a particular city is called "Paris" in some speech community. That is an empirical, verifiable, "feature of reality" --- as real as that the city is the capital of France. So I assume you are still restricting the word "property" to some narrow class of attributes predicable of things. That restriction is arbitrary and inconsistent with common uses and understandings of that term. That Alfie is married is a property of Alfie; that he was born in Sweden is a property of Alfie; that he is a philosophy professor is a property of Alfie; that he was named "Alfie" by his parents and is called "Alfie" in a certain speech community are properties of Alfie.

I assume that narrow class of attributes you're willing to call "properties" are the physical properties of a thing --- Alfie's height, weight, hair and eye color, etc. Surely you realize how incomplete and uninformative such a limited description of Alfie would be.

No comment on any of that? Are being married, being Swedish, etc., properties of Alfie or not, in your view?

Ah, a moment of clarity? In what way is naming different from describing?

Assigning a name to a thing is attaching a tag to it. It does not describe it. However, once attached, it adds a new property to its description: "Objects of this class have been tagged with the word 'dog' by English speakers."

Could it be that we decribe things (predicate things of them) by using names of other things? For example, we may describe dogs by saying 'dogs are quadrupeds'. And we don't describe dogs by saying 'we call those things dogs'. Q What are these strange things I'm about to see? A We call them dogs. Q And what are dogs? Can you describe them for me, so I'll recognise them? A Well, one way to recognise them is that we call them dogs. (Thanks.)

Properties that aid or enable recognition are not the only properties things have. Knowing that Alfie is married or Swedish or Catholic will not help me recognize him either. Knowing the tag (name) for the class of canines enables you to talk about them, not recognize them.

His name is now one of his properties, just like his former home, his ears, color, etc.

But you just wrote ' a word, a name, is not a property of a thing'. You seem confused.

That was a shorthand, Peter. A name, as a word, is not a property of anything. That a certain thing has been given a name IS a property of the thing.

List Rosco's properties. Well, for one thing, we call him Rosco.

Why write that as a redundant tautology? If the questioner already knew the name, the answer is redundant. Try this instead:

"List your dog's properties."

"Well, first, his name is Rosco."

A bit more informative?

In a line up of all the Peters, in what way would the name Peter describe or help to identify me? Do the names in the phone book describe the people named?

Yes. The phone book typically gives an address and telephone number. Those are a couple more of your properties, probably sufficient to allow someone to find you.

If I changed my name by deed poll to David, what would be the correlative or corresponding change in me? (Your confusion is profound.)

That change is itself a change in you. Not a change in any of your physical characteristics, but those are not the only properties you have (an arbitrary limitation you can't seem to set aside).

You need to forget all the metaphyical/ontological nonsense surrounding the concept of a "property." A property of a thing is simply any confirmable fact about or observable feature of a thing that helps us identify a thing and distinguish it from other things. There are, to be sure, different classes or categories of properties, such as "local" and "non-local." Local properties are those that can be determined by observing the thing; non-local properties require confirmation of some fact beyond the thing. "Alfie is bald" is an example of the former; "Alfie is Swedish" is an example of the latter.

So you agree that a property is a feature of reality, not the predicate of a declarative clause.

False dichotomy. "Reality" just is what is asserted with true declarative propositions.

Where we disagree is here: you think that our calling some animals 'dogs' and other animals 'cats' is picking out a property of dogs and cats that helps us to identify and distinguish them from each other, and from all the non-cats and non-dogs. And I think that's patent nonsense. You're just wrong. And I simply don't understand why you find the fact that naming is not describing so hard to grasp.

No. Calling some animals "dogs" and others "cats" does not "pick out" a property of those animals, respectively. It gives them, assigns to them, a property that enables us to pick them out from one another and other animals when we communicate about them. Also, you're overlooking the active forms of "identify" and "distinguish." "Identify" can mean "recognize," but it can also mean, "point out." "Distinguish" can mean, ""tell apart," but also, "make distinct."

So you agree that 'the set of domestic canines' isn't self-identifying. And I assume you agree there's nothing canine about the word 'dog', or the word 'canine'. The sign doesn't, in some primitive magical way, contain the signified.

Of course not. Assignments of words to things or classes of things are arbitrary. But once a name is assigned and accepted in some speech community the relationship between the word and that set of things is not arbitrary, and that relationship is empirically verifiable..

There are no categories in reality, but only things that can be categorised in different ways for different purposes.

Agree.

Hooray. The arrow doesn't choose its own target. But if there were a correspondence between the arrow and its target, the two would be united in some magical way, so that the arrow (the name) couldn't but hit that one target.

Well, no. That there is a correspondence between X and Y certainly does not entail that they are "united in some magical way." As with "exists" and "property," you've apparently restricted the scope of "corresponds" to some arbitrarily narrow range of relationships. That the archer aims arrow X at target Y establishes a correspondence/correlation between X and Y.

We could check that reality conforms to the ways we talk about it.

Yes, we can. We can certainly check whether "reality" conforms to what we're saying about it. E.g., we can ask the archer what he is aiming at with this arrow.

1 Naming is not describing.

True. The act of naming a thing does not describe it. But that it has a name accepted and used in a speech community does describe it, and is a property of it.

There's no correspondence between a name and the thing(s) we name with the name - because a one-way relationship isn't a correspondence.

Correspondences are neither "one-way" or "two-way." Those adjectives are inapplicable to the term, and denote nothing meaningful about it. X and Y correspond if there is some sort of 1-to-1 relationship between them (such as that this arrow is aimed at this target).

The correspondence claim - 'snow is white' is true because snow is white - is a tautology.

Not if the claim is phrased as you just phrased it, with the first "snow is white" in quotes. The subject there is a proposition; "snow is white" in the second clause denotes a state-of-affairs. What is denoted by the clause is the subject, not the words denoting it. The state-of-affairs is expressed in a language assumed to be understoood --- a necessary assumption if you wish to avoid an infinite regress.