I think you are getting at something important, however I might quibble with the use of the term "essence" in these cases because it will lead to confusion as respects the employment of essences in metaphysics and the philosophy of nature (your sleet/snow example being a fine one). If everything has an essence/nature, then there can be nothing like Aristotle's distinction early in the Physics of:

A. "Those things that exist by causes," (e.g. a rock or volume of water, a bundle of external causes with a very weak principle of unity as compared to water molecules themselves, hydrogen and oxygen atoms, or especially living organisms), and;

B. "Those things that exist by nature," which are those things that possess an essence and are (more) self-determining, self-organizing, self-governing wholes, that are (more) intelligible in themselves; and we might add,

C. The random stipulated "wholes" dreamed up by later philosophers, such as the pairing of non-continuous different halves of foxes and halves of trout, or the "whole" made up by a person's feet and the ground they are standing on extending in a 8 foot cube beneath.

I add "more" to B because no finite being is wholly intelligible in itself. For instance, one cannot explain what a horse is without reference to any other thing. But one can also avoid the slide in multiplicity/smallism/reductionism or absolute unity/bigism so common in both ancient and contemporary metaphysics and philosophy of physics. We avoid either an infinite regress of smaller parts or atomism, on the one hand, and a bigism where things are only intelligible in terms of their role in the entire universe. A via media is needed, and essence/natures help here.

Holism, in many forms, is itself a sort of bigism as applied to epistemology. Perhaps no knowledge of any thing is entirely intelligible in isolation from other knowledge, but neither must we call in all belief and knowledge to explain any individual instance of knowing or believing.

This is why stipulated games like chess are not great examples for essences. They are by no means arbitrary (i.e., if you create a game with random rules, it will likely suck and lack strategic depth and never catch on) but neither are they akin to carbon, stars, or cats.

But, to your point, language absolutely requires universals which are closely related. Likewise, the things we speak of have a certain quiddity, whatness, in phenomenological awareness, that is essential to language.

It seems to me that the concern that the things of phenomenological awareness might have nothing to do with "real things," only creeps in with the presupposition that what we experience and sense are mental representations, not things—that the "mental" consists in accidental representations of the physical. There is a sort of iron clad dualism at the heart of most modern philosophy in this form, and it leads to the objects of all knowledge being either out own ideas, or in later forms, language.

But then these skeptical explanations are often wildly counter intuitive. I tend to agree with Domingo Soto that, while nominalism is easier to understand, it is much harder to believe.

My preferred solution, as many of you know. I've seen you refer to this as Quine's "joke" about being, but it's about time we took him seriously

Well, I'll reuse a prior post on this issue. Consider:

Brutus: Wow Cassius. I saw your results to my survey. I had always thought you were an atheist and a materialist, but I see here that you marked down that you think that both God and ghosts exist.

Cassius: Well of course they do Brutus. Both can be the subjects of existential quantification! But no, I am an atheist and I don't believe in ghosts.

Well, does Brutus have a right to be miffed over what seems to be sophistic equivocation here?

A different approach! In the most direct behaviourist account, a behaviour is rewarded and so reinforced. Am I correct that in IRL the reward is used to predict the behaviour? Very interesting. The reward function is the hidden variable, the "belief" that is used to predict future behaviour, while remaining hidden.

I'm struck by how similar this is to the discussion here with @J...

"What are beliefs?" — J

Both the reward function an the belief are understood and inferred from behaviour and outcomes. The reward function might indeed be an analogue, model or metaphor for belief. I'm not sure I would call them equivalent, but I might be convinced.

It's those damn Markov blankets again. Thanks for bringing this to my attention.

Maybe belief is a psychological construct. It's something unobservable, but we use it to explain and predict behavior. I think the more complex the behavior is, the more likely it is that we'll explain it in terms of belief. Simple behavior could be instinct, but something like plotting revenge needs propositions for the explanation. — frank

Again, the parallel between a belief and a reward function is striking.

Do beliefs have an ontology? — J

Perhaps they are a folk=psychology term for a reward function being processed in our neural nets...


(this is only half facetious)

does Brutus have a right to be miffed over what seems to be sophistic equivocation here? — Count Timothy von Icarus

No. If Brutus insists on misunderstanding what Cassius is saying, we can't help him, but he doesn't have to. And in fairness, Cassius is obviously trying to get Brutus' goat. He ought to be clearer, and explain that the word "exists" can be used in several different ways . . . which is why I dislike it so much as a keystone philosophical term. But I guess you think it should only be used one way?

At the risk of being repetitive, my complaint is with the term "existence", not the various concepts associated with its use.

I agree, but I suspect that this thread will not be the place to progress these ideas - others here appear to be misconstruing Quine's account. We are a long way from the Gavagai.

Your reply here is much the same as the one I gave to 's question previously.

...quantification is only one part of the explanation offered - it includes predication and equivalence and domains of discourse. Quantification tells Brutus and Cassius that we can talk about ghosts. Predication might be used to further say that ghosts are immaterial, imaginary or superstition. Cassius is mistaking quantification for predication. — Banno

On holism...

Quine's student, Davidson, supplemented Quine's holism by introducing charity, while moving form a descriptive to a normative position. Charity deals with the indeterminacy of meaning by assuming a shared background of congruent beliefs.

In particular, I'm still troubled by background beliefs. If I say, "I [background] believe that the earth is round," what am I claiming? — J

I'm reasonably happy with two sorts of background beliefs. First, those that are constitutive, Searle's status functions - the "counts as". Second, and less securely, certain first-person beliefs, such as that you have a pain in your foot or a love of Vegemite. These appear indubitable.

There may well be others. But I'm not sure "the earth is round" is amongst them.

And having said that, I have also previously argued that being a background belief is a role taken on by a proposition in a language game rather than a property of some beliefs. So for certain purposes - navigation, perhaps - that the world is round is taken as indubitable.

This would be worth exploring further, particularly in this context in relation to Quine's rejection of the analytic/synthetic distinction - a similar argument might be found against certain propositions always being foundational.

Yes, we have many different concepts surrounding existence. Ens vs esse, various uses of "is," "being is said many ways," ens rationis vs ens reale, existence vs subsistence, relationes secundum dici vs relationes secundum esse, real versus virtual, appearance vs reality, abstract vs physical, the inclusion of possibilia, impossibilia for Meinongians and in Hindu thought, "existence is not a predicte," etc., etc.

It would be nice to have an easy catch all solution that deals with them, but I am not seeing how this "solution" resolves any of them. None of those difficulties surrounding existence involve people being confused over whether or not it is possible for someone to say "there is at least one of...(insert anything utterable)." But there does seem to be an issue in kicking existence out to predication in that a diverse group of thinkers from Kant to St. Thomas have rejected being as a predicate.

I think where the solution may have value is vis-á-vis metaontology and mapping ontological commitments. But these are not the contexts where it is being introduced.

On topic, this reminds me of one of the most infamous and consequential strawmen of all time, Ockham's claim that the via antiqua claims that:

a column is to the right by to-the-rightness, God is creating by creation, is good by goodness, just by justice, mighty by might, an accident inheres by inherence, a subject is subjected by subjection, the apt is apt by aptitude, a chimera is nothing by nothingness, someone blind is blind by blindness, a body is mobile by mobility, and so on for other, innumerable cases.

William of Ockham - Summa Logicae, part 1, c. 51

But this solution seems like it will be of little help here in this, one of the most famous and long running debates on ontological commitments (one you'll still find in contemporary surveys of metaphysics), because it is about disagreements over implicit commitments. E.g., does quantifying over employers commit us to employees? Does "I have at least one wife," imply "there is at least one marriage?" Should predicates not include an ontological commitment, and if not aren't we biased against realism (for the realist often wants to show that opponents are ontologically committed to universals)?

Likewise, "if you claim 2 and 3 are prime then you are committed to numbers," is only informative in a conversation about platonism vs immanent realism vs ens rationis, etc. if there is someone actually claiming that numbers don't exist tout court (which I imagine is exceedingly rare).

Arcane Sandwich's point re permissivism and eliminativism is about what we ought to be ontologically committed to more than what we are committed to.

And IIRC, Quine himself was willing to grant that his metaontological approach wouldn't cover "existence" and punted on that whole aspect.

I appreciate your thoughtful response to this.

But there does seem to be an issue in kicking existence out to predication in that a diverse group of thinkers from Kant to St. Thomas have rejected being as a predicate. — Count Timothy von Icarus

The point I'm making about the word "existence" necessitates a kind of viewpoint shift that may not come easily. Let me rephrase it a little: The only thing that all the various uses of "existence" have in common is that they introduce the term as referring to something we can talk about, something we can quantify over. So if we insist on using "existence" and asking what it means for something to exhibit this feature, all we can do is point to the one characteristic they have in common, "being the value of a bound variable." Now I completely agree that this tells us next to nothing. (In particular, it is neutral about some of the uses of "exist" that traditional metaphysics wants to privilege as "real existence" or "what being means" or some such.). But nor should it be controversial. As some like to say on TPF, it's just common sense.

When you say that my modest proposal entails "kicking existence out to predication," that doesn't capture the emphasis I'm placing on language. I'm not saying that the word "existence" be used to cover one sense of existence but not another; I'm recommending we drop the word entirely. (And as I've probably said before, I know this will never happen; but a fellow can dream!). The various grounding and entailment relations that legitimately exist among the various types of being will remain unchanged. A traditional, metaphysically conservative philosopher has nothing to fear here.

The question of existence as a predicate, and Kant's opposition to it, has, I believe, been settled, or at least stabilized, in modern logic. See @Banno's response, above, for a short version. Quantification and predication are two different things. The hot issue here is quantifier variance, but that is (and was) another thread.

Should predicates not include an ontological commitment? — Count Timothy von Icarus

The terminological problem raises its head again, in different guise. Let's say we answer, Yes, they should include such a commitment. What, then, are we committed to? How are we using "being" in a way that clarifies, rather than merely reveals our preferred usage?

(this is only half facetious) — Banno

Aargh! Let's keep it completely facetious! No psychologism :razz:



These are good descriptions of how background beliefs might function, and indeed, I don't think there's a problem with understanding what we mean by them, and how they show up in ordinary life. My worries begin when we try to put them under the same umbrella as "belief" understood as a propositional attitude. Maybe I should just stop there and ask, If I say of Joe, 'He believes that water is H20', when "believes" is understood to refer to background belief of Joe's that he is not currently entertaining, am I ascribing a propositional attitude to him?

Incidentally, I think switching to 3rd person makes the issue clearer, because when we say "I believe such and such .... " it's tempting to say that I couldn't both make the statement and be unaware of the belief. In 3rd person, the believer is not the one doing the stating.

So if we insist on using "existence" and asking what it means for something to exhibit this feature, all we can do is point to the one characteristic they have in common, "being the value of a bound variable." — J

But this is clearly wrong. Consider, "There are things that exist which we have never conceived." To be is not to be the value of a bound variable. There are things that exist which are not attached to any variable. Existence is not bound by our minds. The vista of reality is not nearly so narrow as that.

I'm recommending we drop the word entirely — J

but I am not seeing how this "solution" resolves any of them — Count Timothy von Icarus

The constant problem with Analytical philosophy is that it wants to throw out language and substitute something far inferior: an artificial and brittle system. "We're going to throw out the word 'existence' from the language and replace it with stuff that is more suited to my idiosyncratic philosophy." Nevermind that within a decade or two it always turns out that the newfangled philosophy had surprisingly little to offer. The arrogance of someone who decides to revise language itself in favor of their "systems" is really quite breathtaking.

quantifier variance — J

Quantifier variance is itself proof that existence goes beyond quantification. The domain of the real differs from person to person, and therefore if we are not to be solipsists then we must engage an understanding of existence that goes beyond our own narrows ideas (i.e. we must involve ourselves in ampliation).

This comes back to the point on page 12:

From this Buridanian perspective, one cannot make claims about the relationships between language and reality from some external, God-like position, from the position of the user of a meta-language, who has a certain “context-free” or “context-neutral” access to the object-language and “the world”, both as it is in itself and as it is conceived by users of the object-language, that is to say, the totality of semantic values of items in that language. — Gyula Klima, Quine, Wyman, and Buridan: Three Approaches to Ontological Commitment, 10

Quine's idea that we have independent access to the meta-language and the object-language is absurd, and it underlies all of this. There is no objective-quantification apart from subjective-quantification. We do not possess the language of God, which would overcome all individual disagreements and force existence into our personal, solipsistic horizon.

Here's what I suspect the underlying problem is, in this discussion and others like it:

Most people want to avoid the thesis that existence is a property, and that it can be represented with a first-order predicate, such as "E", instead of the existential quantifier, "∃".

And why do most people want to avoid that thesis? Because they somehow believe that to treat existence as a property is naive, if not outright scholastic. After all, didn't Kant refute the ontological argument by pointing out that existence is not a predicate?

But to say that is to make a conceptual mistake, because it is not the case that properties are identical to predicates. Even if Kant succeeded in demonstrating that existence is not a predicate, it does not follow from there that existence is not a property, nor does it follow that existence has anything to do with quantification.

In simpler terms: no one wants to be accused of being a naive, scholastic, pre-critical philosopher, in the manner of Anselm, Aquinas, or even Descartes. We all want to be "the cool kids", and it seems that the only way to be "the cool kids" is to nod approvingly towards Kant's confused identification of properties and predicates, and to declare that matters of existence are matters of quantification.

Such views are nonsensical, if only because there are 20th Century thinkers like Mario Bunge, who conceptualize existence more or less like Aquinas did (as a property, not as a quantity) without being religious. Bunge was an atheist. So what's the big deal here? To imply, between the lines, that one is a Thomist if one conceptualizes existence as a property, is like saying that one is a Cartesian if one believes that a physical thing such as this table is a res extensa.

How about we start by analyzing these completely irrational themes that underlie these sorts of discussions, instead of digging our heels and just blurting out nonsensical accusations such as "You don't really understand Quine's point."

How about we start by analyzing these completely irrational themes that underlie these sorts of discussions, instead of digging our heals and just blurting out nonsensical accusations such as "You don't really understand Quine's point." — Arcane Sandwich

Agreed, and I think the "cool kids" point is spot-on.

Here's another of the irrational themes:

...I think it is worth noticing in the second [criticism] the smooth transition from “the description has/does not have a referent” to “the referent of the description does/does not exist” [...] What is interesting in the smoothness of this transition is how easy it is nowadays to have an unreflected, and accordingly deep conviction that whatever more restricted meanings existence may have, the full scope of being is that of the possible range of reference of the expressions of our language.[7] — Gyula Klima, St. Anselm's Proof: A Problem of Reference, Intentional Identity and Mutual Understanding, 2

Why does Analytic philosophy think that if a description has a referent, then that referent must exist? Why does it tie up reference with existence? I realize the idea goes back through Russell, but I don't see much merit to it. We handle references all the time without assuming that reference and existence go hand in hand, whether with opinion-claims, theory-claims, fiction-claims, goal-claims, history-claims, imagination-claims, etc.

And why do most people want to avoid that thesis? Because they somehow believe that to treat existence as a property is naive, if not outright scholastic. After all, didn't Kant refute the ontological argument by pointing out that existence is not a predicate? — Arcane Sandwich

To add to this: folks on this forum don't know how to argue. Many of them don't even properly understand what an argument is, and therefore to get them to give real arguments is like pulling teeth. What stands in the place of argument? Appeals to "the cool kids." In order to prove a point, one simply cites a well-known philosopher they have never read and tacks a label onto their opponent: "naive/direct realist," "idealist," "communist," "essentialist," "Analytic," "Continental." Take your pick.

That is what popular/online philosophy has become, "Appeal to popular authorities and never accept the burden of proof."

(This is why I resisted your criticism of Bob Ross' long argument. Ross is one of the rare members who gives arguments, so I don't mind if they are a bit unwieldy.)

In many places today, for example, no one bothers any longer to ask what a person thinks. The verdict on someone's thinking is ready at hand as long as you can assign it to its corresponding, formal category: conservative, reactionary, fundamentalist, progressive, revolutionary. Assignment to a formal scheme suffices to render unnecessary coming to terms with the content — Joseph Ratzinger, Conscience and Truth

"There are things that exist which we have never conceived." — Leontiskos

There is an x such that x has not been conceived. Clearly a quantification.

There are things that exist which are not attached to any variable. — Leontiskos

There is no need to presume that in order to be "attached" to a variable, a thing must first be "conceived of". To "conceive of" things that have not been conceived of is to make them available for "attachment".

Scare quotes, because these are your words, used in order to track your argument. In more logical terms terms we might say that we can include in our domain of discourse things of which we have not yet conceived; or in "common sense" terms, we can talk about things we might not have yet though of, if only in the most general sort of way.

"But how can you include something in the domain if you haven't even conceived of it?" Well, we just did. Notice that we haven't predicated anything else to such "unconceived" entities, nor do we need to in order to say that they are "unconceived".

Analytic procedures give us formal structures that set out how our language hanges together consistently. I just gave an example of how this works.

...we must engage an understanding of existence that goes beyond our own narrows ideas — Leontiskos

Which Quine builds in to his account, using Holism, and which Davidson extends with Charity.

Quantifier variance is an issue within the scope of quantification. To enter into that discussion is to already accept that existence can usefully be thought of as quantification. At issue is how we might think about differing domains, in particular if they are commensurable, as Carnap might argue, or incommensurate, as Hirsch posits. Ontological commitment becomes a discussion of what is and isn't included in the domain - of what we are talking about. And the discussion is ongoing. The recent discussion is not so much about whether domains are commensurable as what are the consequences if they are or are not. And again, this is an example of analytical procedures give us formal structures that enable us to understand our natural languages better.

No psychologism :razz: — J

Form the last few post, it's too late.

If I say of Joe, 'He believes that water is H20', when "believes" is understood to refer to background belief of Joe's that he is not currently entertaining, am I ascribing a propositional attitude to him? — J

Well, if Joe is consistent, he will agree that water is H2O. Perhaps he will say something like "I know water is dihydrogen monoxide, but it's not H2O"? In which case the issue is not with his belief about water but his belief about the equivalence of "dihydrogen monoxide"and H2O. And we are back to the extensional opacity of beliefs.

What Joe believes is not the proposition, but the fact. So "ascribing a propositional attitude" is problematic.

My thought is that a belief can manifest in various ways, but that in order to count as a belief, one should be able to set out what it is that is believed - some truth, and hence some proposition. So, at the risk of opening yet another can of worms, the cat cannot hold some proposition to be true, and yet believes the mouse is behind the cupboard. We can put its belief in a propositional form.

I'm interested in working out the implications of this.

A agree with your point as to the third person.

My thought is that a belief can manifest in various ways, but that in order to count as a belief, one should be able to set out what it is that is believed - some truth, and hence some proposition. So, at the risk of opening yet another can of worms, the cat cannot hold some proposition to be true, and yet believes the mouse is behind the cupboard. We can put its belief in a propositional form. — Banno

I think propositions are part of the human form of life. Language is so central to what we are that we interact with the world using a linguistic format. When we ask the world questions, we anticipate true propositions, as if the world has a narrator. That's why propositions are generally in third person.

In analyzing the way a cat interacts with the world, we translate it into human.

In analyzing the way a cat interacts with the world, we translate it into human. — frank

Yep.

It seems a simple point, but quite a few folk have misunderstood it in various contexts in these forums.

Gadamer says this, in Truth and Method:

That which can be understood is language. — Truth & Method, 442

This is a little cryptic, taken out of context, but what he means is that there are many things we experience that aren't candidates for understanding -- not everything conveys truth or meaning or even comprehension. But what can be understood is language. Nor does this mean that there's nothing but language, or that in understanding language we are only understanding words and symbols. He means, I believe, that we "do understanding" using language, it is our human mode of interpreting the world.

Quite similar, really, to the end of the Tractatus. I know opinions differ about this but I always took Witt to be saying only what is obvious: There are plenty of things we can't talk about -- entire worlds -- but therefore we have to hold our peace and not try to force what can't be articulated into words that we've stipulated can't express it.

I think Quine is just massively overthinking it — Darkneos

This is most philosophy. I think the opposite can be true, to a risible degree though (see: Searle, Austin). We need the mean (thanks, 'Stotle).

"But how can you include something in the domain if you haven't even conceived of it?" Well, we just did. — Banno

All you did was pretend to do something. I could do the same thing, "There is an x such that x is the king of France. Clearly a quantification." There is no work being done here.

We could examine Quine's statement as if it is a definition of existence, but I take it that it is uncontroversial that it will fail as a definition (hence the "joke").

...that an entity can figure as a value of a bound variable in his theory is, according to Quine, equivalent to the assumption that such an entity exists; it is impossible to quantify over entities of which existence is not, eo ipso, assumed. Put more precisely: according to Quine the notion of existence just means the capability of featuring as a value of a bound variable. To assume that something exists is to assume nothing less, and nothing more... — Lukáš Novák, Can We Speak About That Which Is Not?, 159

But a simple problem here is that we do not understand existence to be parasitic upon our language or our accounting. Accepting for the sake of argument, <If I quantify over X, then I am taking X to exist>, no actual relation to existence is at stake. Quantification does not entail existence, because we often quantify over things that do not exist (intentionally and unintentionally).

The only truth in Quine's claim is this: when someone uses variables within a sentence which presupposes the existence of its variables, they are presupposing the existence of these variables. This is little more than, "If you think something exists, then you think it exists." It has no traction on real existence.

Other problems:

Thus theories that allow their variables to take non-existent individuals as their values are automatically understood as possibilist, to the effect that those who share Quine’s dislike towards the overpopulated Meinongian slum feel under pressure to construe their theories so that they enable reference to actual entities only. That results in various technical problems (the Barcan Formula[50] and the like) requiring sophisticated workarounds, which however tend to introduce various ersatz-entities into the actualist systems like individual essences (Plantinga) or bare individuals “in limbo” (Transparent Intensional Logic), in effect barely distinguishable from the abhorred possibilia.

[50] If it is possible that there is an F, then (actually) there is something that is possibly an F: ◇∃x(Fx) → ∃x◇(Fx). — Lukáš Novák, Can We Speak About That Which Is Not?, 159

We need the mean, but disagree upon what the mean is.

Ari's "golden mean" is something of a fallacy when taken out of context -- the middle between extremes isn't going to be true or false just cuz it's in the middle.

For instance, if one were to take the mean between eating shellfish and not-eating shellfish -- where some shellfish are ok to eat some of the time -- that does not thereby make it true. It makes it reasonable-ish sounding to the two extremes, but reasonable-ish sounding isn't a condition of truth, or even good inference.

EDIT: Ought say this is super off-topic; just sparked a thought.

We need the mean, but disagree upon what the mean is. — Moliere

The first point about the mean is that if you think you are identifying it then you must be able to point to both extremes. Many people can only point to one.

The first point about the mean is that if you think you are identifying it then you must be able to point to both extremes. Many people can only point to one. — Leontiskos

True. At least, the way I'd put it, many people identify a Big Bad without identifying the opposite; and also for the Good, when I think about it.

Still, I stand by what I said -- the golden mean sounds good a lot of the time, but that does not thereby make it true, or false.

Sorry for diverting the thread too much, tho -- this has nothing to do with reference.

Sorry for diverting the thread too much, tho — Moliere

Threads with a two-sentence OP are usually a runaway train after the first dozen posts. Nothing to divert. ;)

:grin:

I believe I'm mostly on the side of what @Banno and @frank have been saying, though -- reference is inscrutable.

So with "gavagai", to use the example -- I can't tell if "gavagai" references this post, my memory of my bike when 9, the rabbits foot that I'm looking at right now, or some cultural practice.

"gavagai" ought be understood with respect to translating a totally unknown language, at least by the story. If I don't know how the natives speak and yet I know that "gavagai" is a noun, I will not thereby be able to point to something in the world -- what we might be tempted to call a fact -- to say that this noun in a foreign language refers to this or that.

reference is inscrutable — Moliere

Do you believe that we are successfully communicating with each other right now? Because it seems to me that if reference were inscrutable, then this would be impossible. And if a foreign word were inscrutable, then we would never be able to learn foreign languages. But we are successfully communicating with each other, and it is not impossible to learn foreign languages, therefore reference is not inscrutable.

(See my post <here> or Arcane Sandwich's posts)