To be, and not to be - that is the quantifier

ResearchBlogging.org I don't read a lot of logic, having been sufficiently innoculated as an undergraduate to avoid further infection, but occasionally something pops up that is interesting way beyond the boundaries of that intersection of philosophy and mathematics. Siris points us at a paper in a new journal, Review of Symbolic Logic, which appears to be Open Access, by the redoubtable Graham Priest, on whether existential quantifiers imply existence. Why this matters is a deep issue in metaphysics: on the one hand we have the noneists, following Meinong, who hold that a quantifier in a formal sentence (and natural language) does not imply that some thing is actual, but that we can talk about, say, Santa Claus without needing to make an assertion about his existence. On the other hand we have the analytic tradition as it flowered in Quine, whose slogan is to be is to be the value of a bounded variable.

Roughly, the debate is over what a true sentence commits us to. If I say (truly) that Santa Claus lives at the North Pole, or that Sherlock Holmes lived at 221B Baker Street, London, am I committed to the inference that Santa Claus exists or Sherlock Holmes was an actual person? Meinongians and noneists say no - this is a matter of "exists-within-a-domain-of-discourse" that everyone is capable of understanding. The Quineans say that it does. If you think it's true as a sentence, then you must think that what is existentially (or as noneists seem to prefer, the particularist quantifier, I presume because it picks out a particular object in the sentence) bounded must exist.

Why this matters to me, is that it goes towards what is and what isn't a theoretical object. On the Quinean view, something is theoretical if some true sentence in that theory quantifies over it. But I think that there are things in theoretical sentences - like species - that are not existential commitments of that set of theoretical sentences, but of a larger domain of discourse. It's not quite the same problem, but close enough...

Priest, G. (2008). The closing of the mind: How the particular quantifier became existentially loaded behind our backs. The Review of Symbolic Logic, 1(01) DOI: 10.1017/S1755020308080015

More like this

I would really like to discuss your idea of "species" with you one time as I feel it would be of great benefit to me. I've written a few essays on the idea that defining a group of organisms as a "species" is useful only for purposes of organization and illustration while being both too specific for some research and too narrow for other fields. It fails to provide the specific mutations which may affect the outcome of experiments while simultaneously leaving out organisms which may, in terms of common ancestry, be closer related to one another than two individuals from the same "species."

Cool! This is one of my main interests outside my real expert area (there should be a single word for that whole phrase), so if you ever want to co-write something on existence and species, think of me.

I read some of Quine's 'From a logical point of view' I was referred by a guy who's doing his Master's in Philosophy and is a bit of Kantian because I'm a bit partial to Hume and I guess he wanted me to read 'two dogmas of empiricism' because it obviously attacks empiricism, but which to me seemed more of an attack on logical positivism than anything else. Anyway, those things about existential quantifiers that Quine proposed in 'on what there is' (I think) seemed to me to be making at least as much as a dogmatic statement as Hume's division of perception into impressions and ideas or Kant's dogmatic insistence on rationalistic ideas like intuition and categories. (I'm not saying I understand Hume and Kant well, the above is just my impression on of them).

I'm not a Platonist or much else. I'm happy to talk about the answer to a power series existing without meaning it has 'real being'. I think the biggest problem is not with logic, which we must use to communicate, but with people assuming that logic metaphysically is necessarily coordinate with the universe. As an evolved species, we have the wherewithall to survive in our original evolutionary environment (OEE). That most probably doesn't include a way of thinking or rationalizing that understands the universe perfectly and why would it? We obviously don't see quarks or EM waves, but as far as we can tell from indirect evidence they exist. They're just not necessary for an animal that evolved in our OEE.

By Brian English (not verified) on 23 Aug 2008 #permalink

Thanks for the tip the article is indeed interesting particularly for my Professor who is an expert on the history of quantification. (I passed the tip on to him) The Penelope Maddy article HOW APPLIED MATHEMATICS BECAME PURE is something that interests me very much, its the article directly before Graham Priest's one, so thanks for that too.

Fuller writes:

Newton is supposed to have "presented his mathematical physics as the divine plan that was implicitly written into the Bible [emphasis added]" (p. 54). Fuller must have access to an otherwise unknown veridical edition of the Principia."

No, of course I do not have access to any such edition. However, once one adds some context - Newton's correspondence, successive editions of the long interpretive essay he attached to the Principia, called the General Scholium, as well as his other major work, the Opticks - it becomes clear that Newton intended his physics to be a decoding of hidden biblical truths.(1) Again this point should be obvious to anyone schooled in HPS.(2) Such a person would be mindful of the tricky 17th and 18th century conventions concerning the expression of theological opinions in scientific tracts.

Fuller gives no justification for his statements so I shall not either, other than to say this is a subject on which, although I would never call myself an expert, I am very well and very extensively read.

My comments:

(1)No! He didn't intend that at all and nothing in his writings or the context gives any grounds what so ever for making this claim.

(2)No it wouldn't because it's simply not true.

this is a matter of "exists-within-a-domain-of-discourse" that everyone is capable of understanding. The Quineans say that it does. If you think it's true as a sentence, then you must think that what is existentially (or as noneists seem to prefer, the particularist quantifier, I presume because it picks out a particular object in the sentence) bounded must exist.

This seems to have obvious connections to various QM interpretations - the noneists with interpretations that collapse the wavefunction, and Quine with many-worlds and related interpretations, as well as modal realism. Not to say that QM has anything to do with species...

"this is a matter of "exists-within-a-domain-of-discourse"

Taken literally, "exists within a domain of discourse" asserts existence openly, and is as much at fault as the Quinean interpretation, nevermind that 'a domain of discourse' now needs to be explicated as something within which another thing can exist.
But that's obviously not what the noneist means: he means "within a domain of discourse" to modify the notion of existence in question. Therefore the noneist accepts here that the proposition implies existence, but says that there's more than one sort of existence to choose from. Existence is here discourse-relative, perhaps in contrast to absolute conceptions of existence.
But is that really so different from the Quinean view? If "domain of discourse" is sufficient to ground the relative existential commitments of certain sentences and make them true, then I can't see that it differs significantly from "theory" as the relative truthmaker for its own theoretic objects. A theory is just a formalization of a domain of discourse.

Dace, you raise some good questions, and I really don't have much more of a response but this: Domains of discourse, at least if you are a realist, are not determinants of what is actual. There is a different meaning to "truthmaker" when sentential structures make truth from when sentences refer (whether we know it or not). So if a domain of discourse implies the existence of some class of objects Quine-style, that doesn't thereby mean these objects are actual, although if someone has commitment to that domain/theory, it might (but not must) imply they believe it's actual.

But I do think, although I can't defend this in the professional manner of a philosopher of language, that not all our ontological commitments derive from our theory, nor that everything that is implied in our theory need be an ontological commitment. For example, species are (I think) observed, not defined by theory, while I am still uncommitted to the Higgs boson. The former is a case of some class that is not derived (if I am right) from theory, while the latter is something derived from theory that I don't yet believe in (supposing counterfactually that I actually understand this theory/domain).

The Quinean view requires that one must believe in whatever one's best theories quantify over, and that nothing else is to be accepted. Actualism of the kind the Quineans assert seems to me wrong, which is why I find Priest's noneism appealing. A belief is not a determinant of a reality, in other words. But, as I say, this is a novel area for me...

I think the biggest problem is not with logic, which we must use to communicate, but with people assuming that logic metaphysically is necessarily coordinate with the universe.

Well that would be quite a big mistake wouldn't it? To confuse the language we use to describe the Real with reality itself. Very common these days though.

John:
"Domains of discourse, at least if you are a realist, are not determinants of what is actual."

Well, if you're a Quinean realist, and if 'domains of discourse' do not determine the actual, then you shouldn't grant them existence at all. But, on the contrary, I think a Quinean realist should accept that 'domains of discourse' do determine the actual, since he will have to provide for the truth of verbal reports anyway (e.g."John said that Mary was stubborn"). The way he does this is to say that domains of discourse have existence through verbal events, and verbal events are ultimately explicable in terms of the physical. Hey presto! - domains of discourse pay their way like any other referrent.
So the Quinean will not, I think, admit any distinction in truthmaker kinds. It is up to the noneist to preserve the distinction, but I really can't see how - his view just seems a vaguer version of the Quinean one.

"The Quinean view requires that one must believe in whatever one's best theories quantify over, and that nothing else is to be accepted."
I'm not sure how Quine understands this, but this does not seem to follow from existential quantification. Rather, he appears to be laying down norms for rational scientific belief. It follows from existential quantification that if one believes a theory, then one should also believe what it quantifies over (i.e. the ontological commitments of the theory); what seems to be added is that one should believe the best theories available. Nothing else is to be accepted because an ontological commitment to something outside theory is vacuous - without a theory, we don't really understand what it is we are committing to.
But in any case, it's my opinion that you need not accept his norms to accept existential quantification.

On the matter of species: It strikes me that species labels, if phenomenal, function like predicates, albeit complex ones. 'Ape' is then perhaps a more strict version of the predicate 'ape-like', and both are similar to 'red', 'piercing', and 'shivered' in that they ascribe phenomenal features to subjects or events when they appear in sentences. You might use Russell's theory of descriptions to seperate the 'ape' from the thing which is described by the term. So species need not commit us ontologically at all.

But that's obviously not what the noneist means: he means "within a domain of discourse" to modify the notion of existence in question. Therefore the noneist accepts here that the proposition implies existence, but says that there's more than one sort of existence to choose from. Existence is here discourse-relative, perhaps in contrast to absolute conceptions of existence.

This is an interesting argument. While there might possibly be exceptions, I don't think noneists do take 'within a domain of discourse' to modify the notion of existence in question; rather, they usually take 'exists within a domain of discourse' as a figure of speech, convenient primarily for its use in letting you use the same expression in whatever domain you please. That is, most noneists think that ontological commitments, truthmakers, and the like, are simply beside the point; for the most part, they contribute nothing of importance to the logic or reasoning. Any possible commitments you have, ontological or otherwise, you implicitly set out in defining the domain, well before you quantify; and you may not be committing to much at all, e.g., if your domain is the domain of impossible objects or the domain of purely fictional characters. In other words, the noneist usually would say (it seems to me) that you wouldn't think any of these things particularly relevant to quantification unless you were already a Quinean. Accounts of truthmakers or existence may be interesting in their own right; but they don't contribute anything to our understanding of rational discourse in general. The noneist, one might say, holds of logic (as far as quantification goes, at least) what Cantor somewhere says of mathematics -- that the essence of it is its freedom.

...orogenesis

John: I'm not sure whether that question is directed at my comment on species or Quine's norms. Regardless, I would put it this way: 'mountains' are phenomenal objects,
and so are combinations of predicates, which may or may not be substantiated outside my own head. Do they exist? As phenomenal objects they do, but this is to say no more than I use the term 'mountain' to describe my experiences on occasion. As mind-independant objects, well, I need a theory, even if that theory is a naive realism, because if I don't have one then I am not saying anything more by ascribing existence then by describing experience.
Since I'm not a naive realist, but a scientific realist, I think that some of the predicates which describe 'mountains' will reduce to others (e.g. colours reduce to microtexture). Ultimately, as these are physical objects, they yield to the reduction which physics prescribes, in terms of mass, energy, and particle.
So, insofar as mountains exist, they are described by the theories of physics. The word 'mountain' collects predicates together, but does not proclaim their existence.

Brandon: Aren't noneists ruining the good ol' word 'exist'? ;)
Well, I'm sure you're right to say that noneists take 'exists within a domain of discourse' to be a figure of speech, but the question becomes what it is they are claiming by using it. You seem to suggest that the claim might be to membership of a set, and that the work with these progresses as well with sets which represent reality and those that don't.
Well, fine, but if membership of a set is a claim at all, then there is a distinction between the truth and falsehood of that claim. If sets are just human constructions, then there seems to be no reason why a set of fictional characters should contain the member 'Sherlock Holmes' rather than exclude him. We can construct sets at will, we can lay down our own rules for inclusion and exclusion, even if this does violence to the ordinary understanding of the label we give to that set - these are, after all, technical abstractions.
So somewhere along the line, the set has to be pinned to reality to satisfy this demand for a distinction between truth and falsity. And, despite the philosopher's abstractions, they almost always are: Sherlock Holmes belongs in the set of fictional characters because the set of fictional characters is derived from actual works of fiction, which are physical objects. Sure, we can ignore reality and just work with sets, but the correct use of the label 'true', if it has sense, is provisional upon the grounding of that set in reality.

Very interesting...

By Cesar Daniel G… (not verified) on 30 Jan 2009 #permalink