Existence without Possible Worlds Drucken E-Mail
Dienstag, 11. Mai 2010 um 20:36

[Download PDF] 


1. Introduction

It is quite challenging to find a statement about which all philosophers

agree, since one of the major methods of philosophizing is to doubt

everything. Although his empiricism and naturalism are highly disputed,

one statement by William Van Orman Quine would be acknowledged by

almost all philosophers, most likely because it is about the general task

of philosophy: “I think of philosophy as concerned with our knowledge

of the world and the nature of the world.”1

The disagreement begins when philosophers attempt to define

the terms of this statement, especially when they try to find a definition

of ‘world’: it can be used as a synonym for reality, sometimes in the

narrow sense of physical reality – whatever ‘physical’ means (classically

‘experienceable’ or comprising objects that are studies by natural

sciences) –, sometimes in a wider sense as a combination of physical and

mental reality. The latter one can only be used as far as the existence of

mental entities is at all accepted. Depending on the standpoint from

which one looks at this concept, reality can be seen as the sum of either

all things (this would be a materialistic standpoint) or as the sum of all

facts (this would be a ‘mental’ standpoint, because a fact or true

statement is a mental entity). In his ‘Tractatus Logico-Philosophicus’

Ludwig Wittgenstein argued for the latter, saying that the world is

everything that is the case – the sum of all facts.2

Quine argues for the former concept – reality as the sum of all

things. Advancing philosophical extensionalism, he argues that

universals are not more than classes, and that a class is not more than the

sum of its members. Statements do not have an independent existence

(namely the ‘thought’ – a mental entity expressed by the statement), but

expressions merely have to be regarded as physical objects. Rejecting

Frege’s distinction between meaning (as intensional) and reference (as

extensional), Quine denies the existence of intensionality, with the

consequence that the meaning of a sentence is inseparable from its

reference to the physical world: The meaning of a sentence is the fact

making it true or false; the meaning of name is the object(s) to which it

refers; the meaning of a predicate is the property which is truthfully or

wrongly assigned to an object.3

But what can be regarded as a thing, an object, an existing

entity? In the following chapter I would like to analyze Quine’s usage of

the term ‘existence’.


2. Existence

“There are the ontological questions, as they might be called: general

questions as to what sorts of things there are, as well as what it means

to exist, for there to be something.”4

W.V. Quine

As pointed out, Philosophy is concerned with everything that exists,

specifically within the sub area metaphysics or ontology. Originating

from the Latin verb ‘ex-sistere’, a thing can be said to exist by ‘standing

out’ or ‘appearing’ – very much connected to sensory experience. That

does not necessarily mean that only the things we see exist. A radical

empiricist would rather say that only those things we can potentially see

or experience can be recognized as existing or non-existing objects.

However, this view stands on shaky ground, since in modern physics one

can no longer clearly distinguish between what can and cannot be

experienced: Molecules, atoms, electrons, strings – these entities can

only be indirectly observed, and the definition of what counts as an

indirect observation highly depends on the physical theory the observer


If one applies the methods of critical empiricism to ontology,

every knowledge about the existence or non-existence of an object has to

be regarded as preliminary, as it is the case for any hypothesis in natural

science. Even if one observed a dragon oneself, the existence of dragons

can be doubted, alleging that the observation has been a dream or that

one lived in the Matrix. Non-existence is even harder to claim, since –

similar to natural laws – one relies solely on the inductive method: I have 

never seen dragons, nobody I know has ever seen dragons, almost every

part of this world has been explored and if somebody had ever found

dragons, it would be very likely that this information somehow became

public and I or somebody I know would have heard about it. Thus, it is

very likely that dragons do not exist.

Kant’s critique of the ontological argument triggered an ongoing

discourse among philosophers about the nature of the predicate ‘exists’.

Usually opposing views are formulated using the question whether

existence predicates are first-order predicates or not; a first-order

predicate only takes individual entities as arguments and constitutes a

definite subset. There are several arguments that favor existence as not a

first-order predicate: (1) If it were, its negation would be a first-order

predicate, too; it would be self-contradictory to talk about someone’s

non-existence, since forming a subset of non-existent human beings

already requires a set of all existing human beings, of which the nonexisting

people would be a subset.5 (2) If ‘exists’ were a first-order

predicate, it would be true of everything; therefore ‘not exists’ would be

true of nothing – false of everything. No meaningful statement “… does

not exist” could be formulated. (3) As Kant argued, existence does not

add anything to an object or individual: “One hundred real talers do not

contain the least more than a hundred imaginary talers.”6 Intrinsic

predicates define a concept, but it is independent from the fact as to

whether the concept is currently actualized in this world. Thus, the

existence of an object is separate from its essence.

In order to avoid these problems, Gottlob Frege suggested

treating ‘to exist’ as a secondary-order predicate. These do not quantify

over individuals but over classes of individuals, namely first-order

predicates. Then to speak about the existence of a concept means to

acknowledge that the concept is instantiated at least once. “Dragons do

not exist” therefore means that the concept of dragon is not actualized –

at least not at this time in this world. The problem that arises here is that

when talking about individual persons we usually do not regard them as

instantiations of a certain concept. Using Quine’s example, how could

one say that Pegasus, Poseidon’s flying horse, does not exist? The failure

of applying a second-level existence predicate to individuals and the

failure of applying a first-level existence predicate to concepts is the

main reason why followers of Frege’s view usually see existence as an

equivocal concept, having two distinct meanings, one expressed when

applying it to individuals, one when applying it to properties (kinds).


2.1 A Univocal Theory of Existence

Quine tries to give a clearer definition of existence. Similar to Frege and

Russell, he did not accept an equivocal concept or two-sense theory of

existence. How is this way tenable? The existence predicate must be only

first- or only second-order. Because the problems of seeing it as a merely

first-order predicate (like meaningfully claiming the non-existence of a

thing) can only be avoided by introducing second-order existence, the

only way to maintain a univocal theory of existence is to rule out firstorder

existence predicates. This has an enormous ontological

consequence: properties are now ontologically primitive – individuals are

only the sum of their properties. Quine’s quotation defining

extensionalism “[T]he universal is no more than the sum of its

particulars”7 actually applies not only to concepts, but to what we

recognize as individuals. It doesn’t mean that a human person is not any

more than the sum of its physical parts, rather that our reference to a

specific human person is not more than the sum of all properties we use

to distinguish this person from the rest of the world. Individuality seems

to be a non-primordial attribute: the assumption that a class of properties

used to refer to an entity is so complex that it is only instantiated once.

Two major problems are created when eliminating first-order

existence. The first one is the issue of individual life forms. Quine solves

this quandary by reducing names to properties. “Socrates exists” is

therefore an ordinary-language version of the sentence “There is an x

such that x socratizes.” Socratizing is the sum of all known properties

that can be applied to x, and by creating a name-property ‘to socratize’

we assume that there exists only one instantiation of this property,

namely x. The second problem arises when talking about fictional

characters. How is it possible to state meaningful sentences about

fairytale characters when meaningfully referring to fairytale characters is

not possible?

Quine had claimed that all proper names can be construed as

definite descriptions. With ‘Designation and Existence’ (1939) he wrote

an essay attempting to explain how one can deny the existence of

Pegasus. His major argument is that one does not commit oneself to an

ontology containing Pegasus when making the statement about Pegasus

that it does not exist. Inspired by Russell, Quine proposed to translate

proper names into description and define existence claims of singular

terms as Ø$x(Wx Ù "y(Wy ® y = x)) . In this case, ‘Pegasus’ would

have to be translated into a predicate that is definite enough to ensure

that it can – if it exists – only be actualized once; Quine uses ‘the winged

horse that was captured by Bellerophon’8. Acknowledging Pegasus’

existence is therefore a simplification of saying “There is at least one

entity x such that it is a winged horse captured by Bellerophon and there

is no entity y different from x that is a winged horse captured by

Bellerophon. Likewise, denying Pegasus’ existence is a simplification of

saying “For all existing entities x it is the case that x does not have the

property of being the winged horse captured by Bellerophon (and if it

were not the case, all other entities fulfilling this property would be

identical to x)”.

The problem with the elimination of proper names is that the

substituted descriptions are often not any more ‘basic’ than the original,

since they again contain proper names that would have to be eliminated

as well. This is well shown in Quine’s own example, in which the term

‘Pegasus’ is exchanged with a description containing the proper name

‘Bellerophon’. Being aware of these difficulties, Quine ceased to use

Russell’s type theory to eliminate proper names in his later works.9 But

he didn’t change his mind about exchanging proper names with

descriptions, now substituting Pegasus with the description of being-

Pegasus which cannot be further analyzed.10 Hence, “Pegasus does not

exist” is simply treated as “Nothing is Pegasus”11: Ø$x IsPegasus (x) .

By eliminating the first-order predicative use of existence in this way,

Quine establishes a univocal theory of existence, limiting the meaning of

existence to ”what existential quantification expresses.”12 Quine means

“’exists’ to cover all there is”13, but the question on what there is cannot

be answered within an ontological theory:14 it makes sense to talk about

the universe of a theory “only relative to some background theory”15 This

ontological relativity has severe consequences for epistemology.


2.2 Ontological Commitment

Quine rejects the notion of epistemology as the search for a ‘first

philosophy’ that can create an ontological foundation from which

everything else can be deduced.16 Different ontologies should rather be

treated similarly to scientific theories: Theories are provisionary, they

can be falsified by empirical evidence, they can be modified. But they

cannot be compared ‘from the outside’; like Feyerabend’s and Kuhn’s

assumptions in philosophy of science – the incommensurability of

theories –, Quine advances a view that there is no ‘vantage point’ that

can be used to objectively compare theories and say one is (from an

epistemological point of view) better than the other.17 This is what Quine

calls naturalized epistemology:

“[There is no] vantage point outside the conceptual scheme that [the

philosopher] takes in charge (…). He cannot study and revise the

fundamental conceptual scheme of science and common sense

without having some conceptual scheme, whether the same or

another no less in need of philosophical scrutiny, in which to work.

He can scrutinize and improve the system from within.”18

So how do we decide what ontology we should use? If there were no

criterion upon which to argue in favor of a specific ontology, one would

have to become a radical skeptic. This might be the reason why one of

the major criticisms of Quine is that he is a relativist. But this accusation

is unsubstantiated. Although naturalized epistemology gives up the quest

for a theory of knowledge19, it is very concerned with the justification of

an ontology, determining “one belief as firmer or more certain, [relative]

to the believer’s mind”20.

For Quine all language necessarily presupposes some ontological

commitment: “Conceptualization on any considerable scale is

inseparable from language.”21 So how do we choose a certain language

and with it an ontology? Well, we actually did not choose our language,

but rather acquired it from our parents and the society we live in. In a

naturalistic theory, language is a product of evolution: talking and

referring to possible harms and telling stories about how to avoid lifethreatening

situations or how to treat illnesses gave people an advantage

in natural selection. This means we speak a certain language for practical

reasons; the ontology that we adapted with this language is therefore also

a practical ontology. It includes simplifications, abstractions and

postulated entities that simplify our sensory input more economically in

order to transfer life-saving behavioral information from one generation

to the next.

The most fundamental concept in our language is the notion of

external objects, which is used as a basis to organize sensory data.22

“Physical objects are postulated entities which round out and simplify

our account of the flux of experience, just as the introduction of irrational

numbers simplifies laws of arithmetic.”23 This does not mean that our

notion of physical objects is objective, but that it is epistemologically

foundational, because it is the basis of an ontology that every human

language we know is – primarily for practical reasons – committed to.24

Quine explains in ‘On what there is’ that the “physical conceptual

scheme simplifies our account of experience because of the way myriad

scattered sense events come to be associated with single so-called

objects.”. In ‘Two dogmas of empiricism’ he writes:

“The myth of physical objects is epistemologically superior to most

in that it has proved more efficacious than other myths as a device for

working a manageable structure into the flux of experience”25

If a language guarantees the intelligibility of statements uttered in this

particular language, every entity we can talk about is part of our

ontology. That is true except if we were able to reformulate a sentence

containing this object into another sentence that is more basic, which is

Quine’s method to avoid abstract terms and proper names within his

ontology. Every object in a language that cannot be paraphrased is part

of the speaker’s ontology. Ilham Dilman formulates Quine’s criterion of

ontological commitment as following:

“An entity is presupposed by a theory if and only if it is needed

among the values of the bound variables in order to make the

statement affirmed in the theory true”26

Existence (‘presupposing an entity’) therefore is theory-dependent and

can, as a second-order predicate or quantifier, be applied to all entities

which can be quantified over. Dilman continues:

“[W]hat a language guarantees is the intelligibility of statements

made in that language, including existential statements made in it;

and that means guarantees the possibility of their truth.”27

If the notion of existence is based on the theory of physical objects, then

what sense does it make to say that physical objects exist? To make this

proposition meaningful, one must have a theory of existence outside the

theory of physical objects. But according to Quine, this is not necessary:

“Physical objects exist” can simply be paraphrased as “physical object

propositions make sense and therefore can be true.”28 Thus, by saying in

ordinary language that something exists or does not exist, one does not

automatically commit oneself to an ontology containing the object that is

referred to. This is exactly how Quine solves the Pegasus issue:

We commit ourselves to an ontology containing Pegasus when we

say Pegasus is. But we do not commit ourselves to an ontology

containing Pegasus when we say that Pegasus … is not.”29

As I have shown above, to deny Pegasus’ existence means claiming that

the property Is-Pegasus is not actualized at this time. Since Quine does

not accept mental entities, we cannot see this property as an entity itself,

but rather we have to see it as a counterpart to a specific neurological

pattern in our brain which is formed by cultural influence (books and

movies about Greek mythology) and which would cause us to regard

“Pegasus exists” as a true statement as soon as sensory data is interpreted

as fulfilling this specific pattern. By this example one can see why

Quine’s theory of reference is often associated with behaviorism.


2.3 Abstract Entities

According to Quine, most abstract entities can be paraphrased into

sentences that only contain descriptive terms. Accordingly, substantiated

properties are not treated as abstract universals, but in a behaviorist way,

meaning that for example ‘redness’ is not an entity, but to associate the

property Is-Red to some objects but not to others is a conditioned

behavior, originating in the basic experience of similarity within one’s

sensory input and in the adaptation of cultural norms.

But for Quine the non-existence of abstract entities is not a

dogma, as it is in classical nominalism. He is at the same time a naturalist

and fallibilist, so he does not believe in “any evidence, any avenue to

truth higher than or more fundamental than ordinary scientific method

itself”30 and he “recognizes that science changes over time and that

someday science could conceivably withdraw its support for physicalism

and/or empiricism.”31 Thus, every entity that is indispensable for our

scientific practice needs to be included into one’s ontological

commitments. ‘To exist’ does not mean ‘occupies a spatio-temporal

region’.32 Today this implies: Since mathematics is as indispensable for

science as the notion of physical objects33, and all of mathematics

including numbers can be reduced to set theory, Quine acknowledges the

existence of sets – which are abstract non-mental entities – as stated in an


“(…) but you do believe in the existence of certain abstract nonmental

entities.” – “Yes, numbers notably. (…) Assuming sets, or

classes, is on an equal footing with assuming molecules, atoms,

electrons, neutrons, and the rest; all these are objects, concrete and

abstract, that are assumed by the network of hypotheses by which we

predict and explain our observations of nature.”34

In the same interview, Quine summarizes his ontological commitments:

“I hold that physical objects are real, and exist externally and

independently of us. I don’t hold that there are only these physical

objects. There are also abstract objects: objects of mathematics that

seem to be needed to fill out the system of the world. But I don’t

recognize the existence of minds, of mental entities, in any sense

other than as attributes or activities on the part of physical objects,

mainly persons.”35


2.4 Comparing Theories

How can one theory of reality be determined to be better than another?

What makes it rational to commit one’s ontology to the existence of

certain entities and to deny the existence of others? Quine showed that

there are some entities that everybody is committed to – simply by using

language. These include physical objects, sets, and maybe even some

properties of physical objects like space, time and causality (similar to

Kant’s ‘a priori’ judgments). But what makes Quine prefer

conceptualism over Platonism, physicalism over Cartesian dualism,

empiricism over rationalism?

For an empiricist, rationality is not objective, but consists in a set

of rules that one learns as a foundation of one’s language and thinking.

‘Our’ rationality is often split up into the criteria consistency, coherency

and parsimony. Logical consistency is a fundamental part of language;

logical rules define how sentences have to be structured, and are –

according to Quine – not objective and necessary truths (analytical) but

relative to one’s language. Coherency usually means compliance of the

model’s implications with sensory inputs and other accepted models.

According to Karl Popper, a theory is valid as long as it is not disproved

by experiment. However, particular sentences can never be verified or

falsified in isolation, since every theory can be modified to comply with

unpredicted sensory data (Duhem-Quine-Thesis). Thus, neither

consistency nor coherency are sufficient as clear principles of rationality.

The principle of parsimony states that – out of two incoherent but

consistent theories both having equal strength of explanation and

prediction – it is rational to accept the one with less ontological

assumptions and irrational to accept the other.

Equating the methods of natural science and the rules of

rationality that are applied to ontological commitment, Quine uses both

the ‘aesthetic’ rule of ontological parsimony and the rule of minimizing

brute facts. His major argument against Platonism is, therefore, that

accepting the existence of universals suggests a rather great amount of

ontological entities, although many of them are avoidable without

causing a loss of explanatory value. “[The Platonist’s] over-populated

universe is in many ways unlovely. (…) [Its] slum of possibles is a

breeding ground for disorderly elements.”36 Quine proposes that we

should “limit modalities to whole statements”37. The words ‘possible’

and ‘necessary’ should therefore be used rather as an adverb than as an

adjective: Instead of saying “There is a possible entity Pegasus” we

should claim “Possibly, Pegasus exists” or, more exactly, “Possibly,

there is an x that pegasizes.” When we speak of possibility, we speak

about our own ignorance of the truth-value of a certain proposition, but

not about any object.

Here we see that through changing ordinary language into a

special grammatical form, Quine introduces a way to reduce language to

an ontology that doesn’t contain possible entities. Likewise, Quine

rejects possibly but not actually existing objects (‘Meinongian objects’)

as well as any other kind of mental entities like meanings or



2.5 Critique

According to Quine’s extensionalism, the meaning of a sentence is the

fact that it makes it true or false, opposing Frege’s distinction between

sense (Sinn or meaning) and reference (Bedeutung). But I do not think

this distinction can be easily eliminated. To use Frege’s example, the

sentences “The morning star is Venus” and “The evening star is Venus”

are made true by the same fact, namely the planets Venus and Earth

being in a constellation so that Venus appears as a bright star every

evening and every morning. But the meaning of these sentences is

different: I can imagine a possible world in which the references of

‘morning star’ and ‘evening star’ are not identical.

Especially intensional contexts create problems for Quine’s

extensionalism; these contexts are constituted by the use of modal

expressions (as in ‘it is necessary that x’) and propositional attitudes (as

in ‘to belief that p’, ‘to hope that p’, ‘to know that p’ or ‘to wish that p’).

If one accepted quantification over propositional attitudes, one would

have to accept mental entities. Denying the existence of such purely

mental states, Quine does not allow quantification over them, drawing

off any meaningfulness from propositional attitudes. This is what Quine

calls referential opacity (opaqueness), as opposed to referential

transparency.38 A “referential opaque context is one that cannot be

quantified into (with quantifier outside the context and variable

inside).”39 As a result, the words within an intensional context no longer

refer to their original referees.

In his paper from 1956 ‘Quantifiers and Propositional Attitudes’,

Quine defined the distinction between notional and relational senses of

‘believes’40, now commonly referred to as ‘de re’ and ‘de dicto’ believes.

Accordingly, the following sentences are not identical: “There is an x

such that Ralph believes x is a spy.” and “Ralph believes that ‘There

exists an x such that x is a spy’.” Quine rejects the first use, claiming that

“quantifying into a propositional idiom from outside … [is] a dubious

business"41, because their formal treatment leads to contradictions. But it

is problematic to reduce the first sentence to the second, because they

don’t have the same meaning. Quantification into intensional contexts

appears meaningful to us; when I say “I belief Ralph to be a spy”, I also

belief that ‘Ralph’ and ‘spy’ are not confined into a referentially opaque

context, but that they have the same referential meaning as when I say

“Ralph is a spy”.

Quine seems to be quite the opposite of an ordinary-language

philosopher. If there are sentences that ‘ordinary people’, maybe even the

majority of philosophers, claim to be meaningful, he would say “No, you

actually mean something different. What you suppose to mean cannot be

logically formulated without contradictions, so your intuition must be



3. Alternatives

3.1 Meinongianism and Parsimony

In ‘On what there is’, Quine leads a fictitious dialogue with a

philosopher called Wyman. Wyman advances Meinongianism, a nonunivocal

theory of existence. According to Alexius Meinong’s42 theory,

things can either exist (existieren), subsist (bestehen) or absist (gegeben

sein).43 Existence denotes the spatio-temporal being of an object;

subsistence, on the other hand, is a state of possible existence, as held by

unicorns, flying elephants, or mathematical theories. Absistence is an

ontological state that is taken by logically inconsistent objects like round

squares; however, these are still considered objects since we are able to

talk about them.

The distinction between metaphysical or logical existence and

physical existence is not uncommon in contemporary ontology, for

example used by Edward N. Zalta.44 Quine’s critique of Meinongianism

and related theories includes the dubious ontological state of possibilia

(“No entity without identity”), the fact that self-contradicting concepts

cannot be meaningfully talked about, as well as the ontological

profligacy (non-parsimony) of this theory.

Parsimony, in turn, cannot be seen as the top distinctive criterion

when comparing theories. Ontological parsimony, according to William

of Ockham, is only a distinctive criterion when comparing two theories

of equal explanatory power: entities should not be multiplied

unnecessarily. Thus, if it were possible to reduce all thinking processes

to physical (neurological) processes, as supposed by Quine, mental

entities would not be necessary and could therefore be eliminated from

our ontology.

But as the example of quantification over propositional attitudes

showed, an ontological theory does not maintain its full explanatory

power after eliminating intensions and mental states. This is why the

criterion of parsimony does not apply here. However, the dubious

ontological state of possibilia is a sincere argument. Why do unicorns or

flying elephants have to have a certain kind of existence? Why do sets or

numbers? Why is it not possible to simply regard them as ideas, as

creations of the human mind? Quine himself gives a partial answer to

this question in ‘On what there is’:

“We may for the sake of argument concede that there is an entity, and

even a unique entity (…) which is the mental Pegasus-idea; but this

mental entity is not what people are talking about when they deny


If words only refer to concepts in our mind, there is no relation between

the concept and the external object it supposedly refers to. Quine solves

this problem using his extensionalism, according to which ideas of

objects are not more than the sum of the ideas of all the objects’

predicates. But, as showed, this conception seems to fail when one talks

about propositions within intensional contexts.

When one claims a proposition to be a necessary or possible

truth, they do not always, as Quine suggests, acknowledge to be referring

to the level of certitude of their knowledge. When one expresses a hope

or fear, they imagine a possible world which is not referentially opaque,

but their concepts refer to objects within certain possible worlds as well

as to objects in present-day reality.

A univocal theory of existence calls for a high price: the

negation of the existence of mental entities and the referential

opaqueness of propositions in intensional contexts. This is why I am

quite skeptical of such a univocal theory. In the following, I would like

to propose an alternative approach to solve the dilemma of existence.


3.2 Own Conception

If concepts are only in our head – as presupposed by any form of

conceptualism –, then they are not definite, but rather blurry at their

edges. Wittgenstein, for instance, showed this using the theory of family

resemblance. I claim that establishing concepts in our mind requires the

thought of possible worlds from the moment they are created. Thinking

only works using modalities.

Concepts are open sets in two ways: They are open because we

do not know yet how many objects are contained in the set. They are also

open because we do not even know the specific meaning of our concepts.

A concept can be narrowed / reduced by adding another distinctive

property, or it can be extended by taking away one non-distinctive

property from its meaning. When we create or learn a concept, we co22

create a large number of possible extensions. When we see a flying

elephant, we can either narrow the original concept to non-flying

elephant-like looking creatures, or we can extend the concept elephant so

that it also covers the flying ones.46 We do not have a problem with this,

because already now, when we think of elephants, we can think of all

kinds of possible worlds where some extended concepts are actualized.

This is how many fairytales were created.

We do not only think of possible worlds, we rate them according

to chances that they might be real, now or in the future. For instance, we

regard the existence of flying elephants as quite unlikely. This is because

it is difficult to imagine how flying elephants could be compatible with

our other beliefs about the world, namely widely acknowledged physical

theories. On the other hand, white elephants, although we might have

never seen one, are thought to be more likely to exist.

Quine thinks that when we state modal sentences, like “It is

possible that elephants fly”, we do not talk about possible objects.

Rather, by saying that something has possible existence, we

acknowledge our own ignorance. That might sometimes be true. But a

constituting issue for most of our concepts are their origin and purpose,

things that the entity can possibly do or what can possibly be done with

it. When we think of an elephant, we think of watching them in zoos,

elephant mothers looking after their children, us riding on them in the

desert, using their tusks to produce piano keys, crossing the Alps with

them to combat Rome and so on. The argument is that the notion of

46 This conception is in my judgment compatible with many of the points Quine

makes for the indeterminacy of translation. Quine says that nobody knows if a

translation of a word comprises an equal concept or a narrower or extended one.

I go even further by saying that this is not even known by the speaker himself.

possibility is even prior to concepts, since thinking of possible worlds is

a constituting method for concepts.


3.3 Conclusion

Quine’s method of philosophy is quite appealing. His thinking is more

epistemological than ontological, because he tries to see ontological

positions as theories that can be argued for or argued against. The type of

evidence applied towards reasoning about ontological theories is not any

different from reasoning about natural science, mathematics or logic.

“I see natural science as continuous with the mathematics that it uses,

just as I see all this as continuous with philosophy. It all goes to make

up our inclusive system of the world.”47

With this in mind, empiricism seems to be rather a method than an

ontological position for Quine. In his famous paper ‘Two Dogmas of

Empiricism’ he presents two never questioned principles of empiricism,

namely reductionism and the analytic-synthetic distinction. If naïve

empiricism is set as an ontological position, it results in positivism – that

all talk about metaphysics is meaningless. But as a method, Quine’s

empiricism demands that the philosopher see the study of the world and

how it really is as one big task comprised of philosophy, mathematics

and natural sciences. Quine sees all these approaches using the same

principles of reasoning, objecting to any ‘first principle’, ‘analytic

statement’ or ‘fact with indubitable certainty’. There is no knowledge,

only beliefs; and all beliefs are thoroughly “pragmatic”.48

Based on such a pragmatic ontology, the existence of possible

worlds and mental entities is rejected by Quine not because they cannot

be experienced or seen (if that was the case, the existence of sets would

have to be denied, too). Rather they are rejected because one of the

principles of reasoning, ontological parsimony, demands this rejection –

provided that the world can be explained equally well without assuming

these entities as can be done with them.

The fundamental use of modalities in basic thinking makes

Quine’s rejection of modalities highly disputable. Eliminating all kinds

of modalities and mental entities, the explanatory power of an

ontological system is reduced, and therefore more and more ‘brute facts’

must be implemented. This is why the rule of parsimony cannot be used

to compare such ontological theories.




BERGSTROM, Lars / FØLLESDAL, Dagfinn: Interview with Willard Van

Orman Quine in November 1993, in: FØLLESDALL, Dagfinn: General,

Reviews, and Analytic/Synthetic, Philosophy of Quine 1, New York

2001, 65-78.

DILMAN, Ilham: Quine on ontology, necessity, and experience. A

philosophical critique, Albany 1984.

GIBSON, Roger: Willard Van Orman Quine, in: ID. (Ed.): The Cambridge

Companion to Quine, 1-18.

MAGEE, Bryan: The Ideas of Quine: Dialogue with W.V. Quine, in:

FØLLESDALL, Dagfinn: General, Reviews, and Analytic/Synthetic,

Philosophy of Quine 1, New York 2001, 13-24.

MORRIS, Michael: An Introduction to the Philosophy of Language,

Cambridge 2006.

QUINE, Willard V.O.: Designation and Existence, in: The Journal of

Philosophy, Vol. 36, No. 26. (Dec. 21, 1939), 701-709.

QUINE, Willard V.O.: Existence and Quantification, in: ID.: Ontological

Relativity and Other Essays, New York – London 1969, 91-113.

QUINE, Willard V.O.: Natural Kinds, in: ID.: Ontological Relativity and

Other Essays, New York – London 1969, 114-138.

QUINE, Willard V.O.: Ontological Relativity, in: ID.: Ontological

Relativity and Other Essays, New York – London 1969, 27-68.

QUINE, Willard V.O.: On What There Is, in: ID.: From a Logical Point of

View. Second Revised Edition, New York et. al. 1963, 1-19.

QUINE, Willard V.O.: Quantifiers and Propositional Attitudes, in: ID.,

The Ways of Paradox, New York 1966, 185–196.


QUINE, Willard V.O.: Quiddities. An Intermittently Philosophical

Dictionary, Cambridge 1987.

QUINE, Willard V.O.: The Time of my Life, Cambridge 1985.

QUINE, Willard V.O.: Things and Their Place in Theories, in: ID.,

Theories and Things, Cambridge 1981, 1-23.

QUINE, Willard V.O.: Three Grades of Modal Involvement, in: ID., The

Ways of Paradox, New York 1966, 185-196.

QUINE, Willard V.O.: Two dogmas of Empiricism, in: ID.: From a

Logical Point of View. Second Revised Edition, New York et. al. 1963,


QUINE, Willard V.O.: Word and Object, Cambridge 1960.

WITTGENSTEIN, Ludwig J.J.: Tractatus Logico-Philosophicus, Frankfurt



Johannes Groessl

1408 Commonwealth Ave, #21

Brighton, MA 02135