Meaning and Naming in “On What There Is”_ Quine’s Response to “On Denoting”

Meaning and Naming in “On What There Is”: Quine’s Response to “On Denoting”

Quine’s essay “On What There Is” (1948) seeks to clarify how one can speak about what there is and how one can speak about nonexistent entities without giving them ontological status. This is done by appealing to the technique used by Russell in “On Denoting” (1905), where denoting phrases, phrases which name something by virtue of their form (phrases that use definite articles like ‘the’, ‘a’, ‘this’, etc.), are recontextualized into propositions which no longer retain the original denoting phrase and thus no longer need to refer to an ontological entity attached to the denoting phrase. However, the way this technique is used changes from Russell to Quine, due to their metaphysical differences surrounding the relation between meaning and denoting/naming (Quine uses the word ‘naming’ rather than ‘denoting’). Whereas Russell was against the meaning and denoting distinction, arguing that the meaning of a word is as a denotation of the entity it represents, Quine divorces the onus of representation from the meaning of a word. In the view of the latter, one can use a word without recognizing its denotation—words don’t have to represent something outside language to be meaningful. In this essay, I will provide an exposition of some of the key arguments Russell uses in “On Denoting” along with some metaphysical differences in their understanding of meaning versus naming/denoting that cause Quine to develop a different use of recontextualization.

On Denoting

In “On Denoting”, Russell seeks to question the distinction between meaning and denotation in denoting phrases, which is a view he attributes to Frege (Russell 483). On this view, denoting phrases always have both a meaning and denotation which are separate. For example, both the phrases ‘the evening star’ and ‘the morning star’ refer to the same celestial body: Venus. Yet each of these phrases has their own respective meaning, or at least are used in different contexts. Russell points out two major problems with this view: (1) cases where a statement is about a nonexistent entity, such as the example “the present King of France is bald”, and (2) the difference between the meaning and denoting of C versus ‘C’. In statements about nonexistent entities, Russell argues that they have meaning but no denotation, since one can make sense of the denoting phrase ‘the present King of France is bald’ yet not be able to point to the present King of France (Russell 483-4).

Russell’s second problem about the difference between a denoting use of ‘C’ versus the meaning of ‘C’ deals with the ambiguity between statements about a denoting phrases’ meaning versus statements about its denotation. For example, a statement like “the Evening Star looks bright tonight” would not just be about the sense of the phrase ‘the Evening Star’ , i.e. its meaning, but it would also be about the way Venus looks in the night, its denotation. If we were to make a statement about the meaning of the phrase ‘the Evening Star’ we would either put it in quotations such as “‘the Evening Star’ was established by some observant Babylonian” or use the phrase ‘the meaning of the Evening Star’ which is interchangeable with the quoted phrase used above. In order to access the meaning of ‘C’, we would have to use a phrase that denotes the meaning of ‘C’, not a phrase that denotes ‘C’ itself (Russel 486-7). This is problematic for Russell because it seems to imply that two entities exist, the denotation of ‘C’ and the meaning of ‘C’, with the former being used in statements about denotation and the latter being used in statements about meaning. However, for Russell, this distinction is untenable because it denies the meaning present in the uses of denoting phrases about ‘C’ terms. His example is the case where George IV wants to know whether Scott is the author of Waverley. If the terms, ‘Scott’ and ‘the author of Waverley’ in this instance, are truth-functionally interchangeable with each other since they both have the same denotation, then the statements “Scott is the author of Waverley” and “Scott is Scott”, should be interchangeable, yet the former is more significant to George IV than the latter (Russell 487-8). The issue for Russell is that these co-referring terms, which should be interchangeable in a statement without changing its truth value, end up turning the statement “George IV wishes to know whether Scott was the author of Waverley” from true to false, if one substitutes ‘Scott’ for ‘the author of Waverley’.

As a solution, Russell recontextualizes denoting phrases by turning them into descriptions— propositions that describe an entity that has the attribute of being the entity named in the denoting phrase. An example of its use is converting the statement “the present King of France is bald” into “there is an entity that is now the King of France and is bald”. The presence of the definite article (the structural marker of the denoting phrase) in the first statement, seemingly refers (denotes) an existing entity, ‘the present King of France’. It is only when it is recontextualized in the second statement that one realizes that it refers to certain predicates (or descriptions) that can be applied to entities in the domain, but are not necessarily a part of the entity being referred to (Russell 490). Because of this there is no need to posit an imaginary present King of France that the statement refers to. This also means that words don’t necessarily need denotations, they can be recontextualized to remove that burden.

In Russell’s view, a denotation occurs when the rephrased proposition is true, such as the statement “there is a difference between A and B” denoting something only if it is true, but there is no necessity for all propositions to denote something (Russell 490). With this, Russell also resolves the issue of the C and ‘C’ ambiguity by referring to ‘C’ as a phrase, not having any meaning of its own, and thus using C to refer to the denotation of a word, which is his collapse of Frege’s distinction between sense and reference: it is both how we use that word and what that word refers to. It is both the meaning and the denotation, at least for descriptive propositions that have a denotation. For descriptive propositions that don’t have a denotation, we can comprehend through rephrasing them in propositions that use terms/descriptions that we are acquainted with the denotation of. This makes the original denoting phrase unnecessary since they can be reduced to other descriptions we can understand (Russell 492). For example, “There is a round square in McHenry Library” can turn into “There is a thing that has the attributes of being both round and a square in McHenry Library”. Even if the term ‘round square’ does not denote anything, it is our acquaintance with objects that are round and squares that allow us to understand the statement.

On What There Is

Quine seeks to answer the problem about the referent of statements about nonexistent entities as Russell did, i.e., without positing the quasi-existence of nonexistent entities. Quine does so by using Russell’s technique of recontextualization but also simultaneously breaks any link between meaning and denoting/naming that Russell attempted to retain. One of the contentions he critiques is the belief that for a statement to be meaningful, it must name something. This he disproves by being able to recontextualize statements in a way that eliminates all singular terms (proper names) and rewords them as being predicates that apply to bound variables. Quine also questions the notion of meaning as something that can be investigated as a universal attribute, putting into question the use of ‘meaningful’.

In his argument In “On What There Is”, Quine wholeheartedly accepts the Russellian claim that names are superfluous, since they can be replaced by descriptions, and uses this against the positions he names “McX” and “Wyman”, philosophers who believe that any meaningful statement with a name has to describe an object being named. With his reduction of naming phrases into descriptions, Quine escapes the onus of having to provide a named entity in the same way Russell did, and thus prevents his ontology from being flooded with things like Pegasuses or round squares. This is because the name itself drops out and instead is replaced by a proposition like “there is an entity which has the attribute of being-Pegasus” or “there is an entity which has both the attributes of being round and a square” which one can just refute as false (Quine 26). Like Russell, Quine also makes the argument that this use of translation is valid on the grounds that a description by itself cannot have meaning outside the context of a proposition (Quine 25-26).

Quine and Russell are in agreement in the use of recontextualization, but Quine understands the notion of a meaning differently than Russell. For Quine, a word cannot refer to a meaning, like the use of ‘C’ in the above example of Russell’s argument. The use of the word ‘meaning’ to Quine is an unnecessary hypostatization. For example, when one talks about meaning, they either talk about things that “have meaning” or they “give meaning” to things. For the former, Quine argues that we can replace phrases like “having meaning” with the words ‘significant’ or ‘insignificant’, while for the latter he argues that “giving meaning” is just an act of clarification, so instead of saying “X is the meaning of Y” we can say “X and Y are synonymous” (Quine 31).

The problem then becomes clear (or the situation becomes more confusing): what does Quine mean by reinforcing the distinction between naming and meaning, if all names can simply be recontextualized into meaningful phrases? We’ve already seen that meaning is a superfluous concept for Quine, to be replaced by ‘significant’ or ‘synonymous’, so how are we to understand what naming is? For Quine, this is the problem of ontological commitment: when we make a statement, we are committing ourselves to the existence of certain entities through paraphrasing our sentences into classical logic, and it is the use of the quantifiers that range over bound variables in classical logic (something, nothing, and everything) that determine the entities that we are ontologically committed to (Quine 32). In situations where ‘nothing’ is the quantifier, there is no referent. In doing so, Quine breaks naming and ontological commitment away from naming phrases, since those can be replaced with descriptive propositions, and instead lets logic decide which entities commit to in our use of statements. In this way, Quine is able to resolve the issue of the ambiguity between the meaning and the denoting of C versus ‘C’ by saying that ontological commitments only matter when they can be formulated in first order logic.

For Quine, one can use words meaningfully without knowing if they have any denotation, or without knowing to which entities one is ontologically committed to in the use of that word. When one uses a term, there is no confusion between a term’s meaning and denotation; the use of sense in Frege for Quine is precisely to explicate the semantic relations a term has outside what physical object it represents. Investigations into denotation and ontological commitment are philosophical problems that can indeed be consequential for other fields like science and math (Quine 32-6), but are irrelevant in discussions about topics like the weather. What this does is precisely sever any notion that for a word to be meaningful it must have a certain correlation to reality. This is because despite ontological differences between two people there is a shared conceptual scheme that allows them to communicate (Quine 35). Explicating ontological commitments is an act of systematically structuring language to make clear the domain of possible entities required for scientific and philosophical systems to function. Quine himself says it clearly: “To whatever extent the adoption of any system of scientific theory may be said to be a matter of language, the same—but no more—may be said of the adoption of an ontology” (Quine 36). The “adoption of an ontology” is relative to one’s conceptual framework but only insofar as it is explicated. If one is not concerned with ontology to begin with, then they have no need to assert their ontological commitments because they can get by just fine without it. Ontologies are what Quine considers “relative myths” (Quine 37), relative in the sense that they are wholly determined by whatever conceptual framework one chooses, whether it be phenomenalistic, physicalist, Platonist, etc. Whatever framework one chooses commits them to entities whose existence they must affirm for the function of that framework, or what entities are indispensable to that framework. What recontextualization does is explicate relations of indispensability between a conceptual framework and a domain of possible entities.

What a phrase actually denotes, comes after one would want to contest ontological claims; it is not the determinant of meaningfulness like for Russell, where even if one uses a proposition that does not denote something, the proposition still must be understood through the denotation of its constituents. By separating denoting from the relevance of a proposition, Quine is able to separate himself from Russell’s assumption that statements must denote something to be meaningful. Quine and Russell share the same goal of trying to rectify language, but for Quine this does not mean explicating the ontological conditions undergirding language as a whole, but making clear to what ontological entities one is committed to in the use of a proposition and in the adoption of a systematic theory. With this, Quine’s approach localizes ontology to conceptual frameworks, unlike Russell’s, which remains committed to a universe of entities that one must acquaint themselves with to know the meaning of propositions.

Conclusion

In this essay, it has been argued that Quine and Russell share similarities in their use of recontextualization as a technique to break away from the assumption that every meaningful sentence must have a named object it denotes. However, the issue of the ambiguity between what a term denotes and what it means is contested between the two and influences their conception of what meaning is for recontextualized phrases. For Russell, to recontextualize a phrase is to specify the denotations of the constituent attributes of the denoting phrase in order to understand what is actually being denoted. To quote Russell, “Thus in every proposition we can apprehend (…) all the constituents are really entities with which we have immediate acquaintance.” (Russell 492) For example, to recontextualize a statement about a unicorn is to understand the unicorn as an aggregate of two other attributes we know the denotations of and thus have epistemic access to: a horse and a horn. For Russell, recontextualization is not just a tool of ontological critique, dispelling all thoughts of round squares and present Kings of France, but it’s a way to specify epistemic access: how do we actually understand statements about nonexistent entities? To Russell, the answer is to look at what ontological entities or attributes we can acquaint ourselves with, and use that to direct our use of meaningful language.

For Quine, recontextualization is an act of asserting ontological commitment. To recontextualize a phrase in first order logic is to precisely determine the domain of entities our language commits ourselves to, nothing more. In this way, Quine holds no epistemic pretenses and is able to establish a way for philosophers to clarify their ontological disputes by clarifying the entities in their ontology by way of bound variables. Quine is also able to simultaneously break with Russell’s solution to the aforementioned problem of meaning versus denoting C and ‘C’, by separating ontological disputes from the ordinary use of language, arguing against both meaning and denoting as we traditionally understand it. Propositions can be meaningful without “having meaning” and denoting more a philosophical consideration for the realm of ontology than a consideration for everyday meteorological discussions. This is because recontextualization for Quine is to explicate ontological commitments for the purpose of establishing a tool to be used in scientific and philosophical disputes, where issues about ontology actually make a difference. The ontological commitments one makes is not something that underlies all modes of life. It is simply an act of recognizing what entities must exist for one’s philosophical framework to function. What there is is relative to what framework we wish to choose to view the world through. Whether it be phenomenalistic or physicalist, the ontology we adopt is whatever relative myth works best for our purposes. For Quine, what there is does not define what can be said like for Russell. What can be said defines what there is. It is within this context that Quine leaves us with his solution for all McX’s and Wyman’s of any form: “To be is, purely and simply, to be the value of a variable '' (Quine 32).

Bibliography:

1. Russell, Bertrand. “On Denoting.” Mind, New Series, vol. 14, no. 56, Oct. 1905, pp. 479–493.

2. Quine, Willard V. “On What There Is.” The Review of Metaphysics, vol. 2, no. 5, Sept. 1948, pp. 21–38.