This paper proposes an account of the semantics of names in fiction (fictional names)

ŞTEFAN OLTEAN¹

Abstract. This paper proposes an account of the semantics of names in fiction (fictional names). In doing so, it addresses the semantics of proper names first, which are seen as unstructured linguistic expressions, with no descriptive content or inherent meaning, whose sole function is to denote an individual directly. Their semantic value lies in their denotation, and they are like constants in the logical form. In contrast, fictional names are seen to function like variables bound by the existential quantifier, the individuals they denote in the worlds of fiction being identified at the actual world by description.

Keywords: fictional names, semantics, individuals.

1. INTRODUCTION

This paper proposes an account of the semantics of names in fiction (fictional names). To this effect, it addresses proper names first, whose semantics furnishes a backdrop against which fictional names are discussed. Out of various accounts of the former – descriptional, nondescriptional, intentional (see D’Angelo and Napoli 2000; Davis 2007; Frege 1960; Kripke 1982 and 2011; Portner 2005; Searle 1983) –, the nondescriptional ones are of greater relevance for this paper². According to them, proper names are unstructured linguistic expressions, with no inherent meaning or sense, without descriptive content, whose sole function is to denote an individual directly. Thus their semantic value lies in their denotation – the unique individual they identify; they are like constants in the logical form. In contrast, fictional names function like variables bound by the existential quantifier, the individuals they denote in the worlds of fiction being identified at the actual world by description.

2. PROPER NAMES

In the view indicated above, a proper name, Walter Scott for instance, would function like a symbol that cannot be analyzed into constituent parts, which would designate the unique individual “Walter Scott;” in other words, it would have no “inherent”, descriptive

1 Department of English, “Babeş-Bolyai” University, Cluj-Napoca; [email protected], [email protected].

2 This paper contains material from my article On the Semantics of Proper Names, in O. Felecan, A. Burgheşiu (eds.), Onomastics in Contemporary Public Space. Newcastle upon Tyne: Cambridge Scholars Publishing, 2013, 50–65.

RRL, LVIII, 4, p. 371–382, Bucureşti, 2013

meaning, or sense.³ But given that the same individual can also be identified by the definite description “the author of Waverley,” whose meaning is compositional, i.e., it is a function of the distinct meaning of the parts, some scholars (e.g., Frege, Russell – see Moeschler 1999: 151–152; and Kripke 1982, 2011) have proposed that proper names are descriptional, i.e., that they have a distinct sense, the exceptions being represented by situations in which the user of a proper name is in direct sensory contact with the individual denoted, in which case the name refers in a direct, unmediated way.

3. DESCRIPTIONAL VIEWS OF PROPER NAMES

An advocate of the descriptional view of proper names is Frege (1960). According to him, proper names are like definite descriptions, they having both sense and reference (see also Moeschler 1999; Davis 2007). If this were not the case, if the meaning of a name were identified with the referent only, we could assert that (1), an identity statement, would be synonymous with (2) (Davis 2007: 105):

(1) Shakespeare was Bacon.

(2) Shakespeare was Shakespeare.

Now, this would be wrong, even if (1) were true, given that we associate different descriptions or ideas with Shakespeare and Bacon. In other words, true identity statements involving names would become as trivial as “N is N” in a purely referential account (see Davis 2007: 105). We indeed come across descriptions that have the semantic content of names, as is the case with the so-called rigidified descriptions (Soames 2002: 43)⁴, which would give support to descriptional accounts. A case in point would be “Gödel, the discoverer of the incompleteness of arithmetic,” discussed by Kripke (1982): if the only thing that many people know about him is that he is the discoverer of the incompleteness of arithmetic, it would not follow that whoever satisfies this description is Gödel. As Soames (2002: 43) indicates, descriptions of this kind are not as reliable as proper names in picking out some individual. Russell, too, gave a descriptional account of proper names, considering them to be abbreviated definite descriptions. Discussing his view, Kripke (2011: 47) shows that if we take sentence (3), we can assert that it expresses a necessary truth, since if Hesperus and Phosphorus are proper names, and Hesperus is Phosphorus (both have the value “Venus”), it is necessary that Hesperus is Phosphorus (ibid.):

3 The distinction between sense “Sinn” and reference “Bedeutung” goes back to Frege (1960 [1892]), who assessed that linguistic expressions have two major sides of meaning – internal and external, representational and referential. His distinction has been reformulated within possible world semantics (see Carnap 1956) into that between intension and extension; the former corresponds to sense, being a function that assigns the extension at each possible world; the latter corresponds to reference, being determined starting from individuals (referential noun phrases), sets (predicates), and truth values (sentences).

4 “Rigidified descriptions” are “rigidified versions” of the descriptions associated with proper names (Soames 2002: 39). Soames indicates that while some scholars consider rigidified descriptions to be synonymous with the corresponding names, this analysis is wrong because these expressions do not have the actual semantic value of names.

(3) Hesperus is Phosphorus.

According to Russell (quoted in Kripke 2011: 47), things “might have come out otherwise”, however, and (3) could be an ordinary contingent, empirical truth: one planet is labeled The Morning Star (Phosphorus) when seen in the morning; the same planet was called The Evening Star (Hesperus) when seen in the evening; later on, it is discovered that it is the same planet. Still, he considers the names Hesperus and Phosphorus not to be genuine proper names or mere labels for an object, but “abbreviations for descriptions”

(ibid.): in this respect, even though Hesperus and Phosphorus are coreferential, one could assert something like Hesperus is seen in the evening, without asserting or believing the proposition expressed by Phosphorus is seen in the evening (Soames 2002: 26), because s/he would associate them with different descriptions. An analogy with the use of names in propositional attitude contexts is relevant at this point: if Cicero is Tully is true, and Tom believes that Cicero is an orator is true, it should follow that Tom believes Tully is an orator is also true; this is not the case, however, since Tom might not know that Cicero is Tully (Kripke 2011). It follows that the substitutivity of coreferential names fails, which is invoked as evidence for their descriptive content, too. Phenomena of this kind lead Russell to conclude that names name an object while also describing it, the truly proper names being names of things that we know by direct acquaintance, i.e., “this” or “that,” the only

“names” that function in the demonstrative mode (ibid.).

A different type of descriptional view is advocated by Searle (1983) in his variable description theory of proper names, who holds that a name expresses different descriptive contents on different occasions, all of which refer to the same object. In other words, names would not have a sense proper,⁵ but would be like indexicals, being “associated with different contents in different contexts” (Davis 2007: 118), but unlike indexicals, which are functions from contexts to intensions,⁶ they would be functions from contexts to

“descriptive intentional contents” (ibid.) and from the latter to extensions.In this respect, sentence (4) could be used to express the propositions (5–6):

(4) Aristotle was a philosopher (Davis 2007: 118)

(5) The author of De Anima was a philosopher (ibid.), or the proposition (6) The husband of Pythias was a philosopher” (ibid.).

This view of Searle’s comes closer to nondescriptional views.

4. NONDESCRIPTIONAL VIEWS OF PROPER NAMES

Nondescriptional accounts of proper names, or direct reference theories, go back to John Stuart Mill and thus form part of the Millian tradition. Representatives of this view (e.g., Kripke 1982; Soames 2002; Salmon 2005) argue that proper names have no descriptive content, that they are purely referential, unstructured linguistic expressions, with no inherent meaning or sense, and that their sole contribution to the proposition expressed by a sentence is the individual that they pick out. Central ingredients are the notions of

5 “None of the intentional contents associated with a name are what it means” (Davis 2007: 118).

6 See Kaplan’s (1989) notion of “expression character”– a function mapping contexts to intensions.

transworld identity⁷ and the derived notion of rigid designation from possible world semantics, which are to do with the issue of individuals at one world being identical with themselves at every other world.

5. RIGID VS NON-RIGID DESIGNATORS

Even though transworld identity is a logical issue, it is interestingly intertwined with the semantics of linguistic expressions. Among other things, it proves illuminating in an account of referential terms in natural language, which fall into two main categories, depending on whether they denote the same individual(s) in every possible world, or different individuals in different worlds. The one who addressed this issue is Kripke (1982;

2011), who classified the referential terms into rigid and, respectively, non-rigid designators, the former designating an object rigidly with respect to every possible world in which that object exists, and denoting nothing else with respect to worlds in which the object does not exist. The notion was further refined by Soames (2002) and Salmon (2005).

Proper names of individuals are the most typical case of rigid designators⁸; they are singular terms, “rigid with respect to tense and modality” (Cocchiarella 2005: 158), denote a single individual with respect to a given possible world, or “a class as many of one object […] identical with that object” (ibid.: 178), where “there is no empty class as many” (ibid.).

In other words, they introduce a constant – a, b, c, the entity associated with the name – at the level of logical form (LF). Of course, we can wonder, what is their contribution to meaning, as long as they have no descriptive content? Is it the individual they denote or is it some kind of concept? I believe it is the individual, even though this is still an open question, since LF is a level of interpretive representation, with constants having a conceptual nature (see Cocchiarella 2005, for referential concepts; and Carnap 1956 for individual concepts).

Kripke includes into the class of rigid designators certain common names as well, such as tiger, water, heat, blue; they are general terms, since they are applicable to any quantity or number of individuals. The latter are proper names of a special kind, namely

“proper names of universals” (D’Angelo and Napoli 2000: 203)—the universal meant by the common noun—, since they introduce a “higher order constant” at LF (ibid.).

Descriptions, be they definite or indefinite, like “the author of Waverley” (“the NP”) or “a dog” in “a dog is chasing a cat” (ibid.), are non-rigid designators, since they pick out other entities in different worlds. They can contain words of different categories, their meaning is compositional, and the entities they identify are not conventionally associated with them.

They introduce a variable at LF and “the universal or existential quantifier” binding it (Russell, quoted in D’Angelo and Napoli 2000: 207), which can stand, e.g., for Walter Scott in worlds in which he writes Waverley, for another individual in worlds in which someone else writes the book for him, an arbitrary, non-specific dog (“quantificational description”), or a specific dog (“non-quantificational, referential, description”) (ibid.: 203).

7 See, e.g., Kaplan (1979) and Salmon (2005) for the notions of transworld identity and rigid designation.

8 Of course, several individuals can bear identical names. This, however, does not challenge the claim of uniqueness associated with proper names. According to D’Angelo and Napoli (2000), in this case we have to do with “many different, though homophonous names, rather than with one shared name”.

6. PROPER NAMES AS PURELY REFERENTIAL EXPRESSIONS

In his work dedicated to names and naming, Kripke (1982) questions the relevance of descriptions or properties for an account of proper names, and shows that the meaning of the latter lies exclusively in their denotation. He expresses his view in his causal theory of reference, in which he strongly argues that these names are nothing more than mere labels attached to individuals within a ceremony of baptism, a link being thereby established between the name and its bearer. The link thus established becomes a necessary one which is retained and propagates within the community of speakers. It is not a consequence of the features of the individual, nor is it affected by the individual’s life history; thus, Walter Scott, who can be described as “the author Waverley”, would denote “Walter Scott” even if the latter had not written Waverley. Theoretically speaking, we could imagine a genuine causal chain of links from name user to name user that extends from Scott’s baptismal ceremony down to us today, our use of the name being grounded on the original act of naming. Now, with regard to what determines the identity of an individual, or what makes Walter Scott be “Walter Scott”, the society relies on experts, e.g. his parents or authorities.

Of course, one can wonder what kind of knowledge a speaker deploys when s/he uses such a name. As Chierchia and McConnell-Ginet (2000: 107) argue in an interpretation of Kripke’s theory, speakers do cognitively represent concepts associated with proper names (e.g., Walter Scott, Cicero) and use them in processing these words. The two scholars express, nevertheless, doubts as to what semantic role these concepts have or if they have any such role, since, according to Kripke’s causal theory, cognitive representations do not enter the formulation of truth conditions: “What is crucial for truth is the referential link itself, and that is a matter of the causal history of the world (…) rather than the conceptual structure” (ibid.). Names are thus linked to their extensions without the mediation of some descriptive content, and their meaning displays “an inescapable demonstrative component”

(Chierchia, McConnell-Ginet 2000: 107) ⁹; they are directly referring linguistic expressions.

The major arguments put forward by Kripke – that proper names contribute the entity they refer to and have no descriptive content – are semantic, epistemic and modal (Soames 2002: 19). The first ones concern the fact that the reference of a proper name is, in principle, independent of any description. A case in point are classical names, such as Aristotle or Cicero, about whom some people may know very few things if anything (e.g.,

9 Names can nevertheless change their reference, as Portner (2005) has shown. He cites an example from Gareth Evans: Madagascar originally was applied to part of the African mainland, but Marco Polo mistakenly used it for the island and gradually its meaning became that of an island. Thus the causal chain does not go back to an original act of naming, e.g., by a speaker who named the island, but reflects (mistaken) speaker meaning. It is thus grounded on the beliefs that the members of the community have in connection with the name: Madagascar refers to “Madagascar” because our beliefs about the name originate from the island and not from African mainland. Things would be similar with Walter Scott: this name denotes “Walter Scott” because our beliefs about him originate in the texts he authored, and even if he hadn’t written Waverley, we would still believe that he did, since our beliefs are associated with his actions of publishing the novel as his own. While we have reason to believe that this view is correct, we consider that Kripke’s causal theory is nevertheless preferable, given that it is possible to imagine a scenario, suggested by Portner (2005), in which authorities decide to switch to Madagascar’s original name because the present name is the result of a historical mistake. This highlights the importance of the original use of the name and the relevance of Kripke’s causal theory. In both cases, however, names emerge as rigid designators.

that the former refers to an ancient Greek philosopher, but know nothing else about him), but they are able to use the names successfully to denote some specific individual. Kripke’s contention is that the causal chain, which secures the gradual propagation of the link between the name and the individual from the time of the baptism down to us, is responsible for this. (In such cases, upon hearing the name, the speaker will use it in accordance with his “sources” [Soames ibid.]). The epistemic arguments are to do with the fact that the substitution of a proper name in a sentence by a description coupled with it yields a sentence compatible with a different knowledge or belief set as compared to the former. Thus, Columbus is commonly identified as the first European who discovered America. It is, however, much likely that the Vikings had discovered it long before. Now, those who associate the description “discoverer of America” with Columbus do not use the name “Columbus” to refer to some individual from northern Europe, but to the historical figure Columbus. They entertain false beliefs about him, but such beliefs do not change the denotation of the name. It follows that (7) is not an a priori, necessary truth, and that sentences containing “Columbus” are different semantically from sentences containing “the first European who discovered America” (see ibid.):

(7) Christopher Columbus is the first European who discovered America.

The modal arguments, in exchange, are centred around the fact that sentences containing proper names and the corresponding sentences containing descriptions have different truth conditions and take different truth values at different worlds (see ibid.). I will try to illustrate this point by examples based on models suggested by Soames (2002) and Portner (2005):

(8) Walter Scott is the author of Waverley.

(9) Walter Scott is Walter Scott.

(10) The author of Waverley is the author of Waverley.

These sentences should be semantically equivalent, as long as the proper name Walter Scott is replaced by the purportedly synonymous definite description “the author of Waverley”, and vice-versa. However, this is not the case, since (8) does not mean exactly the same thing as (9) and (10): the first is a contingently true and informative sentence, i.e., it is true at the actual world (wa) if and only if the individual bearer of the name Walter Scott authored Waverley in wa;¹⁰ the second (9) and the third (10) are necessarily true and uninformative identity statements¹¹, i.e. they are true everywhere. Let us now take (11) and (12), the latter being modelled on an example due to Russell, quoted in Kripke (2011: 45):

(11) Walter Scott is not the author of Waverley.

(12) The author of Waverley might not have written Waverley.

10 A sentence α is F is true at some world if and only if the value of α at w is the extension of F at w. So, for any world w, Walter Scott denotes in w the individual who was Walter Scott in wa) (see Soames 2002: 24).

11 We do not address here the issue of the non-literal, speaker or pragmatic meaning expressed by (9) and (10), explainable within a framework of Gricean pragmatics. Ours is a semantic account, being concerned with the literal meaning.

Walter Scott wrote Waverley, according to what we know, but (11) expresses the fact that there are worlds in which he did not author this work, but someone else did, e.g., a person he hired to write the book for him; the sentence is informative, but false at wa; however, it can be true at other worlds, i.e., there are logically possible worlds in which he is not the author of Waverley. Now, (12), which might seem a contradictory statement, expresses the fact that there are worlds in which the author of Waverley has the property that he does not write Waverley.

These indicate that the name Walter Scott and the description “the author of Waverley” do not have identical sense and reference, as they should if they were equivalent:

the first designates “Walter Scott” at all worlds, while the second designates whoever the author of the novel is, its denotation being determined by whatever individual has the respective property at different worlds. It follows that the definite description is not such that it identifies the individual Walter Scott. They are not synonymous either, since if “the author of Waverley” were synonymous with “Walter Scott”, and (11) were true, it would follow that (13) is also true:

(13) The author of Waverley is not the author of Waverley.

(13) is necessarily false. Consequently, the meaning of Walter Scott does not lie in some essential property or feature semantically associated with the name. The denotation of the former is not mediated by some descriptive content or feature ascription, while such content is relevant for identifying the referent of the latter. In other words, proper names only denote, and do so rigidly, designating the same individual in all possible worlds, while descriptions are nonrigid, designating different individuals in different worlds. It follows that proper names and descriptions are semantically different, or that, as Soames (2002: 53) asserts, “[T]here is little or no specific descriptive information that a speaker must associate with a name in order to understand it, or to be a competent user of it; hence there is little or no descriptive information that is part of the semantic content of such a name”¹².

Thus, proper names of individuals are non-descriptive, rigid designators, whose meaning lies exclusively in their denotation. As such, their contribution to the semantic content of the sentence are the individuals they denote, and not descriptive contents.

7. NAMES IN FICTION

If this characterization of proper names appears to be correct for names like Walter Scott, Aristotle or Columbus, there are names that seem to defy this rule, such as the names of individuals in fiction. They seem to contradict our thinking about proper names, since they lack denotation and thus contribute no entity to the meaning of the sentence. As a result, we cannot identify the individual and we do not know what is said by a sentence in which the name occurs, as (14) illustrates:

(14) Susan Rawlings had four children. (To Room Nineteen by Doris Lessing 1981 [1963])

12 This does not exclude the possibility that the reference of the name Walter Scott be fixed by the descriptive phrase “the author of Waverley”, or that the latter can be used as a criterion of identification of an individual as the referent of the name Walter Scott, the more so because the referent is not given to us through perception.

In (14), “Susan Rawlings”, the name of the main character in Lessing’s short story, is supposed to denote some individual, but it is unclear who this individual is, since it cannot be identified by ostension. It follows that this is an empty proper name (Currie 1990: 129) or a vacuous name. Now, as long as the contribution of a name to the meaning of a sentence is the individual it denotes, not identifying this individual would entail not knowing what proposition is expressed. Still, “Susan Rawlings” is meaningful, in that it contributes to the meaning of (14) the subject argument of a predicate.

An analogy with Russell’s treatment of definite descriptions is illuminating at this point (15 below, quoted from Currie 1990: 146); the latter pose problems, too, if they have no reference, being truth valueless or false:

(15) the F is G [something exists with property F, there is only one such thing, and this unique thing has the property G]

The formula (Currie 1990: 146; see also Salmon 2002: 45) is clearly wrong for (16 below), since there is no x with property F (individual being king of France – the name lacks reference) to fulfil the existential clause and the proposition “x is G” (“x is bald”) remains false for all values of x. Consequently (16) is either truth valueless or false. In (17), however, which contains a genuine proper name, things are different from (14): there is a historical individual identified by the name “Walter Scott” that satisfies the description “the author of Waverley” and the sentence can be assigned the truth value true:

(16) The king of France is bald.

(17) Walter Scott is the author of Waverley.

Now, even if someone may know next to nothing about Walter Scott, it can be argued that the causal chain would allow him or her to use the name successfully to denote the individual in question, in accordance with his or her “sources” (Soames 2002: 19).

Likewise, the sentence would give a clue as to how the referent can be identified, namely by ascribing to the denoted individual the property that he is identical with the individual who wrote Waverley.

8. HANDLING THE PROBLEM OF FICTIONAL NAMES

A philosopher who tried to look for a solution to this problem is Alexius Meinong (quoted in Crittenden 1991). He distinguished between existing and non-existing objects, and contended that words describing non-existing objects, e.g., “unicorn”, “fairy”, “the golden mountain” are not meaningless. Quine voices a similar opinion (see Marconi 1996:

11), considering that nondesignatig words of this kind (e.g., “Pegasus”) nevertheless have meaning. As is the case with Frege, what matters in their case is their sense, which allows us to establish that they lack reference. In keeping with such views, “Susan Rawlings” in (14) could be treated like a non-actual existent. Russell, however, reacted against this view, asserting that the notion of non-existing objects violates our “robust sense of reality”

(Crittenden 1991: 19). Currie (1990), likewise, rejects this notion, considering that it confronts other problems as well.

The assumption that speaking about non-existing objects should be a focus of inquiry in semantics is central to Lewis’s account (see 1979a, 1979b), whose work constitutes a breakthrough from a possible worlds perspective. His conception is characterized as radical actualism (Loux 1979: 46), in view of his distinction between what exists and what is actual, between existents and non-actual existents (these exist in worlds other than the actual world), or his suggestion that the latter are on a par with the former¹³. These distinctions have proved, nevertheless, seminal in an account of fictional names, since they allow us to assume that the individuals denoted by such names are located not in the actual world, but in worlds that the text of fiction describes. In this respect, “Susan Rawlings”

could be treated as a non-actual existent, and her actuality would be an indexical matter, in that it would depend on the context of the fictional utterance. She would thus exist in some world, w, compatible with the fiction, even if she cannot be identified by ostension, but only by description: the person who is and does such and such.

According to Currie (1990), however, a description at wa is not safe for identifying and fixing the reference of such a name, since there may be different individuals in different worlds of the fiction with the same name, as the following scenario, adapted to our example, illustrates. Let us suppose, for instance, that in one world of the fiction To Room Nineteen (w1) “Susan Rawlings” is individual a, who has four children; that in another world (w2), “Susan Rawlings” is individual b, the only one with four children; in still another world (w3), “Susan Rawlings” is also individual a, but no one in this world has four children. Sentence (14) would thus express different propositions (p1) in w1 and w3, on the one hand, and w2, on the other (p2), with truth values as shown below:

p1 = [w1 → 1], [w3 → 0] (p1 is true in w1, false in w3) p2 = [w2 → 1] (p2 is true in w2)

It follows that the unique object (a) in w1 having property P (has four children) is different from the unique object (b) in w2 having property P, and is still different from the unique object (a) in w3 having property P’ (does not have four children). Furthermore, the object denoted by “Susan Rawlings” in w1 and w3 displays conflicting properties (a ≠ a), which leads to inconsistency. As a result, the individual called “Susan Rawlings” is not identical with herself at other worlds: it denotes different individuals or individuals with conflicting, impossible features. What may be counterintuitive indicates, nevertheless, that the name emerges as a non-rigid designator and, as such, it is not a genuine proper name.

Now, “Susan Rawlings” designates nothing at wa, and as long as proper names should have reference in order to be meaningful, no sentence in which this name occurs expresses a true proposition. But (14) is fictional, and the reader knows this. As a result, s/he will not take it

13 To handle these objects, he proposes two senses in which the existential quantifier can be used: a restricted sense, in which it ranges over things that exist and in an unrestricted sense (absolute quantification), in which it ranges over all possible objects, existents and non-existents (quantifier with a modifier attached: ∃a). While this distinction preserves consistency, most philosophers find the suggestion unilluminating, since “Lewis’s theory requires world-bound individuals and counterparts that fail to accommodate modal intuitions and seem to result principally from his possibilism […]

[A]nother, namely actualist, interpretation is […] to be preferred” (Jay W. Richards, The Untamed God: A Philosophical Exploration of Divine Perfection, Downers Grove: InterVarsity Press, 2003:

72); likewise, anything that exists in the actual world, is “greater” than what exists in a possible world.

to be true at wa but will make-believe it to be true at some story world (Currie 1990: 72), make-belief being, like truth, a property of propositions, i.e., a function taking propositions and giving back truth values, not in the actual world but in the worlds of the fiction (ibid.).¹⁴ So, while (14) does not express a proposition, (18) does.

(18) It is fictional that Susan Rawlings had four children / In the fiction To Room Nineteen Susan Rawlings had four children. [the sentence is in the scope of the fictional operator “Fs(P)” (“P is true in fiction S…,”¹⁵)].

Sentence (18) is thus taken to be true in the worlds of the fiction, but it can express different propositions at each world (w1, w2, w3) compatible with the fiction (see above):

(had four children)(a), (had four children)(b), but no proposition at wa. Now, while (18) is relevant for the way the reader makes sense of the fictional utterance in the story, it still does not capture how s/he understands sentence (14) at wa, since “Susan Rawlings” does not pick out some particular individual in the actual world. Here is an answer furnished by Currie (1990): the reader will not pick out some particular individual and make believe this individual is involved in the events of the story, but will pick out those worlds (alternative worlds) from the set of possible worlds in which there is someone who is and does everything that, e.g., Susan Rawlings in our example, is described to do in the fiction (see 19, 20 and 21 below): this individual is called Susan Rawlings, has four children, lives in London in the 1930’s and so on, but is different in each distinct world.

(19) Someone has four children (“someone” is a variable bound by the existential quantifier)

(20) ∃x [person(x) & has four children (x)]

(21) ∃x [Susan Rawlings(x) & has four children (x)] (name replaced by a variable bound by the existential quantifier ranging over objects in some fictional world)

So, the name being empty at wa, it is understood the way variables bound by the existential quantifier are understood, and identification at other worlds is by description: the individual who is Susan Rawlings and does all the things Susan Rawlings is and is said to do in the fiction (see Currie 1990). In keeping with these, I propose in (22) a tentative formula for the name Susan Rawlings (“S.R.”) in (14). In elaborating it, I have been guided by the assumption that the assignment of properties in fictional worlds does not differ from that in the actual world, and that understanding a fictional name implies considering all the descriptions associated with it in the fictional text¹⁶, i.e., it involves a maximal mental representation.

14 Make-belief is an act of fiction making, a kind of pretense, different from assertion (Currie 1990).

15 A proposition P in fiction S is non-vacuously true if and only if there is a story world where the fiction is told as known fact and P is true; it is vacously true or untrue if and only if there are no story worlds where the fiction is told as known fact Lewis (1983 [1978]: 270).

16 “[N]o proper part of a fictional story containing fictional names is semantically independent enough to express a proposition on its own. Only the whole story expresses a proposition […] ” (Currie 1990: 155). Accordingly, the English scholar proposes a Ramsey sentence in his account of fictional names (from Frank Ramsey’s formula for the structure of scientific theories: T [t1…tn],

(22) [[S.R.]]^FS = ∃xλP [Px & R[P (P ⊆ Q), x (x is called S.R.)]] – where FS is fiction To Room Nineteen, P is the set of properties associated with S.R.; Q is the full set of descriptions in the fiction, of which P is a subset, and R is a relation linking the set of properties P to the individual S.R.¹⁷ (the property of having four children, of being called S.R, etc.)

The formula captures the fact that the set P of properties associated with some individual (Susan Rawlings) in the fiction is a subset of the set of properties Q that the fiction describes (a story features, as a rule, several individuals: x1, x2, x3...xn). According to this formula, a fictional name takes a variable (an x) bound by the existential quantifier as its value (not a constant: a, b, c); as such, the individual is not unique, s/he satisfies all the descriptions in the text of the fiction, but can be different in each different world of the story. The name is therefore non-rigid, from the actual world perspective. Furthermore, the identification of the individual is by description, it involves considering all the descriptions associated with it in the story¹⁸. Fictional names therefore emerge as non rigid descriptive terms that enable readers to make-believe the individuals in fictions in keeping with the descriptions in the stories. They do not contribute individuals to the logical form, but variables.

9. CONCLUSIONS

This paper has proposed an account of fictional names from the actual world perspective, according to which the meaning of these names is granted by the sum of descriptions associated with individuals in fictions. From the perspective of the worlds of fiction, however, fictional names denote individuals; since the latter cannot be identified by ostension, it is unclear who these individuals are, and we are left with making sense of them also in terms of properties / descriptions.

REFERENCES

Carnap R., 1956 (1947), Meaning and Necessity: a Study in Semantics and Modal Logic, Chicago, University of Chicago Press.

Chierchia, G., S. McConnell-Ginet, 2000, Meaning and Grammar: An Introduction to Semantics, Cambridge, MA: MIT Press.

Cocchiarella, N. B., 2005, “Denoting Concepts, Reference, and the Logic of Names, Classes as Many, Goups, and Plurals”, Linguistics and Philosophy 28, 135−179.

where ti’s are terms [see Currie 1990: 150]), which looks like ∃x1…∃xn[F(x1…xn)] and which he applies to the structure of stories. In this formula, the xi’s are variables bound by an existential quantifier that replace fictional names, each having a set of properties, including relations with the others: there is an n-tuple of things that satisfies the properties and relations specified in the story (ibid.: 150).

17 Such a relation was suggested to us by K. von Heusinger and J. Wespel (2006), who use C for specific types of contextual relations linking individuals to properties.

18 See also ftn. 16 for Curry’s Ramsay sentence, which is meant to capture the structure of stories in terms of relations holding among individuals.

Crittenden, C., 1991, Unreality. The metaphysics of fictional objects, Ithaca and London, Cornell University Press.

Currie, G., 1990, The Nature of Fiction, Cambridge, Cambridge University Press.

D’Angelo, M., and E. Napoli, 2000, “Proper Names, Descriptions and Quantifier Phrases,” in:

D. Marconi (ed.), Knowledge and Meaning, Vercelli, Edizioni Mercurio, 195−234.

Davis, W. A., 2007, “Intentionalism, descriptivism, and proper names,” in: S.L. Tsohatzidis (ed.), John Searle’s Philosophy of Language. Force, meaning and Mind, Cambridge, Cambridge University Press, 102−124.

Frege, G., 1960 (1892), “On Sense and Reference”, in: G. P. Geach, M. Black (eds.), Philosophical Writings of Gottlob Frege, Oxford: Blackwell, 1960, 25−50.

Kaplan, D., 1979 (1978), “Transworld Heir Lines”, in: M. J. Loux (ed.), The Possible and the Actual:

Readings in the Metaphysics of Modality, Ithaca and London, Cornell University Press, 1979, 88−109.

Heusinger, K. von, J. Wespel, 2006, “Indefinite Proper Names and Quantification over Manifestations”, in:

E. Puig-Waldmüller (ed.), Proceedings of Sinnund Bedeutung 11, Barcelona, Universitat Pompeu Fabra, 332−345.

Kripke, S., 1982 (1972), Naming and Necessity, Oxford, Blackwell.

Kripke, S., 2011, “Identity and necessity”, paper presented at the University of Bucharest, 2011 (reprinted by permission of New York University Press from Milton K. Munitz (ed.), Identity and Individuation, New York, New York University Press, 1971, 135−164.

Lewis, D., 1979a (1969), “Counterpart Theory and Quantified Modal Logic,” in: M. J. Loux (ed.), The Possible and the Actual: Readings in the Metaphysics of Modality, Ithaca and London, Cornell University Press, 1979, 110−128.

Lewis, D., 1979b (1973), “Possible Worlds”, in: M. J. Loux (ed.), The Possible and the Actual:

Readings in the Metaphysics of Modality, Ithaca and London, Cornell University Press, 1979, 182−189.

Lewis, D., 1983 (1978), “Truth in Fiction”, in: Philosophical Papers, vol. I, New York, Oxford, Oxford University Press, 261−280.

Loux, M. J. (ed.), 1979, The Possible and the Actual: Readings in the Metaphysics of Modality, Ithaca and London, Cornell, Cornell University Press.

Marconi, D., 1996, Lexical Competence, Cambridge, MA, MIT Press.

Moeschler, J., and A. Reboul, 1999 (1994), Dicţionar enciclopedic de pragmatică, Cluj, Editura Echinox.

Oltean, Ş., 2009, “On the semantics of proper names and common names”, Studia Universitatis Babeş-Bolyai, Philologia, LIV, 4, 349−359.

Oltean, Ş., 2013, “On the Semantics of Proper Names”, Onomastics in Contemporary Public Space, Newcastle upon Tyne, Cambridge Scholars Publishing (to appear).

Portner, P. H., 2005, What Is Meaning? Fundamentals of Formal Semantics, Oxford, Blackwell.

Rescher, N., 1979 (1973), “The Ontology of the Possible”, in: M.J. Loux (ed.), The Possible and the Actual: Readings in the Metaphysics of Modality, Ithaca and London, Cornell University Press, 1979, 166−181.

Salmon, N., 2005, “Are General Terms Rigid?”, Linguistics and Philosophy, 28, 117−134.

Searle, J. R. 1983, Intentionality. An Essay in the Philosophy of Mind, Cambridge, Cambridge University Press.

Soames, S., 2002, Beyond Rigidity. The Unfinished Semantic Agenda of ‘Naming and Necessity, Oxford, Oxford University Press.