(PDF) Names are not (always) predicates

Submitted article Names are not (always) predicates Laura Delgado Centre of Philosophy, University of Lisbon (Requested info) Correspondence: Laura Delgado, Centre of Philosophy, University of Lis- bon, Alameda da Universidade 1600-214 Lisbon, Portugal Email:

Abstract A main selling point of Predicativism is that, in addition to accounting for predicative uses of proper names, it can successfully account for their referential uses while treating them as predicates, thus providing a uniform semantics for proper names. The strategy is to postulate an unpronounced determiner that is realised with names when they appear to function as singular terms, making them effectively a concealed determiner phrase. I argue against the thesis that names are really predicates in referential uses. I discuss four different environments where names do not behave like the determiner phrases that are thought to embed them. Key words philosophy of language, meaning, reference, proper names, predicativism 1 INTRODUCTION Although on the orthodox post-Kripkean tradition, proper names have been treated as referential singular terms, this orthodoxy has come under attack in the work of several philosophers who argue that proper names are in fact predicates: most prominently Burge (1973); and more recently, Elbourne (2005), Matushansky (2006, 2008, 2015), Fara (2011, 2015a).1 On this view, proper names are not referential expressions but treated as general terms instead. The idea that names are predicates starts with the observation that on some occurrences, proper names behave like common count nouns, that is, they occur in the plural, and take indefinite and 1 Other defenders of the predicate view include Sloat (1969); Hornsby (1976); Higginbotham (1988); Larson and Segal, 1995; Elugardo (2002); Sawyer (2010); Izumi (2012); and Tayebi (2018) 1 definite articles, quantifiers and other determiners. Here are a few examples, first emphasized by Burge (1973, p. 429):2 (1) a. There are relatively few Alfreds in Princeton. b. An Alfred Russell joined the club today. c. The Alfred who joined the club today was a baboon. d. Some Alfreds are crazy; some are sane. Burge argues further that given that these modified occurrences (i.e., predicative) should not be set aside as special and, moreover, that there is no reason to considered them as semantically independent from the more typical unmodified occurrences (i.e. referential), then the best theory would be a uniform theory that is able to account for both types of occurrences within a single semantics. And this is possible, according to Burge, because modified occurrences of proper names have the same conditions of literal application to an object as singular, unmodified occurrences have: “A proper name is (literally) true of an object just in case that object is given that name in an appropriate way” (Burge, 1973, p. 430). Thus, the meaning of a name N would have as semantic value either a property, that is, being called “N”; or an extension consisting in those individuals called “N”. While in general referentialists accept that this is the right semantics for predicative uses of names, they deny that referential uses are also predicative.3 Clearly, if proper names are still predicates in these referential uses, the predicativist needs to explain two things. The first is the semantic ability of a proper name, when appearing as a bare singular in argument position, that is, “Alfred is happy”, to denote a particular individual, while common count nouns cannot denote a particular individual unless they are modified by some determiner, that is (*)“Student is happy”. The second thing to explain is the syntactic fact that proper names can appear as bare singulars in argument position while common nouns cannot; in argument position common nouns must take some determiner.4 Predicativists explain both things by postulating an unpronounced determiner to go with bare names (either a demonstrative (Burge, 1973; Sawyer, 2010), or a definite article (Elbourne, 2005; Fara, 2015a; Matushansky, 2006, 2008)). Names would then be like any other count noun occupying the nominal position in either a definite description or a complex demonstrative. A sentence like 2 Burge mentions Sloat (1969) as a predecessor of his view. Sloat had a linguistic theory that proper names were a kind of count noun, based precisely on the distributions of proper names and determiners. 3 The most important works defending Referentialism against Predicativism include Delgado (2011, 2019); Leckie (2013); Rami (2014, 2015); Jeshion (2015a, 2015b, 2017, 2018); Hinzen (2016); Schoubye (2016a, 2017); and García- Carpintero (2017). More recently also Lee (2020b); Lee (2020a) and Delgado (2022). 4 Unless a mass, kind-denoting or otherwise abstract interpretation is intended. 2 “Alfred is happy”, where the name seems to function as a referential singular term, would be analysed as either “The Alfred is happy”, or “That Alfred is happy”. Thus, in both predicative and seemingly referential uses proper names will have the same meaning, and the same contribution to propositional content, that is, the property of being called “N”, or something of the sort.5 This is supposed to give superiority to the predicative view. Referentialists cannot provide a uniform semantic account for both referential and predicative uses because in predicative uses proper names clearly cannot be interpreted as singular terms or individual constants.6 So they must hold separate accounts for referential and predicative occurrences of proper names. Predicativists argue that since their account can accommodate the putative referential uses of names while still treating them as predicates, their view has the advantage of providing a uniform account.7 Clearly, the success of the predicate view – and its superior ability to provide a uniform account of proper names – hinges on whether it can adequately accommodate referential uses of proper names. In this paper, I argue against the thesis that proper names are really predicates in their typical referential uses. In the following sections I will discuss four environments in which proper names (when bare singulars in argument position) appear not to behave like the definite descriptions or the complex demonstratives that are thought to embed them. I will begin in section 2 by discussing a problem raised by King (2006) that bare proper names do not interact with certain anaphoric uses of “one”, to show that this problem also extend to versions of The-Predicativism. In section 3, I revisit the problem of binding that has been discussed in the literature, to give a response to Fara’s 5 All predicativists endorse some version of the being-called condition: BCC ‘N’ (when a predicate) is true of a thing just in case it is called N (Fara, 2015a, 64). See for example Elbourne (2005, 172–74), and Matushansky (2008, 599). 6 Referentialist could however reject predicative uses as non genuine (and therefore reject the need to account for them within the semantics of proper names). One option is to treat the so-called predicative uses as involving ellipsis where the name is mentioned but not used. Then “An Alfred joined the club” would be elliptical for “A person called ‘Alfred’ joined the club”. Support for this explanation can be drawn from the fact that speakers would indistinctly use both the full “an x called N ” and the elided “an N ” forms, and sometimes would even offer the full form as clarification when the elided form sounds odd (for example, due to the morphology of the name, i.e., “An Ann”) or ambiguous (i.e., “That Rose was beautiful”), or unclear in some other way. However, is not clear that the view copes with the plural, that is, “Two Alfreds”, since it is not simply that some material is elided, but some transformation is involved to get the plural marking. Or consider verbal uses of names that can be conjugated, for example, “John googled the information”, where is not clear how to get to the past tense from some full form that is elided (“search using the search engine called ‘Google” ’). It seems likely that for the case of verbal uses of names some process of conversion and lexicalisation takes place. But then it seems likely that at least to some degree the same applies to other uses of proper names, including the predicative use (cf. Leckie (2013); Delgado (2019)). Referentialists in general accept that proper names have predicative uses. 7 The Uniformity Argument is generally appealed to by predicativists in some way or another: Burge (1973, 437), Hornsby (1976, 229); Elugardo (2002, 474–83); Elbourne (2005, 170–71); Sawyer (2010, 207); Fara (2011, 9) and Fara (2015a, 109–12). 3 treatment of it Fara (2015b). I also provide cross-linguistic evidence that the problem persists even for languages, or versions of languages, that allow the definite article to go with names. In section 4, I bring an argument given by Longobardi (1994) about the expletive definite article in Italian to bear on this debate, showing that the argument tells against the unpronounced determiner thesis for English as well as other romance languages (Catalan, Portuguese, Spanish). Lastly, in section 5 I provide an argument that bare proper names cannot generate generic sentences and this put pressure on the idea that a bare occurrence of a proper name is effectively a determiner phrase. It has been argued that the definite description hypothesis (“The-Predicativism” henceforth) enjoys cross-linguistic support from the fact that in some languages (or in some variants of them) names appear with the definite article.8 However it has never been shown that this fact does support The-Predicativism, that is, their proponents have never tested whether “the+N ” in these languages really work like ordinary descriptions.9 I will show throughout the paper that in many of those languages that permit or require the use of the definite article with names (i.e., Catalan, Greek, Portuguese and versions of German, Italian, and Spanish), the construction the+N does not behave like an ordinary definite description in every context.10 8 Elbourne says that a “very powerful’ reason for taking the unpronounced determiner to be a definite article is that one gets automatic cross-linguistic support for it, since in other languages proper names take the definite article (he mentions Greek and some dialects of German) (Elbourne, 2005, 172). Matushansky claims that “the default is instantiated by languages that do have definite articles with proper names in argument positions” (Matushansky, 2006, 285), citing in particular Portuguese, Greek, dialects of German and Italian, as well as languages with preproprial articles (i.e., Catalan, Icelandic, Northern Norwegian and Tagalog among others). Fara mentions Portuguese, Modern Greek, French, German, Italian, and Spanish (Fara, 2015a, fn21). 9 Hinzen (2016) on a similar vein discusses the case of Catalan that, he notes, is often cited “yet rarely explic- itly analysed” (p.598). Hinzen also mentions Longobardi’s discussion of Italian (Longobardi, 1994), to conclude that “Languages in which proper names under referential readings co-occur with determiners falsify predicativism." (Hinzen, 2016, 599). 10 It is important to note here some nuances regarding the use of the definite article with proper names in these languages. The article is obligatory with names in Catalan, Greek and Portuguese; however there has been some discussion that for these languages this use of the article is an expletive (Martí (2017, fn5) suggests this about Catalan and cites Coromides (1982, p. 309) for a similar explanation of the origin of the use of the article with names in Portuguese. Also Hinzen (2016) argue that the article with proper names in Catalan is an expletive, and Longobardi (1994) suggests the same for both Catalan and German, in addition to Italian. See Lekakou and Szendroi (2012) for the Greek case). Also, proper analysis of the Catalan case needs to include mention of the fact that Catalan has a preproprial article and there are regional variations regarding its use (it is obligatory in Balearic Catalan, but optional in both Catalonian and Valencian Catalan). With respect to German, Italian and Spanish, the use of the article with proper names varies regionally: The article is not used in the north of Germany but is common in the south and even preferred in Switzerland. For Spanish, the article is typically absent in most countries but is customarily used in Chile and the regions of Catalonia and Murcia in Spain; and is not common but used in certain contexts in some other countries. For Italian, in addition to regional variations, the use of the article varies according to gender (in most regions where it is used, it is only with female names). See Delgado (2018) for more details. In this paper, I discuss examples from speakers of those variants of these languages that do normally use the article 4 2 ONE-ANAPHORA AND NOUNS VS NAMES The first problem for the view that a proper name N is equivalent to the complex demonstrative “that N ”, as noted by King (2006, 149), is that unmodified proper names do not interact with certain anaphoric uses of one whereas count nouns do.11 Consider the following pairs of sentences: (2) a. That Martin is happy but this one isn’t. b. (*) Martin is happy but this one isn’t (King, 2006, 149, slighly modified). The thought is that if in (2.b) the demonstrative is present – albeit covertly – both sentences should work fine. But (2.b) does not. To reinforce the point against a possible rejoinder that what is preventing the anaphoric “one” from interacting with the proper name is the fact that the determiner is implicit, King argues that even in cases where bare common nouns are thought to carry an implicit determiner – that is, the generic determiner Gen – the analogous to (2.b) works fine: (3) a. Dogs are kind. b. Dogs are kind, but this one isn’t (King, 2006, 149). The anaphoric one would interact with proper names in their predicative uses – that is, (2.a) – but not when the proper name occurs unmodified. This difference suggests that proper names in their unmodified uses do not involve an implicit demonstrative and are not functioning as predicates.12 King’s one-anaphora argument was raised against That-Predicativism. I would like to argue that the problem also exists for The-Predicativism, at least in some versions of it. For example, in both Elbourne’s and Fara (2011)’s versions, proper names carry an unpronounced restrictor or predicate modifier in addition to the unpronounced definite article.13 . So bare occurrences of N in argument position are analysed as either “TheF N ”, or “The N F ”. This unpronounced restrictor or predicate modifier serves to single out an individual from the extension of N. But if this were so, then there should be no difference between (4.a) and (4.b), where (4.a) makes the putative with names. 11 I said “certain” because there is at least one type of construction where one is anaphoric on a bare proper name: “I met David. Which one?”. However, the claim is not that one can never be anaphoric on a bare name, or for that matter, on a referential term in general (witness: “I want to buy it/ I want that. Which one?”). The claim is rather that on those cases where one requires a count noun as anaphoric anchor, bare names in argument position cannot do the job, suggesting they are not count nouns on those occasions. 12 Similarly, Rami (2014, 853) mentions the one-anaphora problem to argue that referential names have no descrip- tive content (what he calls the objection from descriptive austerity). 13 With the difference that while Elbourne believes that this unpronounced restrictor is an index attached to the definite article in its underlying structure (Elbourne, 2005, 95, 173), Fara (2011) thinks it is a predicate modifier implicit in the NP, that gets assigned a value in context (Fara, 2011, 8–12) 5 underlying structure of (4.b) explicit: (4) a. The Martin I know/from Alaska/that we met is tall but this one isn’t. b. (*) Martin is tall but this one isn’t. In (4.a), the proper name can serve as anchor for the anaphoric one, since it is been used predica- tively, just like any other count noun would, for example, “The teacher (I know/from Alaska/that we met) is tall but this one isn’t”. But the fact that the proper name does not interact with the anaphoric one in (4.b) suggests that the proposed underlying structure is not correct.14 Further, I would also argue that the problem arises more generally for The-Predicativism. Con- sider the case of a stressed the, which arguably makes the proper name function like a predicate. This allows the proper name to interact with the anaphoric one in (5.a), whereas the bare name in (5.b) does not:15 (5) a. THE Marc Jacobs is rich but this one isn’t. b. (*) Marc Jacobs is rich but this one isn’t. The examples show that definite descriptions containing proper names (whether incomplete or complete) can interact with the anaphoric one, suggesting that it is nothing about the proper name as a class of word but about its syntactic position and corresponding semantics, that is, when bare in argument position, that disallows one-anaphora anchoring. 3 THE ARGUMENT FROM BINDING A second problem for both types of predicativism is that bare proper names behave differently from complex demonstratives and definite descriptions when embedded in quantified sentences.16 14 Schoubye (2017) also argues that King’s one-anaphora problem carries over to The-Predicativism. He offers an example with an incomplete definite description containing a count noun that can serve as anaphoric anchor for a one-anaphor, that is, “I left the duvet outside, but there is another one upstairs” (Schoubye, 2017, 757, my highlight), whereas a bare name in it place would not: “(*) I left Paul outside, but there is another one upstairs”. But anaphora on an incomplete description does not always work well: “(?) I met the student, but there is another one outside”; “(?) The doctor is taller than the one I usually see”; “(?) The player is good but the one I met isn’t”. And this perhaps allows for an alternative explanation of what prevents bare names from being anaphoric anchors on one-anaphora. Thus, while I agree with Schoubye in general, I argue for the same point appealing to different cases. 15 Note that the cases I flagged as marked in fn14 work better with the stressed the: “I met THE student, but there is another one outside”, “THE doctor is taller than the one I usually see”, and so forth. 16 This problem was originally raised by Elia Zardini to John Hawthorne and David Manley on a seminar in Arché (St Andrews) in November 2008, for a early version of The-Predicativism they were defending. This has been taken up in Delgado (2011) for That-predicativism, and is discussed in Hawthorne and Manley (2012), and Schoubye (2017) with respect to definite descriptions 6 Consider first the case with complex demonstratives: (6) a. Every philosopher who admires a logician will remain fan of that logician forever. b. Every philosopher who admires a David, will remain fan of that David forever. c. We know a Martin did it but we do not know which one. So, everybody who knows a Martin should call that Martin immediately. (Delgado, 2011, 51) On the most natural reading of the sentences in (6), the demonstrative phrases have narrow scope readings with respect to the quantifier phrases that precede them, so that their semantic value co-varies with the preceding quantifier phrase. That is to say, the demonstratives “that logician”, “that David”, and “that Martin” would pick up different individuals for every philosopher or person who knows a Martin, respectively. A wide scope reading with respect to the quantifier is possible too, but less salient: for example, (6.a) has a reading where every philosopher that admires one logician or other will remain fan of one particular demonstrated logician (referred to by “that logician”). However, bare proper names in place of the demonstrative phrase in these sentences do not naturally have narrow scope readings – indeed the converse happens. Narrow scope readings are at best hard to access for (7.a)-(7.b), while wide scope readings are most salient: (7) a. Every philosopher who admires a David, will remain fan of David forever. b. We were searching for a Martin. So I requested to everyone who knew a Martin to call Martin immediately. But this is surprising if according to That-Predicativism, names are covert demonstrative phrases. Definite descriptions also have very natural co-varying readings according to preceding quantified phrases: (8) a. In every conference we hold, it is the organizer who writes to the speakers. b. In every presidential debate, the democrat won. On the narrow scope reading, the value of “the organizer” or “the democrat” will co-vary with the quantified expressions “in every conference” and “in every presidential debate” respectively. However, when a bare name occurs in place of the definite descriptions in (8.a)-(8.b), even if a special setup is provided, a narrow scope reading is not available:17 (9) a. In every conference we hold, it is David who writes to the speakers (assume that in 17 Schoubye argues that if we replace the bare proper name in (9.a)-(9.b) with the alleged semantic correlate “the person called N ”, we get the narrow scope reading all right (Schoubye, 2017, 760). This shows (in my opinion, conclusively) that bare proper names do not have the predicativist’s alleged semantics on these occasions. 7 the faculty there are three Davids and as it happens, them and only them organize the conferences, so there is a salient relation “conference-David”). b. In every presidential debate, Clinton won (assume that a different Clinton participated in different debates).18 This shows that bare proper names typically behave differently from definite descriptions in quan- tified sentences, suggesting that at least in these occasions they are not incomplete definite descrip- tions.19 One could resist this conclusion by noting that while adding the definite article to the bare name in (9.b), that is, the Clinton, makes the bound reading available, adding it in (9.a), that is, the David, does not, suggesting that incomplete definite descriptions involving names do not always have bound readings. Perhaps the difficulty here has to do with the fact that this construction (definite article plus name) is generally judged ungrammatical, and judgements of ungrammaticality clearly can interfere with what readings we get for a sentence. Whether the+N is actually ungrammatical or not (Jeshion (2015b, 2017) has forcefully argued for the grammaticality of the+N ) the agreement is that a special set up is needed to hear the+N as grammatical. Consequently, the point can be made circumscribed to these contexts. Wherever the construction the+N is judged grammatical and it has a natural bound reading (a-sentences) the difficulty of binding into bare names remains (b-sentences): (10) a. In every presidential debate, the Clinton won. b. In every presidential debate, Clinton won. (11) a. From then on, at every party, Colin would buy the Katherine a drink (adapted from Jeshion (2017, 231)).20 b. From then on, at every party, Colin would buy Katherine a drink. I would like to add to this discussion a widely overlooked datum: in those languages or variants of languages where proper names can take the definite article (and so there are no ungrammaticality judgements interfering with getting the relevant readings) they still resist a co-varying reading. 18 A similar case is mentioned in Hawthorne and Manley (2012, 237) who claim that “In every presidential debate, the Clinton won” is far more acceptable to describe the situation (my italics). 19 I say “typically” because there are some (rather marginal) examples of proper names co-varying with quantified phrases. I will come back to this in section 3.1 20 Jeshion introduced several examples known as “The Katherine sentences”, taking inspiration from John Green’s novel An Abundance of Katherines in which the protagonist, Colin Singleton, has a proclivity for Katherines. In a context such as that of the novel, where being called Katherine is a salient property, a sentence such as “Where is the Katherine?” would be judged grammatical. See Jeshion (2015b, 396) and (2017, 227) for the exposition of this claim. 8 Consider the English sentence “In every class, Maria passed the exam” and its equivalents in other 2122 languages: (12) a. A totes les classes, la Maria ha aprovat l’exàmen. (Catalan) in all the classes, the Maria has passed the-exam “In every class, (the) Maria passed the exam” 23 b. Se kathe taxe he Maria perase ten exetase. (Greek) In each class the Maria passed the exam “In every class, (the) Maria passed the exam” c. In jedem Kurs, die Maria hat bestanden. (German) In every class, the Maria have passed “In every class, (the) Maria passed the exam” d. In ogni classe, la Maria ha passato l’esame (Italian) In each class, the Maria have passed the-exam “In every class, (the) Maria passed the exam” e. Em cada disciplina, a Maria passou o exame. (Portuguese) In each class, the Maria passed the exam “In every class, (the) Maria passed the exam” f. En cada clase, la Maria aprobó el examen. (Spanish) In each class, the Maria passed the exam “In every class, (the) Maria passed the exam” The sentences in (12) do not naturally have a reading in which in every class a different Maria passed the exam (the co-varying reading is not impossible, but it is hard to access and would require a special set up).24 This indicates that proper names in these languages, despite the fact that they 21 Note that in these examples, the sentences are ordered to give the quantified expression the widest scope more naturally, in order to facilitate the reading in which a different person in every class passed the exam. 22 For these data, I am partly relying on my own competence in Spanish, Italian, and Portuguese. However, I have also consulted native speakers: for Catalan (data due to Sonia Roca Royes), for Greek (data due to Ira Kiourti), for German (data due to Barbara Vetter), for Italian (data due to Elia Zardini), and for Portuguese (data due to Teresa Marques) (all personal communication). Note also that I have consulted speakers for which the use of the article is common. 23 Note that I have bracketed the definite article before the name in the translation, to indicate that, in my opinion, the natural translation of these sentences into English would not include the definite article. 24 To be clear, if “the Maria” is replaced with some ordinary definite description in these sentences, that is, “In every class, the orphan/the foreigner passed the exam”, the co-varying reading emerges naturally. 9 take the definite article, do not automatically function like definite descriptions. The thought here is not that incomplete definite descriptions involving names in these languages are special in that they resist bound readings. Rather, it is likely that the article here is an expletive, and so the name still functions as a referential term (cf. Longobardi (1994); Lekakou and Szendroi (2012); Hinzen (2016); and Martí (2017)). In sum, the fact that proper names behave differently from complex demonstratives or definite descriptions in quantified sentences suggests that bare proper names are not equivalent to them on these occasions. Elbourne, however, believes that the inability of proper names to be bound in quantified sentences is not a problem for his approach: “It is quite open to us to say that, while they have the structure of definite descriptions, proper names also have some additional features that prevent their occurrence in [these] sentences” (Elbourne, 2005, 180). But he also says that he would not attempt to give an explanation of these additional features without which, I think, his remark is ineffective. Note that although it has been argued that proper names can have bound readings (cf. following section), the point is that proper names generally resist to be bound in those quantified sentences in which ordinary count nouns can be bound. 3.1 Fara’s reply to the argument from binding Fara replied to this objection arguing that in fact this inability of bare proper names to be bound in quantified sentences is not a problem for The-predicativism and her account can actually predict this (Fara, 2015b, 362). The idea is that the contrast between (13.a) and (13.b) below is explained by the hypothesis that the noun in (13.a) is subject to nominal restriction (as has been argued by Stanley and Szabó (2000) and Stanley (2000, 2002, 2005)),25 while the proper name in (13.b) is not. She volunteers (13.c) as a parallel to (13.a) where the proper name would be subject to nominal restriction as well. (13) a. In every race, the colt won. b. In every race, John won. c. In every race, the John won. Nominal restriction is the mechanism by which each nominal expression carries a variable syntac- tically attached to it. This variable can be assigned in context a property or set as value that intersects the property or set of the noun, yielding this intersection as denotation. This variable can also be bound by a previous quantified noun phrase, thus explaining the co-varying reading 25 Nominal restriction is arguably a controversial theory. However my discussion of Fara’s proposal does not depend on specific problems nominal restriction theory might have. For the purpose of this discussion I would presume that the best formulation of this theory is in place. 10 that (13.a) allows (i.e., where different colts won in different races). Thus the analysis of (13.a) would be: In every race [x], the colt [in x] won Fara then explains the difference between (13.b) and (13.c) by the rules governing the article appearance she provides in her theory, summarized as follows (Fara, 2015b, 367): Where ‘∅the ’ When a name occurs in a definite description, the definite article must be unpro- nounced when the name is its structural sister, unless the definite article is stressed. Given this rule, if a proper name carried a covert nominal restriction syntactically attached, then it would not have the definite article as its structural sister, and therefore the article would be required to appear, just like with overt restrictions. According to her, this explains the co-varying reading (13.c) has. This allows her to conclude that since in (13.b) the article is unpronounced, there is no nominal restriction at play, and this explains why the name in (13.b) cannot be bound into by the previous quantified noun phrase (Fara, 2015b, 370). There are a number of worries about Fara’s argument. As discussed above, it is not clear that (13.c) is grammatical, that is, that “the John” can appear without overt restriction (or stress).26 If (13.c) is not grammatical, then Fara’s explanation fails: either proper names are not subject to nominal restriction just the same as count nouns in definite descriptions are, or this nominal restriction does not forces the appearance of the article (and so the Where ‘∅the ’ rule is not correct, at least with respect to covert restrictions). On the other hand, if (13.c) is grammatical, we should find similar cases to be grammatical too. But note that there are other cases where definite descriptions show quantificational variability but where parallel sentences with proper names sound as bad as (13.c) does. Consider the following: (14) a. Whenever executions occur in this country, the executioner is not proud of his job. b. (??) Whenever presentations occur in this country, the Adolf is not proud of his name. (15) a. Because they use alphabetic order, the anthropologist is always the first to be called. b. (??) Because they use alphabetic order, the Aaron is always the first to be called. The conjunction “whenever” and adverbs like “always” may be considered as a kind of quantifier ranging over events, times, or situations. These expressions would bind a variable in the noun of the definite description.27 If the same phenomenon was at play with the proper names in the b-sentences above – and Fara’s argument is right – we will expect the definite article showing up 26 Some people hear (13.c) as marked even in the right set up. 27 This is proposed by Stanley (2002, 23–24) and adopted by Fara (2006, 82) with some modifications. 11 with unmodified proper names to be perfectly grammatical. But I submit that these sentences are at best marked. A second problem with Fara’s explanation is that discussions on the possibility of binding into proper names do not cite examples like (13.c); rather, the discussion focuses on few examples where bare proper names apparently do have co-varying readings: (16) a. If a child is christened ‘Bambi’, and Disney Inc. hear about it, then they will sue Bambi’s parents (Geurts, 1997, 322). b. Every woman who has a husband called John and a lover called Gerontius takes only Gerontius to the Rare Names Convention (Elbourne, 2005, 181). c. If John insists on calling his next son Gerontius, then his wife will be annoyed and Gerontius will get made fun of because of his name (Elbourne, 2005, 182). It is significant that Fara herself proposes some of these examples as cases of co-varying uses of names, where the name can be interpreted as bound by an antecedent noun phrase (Fara, 2015a, 105). But now we have a dilemma: either (16.a)-(16.c) have co-varying readings as claimed or Fara is right. On the one hand, if (16.a)-(16.c) have co-varying readings as claimed, then Fara cannot be right, because if nominal restriction is what explains this co-variation, the Where ‘∅the ’ rule is false, given that (16.a)-(16.c) are not ungrammatical and the names in them appear bare.28 . On this horn of the dilemma, (13.b) should also have a co-varying reading – since the name would carry a nominal restriction, like any noun. However, this is the point of the initial objection: that (13.b) does not have a co-varying reading, even if (16.a)-(16.c) do. If, however, the Where ‘∅the ’ rule is right, then nominal restriction is not what explains the co-varying readings that (16.a)-(16.c) have; but then it is unlikely that nominal restriction explain anything regarding (13.b) and (13.c). On the other hand, if Fara is right then, contrary to the intuitions, (16.a)-(16.c) do not have co- varying readings. It seems clear that we cannot have (13.c) and (16.a)-(16.c) all being grammatical and also their alleged co-variation being explained by the same strategy. Therefore either Fara’s explanation fails, or her proposal has counter-intuitive results with respect to (16.a)-(16.c). The third problem I see with Fara’s argument is that nominal restriction as proposed by Stanley (in particular in Stanley (2002, 2005) is supposed to apply to common nouns more generally, not just when they appear to be bound by a quantifier.29 But, if count nouns are generally restricted, 28 Note that if anything is wrong with (16.a)-(16.c), they certainly do not improve by adding the article to the proper name (See Delgado (2022) for discussion of the grammaticality of the+N ) 29 Stanley (2002) argues for nominal restriction on nouns used with comparative adjectives, and on mass terms, for 12 proper names are also generally restricted, given that, according to the predicativist, they are just count nouns. But then Fara’s account makes the wrong predictions: proper names should never appear bare, since by Fara’s rule, when the name is restricted, it is not longer the structural sister of the definite article, and this forces the article to appear. But it is clearly not the case that proper names never appear bare. Further, it is not clear that Fara’s proposal can be restricted to apply to proper names just when they appear to be bound by a quantifier, or that it would not be ad hoc to do so.30 One reason it would seem ad hoc to reject nominal restriction as applying generally is that Fara has good reasons to adopt it: this is because proper names still need to be restricted to a single denotation when appearing in sentences like (13.b) above, since they are supposed to be predicates of multiple application.31 Thus, given this general need of a way of restricting proper names, so that they get a single denotation in these uses, it would seems unjustified to reject nominal restriction in general while adopting it for certain cases only. Finally, it also seems ad hoc to adopt nominal restriction only to resolve the contrast between (13.a) and (13.b), given that Fara also has good reasons to reject nominal restriction as applying generally. This is because nominal restriction is in conflict with her account on the rigidity of proper names. In order to account for the rigidity of proper names, given that referential uses of proper names are incomplete definite descriptions in her account, Fara claims that incomplete definite descriptions in general are rigid-in-a-context. She argues that when we use incomplete definite descriptions, like “the party” or “the book”, we use them to talk about particular things, without there being a particular way of completing the description which is part of its semantic value (Fara, 2015a, 98). Even though we can use such descriptions to talk about many different things, once its value is fixed in a context, the description is rigid. Fara refuses to take a stand on precisely how incomplete descriptions are rigid. She considers but rejects a number of ways in which an incomplete description can secure a single denotation in a context, including nominal restriction (Fara, 2015a, 99–100). She rejects nominal restriction in particular because the resulting complete description would have both wide and narrow scope readings with respect to a modal operator. Consider (17.b) as one particular way of completing the description “the party” in (17.a) that would have both wide and narrow scope readings: (17) a. Olga might have enjoyed the party. b. Olga might have enjoyed the party I went to last night (Fara, 2015a, 99). example. 30 Schoubye (2016b, 5–6) argues that the proposal cannot be restricted to those cases Fara intends, and that it is not clear that her generalization (the Where ‘∅the ’ rule) can be amended. 31 Thanks to Elia Zardini to bring this detail to my attention. 13 But Fara claims that (17.a) does not have a narrow scope reading. For its evaluation, the only relevant party is the party she is talking about, and not some other parties to which she went at other worlds: “In other words, [17.a] has only the truth conditions that it would have if the incomplete definite description were rigid” (Fara, 2015a, 100). Thus, if nominal restriction was adopted, incomplete definite descriptions cannot longer be considered as generally rigid, and her account of the rigidity of proper names would then fail.32 Therefore, given this reason to reject nominal restriction as applying generally, it seems ad hoc to adopt it only for the purpose of replying to the argument from biding. 4 THE EXPLETIVE DEFINITE ARTICLE Longobardi provides evidence that at least in Italian, the article with proper names must be un- derstood as an expletive. He begins by noting that when two singular NPs are coordinated in argument position, “a single singular determiner is sufficient to impose singular designation to the entire nominal expression, whereas the sum of two singular determiners automatically imposes plu- ral designation” (Longobardi, 1994, 621). He gives (18) as an example of this in Italian (Longobardi, 1994, 620), which I take is true in English as well, as in (19): (18) a. La mia segretaria e tua collaboratrice sta/*stanno uscendo. the my secretary and your collaborator is/are going out “My secretary and your collaborator is/are going out” b. La mia segretaria e la tua collaboratrice stanno/*sta uscendo. the my secretary and the your collaborator is/are going out “My secretary and your collaborator is/are going out” 33 (19) a. The secretary and collaborator is/*are going out. b. The secretary and the collaborator are/*is going out. With respect to proper names, Longobardi notes that the coordination of two singular NPs, the first 32 However, the claim that incomplete definite descriptions consistently take wide scope with respect to the modal operator has been contested in Delgado (2011, 58–62), and in Schoubye (2017, 752). Fara does not deny that some incomplete definite descriptions have no-rigid readings. But she hopes to put aside these as cases of role-type uses of descriptions, as argued by Rothschild (2007) (Fara, 2015a, 103–05). Schoubye has argued however that incomplete descriptions that would not be classified as role-type also have non-rigid readings (or exhibit narrow scope readings with respect to a modal quantifier (Schoubye, 2017, 752). 33 Note that while in the paraphrase in English the nominal phrase is ambiguous between denoting a single individual and denoting two, in Italian the presence or absence of the article with the second NP determines whether the designation is singular or plural. 14 headed by an articled proper name and the second by an articleless common noun, is impossible in Italian: (20) *La Maria e (mia) segretaria è arrivata in ritardo. the Maria and (my) secretary is arrived late “The Maria and (my) secretary has arrived late” suggesting that the proper name does not provide a range for the article to operate over (Longobardi, 1994, 651), unlike common nouns (i.e., (18.a)). This is explained by the hypothesis that the definite article is an expletive.34 Identical ungrammaticality results in Catalan, Portuguese and versions of Spanish:35 (21) a. *La Maria y secretaria ha llegado tarde. (Spanish) the Maria and secretary has arrived late “The Maria and secretary arrived late” b. *A Maria e secretária chegou tarde. (Portuguese) the Maria and secretaria arrived-SING late “The Maria and secretary arrived late” c. *La Maria i secretària ha arribat tard. (Catalan) the Maria and secretary has arrived late “The Maria and secretary arrived late” Incidentally, this phenomenon also suggests that proper names, when bare in argument position, do not carry an unpronounced determiner. A parallel of (19) with a proper name is sharply marked: (22) a. *Mary and collaborator is going out. Moreover, a single definite article cannot impose a singular reading on two coordinated NPs what- ever position the name occupies: (23) a. *The collaborator and Mary is going out. b. *The Mary and collaborator is going out. Constructions like these should be possible at least in contexts where classifying people according to their names is relevant. But (23.a)-(23.b) are sharply marked, suggesting that the name just does not provide predicative material to coordinate with the other NP, or for the article to operate 34 Longobardi also provides other arguments involving other languages like Catalan and German. See Longobardi (1994, Sections 7 and 8) 35 Data due to Teresa Marques (Portuguese), and Pilar Terrés (Catalan), both personal communication. 15 over. The name remains strongly referential. 5 GENERIC SENTENCES AND NOUNS VS NAMES A last piece of evidence that proper names behave differently from the definite descriptions to which they are supposed to be equivalent is that singular definites containing count nouns – that is, “the F ” when F is a count noun – can generate generic readings while bare singular proper names cannot. Generic sentences are classified into two main types: D-Generics and I-Generics.36 D- generic sentences, also called kind-referring generics, attributes some property to a kind or class as a whole rather than to the instances of that kind individually. I-generic sentences, sometimes called characterizing generics, intuitively express a general attribution of properties to many individuals.37 Singular definites containing count nouns can appear in both type of generic sentence. Consider the following examples: (24) a. The potato was unknown in Europe before 1400. (D-generics) b. The potato is a starchy, tuberous crop. (I-generics) (25) a. The tiger is the largest feline. (D-generics) b. The tiger is striped whereas the leopard is dotted. (I-generics) (26) The ballpoint pen was invented by L. Biro. (D-generics) (27) The pencil contains graphite. (I-generics) Sentences (24.a), (25.a) and (26) are examples of D-generics because the property attributed to the noun in the NP is not a property individually had by every potato, tiger, or ballpoint pen respectively, but by the kinds potato, tiger, ballpoint pen as a whole. Conversely, sentences (24.b), (25.b) and (27) are examples of I-generics because the property attributed to the noun in the NP is (generally) had individually by the members in the extension of the noun, that is, every potato is a starchy tuber, almost every tiger has stripes, and most pencils contain graphite. Singular definites containing count nouns can in principle generate generic sentences of either 36 See Krifka et al. (1995), for definitions. 37 Henceforth, “generic sentence” abbreviates “sentence that has a generic reading”. Of course these sentences may have other readings as well. 16 type.38 But bare singular proper names cannot: “Sarah is scary” cannot possibly be read as a generic sentence either attributing a property generally true of every Sarah, or attributing a property to the set of Sarahs as a whole. In other words, unlike count nouns, bare proper names cannot be understood as ordinary predicates, denoting an extension that intersects the extension of the predicate in the VP, as to generate I-generic sentences. Likewise, a bare proper name cannot be used to refer to the kind, or a class, consisting in the individuals in its extension, as to generate D-generic sentences. This suggests that bare singular proper names do not function as predicates, at least in these occasions. Contrast this with predicative uses of proper names that can generate generic sentences all right: (28) a. Sarahs are widespread. (D-generics) b. Sarahs from the north are scary. (I-generics) c. A Sarah is an individual with a short name. (I-generics) d. Every Sarah is scary. (I-generics) Sentence (28.a) makes reference to a certain kind, or set of individuals, and attributes a property to it as a whole (no individual is widespread), just like the sentence “Dogs are widespread”, containing an ordinary common noun does. Sentences (28.b) and (28.c) make a general attribution of a property (being scary, or, having a short name) to every (or most) members of a set, that is, the 38 However, not every definite (the+count noun) can generate generic sentences. It has been noted in the literature (see for example, Krifka et al. (1995); Vendler (1967)) that it is hard for a definite to generate a generic sentence if the noun does not denote a well–established kind or is too generic (where being “generic” here means being like a genus that subdivides into more specific kinds). For example, a noun like “vehicle” is too generic for the definite “the vehicle” to generate a generic sentence; however nouns denoting specific kinds of the genus vehicle) can, for example, “the motorbike”, “the cart”, “the airplane”, and so forth. But is it not clear that a proper name qua predicate is too generic for the purpose. It denotes a class of people identified by the name they have (i.e., people called “N ”), much like other count nouns that denote classes of people identified by other things, like profession (i.e., the mathematician, the teacher, the fireman); religion (i.e., the Christian, the Buddhist); political inclination (i.e., the feminist, the socialist, the conservative); or other social categories or roles (i.e., the patient, the student, the widow, the professional). Definites with such count nouns that denote classes of people can generate generic sentences all right: “The mathematician typically feels superior”; “The feminist believes in the equality of genders”; “The patient does not like to wait”. What a proper name denotes as a predicate (i.e., Martin denoting the set of Martins/people called “Martin”) seems to be neither more, nor less, general than what count nouns denoting classes of people denote (i.e., “widow” / “socialist” denoting the set of widows/socialists, respectively). So, I suspect that what prevents a bare name from generating a genetic sentence is not that its meaning is too generic, but that its meaning is too specific, indeed, referential on a single individual. 17 set of Sarahs. Note that, apart from ordinarily functioning like predicates, count nouns can also be used to denote the kind, or class, consisting of the individuals in their extension. This is a general property of count nouns. When proper names are functioning as predicates, they can also denote the kind or class consisting of the individuals that bear that name, as witness by the sentences in (28). But unlike definites with count nouns, bare singular proper names cannot be use to refer to the kind or class consisting in the individuals called that name. This suggests that bare singular proper names are simply not equivalent to a count noun with a covert definite article. To press the point further, I would like to show that in some of those languages that permit or require the use of the definite article with singular proper names, the resulting the+N construction cannot generate generic sentences either (e.g. (29.a) and (30.a)); whereas singular definites with count nouns can (e.g. (29.b) and (30.b)). In contrast, predicative uses of proper names can generate generic readings (e.g. (29.c)-(29.d) and (30.c)):39 Spanish (29) a. La Maria es astuta, mientras que la Marta no. (no generic) The Maria is clever, while that the Marta not “(the) Maria is clever while (the) Marta is not” b. El gato es astuto, mientras que el perro no. (I-generic) The cat is clever, while that the dog not “The dog is clever while the cat is not” c. Las Marias son astutas, mientras que las Martas no. (I-generic) The-PLU Marias are clever, while that the-PLU Martas not “Marias/The Marias as are clever while Martas/the Martas are not” d. Todas las Marias son astutas. (I-generic) All the-PLU Marias are clever “Every Maria is/ All the Marias are clever” Portuguese (30) a. A Anacleta detesta o seu nome. The Anacleta hates the her name 39 Data due to Teresa Marques for Portuguese, and Pilar Terrés for Catalan. Identical examples can be written in Italian (data due to Elia Zardini), all personal communication. 18 (no generic) “(the) Anacleta hates her name” b. O varredor de rua detesta o seu trabalho. (I-generic) The sweeper of street hates the his job “The street sweeper hates his job” c. As Anacletas detestam o seu nome. (I-generic) The-PLU Anacletas hate the their name “The Anacletas/Anacletas hate their name” Catalan (31) a. El Joan és arrogant. (no generic) The John is arrogant “(the) John is arrogant” b. L’enginyer és arrogant. (I-generic) The-engineer is arrogant “The engineer is arrogant” c. Els joans són arrogants. (I-generic) The-PLU Johns are arrogant “Johns are arrogant” 6 Conclusion The success of the predicate view – and its superior ability to provide a uniform account of proper names – hinges on whether it can adequately accommodate referential uses of proper names. This crucially involves bridging the apparent differences between predicates and proper names. One of these differences is that unlike common nouns, proper names seem to be referential and singular terms, that is, in their typical uses, they seem to simply stand for the objects they refer to, and to have an individual as semantic value, as oppose to designating a property or an extension. The unpronounced determiner thesis is supposed to resolve this difference and provide semantic uniformity between predicative and referential uses of proper names: names would be predicates in all their occurrences. The appearance of referentiality and singularity that seem to be the main characteristics of proper names would be explained by their putative underlying structure as complex demonstratives or definite descriptions. 19 But the unpronounced determiner thesis fails. I have argued that the significant semantic differ- ence between count nouns and proper names persists notwithstanding the predicativist’s treatment of referential proper names as definite descriptions or complex demonstratives. The evidence pre- sented above shows that proper names behave differently from the complex demonstratives or the definite descriptions that are thought to embed them. In addition, I showed that proper names behave differently from definite description even in some of the languages that use the definite article with proper names, and therefore this aspect of the cross-linguistic data does not support The-Predicativism, as claimed. As it stands, therefore, Predicativism cannot be the right account of the semantics of proper names. Acknowledgements This is a paper with a log history that originated from my Bphil dissertation in 2011 and I thank Ofra Magidor for her help at this stage. The material later grew into a paper and eventually made a chapter of my PhD dissertation (2018). I am indebted to Manuel García- Carpintero for his unfailing support throughout those years and for detailed comments on different versions of the manuscript. Special thanks to Elia Zardini, David Yates, John Horden, and three anonymous referees whose comments helped me to improve the paper. I would like to thank Sarah Sawyer whose spontaneous encouragement regarding my work on names boosted my confidence at a crucial time of working on the present form of the paper. Early versions of the paper have been presented in 2014 at the Graduate Reading Group Seminar (University of Barcelona); in 2015 at the Metaphysics, Epistemology, Logic and Language Seminar (University of Lisbon), and at the Veritas Research Seminar (Yonsei University, Seoul). I’d like to thank all these audiences for stimulating comments and discussion. I’m also very grateful to all the informants who shared their knowledge of their native languages with me. At different stages during the development of this paper I have benefitted from an AHRC Postgraduate Studentship Award; from a Doctoral Award from the programme of grants to new researchers (FI-DGR) of the Catalonian Government; and from a Junior Research Fellowship from the Portuguese national funding agency for science, research and technology (FCT). References Burge, T. (1973). Reference and proper names. The Journal of Philosophy, 70 (4), 425–439. Coromides, J. (1982). Diccionari etimològic i complementari de la llengua catalana (Vol. 3). Barcelona: Curial. Delgado, L. (2011). Between singularity and generality: The semantic life of proper names (Master’s thesis). Oxford University. 20 Delgado, L. (2019). Between singularity and generality: The semantic life of proper namesy and generality. Linguistics and Philosophy, 42 (4), 381–417. Delgado, L. (2022). Names vs nouns. Philosophical Studies, 179 (11), 3233–3258. Delgado, L. (2018). David, some davids, and all davids: Reference, category change, and bearerhood of real-life names (Doctoral dissertation). Universitat de Barcelona. Elbourne, P. D. (2005). Situations and individuals. Cambridge, MA; London, MIT Press. Elugardo, R. (2002). The predicate view of proper names. In G. Preyer & G. Peter (Eds.), Logical form and language (pp. 467–503). Oxford, Oxford University Press. Fara, D. G. (2006). Descriptions with adverbs of quantification. Philosophical Issues, 16 (1), 65–87. Fara, D. G. (2011). Names as predicates [Unpublished manuscript]. Fara, D. G. (2015a). Names are predicates. Philosophical Review, 124 (1), 59–117. Fara, D. G. (2015b). A problem for predicativism solved by predicativism. Analysis, 75 (3), 362–371. García-Carpintero, M. (2017). The mill-frege theory of proper names [(fzx010)]. Mind. Geurts, B. (1997). Good news about the description theory of names. Journal of Semantics, 14 (4), 319–348. Hawthorne, J., & Manley, D. (2012). The reference book. Oxford, Oxford University Press. Higginbotham, J. (1988). Context, models, and meanings: A note on the data of semantics. In R. Kempson (Ed.), Mental representations: The interface between language and reality (pp. 29– 48). Cambridge, Cambridge University Press. Hinzen, W. (2016). Linguistic evidence against predicativism. Philosophy Compass, 11, 591–608. Hornsby, J. (1976). Proper names: A defence of burge. Philosophical Studies, 30 (4), 227–234. Izumi, Y. (2012). The semantics of proper names and other bare nominals (Doctoral dissertation). University of Maryland, College Park. Jeshion, R. (2015a). Names not predicates. In A. Bianchi (Ed.), On reference (pp. 225–50). Oxford, Oxford University Press. Jeshion, R. (2015b). Referentialism and predicativism about proper names. Erkenntnis, 80 (S2), 363–404. Jeshion, R. (2017). ‘‘the” problem for the-predicativism. Philosophical Review, 126 (2), 219–240. Jeshion, R. (2018). Katherine and the katherine: On the syntactic distribution of names and count nouns. Theoria, 33 (333), 473–508. King, J. C. (2006). Singular terms, reference and methodology in semantics. Philosophical Issues, 16 (1), 141–161. Krifka, M., Pelletier, F., Carlson, G., ter Meulen, A., Link, G., & Chierchia, G. (1995). Genericity: An introduction. In G. Carlson & F. Pelletier (Eds.), The generic book (pp. 1–124). Chicago, University of Chicago Press. Larson, R., & Segal, G. (1995). Knowledge of meaning. Cambridge, MA, MIT Press. 21 Leckie, G. (2013). The double life of names. Philosophical Studies, 165 (3), 1139–1160. Lee, J. (2020a). Against predicativism about names. Philosophical Studies, 177 (1), 243–261. Lee, J. (2020b). On the asymmetry between names and count nouns: Syntactic arguments against predicativism. Linguistics and Philosophy, 43 (3), 277–301. Lekakou, M., & Szendroi, K. (2012). Polydefinites in greek: Ellipsis, close appositionand expletive determiners. Journal of Linguistics, 48, 107–49. Longobardi, G. (1994). Reference and proper names: A theory of n-movement in syntax and logical form. Linguistic Inquiry, 25 (4), 609–665. Martí, G. (2017). Names, predicates, and the object–property distinction. In M. Ponte & K. Korta (Eds.), Reference and representation in thought and language (pp. 14–19). Oxford, Oxford University Press. Matushansky, O. (2006). Why rose is the rose: On the use of definite articles in proper names. In O. Bonami & P. Cabredo Hofherr (Eds.), Empirical issues in syntax and semantics (pp. 285– 307). Matushansky, O. (2008). On the linguistic complexity of proper names. Linguistics and Philosophy, 21, 573–627. Matushansky, O. (2015). The other francis bacon: On non-bare proper names. Erkenntnis, 80 (S2), 335–362. Rami, D. (2014). On the unification argument for the predicate view on proper names. Synthese, 191, 841–862. Rami, D. (2015). The multiple uses of proper nouns. Erkenntnis, 80, 405–432. Rothschild, D. (2007). Presuppositions and scope. Journal of Philosophy, 104 (2), 71–106. Sawyer, S. (2010). The modified predicate theory of proper names. In S. Sawyer (Ed.), New waves in philosophy of language (pp. 206–226). Palgrave Macmillan. Schoubye, A. J. (2016a). The predicative predicament [Forthcoming]. Philosophy and Phenomeno- logical Research, 96 (3), 571–595. Schoubye, A. J. (2016b). A problem for predicativism not solved by predicativism. Semantics & Pragmatics, 9 (18), 1–11. Schoubye, A. J. (2017). Type-ambiguous names. Mind, 126, 715–767. Sloat, C. (1969). Proper nouns in english. Language, 45 (1), 26–30. Stanley, J. (2000). Context and logical form. Linguistics and Philosophy, 23, 391–434. Stanley, J. (2002). Nominal restriction. In G. Preyer & G. Peter (Eds.), Logical form and language (pp. 365–388). Oxford, Clarendon Press. Stanley, J. (2005). Semantics in context. In G. Preyer & G. Peter (Eds.), Contextualism in philos- ophy: Knowledge, meaning, and truth (pp. 221–255). Oxford, Oxford University Press. 22 Stanley, J., & Szabó, Z. (2000). On quantifier domain restriction. Mind and Language, 15 (2), 219– 361. Tayebi, S. (2018). In defense of the unification argument for predicativism. Linguistics and Philos- ophy, 41 (5), 557–576. Vendler, Z. (1967). Linguistics in philosophy. Ithaca, London, Cornell University Press. 23