A Semantic–Pragmatic Theory of Proper Names

WILLY VAN LANGENDONCK University of Leuven Fields of research: synchronic linguistics, Construction Grammar and dependency syntax, reference and semantics, onomastics, name theory, markedness, prototypicality, iconicity, word-order, grammatical categories such as definiteness, genericness, number, grammatical relations, prepositions. A SEMANTIC-PRAGMATIC THEORY OF PROPER NAMES Semantinė pragmatinė tikrinių žodžių teorija ANNOTATION This article aims at an extended semantic-pragmatic analysis of proper names in reaction to the reductionist approach of Richard Coates’ Pragmatic Theory of Properhood (ALL 2012). It is argued that names have at least two kinds of presupposition: grammatical features (especially definiteness) and a categorical (basic level) presupposition, which makes subcategorization of names possible. All this is connected to the claim that names do not only refer uniquely in language use (a frequent function, though not the only one) but also denote uniquely at the level of established linguistic convention. To render this acceptable, it is essential to distinguish between name and name lemma. This distinction is also relevant to the realm of possible connotations and the issue of name translation. Grammatical, philosophical, psycho- and neuro-linguistic evidence is adduced. ANOTACIJA Šio straipsnio tikslas – pateikti išsamią semantinę pragmatinę tikrinių žodžių analizę reaguojant į redukcionistinį požiūrį, kurį Richardas Coatesas išdėstė pragmatinėje tikrinių žodžių teorijoje (ALL 2012). Straipsnyje siekiama įrodyti, kad tikriniai žodžiai remiasi bent dviejų rūšių prielaidomis: gramatiniais požymiais (ypač apibrėžtumu) ir kategorine (pradinio lygmens) KEYWORDS: unique denotation, reference, name lemma, presupposition, definiteness, basic level meaning, established linguistic convention, grammar, translation. ESMINIAI ŽODŽIAI: unikali denotacija, referencija, tikrinio žodžio lema, prielaida, apibrėžtumas, pradinė reikšmė, nustatyta lingvistinė konvencija, gramatika, vertimas. Straipsniai / Articles 99 WILLY VAN LANGENDONCK prielaida, kuri sudaro sąlygas tikrinių žodžių subkategorizavimui. Visa tai yra susiję su teiginiu, kad tikriniai žodžiai ne tik nurodo unikalų objektą kalbos vartosenoje (dažna, tačiau ne vienintelė funkcija), tačiau ir pažymi jį nustatytos lingvistinės konvencijos lygiu. Kad tai taptų priimtina, svarbu išskirti tikrinį žodį ir tikrinio žodžio lemą. Šis išskyrimas taip pat susijęs su įmanomų konotacijų sritimi bei tikrinių žodžių vertimo klausimu. Straipsnyje pateikiami gramatiniai, filosofiniai, psicholingvistiniai ir neurolingvistiniai įrodymai. 0. INTRODUCTION In certain onomastic quarters, pragmatics – mostly in the narrow sense of language use – appears to be a fashionable topic for the understanding of proper names [henceforth – names]. Names are then no longer defined a priori, but interactively, from discourse, from conversation (De Stefani and Pepin 2006: 131). Coates (2006b: 38) states: “... in any case, what is a name for one language user may not be for another”. This seems to mean that the speaker decides on the characterization of ‘properhood’. However, while it is true that a pragmatic view of language and name research encompasses the study of context, discourse, talk in interaction etc., it does not necessarily imply that the language user should decide on the status of proper names as such. In its extreme form, this postmodern point of view would entail that the name theorist is threatened by unemployment. Apparently, our fear turns out to be premature since Coates (2000; 2005; 2006a; 2006b; 2009; 2012) unfolds a full-fledged pragmatic name theory, which is in fact as a priori as other theories: the Pragmatic Theory of ‘Properhood’ (PTP). (see already Van Langendonck 2007a: 65–71). The PTP interprets pragmatics as the study of language use and name theory as the study of name usage, both in a rather narrow sense. As a consequence, grammar and even semantics are hardly taken into account. Thus, the PTP seems to be eager to reduce the number of concepts needed to characterize names. However, more elaborate definitions of the notion of pragmatics have been provided in the literature, for which we need not use Occam’s razor. On the whole, it seems safer to simply speak of a linguistic theory, in which semantics, pragmatics and grammar are incorporated. In this case, pragmatics and semantics are fairly intertwined, for some linguists even indistinguishable. For instance, the distinction between an asserted and presupposed meaning is pragmatic as well as semantic. Take the much discussed sentence The King of France is bald, where the king’s existence is presupposed (albeit in a mental, fictional world), but the baldness is asserted. In a semantic approach, the presupposition of existence is preserved under negation, whilst the assertion is not: The King of France is not bald. There is a more pragmatic version of the notion of presupposition, as a 100 Acta Linguistica Lithuanica LXIX A Semantic-Pragmatic Theory of Proper Names rule compatible with the semantic one (Horn 1996).1 Last but not least, grammar needs no longer be the poor cousin of pragmatics, neither for names, nor for other word classes. In this paper, I will especially deal with semantic-pragmatic aspects of ‘proper-hood’, and endeavor to arrive at a satisfactory name theory that allows for an integrated theory of all three components: pragmatics, semantics, and grammar, as was first proposed for the whole of linguistics by the philosopher Morris (1938). To this end, I will discuss the PTP, point out its weaknesses, and propose solutions that are based on semantic and pragmatic considerations and corroborated by morphosyntactic criteria. The morphosyntax will be discussed extensively in Van Langendonck (forthcoming). A preliminary point to be made concerns the distinction between name and name lemma. In Van Langendonck (1999; 2007a: 7–8, 95–102), I introduce and elaborate on a distinction between name and proprial lemma.2 A name is an expression assigned uniquely to an entity in established onomastic and linguistic convention (former langue, see below), e.g. in sentential constructions or in a biographical, encyclopedic lexicon of family or other names (e.g. Le Petit Robert des noms propres, 2000). By contrast, a proprial lemma is merely an onomastic dictionary entry in which no specific entity is designated, yet. Mostly, the purpose of such dictionaries (like lists of first name lemmas) is to provide etymologies. These lemmas, “drawn from stock” (Coates 2006b: 39) are meant to give rise to different names, each assigned to a specific denotatum in an ad hoc unique way. In themselves, proprial lemmas neither denote nor refer. But they have this potential, which mostly also contains a prototypical category, such as ‘female’ in the case of Mary. Here, Coates’ words could apply: they “are categorized expectationally...” (Coates 2012: 132, fn. 16). But it should be clear that this does not apply to names in established onomastic convention. These denote and categorize uniquely, ready to be used in discourse to refer, or possibly to address someone. Note 1 Horn (1996: 305) characterizes presuppositional meaning in the following way: “To presuppose something is to take it for granted in a way that contrasts with asserting it”; or: “A proposition is presupposed if and only if it is (treated as) non-controversially true in every world within the working context set.” – Strawson (1950 [1971: 12, 25]) speaks of ‘implications’, and later on (1959), of ‘presuppositions’. 2 Coates (2006b: 28; 2012: 132, fn. 16) appears to confound name-form (i.e. phonological shape), proprial lemma and proper name. That is why for this author, denotation is not considered unique, and names are not categorized absolutely. – Brendler (2008) makes distinctions similar to those in Van Langendonck (2007a): proper name = nomen; proprial lemma = archinomen; name in language-use = nome. Hence, implicitly, he distinguishes between unique denotation and reference. – Even the language philosopher Strawson (1950 [1971: 6–8]) distinguishes between ‘an expression’ and ‘a use of an expression’, where an expression is akin to a lemma, since it is said not to refer yet. Straipsniai / Articles 101 WILLY VAN LANGENDONCK that not all lemmas are proprial. Some are appellative, e.g. gladiator, from which the film name Gladiator was derived. Therefore, I will speak of ‘name lemmas’ to include all possible lemmas. The PTP is built around two ideas (Coates 2012: 119; 120), which are implicitly present in language philosophical work like that of Mill (1843) and Strawson (1950 [1971] respectively, quoted in Coates (2012: 119–120):3 1) Proper names have no sense 2) The essence of being a name is to be found in reference, not in denotation Whilst the first claim is not uncommon, though too absolute, the second is rather controversial, and has far-reaching consequences. I will deal with both claims, and review four special corollaries of the PTP: – Proper names do not form a determinate set – Proper names do not fall into logically secure categories – Proper names cannot be translated – Theorizing the interface between proper and non-proper expressions has neurophysiological (and psycholinguistic) implications 1. PROPER NAMES HAVE NO SENSE To come to grips with the notion of ‘sense’, we can put specific questions asked by Stephen Ullmann and other scholars, such as: What does the word ‘table’ mean? Or What do you understand by ‘table’? If these are questions that make sense, then the word has ‘sense’, i.e. definitional lexical meaning. Indeed, we can give a definition of the word table as found in dictionaries. Usually, such words, in this case the common noun table, show polysemy, i.e. a coherent set of semantic features, of which often one is prototypical. For instance, Webster’s dictionary defines a table as a piece of furniture consisting of a smooth flat slab fixed on legs; this sense is akin to the sense of a tablet or a contents list, and so on. On the basis of these senses, we can find the referents. By contrast, in the case of names, the designation prevails over the meanings to be discussed hereafter. As Ullmann (1969: 33) contends: “One cannot possibly say that one understands 3 Note that neither Strawson (1950) nor Burge (1973) apply a possible distinction between reference and denotation, so it is obvious that they ‘choose’ the concrete notion of reference, and not the abstract concept of denotation. It is the more obvious to Burge, who sees names as predicative structures with a free variable, hence also accepting “modified proper names” like an Alfred (p. 429). But in such a framework, names cannot be said to be senseless. 102 Acta Linguistica Lithuanica LXIX A Semantic-Pragmatic Theory of Proper Names a proper name; one can only say that one knows whom it refers to, whose name it is.4 It does not make sense to ask: What does the word ‘London’ mean? or: What do you understand by ‘London’? This applies to pronouns as well: it does not make sense to ask: What do you understand by ‘he/she’, or ‘this’? So, neither names nor pronouns appear to have sense, i.e. definitional lexical meaning, let alone a polysemous structure.5 1.1. Presuppositional meanings in names Whereas asserted, definitional meaning (or sense) is excluded for names, they can display presuppositional meanings. These should be distinguished from connotations or associations. While connotations are contingent attributes of names or name lemmas (see below), presuppositional meanings are inherent properties of all names, and are mirrored in morphosyntactic features of them. I distinguish grammatical and categorical meaning. A. Grammatical meaning Both names and (at least personal and demonstrative) pronouns have so-called grammatical meanings (against Coates 2006b: 39; 2012: 124), such as definiteness, which indicates a presupposition of existence and uniqueness in the real or mental world. Examples: London, the Rhine; he/she, this vs. indefinite somebody, something. Definiteness can be called a weak form of deixis (see also La Palme-Reyes 1993 below for names). In my view, whilst both names and (personal and demonstrative) pronouns have unique reference in discourse, names have unique denotation as well, but pronouns only have non-unique denotation. As we will see, the PTP rejects unique denotation for any category. Note that the twofold similarity between names and pronouns concerning deixis (definiteness) and the absence of definitional sense, and in the PTP also of unique denotation, entails that this 4 Similar observations were by made by Nicolaisen (1995: 391); for German: Boesch (1957: 32) and Debus (1980: 194). But Brendler (2005: 108–109) rejects the relevance of such statements since he adheres to a kind of maximum meaningfulness theory for names. 5 Coates (2012: 121) mentions the term intension, but I think we can dispense with this term, at least as conceived of in the PTP. In this approach, this notion seems to resemble what many researchers, including myself, call categorical meaning. – In the same paragraph, Coates says: “Names cannot not have sense” by Lyons’ (1977) definition. This must be a typing error and should read: “Names cannot have sense”. Straipsniai / Articles 103 WILLY VAN LANGENDONCK theory cannot differentiate these two nominal classes (see below for reference and denotation).6 Other grammatical meanings common to names and (personal and demonstrative) pronouns are number (the Ruwenzori vs. the Andes; she vs. they), and gender (Latin Julius vs. Julia; English he vs. she ). Furthermore, in contrast to pronouns, names have more presuppositional meaning than just grammatical features. B. Categorical (basic level) meaning Names, but not name lemmas (compare Coates 2012: 125) have an inherent categorical or, more precisely, a basic level presupposition. Geach (1957, section 16) and Searle (1958) claim this categorical knowledge to be necessary for all uses of a name in order to preserve the referent’s identity. In a long footnote, even the ‘tag’ theorist Kripke (1972, comments, p. 351–352, fn. 58) brings up categorical meaning in names. Strawson (1959: 171, 198) sees ‘sortal universals’ in a sentence like Fido is a dog, an animal, a terrier, and speaks of ‘a presupposition of empirical fact’. Some linguists argue in the same vein: Van Langendonck (1968), Kalverkämper (1978: 89, confounding name and name lemma), Kleiber (1996: 582; 2004, section 4.4), Kuryłowicz (1980), Cisl aru (2006: 2012), Leroy (2012). Examples are: Paris is a city, and The Thames is a river. It is true that one can say: The (river) Thames is not a river, but this implies a contradiction between the presupposition that the Thames is a river and the assertion that it is not. Such categorical presuppositions make the subcategorization of names (not of name lemmas) possible (compare Brendler 2004; Debus 2012, and see below). Psycholinguists as well see a categorical, and more precisely, a basic level meaning, in names. La Palme Reyes et al. (1993: 445), establish the formula: [Freddie: dog] = [this: dog], which is to be read as: “Freddie in the kind DOG” is “this in the kind DOG.” Hence, there is a deictic component in names (this), just as in pronouns, but there is also a categorical meaning (dog), which is otherwise a common noun. This situates names in between pronouns and common nouns (see also Molino 1982: 19, and for further evidence, Van Langendonck 2007a: 169–171). 6 Coates (2006b: 41) claims that pronouns do have sense. For instance, in this/these vs. that/those, the notion of proximity indicates the difference. First, this is not (defining) sense, but a grammatical feature. I cannot reasonably ask: “What do I understand by this?” Second, proximity is almost only relevant where this and that are opposed, e.g. this, not that! Compare the phrase That George Bush in sentence (4), where that is used emotively. Third, in other pronouns, proximity does not play a part. 104 Acta Linguistica Lithuanica LXIX A Semantic-Pragmatic Theory of Proper Names Neurolinguistic evidence is provided by Bayer (1991). A German aphasic with deep dyslexia could not read names aloud, but she was able to specify elementary features like ‘masculine’, ‘feminine’, ‘city’, or ‘country’ belonging to a name. In fact, these are basic level presuppositions. Bayer distinguishes them (with a different Dutch term) from associations (connotations, see below). He argues that associations belong to the ‘episodic memory’, whereas the categorical meaning belongs to the ‘semantic memory’ (see Van Langendonck 2007a: 110–113). It will be clear that a strictly Millian approach (Coates 2009) cannot account for the conclusions drawn from these neurolinguistic phenomena. For grammatical evidence concerning categorical meaning, see section 3, b). That may suffice for presuppositional meanings. A different kind of onymic meanings concerns connotative meanings. 1.2. Connotative or pragmatic meanings in names or name lemmas Van Langendonck (2007a: 81ff.) distinguishes between two kinds of connotative meanings: some of them pertain to the referent and others to the name lemma or the etymology. Connotative or associative meanings pertaining to the referent of names are contingent (ancillary, cf. Coates 2012: 133–138). What meanings, if any, are activated, depends on the context and real world knowledge. Other associative meanings pertain to the name lemma on a synchronic basis or on the basis of the etymology as far as this is still transparent. Both kinds of connotative meaning are pragmatic par excellence. A. Pragmatic meaning pertaining to the name lemma or the etymology There are names whose lemma meaning is synchronically transparent, as in bynames, in literature, or in magical names. This transparency could have gradually faded away so we have to speak of etymological, historical meaning. - Neutral connotations can be synchronically present in name lemmas. For instance, the name the Rocky Mountains still says what it is about. The meaning ‘rocky mountains’ can be activated in a context or not. - Emotive meaning appears as a feature of certain synchronically transparent name lemmas, e.g. of diminutive (e.g. Dutch Jan-tje, Marie-ke) and augmentative Straipsniai / Articles 105 WILLY VAN LANGENDONCK forms (e.g. Flemish/Dutch dialectal Bert-en, Miel-e) of given names, expressive alternative name lemmas, especially bynames or nicknames, such as Flemish/Dutch den Dikke ‘the fat one’, de Witte ‘the white one’, den IJzeren ‘the iron one’, Belgenland instead of België ‘Belgium’, Absurdistan as a nickname for Afghanistan (see Van Langendonck 2007a: 22, 195–197; compare Palmer 1976). I note here that people may lay emotive connotations in the name-form of name lemmas, such as first name lemmas. For instance, some name-givers like Italian names for girls, e.g. Gigliola; others dislike name-forms such as Detlev, etc. As it is well-known, nowadays, Western first names should sound nice (see Van Langendonck 2010b). – Fiction, literature and names. Charactonyms Connotations connected with the name lemma are used a lot in fiction, e.g. Snow White. We should especially think of charactonyms in literature, like Doll Tearsheet, Becky Sharp, Titus Groan, etc. (Coates 2012: 136). – Names and magic Old Germanic personal names like Bern-hard ‘strong as a bear’ show a magical wishing character. Here it is particularly clear that the meaning belongs to the lemma, since numerous persons could have a name with the same name lemma. Also, we know that these meanings have gradually faded away from about the year 1000. Nowadays, only the etymologist knows what the first name Bernard meant a thousand years ago. There are other cases of etymological meaning. – Etymological meaning There are connotations pertaining to an etymologically transparent word, e.g. the family name Baker may remind us of a baker but not necessarily. Surely, etymological meanings may be misleading: Longbridge is no longer a bridge but a place (Coates 2012: 135). Peak’s Tunnel is no longer a tunnel,but a bridge (see below). B. Pragmatic meaning pertaining to the denotatum or the referent There are connotations pertaining to the denotatum or the referent. We observe a continuum from encyclopedic attributes to less objective, subjective connotations. ‘Facets’ of meaning belong to the set of objective connotations. – We observe a continuum from encyclopedic connotations attached to the denotatum, as in, e.g. Napoleon lost the battle of Waterloo, to truly subjective associations, which will usually pertain to a referent in discourse, e.g. Ann does not seem to like Obama. The neuropsychologist Semenza (2009) emphasizes the difference between the sets of coherent meanings in common nouns and the possible sets of incoherent connotations in names. For example, there is no semantic relation between 106 Acta Linguistica Lithuanica LXIX A Semantic-Pragmatic Theory of Proper Names the statements: Obama is president of the United States, Obama has a wife and children, Obama does not eat hamburgers, etc. Polysemy as in appellatives is lacking here. – Finally, there are these ‘facets’ of meaning, as Cisl aru (2006; 2012) calls them, using French, Romanian and Russian material. Although the basic level meaning of city names is ‘city’, and that of country names is ‘country’, these geographical names often adopt additional meanings induced by metonymy. English examples: (1) Paris elected a new mayor < The citizens of Paris elected a new mayor (2) America decided to declare war on terror < The government of The United States of America decided to declare war on terror Personal names, especially of artists, can stand for the work the artists produced: (3) Rodin se trouve dans la troisième salle du musée (Lemghari 2011) ‘(The work of) Rodin is to be found in the third room of the museum’ When one takes all these ‘meanings’ (presuppositions and connotations) together, and puts them on a par, it is understandable that some people regard names as the category with the most meaning of all word classes. The maximum meaningfulness theory (Van Langendonck 2007a: 39ff., 51, 58; Coates 2012: 137–138) is found already in Jespersen (1924: 66) and more recently in Cognitive Linguistics (in the widest sense). Langacker (1991), and in quite a different cognitive framework, Hansack (2004) and Brendler (2008) are modern proponents of this view. However, it will be clear from the foregoing argumentation that it is highly implausible to put all meanings, in their different function of sense, presuppositions, or connotations on the same level. We can conclude that in the first great debate on the philosophy of names between John Stuart Mill and Bertrand Russell (in the wake of Gottlob Frege), they both are wrong: Mill’s (1843) and Coates’ (2012) ‘tag’ theories rightly reject (definitional) sense but ignore presuppositions (unknown in the 19th c.), especially the inherent categorical presupposition of names (not of name lemmas), whereas Russell’s (1918) description theory seems to allow names (‘shorthand descriptions’) to have definitional sense. Although names can have meanings pertaining to the denotatum or the referent, or to the name lemma and the etymology, all these belong to the pragmatic use of names in discourse. Although I deny definitional sense to names, the above grammatical and categorical presuppositions, and even the connotative meanings, distinguish names from nonsense-words. Rejecting any sense or presuppositional meaning will render it hard to semantically distinguish names from nonsense-words. Coates’ way out is to discern sense from meaning in a rather broad sense. Thus, names have meaning in that “they ‘mean’ individuals when they are used” (Coates 2006b: 41–42). In Coates (2012: 133) we read that “the meaning of any name is established through Str aipsniai / Articles 107 WILLY VAN LANGENDONCK reference”, and further that the core meaning of a name is its denotation. But this amounts to mixing up meaning and denotation or reference (see Strawson 1971: 10 for this misunderstanding). Anyway, such statements are rather confusing. In contrast to English, other Western European languages have only one word to translate sense and meaning. Even English has only one verb for both: to mean (since to sense means something else). Hence, I do not see the relevance of introducing the notion of ‘meaning’ for the purpose intended by the PTP. Finally, the above semantic considerations confirm the Saussurean thesis (in accordance with modern Cognitive Linguistics, see Langacker 1987; 1991) that all linguistic elements have both form and meaning against Coates (2006b: 41), who rejects this statement made in Van Langendonck (2007a: 69). As I remarked there, Coates (2006a: fn. 16) makes a concession in admitting that names are not ‘meaningless’. “Some of these ‘meanings’ would best be called associations.” In this way, the question boils down to the nature of the relationship between ‘sense’ and ‘meaning’. In my view, there is no need to posit several radically different kinds of semantic categories (Van Langendonck 2007a: 69; also quoted in Coates 2006b: 41). However, I added there that all depends on the function of the meanings, which is related to the constructions in which they appear. For instance, the grammatical meaning ‘proximity’, which in certain contexts is relevant in the pairs this/that and these/those (but not in other pronouns), cannot be called ‘definitional sense’ (see note 6). Take another example: the concept ‘female’ is represented lexically in the noun woman, but grammatically in the pronoun she and the first name Mary (if e.g. denoting the mother of Jesus). In the last two instances we use the grammatical term ‘feminine’, indeed (see also Van Langendonck 2007a: 85). 2. DENOTATION AND DENOTATA, REFERENCE AND REFERENTS If we keep with Coates (2012) to Lyons (1977), we can hold: “Reference is the act of picking out an individual referent in a context of utterance”. “Denotation is the range of potential referents of a word or other expression (its extension), that is, it is an abstraction from reference...”⁷ The PTP holds that the essence of names 7 In Van Langendonck (2007a) the term ‘referent’ is used in the sense of ‘denotatum’, as has been a fairly common practice. However, in this paper (and theoretically everywhere), it is imperative to sharply distinguish between ‘denotation’ in established linguistic convention (see below) and ‘reference’ in language use. – Note, by the way, that in the first place, denotation and reference are considered here as mental. Whether the mental denotata correspond to real world entities is a philosophical issue. Snow White is a name, just like Obama. 108 Acta Linguistica Lithuanica LXIX A Semantic-Pragmatic Theory of Proper Names is to refer uniquely in context and apparently to denote as well, though not uniquely (Coates 2012: 121). This thesis poses a bigger problem than the view that names have no sense. Unique reference (that is, in language use) is also provided by definite descriptions (e.g. The Dutch queen abdicated) and definite pronouns (e.g. She abdicated). In this respect, the PTP can distinguish names neither from definite pronouns nor from definite appellative noun phrases. Even denotation does not distinguish pronouns from names or appellatives, since all three nominal categories have non-unique denotation in the PTP’s framework. For example, the pronoun I denotes the class of all speakers, and refers to a particular speaker in an indefinite number of instances. There are more problems with the notion of ‘reference’ as used in the PTP. First, as Strawson (1971: 25) remarks, there are other than definite referential uses, e.g. indefinite referring. In the case of names whose “essential duty is to refer uniquely to individuals” (Coates 2012: 30) we should at least speak of ‘definite reference’. Another problem is this: if the definition of names crucially depends on their referring to an entity in language use only, we could wonder whether this is an adequate characterization of properhood. In Table 1, Coates (2006b: 29) displays the “trinoda-necessitas” of name-duties: reference, vocation and nomination. But if ‘properhood’ is equated with (senseless) referring in language use (Coates 2006b: 30, 42; 2012), we are confronted with the important question of the relationship between reference and denotation on the one hand, and vocation and nomination (naming) on the other. A real problem is that there is no ‘referring’ in the last two cases: strictly speaking, the speaker does not ‘refer’ with the name in addressing somebody, at least not in a proper characterization of ‘uniquely referring use’, which Strawson (1950 [1971: 1]) equates with ‘mentioning, making a statement about, talking about’. In the vocative sentence John, come here!, John is ‘addressed’ in the discourse, not ‘referred to’, i.e. not talked about. Therefore, although ‘referring’ in discourse (senseless or not) is the main duty of names, it cannot define names as such. It would be similar to, for instance, equating names with personal names, which is in fact sometimes done because men and women are the most frequently designated entities. Nomination poses another problem. In the sentence I christen you John, a denotatum is created. Again, there is no ‘referring’ here. Nomination leads to denotation, which is ad hoc and unique for names. In addition, so-called ‘ostensive definition’ (Algeo 1973; Van Langendonck 2007a: 34, 37), such as in This is Jennifer, is not a simple referring act, either. Thus, one must realize that reference in discourse is just one of possible uses, albeit the most frequent. Hence, except for its frequency, it should be considered on a par with vocation, nomination and ostension. By contrast, unique denotation can be said to be the basis for these four duties of names: reference, vocation, nomination, and Straipsniai / Articles 109 W I L L Y V A N L A N G E N D O N C K ostension. In other words, denotation is an abstraction from vocation and ostension as well as from reference, and is often created by nomination (bestowal). Finally, it should be pointed out that these four functions also occur with common nouns, and do not differentiate names from them. For the referring act, see above. Vocation occurs with appellatives and pronouns as well as with names, e.g. Kevin/You/Boy, come here! As for appellatives, nomination is found in the astronomer’s statement Let’s call this pulsating star a ‘pulsar’; ostension in This is a pulsar. Similar to ostension are ‘identification statements’, like That is the man who swam the channel twice on one day, which is different from an ordinary sentence such as That man swam the channel twice in one day (Strawson 1971: 25). I would now like to go into the relationship between reference, denotation and unique denotation. Since properhood is defined as mere ‘senseless referring’, it is strange that the PTP does accept denotation for names at all. Coates (2012: 121) considers the denotation of a name “epiphenomenal upon repeated successful reference achieved through uttering it”. On the other hand, Coates (2012: 120–121; 137) finds the view that properhood is senseless denoting “incoherent and demonstrably false”, an unfortunate statement since what he really means is: “Names denote individuals, but do not denote them uniquely” (Coates 2006b: 37, 42). But, remarkably, unique denotation is accepted for expressions like the Milky Way, and even the Sun (ibid., p. 32), although there are lots of suns in the universe. So, why reject the intuitive and widely accepted uniqueness for names? Strangely enough, Coates (2006b: 29) states: “in the case of human bearers the name acquired becomes a focus for unique individual identity”, a statement that comes close to claiming unique denotation at least for personal names. In the same vein, the question how a name identifies a referent in a context of use is answered by Coates (2012: 136) as follows: “Through shared knowledge of an arbitrary but stable and rigid association of a linguistic form with an individual”. In fact, this is what unique denotation is about. The notion of ‘referent’ is in itself insufficient. In addition, Coates (2006b: 41) argues that as ‘deictic expressions’, pronouns should not be confused with names, “whose reference-potential may be multiple, but in each instance fixed (or rigid, if one prefers Kripke’s term” (cf. Kripke 1972 = 1980). However, the same can be said of definite pronouns, whose reference-potential may be multiple, but in each instance fixed or rigid. In fact, both Kripke and Strawson (1950 [1971: 23]) fail to distinguish names from pronouns, obviously because like most language philosophers, they fail to distinguish between (unique) denotation and reference in the same way as between ‘established linguistic convention’ and ‘language use’ (former langue and parole; see below). That is why Kripke (1972: 345, fn. 16) can easily state: “Demonstratives can be used as rigid designators, and free variables can be used as rigid designators of unspecified objects.” This means that rigid 110 A c t a L i n g u i s t i c a L i t h u a n i c a L X I X A Semantic-Pragmatic Theory of Proper Names designation is not even ‘senseless denoting’, but just (unique) ‘senseless referring’ in language use. Nonetheless, his concept of ‘rigid designator’ seems to be meant precisely to indicate unique denotation in names in the first place. In Kripke’s view, a proper name is a rigid designator in that “in any possible world it designates the same object” (Kripke 1972: 269). For instance, with a name like Nixon we refer rigidly, indicating the same person in any context (or ‘world’ in the philosophical terminology) so we can hold: “no one other than Nixon might have been Nixon” (p. 270), but Nixon might, for instance, not have been the president. It is a necessary truth that Nixon is Nixon, but not that he is or was president of the USA, not even that he was called Nixon. Uniqueness is preserved by assuming that in any context, a causal chain of reference leads to the same entity. Donnellan (1974: 3) argues in favor of what he prefers to call the ‘historical explanation theory’ because causality is not really involved. Coates (2012: 133) comes close to the historical explanation theory where he regards the denotation of a name as the “permanent and rigid, bond which has been established extensionally with its historically established referent(s)”.8 Could the same be said of common nouns or pronouns? Apparently not. It is true that Nixon remains Nixon, but the pronoun he or the definite description the president need not indicate the same man in different situations and discourses. It will be objected that there may be different Nixons. However, this can be explained by accidental homophony. Quine (1960: 130) and Searle (1971: 139) speak of homonymous proper names, although the term ‘homonymy’ is normally used in the case of appellative homophony (compare Strawson 1974: 60).9 Anyway, there is no polysemy here. If we take this into account, it seems obvious that a name has a unique denotation.10 However, this is made possible only by invoking the contrast between the name attributed to an entity and the proprial lemma, or more generally, the name lemma. Since Coates (2012: 122) fails to see the difference between the two concepts, name should be read as name lemma where he writes: “Decontextualized, the name Howard Archer denotes any individual which bears that name”. I would formulate it as follows: the potential of the compound proprial lemma Howard Archer includes the different unique denotata to which this lemma is applied ad hoc. No man so called need 8 McKinsey (2010) argues (against Kripke and others) for an individualistic approach to name reference. 9 Even in 1950, Strawson (1971: 23) sees a name as “governed by ad hoc conventions for each particular set of applications of the word to a given person”, although on the same page, he speaks of “the lingering superstition that a name is logically tied to a single individual”. 10 Brendler (2008) accounts for the difference between common and proper nouns by means of set-theory. Whilst proper nouns (names) indicate a one-element class in the brain, common nouns indicate a more-than-one element class (but see Van Langendonck 2006 [2011]; 2010a). Straipsniai / Articles 111 WILLY VAN LANGENDONCK have anything to do with his namesakes. Another argument for unique denotation comes from the possibilities of translation. Suppose several men are called De Cremer (from medieval Flemish/Dutch cremer ‘merchant’). In one instance it is the family name of Mercator. Indeed, only this famous geographer is called by this name. Other people called De Cremer will not be named this way. Let us have a look at place names. There are at least two cities called Bergen, one in Belgium, the other in Norway. But only the Belgian city name Bergen can be translated as Mons (the French endonym). In Flanders, there are three towns with the name Mechelen. Although all three names have the same etymology, the name of the biggest town is translated in French and English as Malines, the second, Mechelen(-Bovelingen) is Marlinne in French, the third, Mechelen(-aan-de-Maas), is not translated.11 No polysemy is present as in appellatives. Any person addressed with the English appellative word baker can be addressed as Becker in German and boulanger in French, since all of them are supposed to bake bread. The same applies to definite pronouns: any man referred to as he in English, can be referred to as er in German, and il or lui in French. Hence, it is not a question of lexical sense or not, but of unique denotation or not. Therefore, both the sun and the pope are neither names nor “single denotata”, as Coates (2012: 123) contends. There are numerous suns, and there have been numerous popes. In each utterance, their sense is basically the same. The words sun and pope have a clear lexical meaning, say, ‘star’ and ‘pontifex maximus’, respectively. Recently, the pope (Benedictus) abdicated, so a new pope, a new pontifex maximus, was elected. That names are given in a unique ad hoc way is not only intuitively and logically clear, but also indispensable to distinguish names from pronouns and definite descriptions, as well as to preserve identity, as a number of philosophers argue. A number of logicians speak of ‘individual constants’. In the same vein, Anderson (2007: 291) argues: “Pronouns... differ from... names primarily in lacking fixed reference.” Anderson’s term ‘fixed reference’ for names is obviously to be interpreted as ‘unique denotation’. Therefore, Strawson’s (1950 [1971: 23]) statement “that the word ‘he’ may be used on different occasions to refer to different people or different animals: so may the word ‘John’ and the phrase ‘the cat’.” is misleading. It is true only on the superficial level of language use, and ignores the difference between name and lemma, although his own notion of ‘expression’ is well-nigh identical with the notion of lemma (see note 2). 11 The extensions to Mechelen, i.e. aan-de-Maas and Bovelingen respectively, are meant to disambiguate these homophonous names officially, but are not used by the respective inhabitants (see Goossens 1986). Note also that these extensions are ad hoc and fixed, i.e. they cannot be replaced by others, as would be possible with appellatives, even for the same referent, compare: the bakery at the corner may or may not be the bakery at the crossroads, etc. 112 Acta Linguistica Lithuanica LXIX A Semantic-Pragmatic Theory of Proper Names The great frequency of homonymous personal names, like John and Brown in English, is not a universal phenomenon. Proprial personal name lemmas, which produce numerous first and family names, constitute an important subclass of name lemmas in Indo-European anthroponymic onomasticons ever since Christianization (cf. Van Langendonck 2007a).12 Let us elaborate on the relationship between name and lemma. Take the proprial lemma Mary in the English onomasticon. We can say that it usually – not always – gives rise to the denotation of particular women. The unique denotata issuing from this lemma include, among others, Mary, the mother of Jesus, or Mary Stuart, and so on... These denotata can then be used as referents in discourse. Consider the derivation: proprial lemma Mary → names with unique denotatum (e.g., Mary, Jesus’ mother, basic level: woman) → discourse referents In fact, Coates’ adoption of the useful notion of bestowal (Coates 2006b: 29, 40) implies that often a name lemma is explicitly assigned ad hoc to one entity at a time, which is then denoted uniquely. Although name lemmas are often proprial (e.g. John, Mary), they are not seldom non-proprial (e.g. the film Gladiator), or constitute a complex lemma (e.g. the novel The Old Man and the Sea). Thus, unlike Coates (2012: 121) claims, name-bestowal is not “established through acts of reference...”, but rather, as Kripke (1972) contends, by a “baptismal” act, or more generally, a naming act. Surely, names do not always come into being by bestowal. Evolution may be more frequent (see Coates 2005: 3; 2012: 124). Many names gradually develop from definite descriptions or other constructions by repeated unique use in context, in this case “through acts of reference...” Only in this case we do have gradual onymization. The order in the above schema should then be reversed: discourse referents the rocky mountains → name with unique denotatum (the Rocky Mountains, basic level: mountain range) → name lemma the Rocky Mountains This does not take away the fact that a name like the Rocky Mountains can still be considered to have a unique denotatum. The importance of the notion of unique denotation for names is implicitly indicated by psychologists: Valentine, Brennen and Brédart (1996: 179) state: “... proper name phrase nodes [like Jack Nicholson] do not connect directly to the identity-specific semantic system, but only via a person identity node (PIN). The PIN is a pre-lexical node which represents an individual”, i.e. “a particular token or address in the conceptual memory system” (Hollis and Valentine 2001: 99). Thus, a PIN can be equated with a unique denotatum, which can lead to a refer- 12 In other languages, we often find appellative-like lemmas, e.g. see Van de Velde (2006; 2008; 2009). Stripsniai / Articles 113 WILLY VAN LANGENDONCK ent in discourse. As a rule, without a denotatum you cannot refer (but see below). But surely, new names can easily develop since names constitute an open class (against Anderson 2004: 457; compare Coates 2006b: 30). If they stay, they enter the onomasticon as denotata, and discourse as referents or as forms of address. 3. NAMES, ESTABLISHED LINGUISTIC CONVENTION, AND GRAMMAR The view that names have a unique denotation and can refer in discourse is in accordance with Langacker’s (1987) notion of established linguistic convention (formerly langue or competence), forming a continuum with language use (formerly parole or performance). Just as denotation is an abstraction from reference etc., established linguistic convention is an abstraction from language use. Only in established linguistic convention does it make sense to speak of grammar or morphosyntax, which most linguists call the heart of linguistic system. In this way, names can be given a genuine place in grammar as a structural category, like all other word classes. Let’s discuss these distinctions. A) Established linguistic convention Langacker’s concept of established linguistic convention is flexible and useful, also for names. As a part of it, we discern established onomastic convention. Names enter established onomastic convention via bestowal or via gradual ony mization. This allows us to make three observations. 1) Although the PTP admits that acts of reference fix the denotation of proper nouns (Coates 2006b: 39; 2012: 121), it is not clear where this denotation finds its place if names are defined in terms of reference in language use. Clearly, unique denotation pertains to established linguistic convention. The rejection of this uniqueness in the PTP again prohibits a distinction between names and pronouns since in this framework both essentially refer in language use, even if they appear to denote as well, but not uniquely in the PTP. The lack of the notion of established linguistic convention led philosophers like Bertrand Russell ([1918] 1964: 201; 1919: 179) to claiming that genuine names were ‘logically proper names’, i.e. referring words like this or that (compare Kripke 1972: 345, fn. 16; see above). Russell called ordinary names ‘shorthand descriptions’. Surely, referring words like this or that refer uniquely in a certain context, but the reference will differ in another context. However, taking Kripke’s term ‘rigid designator’ seriously, names denote uniquely in any context. 114 Acta Linguistica Lithuanica LXIX A Semantic-Pragmatic Theory of Proper Names 2) There is a continuum from established onomastic convention to the use of names in speech and writing. We may use a name just once, and then forget it, so it does not enter established onomastic convention. For instance, referring to an unpopular guest, we could say: Hitler is coming tonight. In this example, Hitler is a new referent in discourse only, not a denotatum in established onomastic convention, yet. That is one extreme. The other extreme for names is that many have been functioning in society for centuries, for instance family names, or city, country, river names, and the like. 3) The notion of established onomastic convention allows us to recognize that there are well established names known and used in only small communities, such as nicknames in a family, e.g. Dutch Ons Pop ‘Our Doll’, called that by her father. This is an established name in this minimal Flemish community (Van Langen-donck 2007a: 286). The other extreme is that there exist names known worldwide, such as Africa or Mandela. B) Grammar and names Within the framework of the PTP, in which the essence of names is to refer in discourse, a grammar of names can hardly find its place. Coates claims: “properhood is best understood as a pragmatic rather than a grammatical or structural notion” (Coates 2000: 1166; 2005: 3; 2006a: 369). Although Coates (2006b: 39) accepts that there exist ‘proper nouns’ (like Mary, John), this notion is called “epiphenomenal on the notion of onymic reference”, where onymic is not further defined. In accordance with Webster’s definition of ‘epiphenomenon’, that would mean that a proper noun – or a proper noun phrase (Coates 2006a) – is a secondary phenomenon accompanying reference using names and caused by this reference. But since Coates (2006a) contrasts onymic with semantic reference, pronouns are onymic as well, as they lack definitional sense, just like names. In other words, the notion ‘onymic’ and those of ‘name’ and ‘proper noun (phrase)’ remain in need of a genuine definition (see also Anderson 2007: 120, quoted in Coates 2006b: 39). Proper names are words just as others (likewise Anderson 2003: 360; see also Nübling, Fahlbusch and Heuser 2012). Deviating from the general notion of word class puts the burden of proof on the one who deviates. On the other hand, if we assume the PTP is right in essence, why would a similar analysis not apply to all lexemes in language: nouns, verbs, etc.? If so, we are saying that established linguistic convention is an epiphenomenon of language use, or in better conventional terms: established linguistic convention is an abstraction from language use, and this can hardly be called a new insight. What have we gained in the PTP’s reasoning? What it pro- Straipsniai / Articles 115 WILLY VAN LANGENDONCK poses boils down to emphasizing the role of reference and language use, and minimizing the role of denotation and established linguistic convention. However, this inevitably leads to neglecting the role of grammar in defining names and other linguistic categories. For the PTP, names are not even a structural category (see above), although it is notoriously difficult to distinguish them from other nouns: common nouns and pronouns. Instead, we should continue the view that grammar describes established linguistic convention, defines names as having a unique denotatum besides an underlying name lemma, and as a subclass of nouns that take determiners in English, while pronouns do not, e.g. that modifying George Bush in (4), and Britain’s modifying Jeremy Irons in (5): (4)　That George Bush is a nice guy. (Vandelanotte and Willemse 2002: 22) (5)　Britain’s Jeremy Irons was present at the premiere in New York. (Vandelanotte and Willemse 2002: 25) Further, we can differentiate names from common nouns. Close appositional patterns of the form [(definite article) + noun + (definite article) + noun], e.g. Fido the dog, are relevant to the definition of names in English and also to the categorical meaning of the name (Van Langendonck 2007a: 4, 131).13 The unit that does not characterize but identify is a name, i.e. Fido. The appellative dog indicates the categorical presupposition. Grammatical, i.e. morphosyntactic, features are of course language specific.14 While most Indo-European languages seem to contain close appositions (e.g. French la ville de Paris, Dutch de stad Amsterdam, or Polish: miasto Kraków ‘(the) city (of) Cracow’, a number of other language groups do not show parallel structures.15 There is no problem with that as long as we mean the same by ‘name’, i.e. the comparative concept (cf. Haspelmath 2010 for this notion) ‘uniquely denoting nominal expression with a basic level presupposition, but without definitional sense’. Every language makes use of the structures available. Thus, Bantu languages use nominal classes and agreement marking as indications of properhood. For example, in the Bantu language Eton, “proper names, whether they are ancient hydronyms, toponyms or clan names, or improvised nicknames, are always of class 1 a” (Van de Velde 2006; 2008). Names of months appear in close apposition, e.g. 13 Moltmann (2013) deals with ‘sortals’ and close appositions with names from a different perspective. 14 See also Haspelmath (2010): “Descriptive formal categories cannot be equated across languages because the criteria for category-assignment are different from language to language.” 15 Surely, not all close appositions give us the basic level meaning, e.g. President Obama does not say that Obama is necessarily a president. Mostly, the basic level meaning is not overtly expressed, especially not in prototypical names such as personal names, where it is taken for granted. 116 Acta Linguistica Lithuanica LXIX A Semantic-Pragmatic Theory of Proper Names the month of June. That means they are names. This is confirmed by other languages. In the Austronesian language Tuvaluan, names of months have onymic status: when somebody asks about the name of the month, the interrogative pronoun used for names is produced. Likewise, in Rapa Nui, spoken on the Easter Island, names of months are marked by the ‘personal marker’ a, indicating onymic status, e.g. i a hora iti ‘in August’ (Idiatov 2007, taken over by Van Langendonck and Van de Velde 2007: 459–461). Unique denotation of names mirrors the grammatical fact that names take no restrictive relative clauses, as has been argued repeatedly (see also Van Langendonck 2007a: 143). To take a Dutch example, it is ungrammatical to say: (6) *Karel die in Londen woont, komt morgen. ‘Charles who lives in London is coming tomorrow’ In contrast, I can say: (7) De Karel die in Londen woont, komt morgen. ‘The Charles who lives in London is coming tomorrow’ Whilst in (6) the reference to Karel cannot be restricted since a name with a unique denotation is involved, in (7), restriction is possible since the noun phrase de Karel contains an appellativized proprial lemma. As I argued above, even the basic level presupposition is corroborated morphologically and/or syntactically. Names of countries usually display a suffix to indicate country status, e.g. Fin-land, German-y, Chin-a (see Van Langendonck 2007a: 202–206). Whilst close appositions can give us the basic level meaning in most Indo-European languages (cf. Fido the dog, the river Thames), in certain Bantu languages the basic level meaning is indicated by agreement. Van de Velde (2009: 224) states: “Proper Names in Kirundi trigger the same agreement pattern as the common noun that is used to refer to their basic level category”. I refer to Van Langendonck (forthcoming) for further grammatical evidence for name status and basic level presupposition in names. 4. COROLLARIES OF THE PTP Proper names do not form a determinate set There is something to say for this view, if only because no lexical category constitutes a determinate set. The main reason for the fuzzy boundary between name and appellative is the fact that there are prototypical names like first and family names, and place names (both with mainly proprial lemmas at the basis), and less Straipsniai / Articles 117 W I L L Y V A N L A N G E N D O N C K prototypical names such as names of languages, where it wholly depends on the construction whether a noun functions as a name or as a common noun, compare (a) Kalkatungu is an exotic language; (b) I’ll never learn any Kalkatungu (Van Langendonck 2007a: 65, 70, 243). In (a), the language name is truly onymic, in (b) it behaves as a quantified common noun. Trade and brand names constitute a similar case (Van Langendonck 2007a: 235–238; 2007b). Numbers can be seen as names, appellatives or adjectives, again according to the construction (Van Langendonck 2007a: 239 ff.), compare: (a) Three is a sacred number; (b) the number three [apposition]; (c) I saw three lions. In (a) and (b) three is a name, but in (c) it functions as a kind of adjective. So, grammar can help us out, although the PTP never adduces any formal, i.e. morphosyntactic evidence, since grammar seems to be just an epiphenomenon for this theory (see above). Names of months constitute another marginal onymic category. Although these names often occur in apposition, their denotation does not always seem to be unique because of the cyclic character of months, compare: (a) June was fairly hot (non-cyclic) vs. (b) June is always hot (cyclic). That’s why the status of month names is much disputed (cf. Van Langendonck 2007a: 55, 64). It is not inconceivable that such marginal names are construed differently in different languages, e.g. as names or as appellatives (see above). Grammar can help us out as well in the case of such expressions with nonproprial, especially appellative-based lemmas, such as the Old Vicarage (Coates 2006a). Surely, when this expression is used in speech, context will decide on its status, as in the case of ordinary homonymy. However, on the level of established linguistic convention (never mentioned in the PTP), there are two expressions: the name the Old Vicarage, and the appellative-based noun phrase the old vicarage. In the last instance, the expression is a genuine phrase, since it can be penetrated by other elements, e.g. the old but still beautiful vicarage. This is not possible in the name the Old Vicarage, which behaves as [article + onymic word Old-Vicarage]. In addition, here we can appellativize the name by saying: I don’t know of another Old Vicarage. Finally, in names as the Rocky Mountains, the North Sea, the Atlantic Ocean, the basic level meaning is formally expressed. This is not always the case, such as in the PTP’s favorite example Peak’s Tunnel, which is in fact a bridge, not a tunnel (Coates 2012: 125). Therefore, we can assume a continuum from synchronically transparent names like the North Sea to names with an etymologically transparent but synchronically misleading connotation, like Peak’s Tunnel. A final point is that people appear to hesitate in the case of expressions like the universe, or the internet. Up to the present day, they are real single denotata, i.e. singletons, but I would not call them names since no categorical presupposition is available. That they are usually not capitalized is an indication of appellative status, though no real proof. 118 A c t a L i n g u i s t i c a L i t h u a n i c a L X I X A Semantic-Pragmatic Theory of Proper Names Proper names do not fall into logically secure categories This thesis again confounds name (attributed to an entity) and name lemma (not yet attributed). It is essential to realize that uniqueness and basic level meaning are assigned not on the level of the name lemma, but on that of the name designating a unique denotatum. There is nothing illogical about that. Mary is a proprial lemma that can be assigned to women, but surely also to other entities, like ships. However, Mary denotes a unique person after the lemma has been bestowed ad hoc on a particular woman as a name. Not a name, but a name lemma can have more than one denotatum (see above; compare Coates 2006b; 2012: 125). Probabilistic categorization (Coates 2012: 121; 125–129) concerns the lemma with its multidenotative potential, not the name. In all probability, i.e. prototypically, the lemma Mary denotes a particular woman, and obviously more than one, but each time ad hoc and uniquely. In my view, names do not only refer in discourse but constitute a category of expressions that denote uniquely, and subcategorize according to linguistic categories on the basis of the synchronic basic level meaning. That’s also why Peak’s Tunnel need not be a tunnel (Coates 2012: 125). Nowadays, the name denotes a bridge. Semantically transparent historical generics may be misleading, indeed. Thus, in respect of designation and categorization, names differ from pronouns and common nouns (i.e. definite descriptions), which is not clear in the PTP. Proper names cannot be translated According to the PTP, names cannot be translated because they lack sense. However, it will be clear that everything depends on one’s idea of what translation is. Webster’s dictionary defines ‘to translate’ as “to turn into one’s own or another language”. Most dictionaries of other languages also give this broad definition. Thus for Spanish traducir, the Diccionario castellana (1987) has “volver una cosa de una lengua en otra”. The Dutch Van Dale (2005) defines vertalen ‘to translate’ as “van de ene taal in de andere overbrengen” (to transfer from one language to another). Likewise Koenen-Endepols (1970): “in een andere taal overzetten”. The German Duden (1999) defines übersetzen ‘to translate’ as “in einer anderen Sprache [wortgetreu] wiedergeben” (to render in another language). A few dictionaries include the sense (or meaning) in the definition. The French Nouveau Petit Robert de la langue française (2010) gives an elaborate definition: traduire = “faire que ce qui était énoncé dans une langue naturelle le soit dans une autre, en tendant à l’équivalence sémantique et expressive des deux énoncés” (to make sure that what was said in one Straipsniai / Articles 119 WILLY VAN LANGENDONCK natural language is said in another, striving for semantic and expressive equivalence of the two utterances). The Oxford English Dictionary (1991) is ambiguous: to translate = “to turn from one language into another”, or “to change into another language retaining the sense”. If we follow the broader and more common definition of translation, even if this is “a loose way” of defining it (Coates 2012: 129), we can easily include names. The French definition and the second version of the OED are apparently meant for sentences. In that case, names in a sentence form no problem, even if they are not translated. At this point, further distinctions have to be made. First of all, we should again distinguish between names and proprial or other name lemmas. Sometimes, names are translated, sometimes only name lemmas are. 1) Concerning names, a further distinction imposes itself. Names with an underlying proprial lemma have either no translation or an ad hoc translation, without the intervention of sense, but with ‘denotational equivalence’, indeed (for this term, see Coates 2012: 130). For instance, a name like German Mönchengladbach is probably never translated; the Belgian city name Liège translates into German as Lüttich, into Dutch as Luik, i.e. in Coates’ terminology, they “have the same denotatum”. Note that these denotata must be conceived of as unique. If there was another Liège, there would probably be a different or no translation (compare Mechelen above). Names with a transparent underlying appellative lemma may or may not be translated using the transparent content. A name like the Rocky Mountains is seldom translated in German and Dutch but always in French: les Montagnes Rocheuses (Michel Rateau, p.c.). However, the German toponym der Schwarzwald is translated as the Black Forest in English, la Forêt Noire in French, and het Zwarte Woud in Dutch. It can be expected that the ‘translation’ of transparent underlying elements is rather arbitrary, indeed, and may vary from language to language. New York is not translated in French, German or Dutch, but in Spanish, the transparent element New is ‘translated’: Nueva York; likewise in Polish: Nowy York. Lyons (1977) distinguishes between personal names and place names. As a rule, personal names are not translated, e.g. John Major is not translated in French as Jean (le) Majeur. However, the names of important historical figures are translated, e.g. Carolus Magnus (Lat.) = Charlemagne (F) = Charlemagne (E) = Karl der Grosse (G) = Karel de Grote (Du.). Humanists translated their surname into Latin. Mercator was originally called De Cremer (see above). Folk-etymology is a disturbing factor in the translation of names. Coates (2012: 131) cites the example of Hereford, which was rendered as Hairy Foot in the language of Irish Gypsies of the early twentieth century.16 16 For sociolinguistic aspects regarding multiple place-name-giving, see Walther (1968). 120 Acta Linguistica Lithuanica LXIX A Semantic-Pragmatic Theory of Proper Names 2) By contrast, it is sometimes just proprial lemmas that are translated. This is the case of several first names, e.g. John, Jean, Johann, Jan, or Juan go back to the original proprial lemma Johannes. In Flanders, the same man can sometimes be called Jan (Dutch) or Jean (French). Likewise, a number of men called Piet (Dutch) are also called Pierre (French). Another example is Miriam, from which Mary, Marie, Maria, etc. are derived. Of course, it depends on people’s knowledge where comprehensible synchrony ends and incomprehensible diachrony begins (compare Brendler 2008). Thus, via the distinction between name and name lemma, a more refined picture of name translation could be sketched. Sense (or meaning) poses no problem for translation. Theorizing the interface between proper and non-proper expressions has neurophysiological (and psycholinguistic) implications The PTP contends: “semantic reference is mediated by lexical and grammatical meaning, and its processing costs are therefore higher” (Coates 2012: 124); hence, higher in common nouns than in names. Further (p. 132): “names have lower processing costs than fully articulated semantically referring expressions”. This leads to the Onymic Reference Default Principle (ORDP), which says that “the default interpretation of any linguistic string is as a name (i.e. it is free of sense)”. However, what is there to be interpreted if there is no sense at all? In connection with nonsense-words, Coates (2006b: 41) argues himself: “You will not be able to refer to anything using the word tegumai until I tell you what its sense is”. And this is precisely what people want to know. Anyway, no experimental or other evidence is given for the claims of the ORDP. If names have categorical meaning, as I and many others have been claiming, this entails that common nouns are primary since the categories in question are derived from common nouns. Indeed, for Langacker (1991) common nouns are the basic nouns (compare Van Langendonck 2007a: 51, 117). Coates (2012: 132) even speculates that “proper/onymic reference is predicated on a shorter chain of neurones or more efficiently firing synapses”, although neuropsychologists like Semenza and Zettin (1988: 718) just suggest “that proper names, like certain other categories, might be processed independently within the semantic system”. There is no mention of processing costs. From psychology it is well-known that people automatically try to make sense of any string of sounds or letters, which corroborates the linguistic primacy of common nouns with their defining sense. Concerning names, if to retrieve a name, the only clue is the basic level sense (not even present in the PTP), and an optional, or anyway, Straipsniai / Articles 121 W I L L Y V A N L A N G E N D O N C K an incoherent set of connotations, the retrieving process will be hard. In the case of common nouns, however, there is a coherent, polysemous set of senses that can lead much more easily to the identification of an entity. As a matter of fact, psycholinguists now assume that “In comparison to common names [i.e. common nouns, WVL], people’s names are difficult to learn and remember” (Hollis and Valentine 2001: 99). This idea comes from psychological experiments such as the one leading to the Baker–baker paradox (Valentine, Brennen and Brédart 1996: 39). Briefly, this paradox says that, as a rule, people being confronted with a photograph of a man, link an appellative like baker to it rather easily, but the homophonous surname Baker with significantly more difficulty. Valentine, Brennen and Brédart (1996: 39) state: “names were recalled less frequently than professions, even when the same items were involved”, as in the case of B/baker (surname and appellative). It is assumed that there is a direct link in the brain from a man to the surname, but a network of related meanings (bread, bakery, etc.) that allow the subject to retain the relation between a man and the occupation baker better than the relation between a man and the name Baker. This also explains why, as people grow older, they forget names much more easily than common nouns (James 2004). Such psycholinguistic findings contradict the Onymic Reference Default Principle, at least as formulated in the PTP. By contrast, it must be said that these psychological findings do not invalidate the claim that onymic reference is “the default mode of reference” (Coates 2012: 132). In the same vein, Van Langendonck (2007a: 51, 117ff.; 2008) argues that names constitute the prototypical nouns for reference (denotation), displaying unmarked grammatical features like definite, and mostly singular, concrete, and non-generic. Considering this, it will have become clear that we are confronted with the paradox that names provide the prototypical, handiest tool for denotation and reference, but are more difficult to retrieve than semantically fully articulated common noun phrases because of the lack of definitional sense in names. In this connection, Coates (2012: 132) holds “that once onymic reference by an expression has been established, i.e. when it has become a name, there is no going back...” This is likely to be the case. Losing a specific sense will be easier than recuperating it. This claim resembles the one in the huge literature on grammaticalization, where it has been argued that grammatical morphemes have lost their lexical sense and cannot go back to their lexical origin (see among others, Brinton and Traugott 2005). With regard to names, Van Langendonck (2009) concludes “that names are neither grammaticalized nor lexicalized appellative or non-onymic constructions ..., but that at least ‘evolution names’, i.e. the names that did not originate in an illocutionary act out of a proprial (or other) lemma (forenames, film names, etc.) have undergone a desemanticization process that may trigger a number of formal processes, just as in the case of grammatical categories, in which desemanticization takes place as well.” 122 Acta Linguistica Lithuanica LXIX A Semantic-Pragmatic Theory of Proper Names 5. CONCLUSIONS My general conclusions are as follows: the PTP’s thesis that names have no sense should be refined. They have no definitional sense but presuppositions, i.e. grammatical meaning and a categorical or basic level presupposition, and connotative, pragmatic meanings in name lemmas, such as emotiveness, literary and magic content, and finally, connotative, ancillary meanings pertaining to the denotatum or referent of a name, such as encyclopedic or subjective information, and facets of meaning. The assumption of an inherent categorical (basic level) presupposition crucially hinges on a distinction between names and name lemmas, of which proprial lemmas are the most conspicuous. Name lemmas have only a potential for denotation and attribution to a category. When philosophers and psychologists think of a categorical meaning, they apparently attribute this to the name itself, not to its lemma. The claim that the essence of being a name is to be found in reference, not in denotation, should be reversed: denotation is primordial and unique. Otherwise, we cannot even distinguish names from definite descriptions and definite pronouns. In language use, reference, vocation, nomination and ostension should be put on a par. Hence, reference as such cannot define names. Only by accepting unique denotation can we speak of names as functioning in the system of established linguistic convention, and referring in language use. Morphosyntax mirrors this situation, e.g. names take no restrictive relative clauses, and show their categorical meaning in certain close appositions, at least in a number of Indo-European languages, and also in agreement patterns in at least the Bantu language Kirundi. Psycho- and neurolinguistics confirm these analyses. Whether names enter the onomasticon by bestowal or via repeated reference of appellative or other constructions does not change things. Secure synchronic categorization of names is made possible by their inherent basic level meaning and by the assumption that their denotation involves uniqueness. Neither a simple definition of translation nor one involving meaning prevents us from assuming that a number of names, or possibly their name lemmas, can be translated, albeit in different ways. No experimental or other evidence is given for the PTP’s Onymic Reference Default Principle, although onymic reference appears to be “the default mode of reference”, provided ‘unique denotation’ underlies the ‘reference’. Finally, I would like to emphasize that grammar confirms the plausibility of the semantic-pragmatic concepts and theses of the present theory.17 A statement that 17 The present theory does not essentially differ from the theses in Van Langendonck (2007a). However, I have focused here on the distinction between unique denotation and unique reference, refining other notions like that of name lemma, presupposition and connotation, and putting forward a few ideas on name translation. Straipsniai / Articles 123 WILLY VAN LANGENDONCK there is no proper place for a grammar of names in the PTP is incomprehensible because it is in blatant contradiction with the facts of language. In the most diverse languages, names can be distinguished from appellative constructions (definite descriptions) and definite pronouns (see among others recently: Anderson 2007; Idiatov 2007; Van de Velde 2006; 2008; Van Langendonck 2007a, and forthcoming). It is indeed inconceivable that only one linguistic category, i.e. names, has no proper (i.e. just an “epiphenomenal”) grammatical status. REFERENCES Algeo John 1973: On Defining the Proper Name (University of Florida Humanities Monograph 41). Gainesville: University of Florida Press. Anderson John M. 2003: On the structure of names. – Folia Linguistica 37: 347–398. Anderson John M. 2004: On the grammatical status of names. – Language 80: 435–474. Anderson John M. 2007: The grammar of names. Oxford: Oxford University Press. Bayer Joseph 1991: Representatie van algemene namen en eigennamen in het mentale lexicon: neurolinguistische evidentie [Representation of common nouns and proper names in the mental lexicon: neurolinguistic evidence]. – Tabu 21(2): 53–66. Boesch Bruno 1957: Die Eigennamen in ihrer geistigen und seelischen Bedeutung für den Menschen. – Der Deutschunterricht 9(5): 32–50. Brendler Silvio 2004: Klassifikation der Namen. – Namenarten und ihre Erforschung. Ein Lehrbuch für das Studium der Onomastik, ed. by Andrea Brendler, Silvio Brendler. Hamburg: Baar, 69–72. Brendler Silvio 2005: Über den gerechten Tod der Auffassung vom Namen als bilaterales Zeichen. – Proceedings of the XXIst International Congress of Onomastic Sciences (Uppsala 2002), ed. by Eva Brylla, Mats Wahlberg. Uppsala: SOFI. Vol. I, 98–117. Brendler Silvio 2008: Nomematik. Identitätstheoretische Grundlagen der Namenforschung (insbesondere der Namengeschichte, Namenlexikographie, Namengeographie, Namenstatistik und Namenstheorie). Hamburg: Baar, 2008. Brinton Laurel J., Elizabeth Closs Traugott 2005: Lexicalization and language change. Cambridge University Press. Burge Tyler 1973: Reference and proper names. – Journal of Philosophy 70/14: 425–439. Cislaru Georgeta 2006: Propriétés catégorielles des noms de pays. – Onoma 41: 83–113. 124 Acta Linguistica Lithuanica LXIX A Semantic-Pragmatic Theory of Proper Names Cislaru Georgeta 2012: Les facettes des toponymes: des données contextuelles aux modèles sémantiques. – Colloque international: Défis de la toponymie synchronique. Structures, contextes et usages. (Université Rennes 2). To appear. Coates Richard 2000: Singular definite expressions with a unique denotatum and the limits of properhood. – Linguistics 38/6 [370 of the continuous series], 1161–1174. Coates Richard 2005: A speculative psycholinguistic model of onymization. – Naming the world. From common nouns to proper names (RIOn International Series 1), ed. by Brozović–Rončević, Dunja and Enzo Caffarelli. Roma: Società Ed. Romana, 3–12. Coates Richard 2006a: Properhood. – Language 82: 356–382. Coates Richard 2006b: Some consequences and critiques of The Pragmatic Theory of Properhood. – Onoma 41: 27–44. Coates Richard 2009: A strictly Millian approach to the definition of the proper name. – Mind and Language 24.4, 433–444. Coates Richard 2012: Eight issues in the Pragmatic Theory of Properhood. – Acta Linguistica Lithuanica 66, 119–140. Debus Friedhelm 1980: Onomastik. – Lexikon der Germanistischen Linguistik, hrsg. von Hans Peter Althaus, Helmut Henne, Herbert Ernst Wiegand. Tübingen, 187–198. Debus Friedhelm 2012: Namenkunde und Namengeschichte. Eine Einführung. Berlin: E. Schmidt Verlag. De Stefani Elwys, Pepin Nicolas 2006: Une approche interactionniste de l’étude des noms propres: les surnoms de famille. – Onoma 41: 131–162. Donnellan Keith 1974: Speaking of nothing. – The Philosophical Review 83: 3–31. Geach Peter 1957: Mental Acts. London: Routledge and K. Paul. Goossens Jan 1986: Die Namen der belgischen Gemeinden und ihre Schreibung. – Amtlicher Gebrauch des geographischen Namengutes, ed. by Bozen E. Kühebacher, 225–242. Hansack Ernst 2004: Das Wesen des Namens. – Namenarten und ihre Erforschung. Ein Lehrbuch für das Studium der Onomastik, ed. by Andrea Brendler, Silvio Brendler. Hamburg: Baar. Haspelmath Martin 2010: Comparative concepts and descriptive categories in crosslinguistic studies. – Language 86/3, 663–687. Hollis Jarrod, Valentine Tim 2001: Proper-Name Processing: Are Proper Names Pure Referencing Expressions? – Journal of Experimental Psychology: Learning, Memory, and Cognition, Vol. 27, No. 1, 99–116. Horn L. R. 1996: Presupposition and implicature. – The Handbook of Contemporary Semantic Theory, ed. by Sh. Lappin. Oxford: Blackwell, 299–319. Straipsniai / Articles 125 WILLY VAN LANGENDONCK Idiatov Dmitry 2007: A typology of non-selective interrogative pronominals. Doct. diss. University of Antwerp. Online manuscript. James L. E. 2004: Meeting Mr. Farmer versus meeting a farmer: specific effects of aging on learning proper names. – Psych aging 19, 515–522. Jespersen Otto 1924: The Philosophy of Grammar. With a New Introduction and Index by James D. McCawley. Chicago/London: The University of Chicago Press. Kalverkämper Hartwig 1978: Textlinguistik der Eigennamen. Stuttgart. Kleiber Georges 1996: Noms propres et noms communs: un problème de dénomination. – Méta 41(4), 567–589. Kleiber Georges 2004: Peut-on sauver un sens de dénomination pour les noms propres? – Jean-Christophe Verstraete (ed.). Grounding and headedness in the noun phrase. Studies for Willy Van Langendonck. Special Issue of Functions of Language 11 (1), 115–145. Kripke Saul 1972: Naming and necessity. – Semantics of Natural Language, ed. by Davidson, Donald, and Gilbert Harman, 2nd ed. Dordrecht/Boston, 253–355. Kuryłowicz Jerzy 1980: The linguistic status of proper nouns (names). – Onomastica 25, 5–8. Langacker Ronald W. 1987: Foundations of Cognitive Grammar, vol. I: Theoretical Prerequisites. Stanford: Stanford University Press. Langacker Ronald W. 1991: Foundations of Cognitive Grammar, vol. II: Descriptive Application. Stanford: Stanford University Press. La Palme Reyes Marie, MacNamara John, Reyes G. E., Zolfaghari H. 1993: Proper names and how they are learned. – Memory 1(4): 433–455. Lemghari El Mustapha 2011: Complexité sous-catégorielle des noms propres métonymiques et continuum massif vs comptable. To appear in: Proceedings of the 24th International Congress of Onomastic Sciences (Barcelona). Leroy Sarah 2012: Toponymes et constructions toponymiques en français. – Colloque international: Défis de la toponymie synchronique. Structures, contextes et usages. (Université Rennes 2). To appear. Lyons John 1977: Semantics. Cambridge: Cambridge University Press. McKinsey Michael 2010: Understanding proper names. – Linguistics and Philosophy 33/4: 325–354. Molino Jean 1982: Le nom propre dans la langue. – Jean Molino et al. (eds.). Le nom propre. Langages 16 (66), 5–20. Moltmann Friederike 2013: Names, sortals, and the mass-count distinction. Online manuscript. 126 Acta Linguistica Lithuanica LXIX A Semantic-Pragmatic Theory of Proper Names Morris Charles 1938: Foundations of the theory of signs. – International Encyclopedia of Unified Science. Nicolaisen Wilhelm F. H. 1995: Name and appellative. – Name Studies. An International Handbook of Onomastics (Handbook of Linguistics and Communication Science 11), ed. by Ernst Eichler, Gerold Hilty, Heinrich Löffler, Hugo Steger, Ladislav Zgusta. Berlin: Mouton de Gruyter, vol. I, art. 56, 384–393. Nübling Damaris, Fahlbusch Fabian, Heuser Rita 2012: Namen. Eine Einführung in die Onomastik. Tübingen: Narr/Francke. Palmer F. R. 1976: Semantics. A New Outline. Cambridge: Cambridge University Press. Pepin Nicolas, Elwys De Stefani, eds. 2010: Eigennamen in der gesprochenen Sprache. Tübingen: Narr/Francke. Quine Willard V. O. 1960: Word and Object. Cambridge/Mass.: M.I.T. Press. Russell Bertrand 1918 [1964]: The philosophy of logical atomism. – Logic and Knowledge. Essays 1901–1950, ed. by R. C. Marsh. London: Allen and Unwin. Original edition 1918. Russell Bertrand 1919: Descriptions. – Introduction to Mathematical Philosophy. London, chapter 16, 167–180. Searle John 1958: Proper names. – Mind 67, 166–173. Searle John 1971: The problem of proper names. – Semantics. An Interdisciplinary Reader in Philosophy, Linguistics and Psychology, ed. by Danny Steinberg, Leon A. Jak obovits. Cambridge: Cambridge University Press, 134–141. Semenza Carlo 2009: The neuropsychology of proper names. – Mind and Language 24/4, 347–369. Semenza Carlo, Zettin Marina 1988: Generating proper names: a case of selective inability. – Cognitive Neuropsychology 5, 711–721. Strawson Peter F. 1950 [1971]: On referring. – Mind, new series 59 (1950), number 235: 320–344. Reprinted with footnotes in: Strawson, Peter F. 1971. Logico-Linguistic Papers, 1–27. London: Methuen. Strawson Peter F. 1959: Individuals. An Essay in Descriptive Metaphysics. London: Methuen. Strawson Peter F. 1974: Subject and Predicate in Logic and Grammar. London: Methuen. Ullmann Stephen: 1969. Words and their Use. London. Valentine Tim, Brennen Tim, Brédart Serge 1996: The cognitive psychology of proper names: on the importance of being Ernest. London: Routledge. Straipsniai / Articles 127 WILLY VAN LANGENDONCK Vandelanotte Lieven, Willemse Peter 2002: Restrictive and non-restrictive modification of proprial lemmas. – Word 53(1), 9–36. Van de Velde Mark L. O. 2006: Multifunctional agreement patterns in Bantu and the possibility of genderless nouns. – Linguistic Typology 10(2), 183–221. Van de Velde Mark L. O. 2008: A grammar of Eton. Berlin: Mouton de Gruyter. Van de Velde Mark L. O. 2009 [2012]: Agreement as a grammatical criterion for Proper Name status in Kirundi. – Onoma 44, 219–241. Van Langendonck Willy 1968: Review of F. Debus. Aspekte zum Verhältnis Name – Wort (Groningen: Wolters 1966). – Beiträge zur Namenforschung 3, 203–205. Van Langendonck Willy 1999: Neurolinguistic and syntactic evidence for basic level meaning in proper names. – Functions of Language 6(1), 95–138. Van Langendonck Willy 2006 [2011]: Name theory and set theory: the return of the maximum meaningfulness view of names. – Onoma 41, 45–62. Van Langendonck Willy 2007a: Theory and Typology of Proper Names. Berlin: Mouton de Gruyter. Van Langendonck Willy 2007b: Trade and brand names: proper or common nouns? – Names in Commerce and Industry: Past and Present (colloquium Antwerpen 2006), ed. by Ludger Kremer and Elke Ronneberger-Sibold. Berlin: Logos, 2007, 23–33. Van Langendonck Willy 2008: Not common nouns but proper names are the prototypical nouns. – I nomi nel tempo e nello spazio. Atti del XXII Congresso Internazionale di Scienze Onomastiche, vol. 1, ed. by Maria-Giovanna Arcamone, Donatella Bremer, Davide De Camilli, Bruno Porcelli. Pisa, ETS, 815–822. Van Langendonck Willy 2009: Proper names, grammaticalization and lexicalization. – Teoretické a komunikační aspekty proprií. Prof. Rudolfu Šrámkovi k životnímu jubileu, ed. by M. Harvalík, Eva Minářová, Jana M. Tuskova. Brno: Masarykova Univerzita, 89–102. Van Langendonck Willy 2010a: Names and identity. On the nomematic approach. Review article of Silvio Brendler, Nomematik, Identitätstheoretische Grundlagen der Namenforschung (insbesondere der Namengeschichte, Namenlexikographie, Namengeographie, Namenstatistik und Namenstheorie). Hamburg: Baar, 2008. Beiträge zur Namenforschung 45/2, 193–220. Van Langendonck Willy 2010b: Europäische Personennamensysteme: Ein Handbuch von Abasisch bis Zentralladinisch. Hrsg. von Brendler, Andrea & Brendler, Silvio. 2007. Anlässlich der 65. Geburtstage von Rosa Kohlheim und Volker Kohlheim. Hamburg: Baar. – Acta Linguistica Lithuanica 62–63, 179–193. Van Langendonck Willy Forthcoming: Names and Grammar. – Oxford Handbook of Names and Naming. Oxford: Oxford University Press. 128 Acta Linguistica Lithuanica LXIX A Semantic-Pragmatic Theory of Proper Names Van Langendonck Willy, Van de Velde Mark 2007: Naar een universele theorie van eigennamen [Towards a universal theory of proper names]. – Handelingen van de Koninklijke Commissie voor Toponymie & Dialectologie 74, 429–467. Walther H. 1968: Mehrnamigkeit von Siedlungen als sprachsoziologische Erscheinung. – Leipziger namen kundliche Beiträge 2, 19–28. Semantinė pragmatinė tikrinių žodžių teorija SANTRAUKA Šis straipsnis – reakcija į redukcionistinį požiūrį, kurį Richardas Coatesas išdėstė publikacijoje „Eight Issues in the Pragmatic Theory of Properhood“ (Acta Linguistica Lithuanica 66, 2012, 119–143). Joje autorius tikrinius žodžius, kaip specifinės leksikos vienetus, aptaria pragmatinės lingvistikos teorijos aspektu, pagrįstu dviem esminėmis idėjomis: tai, kad tikriniai žodžiai neturi reikšmės ir kad tikrinių žodžių esmė – įvardijimas, o ne denotacija. Tokius teiginius reikėtų vertinti gana kritiškai, nes šios nuostatos ne visai pagrįstos. Tai galima įrodyti atlikus išsamią semantinę pragmatinę tikrinių žodžių analizę. Šiuo aspektu išanalizavus tikrinius žodžius ir pateikus gramatinių, filosofinių, psicholingvistinių ir neurolingvistinių įrodymų galima teigti, kad: 1) tikriniai žodžiai turi specifinę semantiką, kuri susijusi su žmonių psichika, t. y. patirtimi, kalbėjimo situacija, suvokimu ir visu enciklopediniu žinojimu, lemiančiu sėkmingą komunikaciją. Semantinis santykis yra santykis ne tarp žodžio ir daikto, bet tarp dviejų mentalinių fenomenų – žodžio reprezentacijos ir pasaulio reprezentacijos individo sąmonėje; 2) tikriniai žodžiai ne tik įvardija unikalų objektą kalbos vartosenoje, bet ir pažymi jį nustatytos lingvistinės konvencijos lygiu. Kad tai taptų priimtina, svarbu išskirti tikrinį žodį ir tikrinio žodžio lemą. Šis išskyrimas susijęs su konotacijos sritimi. Įteikta 2013 m. balandžio 4 d. WILLY VAN LANGENDONCK University of Leuven, Department of Linguistics Blijde-Inkomststraat 21–3308, BE–3000 Leuven, Belgium [email protected] Straipsniai / Articles 129