Aspects of Definiteness in Greek

44
is is a contribution from Studies in Language, Vol. 33:3. © 2009. All rights reserved. is electronic file may not be altered in any way. e author(s) of this article is/are permitted to use this PDF file to generate printed copies to be used by way of offprints, for their personal use only. Permission is granted by the publishers to post this file on a closed server which is accessible to members (students and staff) only of the author’s/s’ institute. For any other use of this material prior written permission should be obtained from the publishers or through the Copyright Clearance Center (for USA: www.copyright.com). Please contact [email protected] or consult our website: www.benjamins.com Tables of Contents, abstracts and guidelines are available at www.benjamins.com John Benjamins Publishing Company

Transcript of Aspects of Definiteness in Greek

This is a contribution from Studies in Language, Vol. 33:3.© 2009. All rights reserved.

This electronic file may not be altered in any way.The author(s) of this article is/are permitted to use this PDF file to generate printed copies to be used by way of offprints, for their personal use only.Permission is granted by the publishers to post this file on a closed server which is accessible to members (students and staff) only of the author’s/s’ institute.For any other use of this material prior written permission should be obtained from the publishers or through the Copyright Clearance Center (for USA: www.copyright.com). Please contact [email protected] or consult our website: www.benjamins.com

Tables of Contents, abstracts and guidelines are available at www.benjamins.com

John Benjamins Publishing Company

Studies in Language 33:3 (2009), 569–611. doi 10.1075/sl.33.3.03napissn 0378–4177 / e-issn 1569–9978 © John Benjamins Publishing Company

Aspects of definiteness in Greek

Maria NapoliUniversity of Pisa

0. Introduction

This paper primarily deals with the use of the Greek definite article with gener-ics and proper nouns, by concentrating on its development from Ancient Greek,1 where this use is typically described as ‘optional’, to Modern Greek, which compul-sorily employs the definite article with these kinds of nouns. Even though the fo-cus of my investigation is the Greek language, the occurrence of the definite article with generic and proper nouns is examined from a cross-linguistic point of view.

The main goal of this work is to illustrate the contribution that Greek can make to a general theory of definiteness, taking this case-study as a starting point. In order to do this, I shall consider some recent achievements in the field of defi-niteness, relevant to a critical revision of traditional assumptions on the usage of the definite article in Ancient Greek. Moreover, I shall compare data from Ancient Greek with data from modern languages (in particular, apart from Modern Greek, English and Italian, which show a different type of behaviour in the occurrence of the definite article with generics and proper nouns). An exhaustive description of all the functions of the Greek definite article is far beyond the scope of the present paper. I shall concentrate on Classical Greek, and, more specifically, on data drawn from some works by Aristophanes and Plato, relevant to the aim of discussing the issue of definiteness with regard to the case-study illustrated here.

1. The emergence of definite articles: The case of Greek

Definiteness is considered as typically conveyed by articles (see, however, 2.1. on languages without an article system): they can be unbound morphemes, as in Eng-lish, or affixes on nouns, as in Old and Modern Scandinavian (cf. Abraham & Leiss 2007). I shall begin with some introductory remarks about definite articles, from a general point of view, and, more specifically, in Greek.

© 2009. All rights reserved

570 Maria Napoli

Across languages “the most common origin of the definite article is the de-monstrative” (Greenberg 1978: 61). This diachronic observation is consistent with the Greek situation:2 Proto-Indo-European is assumed as lacking definite articles, whereas the Greek language developed a definite article from an inflected form, which, in origin, was a demonstrative pronoun. Ancient Greek belongs to those languages which do not have an indefinite article, but only a definite article, which is inflected for number, gender and case: as a consequence, bare nouns are func-tionally opposed to definite nouns in Ancient Greek (cf. 3.2., 4.2.). On the con-trary, Modern Greek has a definite and an indefinite article.3 With regard to the emergence of the Greek definite article, as is well-known, the Homeric language stands apart: scholars agree in recognizing that in Homer, the definite article as an independent category is not yet fully developed; as noted, among others, by Chan-traine (1953),4 the forms which unambiguously correspond to definite articles in Classical Greek, generally still preserve the function of demonstratives. Both in Ancient and Modern Greek, the definite article can also co-occur with a demon-strative, and it can be “doubly articulated” (Plank 2003: 302 f.), when a noun is used with an adjective: in this paper, I will leave these phenomena aside.

From a cross-linguistic point of view, the traditional theory assumes that, in the initial stage of the emergence of definite articles from demonstratives, deictic and anaphoric uses tend to overlap, to the extent that sometimes it is difficult to distinguish between them.5 A challenge to the received opinion comes from more recent attempts to describe such paths of grammaticalization within the theoreti-cal framework of Centering Theory (cf. Stark, Leiss & Abraham 2007, in particular Abraham 2007a, 2007b and Leiss 2007). On the basis of this approach, which em-phasizes the differences between demonstratives and articles, and the significant role played by discourse properties of theme and rheme in the grammaticalization of definiteness, the core function of definite articles is not the anaphoric reference. More specifically, data from Old Icelandic and Gothic provide evidence that the rise of definite articles starts in the rheme: “a definite rhema is marked by a definite article in order to inhibit the indefinite reading presupposed by rhematic objects” (Leiss 2007: 96; cf. also Abraham 2007a, 2007b). This also means that at the begin-ning of the development of definite articles, definiteness is not overtly marked whenever it is ‘inherent’ (Leiss 2007: 88): this comprises the discourse theme, nor-mally entailing definite readings, proper nouns and anaphora (i.e., reference to known information). Following Leiss (2007: 91), it is worth mentioning that, for instance, in Old High German (750–1050) the definite article is used anaphori-cally, whereas Gothic (4th century) does not document this use.

This theoretical model could offer a stimulating background for the attempt to reconstruct the diachronic phases of the emergence of the Greek definite ar-ticle. However, this attempt would be far beyond the scope of this paper. As a

© 2009. All rights reserved

Aspects of definiteness in Greek 571

consequence, there are some important questions about the development of defi-niteness from Ancient to Modern Greek, as well as about the discourse functions of the anaphoric definite article, that I will not try to address here. As already noted in the Introduction, I shall primarily examine the use of the definite article with generics and proper nouns, and its typological variation (cf. 3, 4 and 5).

In the next section, I shall consider the Ancient Greek state of affairs more in detail.

1.1 The issue of the ‘optional’ article in Ancient Greek

In Modern Greek, the definite article is widely used: it occurs in anaphoric and non-anaphoric contexts, and it is extended to generics and proper nouns (cf. 4.1. and 5.1.). On the contrary, in Ancient Greek, the picture is neither consistent nor systematic: the definite article is regularly found with an anaphoric function, but it seems to be optionally used with some specific kinds of nouns, i.e., generics, proper nouns, nouns denoting unique entities, mass and abstract nouns.

First of all, let me quote an example from Classical Greek, which illustrates the anaphoric use of the definite article:

(1) theĩsai mélainan kúlika megálēn huptían, put.up black-ACC.SG cup-ACC.SG big-ACC.SG hollow-ACC.SG mēlosphagoũsai Thásion oínou stamníon slaughter Thasian-ACC.SG wine-GEN.SG magnum-ACC.SG omósōmen eis ten kúlika me ‘picheĩn húdōr swear-1PL into the-DEF-ACC.SG cup-ACC.SG NEG pour-INF water-ACC.SG ‘We put a big black wine cup hollow up, right here; we slaughter a magnum of Thasian wine into it; and we swear not to pour any water into the cup’. (Ar., Lys. 195–197)

In (1), the noun kúlika ‘cup’, used without any determiner in its first mention (which introduces it as a new referent), is followed by the anaphoric expression ten kúlika, with the definite article. This example is an instance of the so-called ‘di-rect anaphora’: the anaphoric noun phrase entails a referent previously mentioned in the text.

The process presupposed by the linguistic phenomenon of direct anaphora is well described by Hawkins (1978: 107–108): “an object or event unknown to the hearer is presented to him, generally in the form of an indefinite description, whereupon both he and the hearer may refer back to this object or event using a definite description […]. What seems to be going on is that the hearer, upon hearing the indefinite description enters an object into his memory store. The

© 2009. All rights reserved

572 Maria Napoli

subsequent use of a definite article with the appropriate descriptive predicate then signals that the hearer should pick out this object from his memory store. Thus, the act of referring anaphorically involves a form of instruction to the hearer to match the linguistic referent of the definite description with a particular object in his mind, an object which has been entered into his memory store in the course of some previous conversation with the speaker”.6

Direct anaphora should be distinguished from ‘indirect anaphora’ (or ‘associa-tive anaphora’, according to Hawkins’ terminology; cf. also Conte 1996). Indirect anaphora is based on the inferences that the hearer can make because a specific referent has been previously mentioned; the mention of this referent indirectly in-troduces a range of other referents which are related to it (because they are physi-cally part of it or because they are associated with it on a more abstract, intellectual level). This inferred relationship allows the speaker to use the definite article when speaking of one of these referents, even though it has not been explicitly men-tioned before. The following case shows the use of the Greek definite article in indirect anaphora:

(2) eimen oũn oíkade eis toũ Polemárchou […]. go-1PL so house.to to the-DEF-GEN.SG Polemarchus-GEN.SG en d’ éndon kaì ho pater be-3SG PART at.home also the-DEF-NOM.SG father-NOM.SG ho toũ Polemárchou Képhalos […], the-DEF-NOM.SG the-DEF-GEN.SG Polemarchus-GEN.SG Cephalus-NOM.SG tethukos gar etúnchanen en tei aulei sacrifice-PTCP-NOM.SG for finish-3SG in the-DEF-DAT.SG court-DAT.SG ‘So we went with them to Polemarchus’ house […]. And the father of

Polemarchus, Cephalus, was also at home […], for he had just finished sacrificing in the court’.

(Pl., R. 328 b.4., b.8, c.2)

In example (2), the noun aule ‘court’, used with the definite article, has not been introduced before, but there has been previous mention of a house. So, the use of the definite article with aule is possible because of the previous mention of the ex-pression oíkade (‘to one’s house’) and, consequently, because of the inference that a house materially has a court (at least, in Plato’s time). In a case like this, the men-tion of a noun provides “a trigger for the association that familiarizes the definite noun phrase. If this is correct, it is possible for an associative use of the to be based on following as well as preceding information” (Lyons 1999: 5).

To sum up, in Classical Greek the definite article regularly occurs in direct and indirect anaphora. On the other hand, with regard to the so-called optional use of the definite article with generics, proper nouns, nouns of unique entities, mass and

© 2009. All rights reserved

Aspects of definiteness in Greek 573

abstract nouns, it is often assumed that there is no significant difference in meaning between the use of these nouns with or without the article. From this, one could argue that Greek authors used the definite article or omitted it interchangeably, depending on stylistic reasons: for example, Gildersleeve (1890: 483, 485 f.), among others, regarded this use as ‘popular’. As an example of the occurrence of nouns de-noting unique entities, I quote (3.a.), where the definite article occurs with the noun selenē ‘moon’, and (3.b.), where the same noun is used without any determiner:

(3) a. hē selenē d’ exéleipen/ the-DEF-NOM.SG moon-NOM.SG PART desert-3SG tas hodoús […] the-DEF-ACC.PL street-ACC.PL ‘The moon deserted her orbit’. (Ar., Nu. 584) b. ei mēkét’ anatélloi selenē mēdamoũ, if no.more rise-3SG moon-NOM.SG anywhere ouk an apodoíēn tous tókous NEG MOD pay.would-1SG the-DEF-ACC.PL interest-ACC.PL ‘If the moon never again rose anywhere, I’d never pay my interest’. (Ar., Nu. 754–755)

A significant contribution to this topic has been given by Manolessou & Horrocks (2007: 230), who rightly note that neither ancient nor modern studies pay attention to the specific contexts in which the definite article occurs with generics, proper nouns, nouns denoting unique entities, mass and abstract nouns. Nevertheless, as they point out, it is possible to recognize a certain ‘regularity’ in the presence or ab-sence of the article with this range of nouns, which “is strongly disfavoured in belle-tristic/poetic contexts, official/legal discourse, prepositional phrases, parenthetical and reported speech, but much more common in comedy/dialogue (especially in anaphoric use), with names used as the subjects of verbs of saying, with topicalised and dislocated items, and with nominals premodified by an attributive adjective. In other words, it seems that ‘popular/natural’ styles employed the ‘new’ construction much more readily than more conservative styles for marking the pragmatic/dis-course functions attaching to semantic definites (viz. topicalization, salience, ref-erence-tracking, etc.)” (Manolessou & Horrocks 2007: 230). Significantly enough, they add that, apart from stylistic influences, “nonetheless, it still remains the case that only the more important referents routinely have the article, and typically only when reintroduced in alternation with other participants, i.e. when there is a clear switch of topic/focus. Hence the article is routinely omitted even with key partici-pants in the absence of such retopicalisation, as also in parentheticals (which are by

© 2009. All rights reserved

574 Maria Napoli

definition outside the main discourse) and reported speech (which is taken from a different discourse context)” (Manolessou & Horrocks 2007: 230–231).

As a matter of fact, working on the distribution of the definite article in An-cient Greek more in depth, it seems that its being ‘optional’ with some specific types of nouns does not necessarily mean that its use is the consequence of a casual choice, nor that the alternatives (the presence or the absence of the article itself) are completely equivalent. On the contrary, the possibility of such an option turns out to be limited, at least in some contexts, by pragmatic forces. On the other hand, the fact that the article is ‘optional’ in some cases is precisely what makes Ancient Greek so interesting from the point of view of a theory of definiteness, because it reflects its being a language which documents the phasal development from a demonstrative element to a definite article.7

I shall come back to this issue after some general considerations on definite-ness.

2. Towards a typology of definiteness

I will treat definiteness within the theoretical framework of functional-typology, taking into account both its formal expression and its semantic content.

2.1 The formal expression of definiteness

First of all, a preliminary consideration is in order: languages with definite articles are particularly widespread “in Europe, especially western Europe, in a wide belt across central Africa from west to east (though not in the south), in New Guinea and the Pacific, and in Mesoamerica, as well as being scattered in other areas” (Dryer 2005: 155). Nevertheless, the majority of the languages of the world lack a specific marker for definiteness; more precisely, as Table 1 shows, the number of languages in Dryer’s sample with a definite word distinct from a demonstrative is almost identical to the number of languages with neither a definite nor an indefi-nite article:

Table 1. Definiteness markers across languages (from Dryer 2005: 154)

1. Definite word distinct from demonstrative 197

2. Demonstrative word used as marker of definiteness 56

3. Definite affix on noun 84

4. No definite article but indefinite article 41

5. Neither definite nor indefinite article 188

Total 566

© 2009. All rights reserved

Aspects of definiteness in Greek 575

The existence of such a cross-linguistic variety, and, in particular, the existence of languages which completely lack any dedicated morphology for definiteness, has given rise to a discussion of the universality of definiteness. There are differ-ent approaches to this issue. One, represented by Krámský (1972), claims that the opposition between definite and indefinite is universal, and languages can denote it by various grammatical means. Another approach, mainly based on Green-berg (1978), focuses on the expression of definiteness by means of articles, and on their behaviour across languages. An intermediate solution has been proposed by Lyons (1999), who distinguishes between semantic/pragmatic definiteness and grammatical definiteness: what seems to be universal is definiteness as a category of meaning, i.e. the semantic/pragmatic concept of definiteness, which is differ-ent from its grammatical encoding. Nevertheless, according to Lyons (1999: 278), definiteness as a grammatical category “is only present in languages which show overt definiteness marking”.

Given that, let us have a look at languages without articles. Following Krámský (1972), many scholars have argued that in both ancient and modern languages without articles, definiteness can be expressed by different grammatical means or, more in general, by different strategies, such as: word order, extensive use of pos-sessives or demonstratives, quantifiers, opposition between different case markers, prosodic means. In this respect, it is worth mentioning the interesting results of some recent works by Abraham and Leiss (Abraham 1997, 2007a, 2007b, Leiss 2000, 2007, Abraham & Leiss 2007), which have pointed to the existence of a close relationship between morphological case, aspect and definiteness: more precisely, they show how “(in)definiteness effects are achieved in Gothic and early stages of the German languages by combining verbal aspect with alternating cases, the main pattern being as in Russian and other Slavic languages” (Leiss 2007: 87).

In my opinion, the fact that ‘indirect’ expressions of the definite vs. indefinite status of referents are found in languages without articles is undeniable. However, from a methodological point of view, it does not seem to be correct to investigate languages without articles trying to identify the ‘exact equivalent’ of articles in languages which possess them. In particular, it is necessary to distinguish statisti-cal tendencies from obligatory markers: as Jakobson (1971: 492) wrote, “The true difference between languages is not in what may or may not be expressed but in what must or must not be conveyed by the speakers. If a Russian says: Ja napisal prijatelju ‘I wrote a friend’, the distinction between the definiteness and indefinite-ness of reference (‘the’ vs. ‘a’) finds no expression, whereas the completion of the letter is expressed by the verbal aspect, and the sex of the friend by the masculine gender”.

I shall briefly deal with the case of Indo-European languages like Latin and Rus-sian, where word order has sometimes been assumed to convey definiteness: noun

© 2009. All rights reserved

576 Maria Napoli

phrases occurring in the initial or in the final part of the sentence are supposed to be definite or indefinite, respectively.8 However, even though the interaction of word order with definiteness seems to be unquestionable in terms of a general ten-dency, the point has to be cautiously considered. Two remarks are of some impor-tance. First, the initial sentence position tends to be associated with given informa-tion and with definiteness; on the contrary, the final sentence position tends to be associated with new information and with indefiniteness, as is typical of Latin: “in ‘all-new’ sentences the order VS seems to be quite common” (Pinkster 1990: 79). This phenomenon, however, is widespread across languages, independently of the presence of articles: as noted by Trenkić (2002), “the tendency for new referents to avoid clause-initial position is not restricted to languages without articles […]. The general tendency of human communication seems to be for given information to precede new information”. Second, definiteness is only one of the semantic features influencing word order. In a recent paper, Sornicola (2006) has pointed out that the animacy of the referent has a great influence on word order in Russian.

To conclude, the fact that the definite or indefinite status of the basic con-stituents can interact with word order must be distinguished from the existence of definiteness as a grammatical category. Indeed, such an influence does not mean that word order is systematically used to convey definiteness or indefiniteness in languages without articles. Bearing in mind the quotation from Jakobson (1971) mentioned above, we could say that definiteness must not be conveyed in Latin or Russian. But, in languages lacking definite articles, there are interesting phe-nomena concerning the interplay of semantic definiteness with other grammatical categories, which can be investigated only by considering the grammatical systems of these languages in their complexity.

2.2 The semantic content of definiteness

I shall turn now to considering the issue of the semantic content of definiteness. As is well-known, the ‘definition of definiteness’ has been a thorny problem for general linguists, logicians, semanticists and pragmatists. In order to describe defi-niteness, different labels have been proposed, such as ‘uniqueness’ (Russel 1905), ‘familiarity’ (Christophersen 1939), ‘inclusiveness’ (Hawkins 1978), ‘identifiabil-ity’ (among others, Comrie [1981] 19892, Birner & Ward 1998, Lambrecht 1994) or ‘grammaticalization of identifiability’ (Lyons 1999). Some of these labels (like ‘uniqueness’) are mainly based on semantic criteria, others (like ‘familiarity’) on a discourse-pragmatic approach. ‘Identifiability’, which is the most widespread label nowadays, has been defined as a ‘mutual category’, shared by speaker and hearer: when the speaker utters a noun phrase plus a definite article he presupposes that the hearer is able to identify the referent whom he has spoken about.9

© 2009. All rights reserved

Aspects of definiteness in Greek 577

As a matter of fact, it is quite evident that the various uses of markers of defi-niteness across languages flee from a unitary formal characterization. The notions of ‘uniqueness’ and ‘identifiability’, for instance, are closely related in every in-tuitive characterization of the types of noun-phrases which tend to be marked as definite across languages (cf. Map. 1 below). However, neither of them can cover all the uses of definite articles, offering a formal characterization of definiteness itself.10

Any attempt to investigate definiteness seems to impose, primarily, the need to change the point of view and go beyond labels. As Corbett (2005: 25) notes, “we fail to take sufficient care over our terminology and so do not see that phenomena labelled identically are in fact distinct (conversely we miss identities because of dif-ferent traditions of labelling)”. Going beyond labels, we should ask what is relevant to the aim of investigating cross-linguistic variety in the field of definiteness. First of all, it is necessary to take into account the grammatical features which correlate with definiteness in a single language, and how they interact (cf. Givón 1978). On the other hand, focusing on languages with articles, their different cross-linguistic uses cannot be explained without invoking language-specific diachronic developments.

Moreover, a preliminary distinction between two different levels of definite-ness could be useful, as I shall try to show in the next section.

3. Logical definiteness and pragmatic definiteness

In this paper, I shall take Löbner’s (1995) distinction between semantic definiteness and pragmatic definiteness as a starting point. Semantic definiteness is conceived of as dependent on the inherent semantic content of nouns: proper nouns and nouns of unique entities (like the sun, the moon, the weather, the air) are assumed as typical examples of ‘semantic definites’, in their denoting entities which are defi-nite by themselves; on the other hand, pragmatic definiteness is independent of the type of nouns, and has to do with the context.11

Following Löbner (1995), I will adopt here the distinction between two differ-ent levels of definiteness; nevertheless, my proposal is to use a different terminol-ogy: logical definiteness (rather than semantic) vs. pragmatic definiteness. In my opinion, the label ‘logical’ is preferable to the label ‘semantic’, which has been used with various meanings in literature on definiteness.12

3.1 A Map of Definiteness

Starting from the distinction between logical and pragmatic definiteness, in Fig-ure 1, I have tried to sketch a possible Map of Definiteness, which is grounded in

© 2009. All rights reserved

578 Maria Napoli

the most widespread uses of the definite article in European languages (mainly on the basis of Hawkins 1978, Lyons 1999 and Schroeder 2006):13

DEFINITENESS

LOGICAL PRAGMATIC

Cases in whichde�niteness is

logically derived

Cases in whichde�niteness is

logically inherent

Cases in whichde�niteness depends on

the extra-linguistic context

Cases in whichde�niteness depends on

the linguistic context

Nouns witha kind-reading

(Generics)

CommonNouns

of unique entities

Nounswith a referent

‘identi�able’from the situation

Nounswith a referent

physically accessiblein the situation

Nounswith a referent

previouslymentioned

Nounswith a referent

‘indirectly’inferable

ProperNouns

In all the world In a relative world

Figure 1. Map of Definiteness as conveyed by articles

The main distinction is between logical definiteness and pragmatic definiteness. Under the label ‘logical definiteness’ I include only those cases in which definite-ness exclusively depends on the semantics of nouns, rather than on the situation or the context. As already said, logical definiteness primarily concerns proper nouns and nouns of unique entities, referring to “something of which there is only one” (Lyons 1999: 8). ‘Nouns of unique entities’ comprise two types of nouns, which are partly different:14 on the one hand, nouns like the sun, the moon, the atmo-sphere, designating an entity which is the only one in all the world; on the other hand, nouns designating an entity which is the only one in a relative world, i.e. with respect to a specific, more restricted place: for example, people living in the same country speak about the Prime Minister. This distinction could appear trivial; however, in my opinion, it is justified not only for a ‘semantic’ reason, i.e. for the fact that “the sun is not very sensitive as to the context in which it may have unique reference, in fact, it has unique reference in almost all times and places of speech” (Schroeder 2006: 549), but it is also reflected in the behaviour of these nouns with respect to number markers: nouns like the sun are generally incompatible with plural markers (in languages which mark number distinctions like English or Ital-ian), whereas I can conceive of talking about a plurality of things, any one of which is unique only for people sharing it (for ex., Prime Ministers are quite numerous in Europe).15 Nevertheless, from the point of view of definiteness, these two types of nouns seem to be equally treated in individual languages: in Ancient Greek, for instance, when the noun basileús ‘king’ is used as the official name of the King

© 2009. All rights reserved

Aspects of definiteness in Greek 579

of Persia, it can occur with or without the definite article (cf. examples in Gilder-sleeve 1911: 261), as well as nouns like the sun or the moon (see 1.1.).

Definite nouns employed with a kind-reading (generics) are in the domain of logical definiteness too. A remark is in order here. Generics are common nouns used with reference to a class, characterized by some specific properties which make it ‘unique’ (the class of cats as opposed to the class of dogs). As we shall see in Section 4, languages employ different types of noun phrases with a generic value, including indefinite singulars or bare plurals. Clearly enough, the definite-ness of generics is not inherent. Nevertheless, while it is true that “definiteness is not a necessary requirement for generic interpretation” (Schroeder 2006: 550), it is equally true, to my knowledge, that there are no languages in which genericity cannot be expressed by a definite noun phrase (singular, plural or both, as in Ital-ian: cf. 4.1.). This is the reason why generics can be included in a Map of Definite-ness. However, we have to take into account that the definiteness of a noun used with a generic value depends on a specific semantic interpretation, on its being the typical instance of the unique category to which it belongs. In this case, definite-ness is a matter of logical inference, and it can be described as ‘logically derived’, rather than ‘logically inherent’.

Languages with a morphological category of definiteness greatly differ in the use of the definite article in the domain of logical definiteness. I shall come back to this point in Section 3.2. below.

By contrast, all cases in which definiteness depends on the context are in the domain of pragmatic definiteness. In this domain, two main situations have to be distinguished: definiteness may depend on the extra-linguistic context or on the linguistic context. The case in which definiteness depends on the extra-linguistic context includes two distinct subtypes:

a. The definite noun corresponds to an entity not physically accessible in the extra-linguistic context, the identifiability of which depends on the situation: it is inferred by means of shared knowledge, or evidence taken from the situ-ation itself.

b. The definite noun corresponds to an entity physically accessible in the extra-linguistic context.

The last subtype can be illustrated by means of the following example, taken from Lyons (1999: 3):

(4) Just give the shelf a quick wipe, will you, before I put this vase on it.

In (4), the real (extra-linguistic) context in which speaker and hearer act, allows the speaker to use the definite article in mentioning an object that is before their eyes. This example is an instance of Hawkins’ (1978) ‘visible situation use’: “If an

© 2009. All rights reserved

580 Maria Napoli

object is visible to both speaker and hearer in the situation of utterance, and is fur-thermore unique, this permits the use of the definite article” (Hawkins 1978: 110). Hawkins (1978: 112 f.) distinguished this case from ‘the immediate situation use’, which is similar, but not identical, to the visible situation use, because the referent is not really visible to both speaker and hearer, as in the well-known notice ‘Beware of the dog!’: obviously, the dog referred to is the one living in the house where the notice is displayed.16

In order to illustrate the subtype (a), let us consider two more examples from Lyons (1999: 3)

(5) I bought the car this morning.

(6) Put these clean towels in the bathroom please.

A sentence like (5) can be uttered even though the referent is not physically pres-ent, but the speaker and the hearer share the knowledge of a specific situation in which the referent exists as a definite entity: in (5), in particular, what the hearer could share with the speaker is the knowledge of his intention to buy a new car. In (6), the situation itself provides evidence for identifying the referent: the bath-room is obviously the bathroom of the house where the hearer is. To sum up, the extra-linguistic context provides the framework necessary in order to speak of the referent as definite.

Pragmatic definiteness may also depend on the linguistic context. In particular, nouns with a referent previously mentioned are codified as definite: the anaphoric function is quite regularly (even though not exclusively) associated with definite articles. However, the referent can be indirectly inferred and introduced as ‘defi-nite’ because of association within the linguistic context: this is proper of indirect anaphora (see ex. 2 quoted above), but also of those cases in which the presence of a linguistic element determines the presence of the definite article (Hawkins 1978: 132 f.), as in relative clauses (I will lend you the book that I have finished to read this morning), associative clauses of the kind the bottom of the sea, the waves of the sea, noun phrase complements (the fact that…, the reason that…).

3.2 Logical versus pragmatic definiteness: across languages, across time

Across languages, articles can mark all subtypes of definiteness illustrated in Fig-ure 1, or only some of them. In particular, as said above, languages have a greatly different behaviour with regard to logical definiteness: for example, Modern Greek always uses the definite article with generics, proper nouns and nouns of unique entities; in English proper nouns of individuals are obligatorily bare, whereas the definite article occurs with nouns of unique entities and with generics too (even

© 2009. All rights reserved

Aspects of definiteness in Greek 581

though generics may be bare; for more details cf. 4.1, 5.1., 6.2.). On the other hand, languages seem to have a more uniform behaviour as concerns the pragmatic use of definite articles. I shall give two sentences translated in three different lan-guages, English, Italian and Bulgarian (which has a definite article suffixed to the noun):17

(7) a. In the distance, I saw a dog. The dog was slim and dirty. b. In lontananza ho visto un cane. in distance see-1SG a-INDF-M.SG dog Il cane era magro e sporco. the-DEF-M.SG dog be-3SG slim and dirty c. V dalečinata vidiah kuče. Kuče-to beše slabo i mrăsno. in distance see-1SG dog. dog-the-DEF be-3SG slim and dirty

(8) a. In the distance, I saw the dog of our neighbour. b. In lontananza ho visto il cane del nostro vicino. in distance see-1SG the-DEF-M.SG dog of-the-DEF our neighbour c. V dalečinata vidiah kuče-to na našia săsed. in distance see-1SG dog-the-DEF of our neighbour

Some scholars, like Lyons (1999), have proposed that the main distinction is be-tween anaphoric and non-anaphoric uses of the definite article. Evidence seems to be provided by languages in which two articles are distinguished: one is spe-cifically anaphoric, the other covers non-anaphoric uses (cf. Schroeder 2006: 570). For example, Western Fering, a North Frisian dialect, has two articles, di and a, in-flected for number and gender: the first is traditionally described as anaphoric, the second as non-anaphoric, and it is also used with proper names (Lyons 1999: 53 f., 161 f.). Moreover, a special case is represented by those languages in which articles are limited to anaphoric use, as in Mangarryi (isolate; Northern Territory, Austra-lia; cf. Dryer 2005: 154).

From this, one could argue that the distinction between logical and pragmatic definiteness is inadequate for these languages, and worthless from a theoretical point of view. However, it seems that, at least in European languages which have a single definite article, this article can cover all instances of pragmatic definiteness, having not only anaphoric functions, whereas the same languages differ greatly in terms of logical definiteness.

Moreover, the distinction between logical and pragmatic definiteness seems to be consistent also from a diachronic point of view, if we consider that languages with definite articles tend to mark pragmatic definiteness before logical definite-ness. For instance, the use of the definite article with generics and nouns of unique entities is not possible in Gothic and Old High German (cf. Abraham 2007: 242, 250).18 Moreover, the prediction that “a language is only likely to start using the

© 2009. All rights reserved

582 Maria Napoli

article with proper nouns when it already makes extensive use of it” (Lyons 1999), is true, for instance, for Ancient Egyptian and for Ancient Greek. Loprieno (1980) observed that in Ancient Egyptian the definite article was first used in anaphoric contexts; only later was it used with proper nouns and nouns of unique entities, as in the following example (from Loprieno 1980: 11–12):

(9) p’ šw ‘the (God) Shu’, p’ jm ‘the sea’. (d’Orb. 5,7 et. al.; 10, 1 et al.)

As regards proper nouns in Ancient Greek, as already said (1.1.), the article is not obligatorily used with them. According to Manolessou & Horrocks (2007), the same state of affairs is attested in Hellenistic Greek, with not much variation (however, a detailed investigation has not yet been carried out); with regard to Me-dieval Greek, in literary vernacular texts from the 12th–15th c. the article occurs optionally with proper names, whereas it is obligatory with generics; by examining later prose texts, Manolessou & Horrocks (2007: 232) conclude that “the article had become obligatory with names [proper names, M.N.] (other than when used with adpositional modifiers, or in certain traditional formulae of legal texts, which continue to show optionality, see e.g. the acts of Manolis Varouchas 1597–1613) some time before the 15th c.”. This corresponds to a development of the logical components of the Greek article (cf. 5.2. and 6.3.).

Focusing on Ancient Greek, what is interesting is that the definite article turns out to be obligatory not only in cases of anaphoric reference, but, more in general, in all cases of pragmatic definiteness, including those instances in which the refer-ent is identifiable because of the extra-linguistic context. An example is (10.a.), where the speaker shares with the listeners the knowledge of a war in their coun-try: the noun for ‘war’ (pólemos) is used with the definite article, even though it has not been mentioned before; the same noun occurs without a definite article in (10.b.) where it is employed with a generic value (in a gnomic sentence):

(10) a. ethéloit’ an oũn, ei mēchanen heúroim’ ego, be.ready.would-2PL MOD so if plan-ACC.SG devise.could-1SG I met’ emoũ katalũsai ton pólemon? with me-GEN.SG end-INF the-DEF-ACC.SG war-ACC.SG ‘Well, if I could devise a plan to end the war, would you be ready to join me?’. (Ar., Lys. 111–112) b. […] pólemos d’ ándressi melesei war-NOM.SG PART man-DAT.PL be.the business of.shall-3SG ‘War shall be the business of menfolk’. (Ar., Lys. 520)

© 2009. All rights reserved

Aspects of definiteness in Greek 583

On the contrary, the article is optional not only for proper names but in the entire domain of logical definiteness, including generics and nouns of unique entities (cf. 1.1.),19 even though its being non-obligatory is due to different reasons. I shall try to make this point clearer after a more detailed discussion of the data.

What I want to underline now is that a shift in perspective based on these premises, i.e., the distinction between logical and pragmatic uses of the definite article, raised to a theoretical level, may help to capture the complexity of factors which determine the distribution of definite/indefinite markers in a single lan-guage. Moreover, this distinction has to be preliminary to the attempt at examin-ing those uses of the definite article traditionally considered problematic, as in the case of generics and proper nouns. It is to this task that I now turn.

4. Generics and definiteness

As mentioned above, generics are those noun phrases which refer to a class as a whole, or which are employed to convey a generic statement about a class con-sidered in its entirety, as composed of elements sharing some specific properties which distinguish them from members of other classes. In other words, in generic noun phrases, reference is made to kinds (Krifka et al. 1995: 2), rather than to in-dividuals: it must be kept in mind that, as rightly argued by Chierchia (1998: 347), “what counts as kind is not set by grammar, but by the shared knowledge of a community of speakers”.

4.1 Typological considerations

In this section, I shall introduce a couple of relevant typological considerations on generics.20 From a cross-linguistic perspective, there is no dedicated morphol-ogy for the expression of a generic meaning: this means that languages do not show grammatical markers exclusively employed to convey genericity, which is normally expressed by noun phrase types which can also be non-generic. Consid-ering that “no NPs appear to demand unequivocal reference to kinds” (Krifka et al. 1995: 6), the context allows us to distinguish between a generic and a non-generic interpretation. In certain languages, generics are ‘typically’ indefinite or bare, in others they are ‘typically’ definite (Krifka et al. 1995: 6, Lyons 1999: 179). The usage of a plural definite noun seems to be the commonest device of denoting a ‘generic’ category, at least in European languages with articles (Lyons 1999: 192, Behrens 2005: 278). Nevertheless, the relevant point is the following: languages normally use more than one type of noun phrase with a generic value. Let us concentrate on English, Modern Greek and Italian.

© 2009. All rights reserved

584 Maria Napoli

In English, genericity is expressed by bare plurals, definite singulars or in-definite singulars; bare singulars can also be interpreted as kind-referring in spe-cific contexts (cf. Stvan 2007). Definite plurals are possible, but they are rare: in particular, they are limited to some types of nouns, such as nouns of national-ity, or to particular contexts.21 In Modern Greek, generic noun phrases normally correspond to definite plurals, but definite singulars are also possible. Italian has three ways of denoting genericity: definite plurals, definite singulars or indefinite singulars. Let us consider a few examples of genericity from these languages.22 In English, the following generic sentences are equally acceptable:

(11) a. Cats are independent animals. b. The cat is an independent animal. c. A cat is an independent animal.

The equivalent of (11.a.) does not exist in Modern Greek, which uses the definite article with both generic plurals (example 12.a.) and generic singulars (example 12.b.); a sentence like (12.c.), with an indefinite article, is unacceptable:

(12) a. I ghates ine aneksartita zoa. the-DEF-NOM.PL cat-NOM.PL be-3PL independent-PL animal-PL b. I ghata ine aneksartito zoo. the-DEF-NOM.SG cat-NOM.SG be-3SG independent-SG animal-SG c. * Mia ghata ine aneksartito zoo. a-INDF-NOM.SG cat-NOM.SG be-3SG independent-SG animal-SG

In Italian, there are three possibilities: definite plural (13.a.), definite singular (13.b.), or indefinite singular (13.c.); bare plurals with a generic value are not ad-mitted as subjects, as in Modern Greek:

(13) a. I gatti sono animali indipendenti. the-DEF-M.PL cat-PL be-3PL animal-PL independent-PL b. Il gatto è un animale indipendente. the-DEF-M.SG cat-SG be-3SG a-INDF-M.SG animal-SG independent-SG c. Un gatto è un animale indipendente. a-INDF-M.SG cat-SG be-3SG a-INDF-M.SG animal-SG independent-SG

The interesting point is that such different types tend to be synonymous in an in-dividual language, but they are not completely so. More specifically, factors like the semantics of the verb, its aspectual content and its modal function, can determine which choice is more appropriate (cf. Hawkins 1978: 214 f., Chesterman 1991: 32 f., Krifka et al. 1995: 95 f., Lyons 1999: 179 f.); moreover, all these factors act as a constraint on the interpretation of the different types. A well-known example is the following:

© 2009. All rights reserved

Aspects of definiteness in Greek 585

(14) a. Dodos are extinct. b. The dodo is extinct. c. * A dodo is extinct.

The fact the (14.c.) is unacceptable depends on the semantic content of the predi-cate ‘to be extinct’, which does not denote an inherent property, but an accidental property, and it refers to the class as a unit, rather than to its individual members.

4.2 Generics in Ancient Greek

In literature on Ancient Greek, it is usually assumed that “in the generic use the employment of the explicit article was optional” (Gildersleeve 1911). As a matter of fact, Ancient Greek exhibits a full range of possibilities in denoting genericity: generics can be definite plurals, definite singulars, bare plurals or bare singulars. I will try to show that, nevertheless, the concept of ‘optional choice’ is an oversim-plification. For this purpose, I shall examine some examples from Classical Greek, by means of a contrastive analysis of different types of generics.

– Definite plural versus bare singular. In denoting genericity, the definite plural is the preferred form when the reference is to a class in its totality, with the focus on all its members. I quote from Plato:

(15) a. légousi gar depouthen pròs hēmãs tell-3PL for ADV to us-ACC.PL hoi poiētai hóti… the-DEF-NOM.PL poet-NOM.PL that ‘For the poets tell us, I believe, that…’. (Pl., Ion 534 a.7) b. koũphon gar chrema poiētes estin kaì ptēnòn kaì hierón light for thing-NOM.SG poet-NOM.SG be-3SG and winged and sacred ‘A poet is a light and winged and sacred thing’. (Pl., Ion 534 b.4)

In (15.a.), for instance, what is focused on is the importance of the class of poets (oi poietái) inside the social community, and their being inspired by gods. By contrast, in the same passage, a few words later (example 15.b.), a bare singular of the same noun — poiētes — occurs. Here, we find a predicative sentence, and the generalization is about an inherent property of the poet as an entity.

– Definite plural versus bare plural. The following examples illustrate the plural ánthrōpoi ‘men’ used as a generic subject, with the definite article (16.a.) or without any determiner (16.b. and 16.c.):

© 2009. All rights reserved

586 Maria Napoli

(16) a. ãr’ oũn ouch hamartánousin hoi ánthrōpoi then so NEG make.mistake-3PL the-DEF-NOM.PL man-NOM.PL perì toũto in this-ACC.SG ‘Do not [the] men make mistakes in this matter?’. (Pl., R. 334 c.6) b. […] hótan thúōsin ánthrōpoi theoĩs whenever sacrifice-3PL human-NOM.PL god-DAT.PL ‘[…] Whenever humans sacrifice to the gods’. (Ar., Av. 190) c. […] hósa ge dia chrēmátōn e as.many.as-ACC.PL PART because.of property-GEN.PL or paídōn kaì sungenon ktesin child-GEN.PL and kin-GEN.PL possession-ACC.SG ánthrōpoi stasiázousin man-NOM.PL quarrel-3PL ‘The dissensions that arise among men from the possession of property,

children, and kin’. (Pl., R. 464 e.1–2)

In (16.a.), which is another example of a generic definite plural, the emphasis is on the fact that all men generally make mistakes in a certain matter. On the contrary, when the verb denotes habitual or timeless aspect, a bare plural is frequently found with a kind-reading, as in (16.b.) and (16.c.).

– Definite singular versus bare singular. Given that indefinite singulars used as generics normally take verbs denoting inherent or permanent properties across languages, it is perfectly consistent that in Greek a bare singular occurs when the verb is typically stative: example (15.b.), quoted above, is a predica-tive sentence with a bare singular as a subject, and the stative verb eimí ‘to be’ as a predicate. Given that, it is not surprising to find the definite singular (ho ánthrōpos ‘the man’), rather than a bare singular, in (17.a.), where the verb (thneiskō ‘to die’) is punctual and inflected as an aorist (apothánēi). On the contrary, bare singulars are found again with stative predicates in (17.b.) and (17.c.). In (17.b.), in particular, the bare singular is the subject of two typically stative verbs, which denote physical perceptions, and are inflected as presents: horãō ‘to see’ and akoúō ‘to hear’:

(17) a. ennoeĩs oũn, éphē, hóti, epeidan observe-2SG so say-3SG that when apothánēi ho ánthrōpos… die-3SG the-DEF-NOM.SG man-NOM.SG

© 2009. All rights reserved

Aspects of definiteness in Greek 587

‘Observe — he went on — that when the man dies …’. (Pl., Phd. 80 c.2) b. ei oũn tís se hod’ erōtoiē: ‘toi if then anyone you-ACC.SG so ask.should-3SG by.what ta leuka kaì mélana horãi ánthrōpos the-DEF-ACC.PL white-ACC.PL and black-ACC.PL see-3SG man-NOM.SG kaì toi ta oxéa kaì baréa akoúei?’ and by.what the-DEF-ACC.PL high-ACC.PL and low-ACC.PL hear-3SG ‘If, then, anyone should ask you, ‘by what does a man see white and

black colours and by what does he hear high and low tones?’ ’. (Pl., Tht. 184 b.8–9) c. chre gar poiēten ándra/ be.necessary-IMPS for poet-ACC.SG man-ACC.SG pròs ta drámata in.reference.to the-DEF-ACC.PL play-ACC.PL ha deĩ poieĩn pròs/ that-ACC.PL there.is.need-IMPS do-INF in.reference.to taũta tous trópous échein this-ACC.PL the-DEF-ACC.PL behaviour-ACC.PL have-INF ‘to be a poet, a man must suit his behaviour to the requirements of his plays’. (Ar., Th. 149–150)

It is worth noting that a similar distribution, depending on semantic, aspectual or modal properties of the verb, can be recognized in the use of the article with nouns of unique entities. In example (3.a.), quoted at the beginning of this paper, the noun ‘moon’, denoting a unique entity, occurs with the definite article: the sen-tence depicts a specific event in the past, the verb is punctual, as in (16.a.), and it is inflected as an imperfect. On the contrary, the same noun is bare in (3.b.), where we have an if-sentence, denoting a hypothetical change in the habit of the moon.

To sum up, as these few examples show, the fact that the Greek article is ‘op-tional’ with generics is due to the co-existence of different strategies in denoting genericity, which is perfectly consistent with the behaviour of modern languages. In Modern Greek, where the use of bare noun phrases is quite restricted, the defi-nite article becomes obligatory in generic noun-phrases. Ancient Greek admits definite noun phrases or bare noun phrases with a generic value, like English, for instance. More precisely, as seen above, Ancient Greek exhibits four types of ge-nerics: although these four types are similar, they are not identical. A more fine-grained analysis is necessary, which takes into account similarities and differences among them. This point needs further research (cf. Section 6).

© 2009. All rights reserved

588 Maria Napoli

5. Proper nouns of individuals and definiteness

I shall now concentrate on the use of the definite article with proper nouns of individuals. Proper nouns presuppose uniqueness of reference; as a consequence, they have been viewed as ‘definite descriptions’ by themselves. Let me quote from Leonardi (2007: 20): “Proper names, definite descriptions, quantifiers give inverse information about the discourse domain. I speak of Roberto — not an uncommon name nowadays — and I do not specify, I do not say ‘Roberto Rossi’, I do not say ‘Roberto, the student who graduated last year with you’. This shows that I take that in the foreground domain there is only one Roberto — otherwise, I would speak of Roberto Rossi, Roberto R., your Roberto, etc. The same is true of definite descrip-tions. If I say ‘The bottle is empty’ — there are millions, perhaps billions of bottles right now in Italy, and the description in the largest domain is surely improper, perhaps it is improper already in the room we are. But most of the times there is no problem, because we are concerned with what’s in our visual field, where there is the empty wine bottle on our table”.

Since Frege (1892), the issue of proper nouns has been extensively discussed in the field of philosophy, in cognitive psychology and, to a lesser extent, in linguis-tics. In particular, scholars disagree in interpreting the value of the definite article with these nouns. I shall come back to this point in Section 6.

5.1 Proper nouns across languages: A short survey

Examining the use of the definite article with proper nouns from a cross-linguistic point of view, we must distinguish between two major groups of languages:

1. Languages which do not admit a definite marker with proper nouns: for ex., Bulgarian and Rumanian, which have a postposed article, and English (apart from some specific contexts: I shall return to this point).

2. Languages which admit a definite marker with proper nouns. In the last case, three subtypes can be identified:

a. Languages which obligatorily use a definite marker with proper nouns; for example, Modern Greek, Albanian, Turkish (only in the accusative case), Modern Western Armenian (in the accusative, dative and ablative cases). Here are just a couple of examples, one from Modern Greek (Mackridge 1985: 198), one from Turkish (Lewis 1967: 35–36), where the accusative of a proper noun takes a marker of definiteness:23

(18) erthe ho Gerásimos / ho kúrios Ponērídēs come-3SG the-DEF-NOM.SG Gerasimos / the-DEF-NOM.SG mister Poniridis ‘Gerasimos/Mr Poniridis has come’.

© 2009. All rights reserved

Aspects of definiteness in Greek 589

(19) Hasan’-ı hemen tanıdım Hasan-ACC-DEF immediately recognize ‘I recognized Hasan immediately’.

b. Languages with a special article for proper nouns. Corbett (2000: 278–279) cites the case of Kambera, spoken in Eastern Indonesia, where three ar-ticles are employed: one is singular and definite; one is plural and definite; one is for proper nouns. Catalan belongs to this group too (from Colo-mina 2002: 548):24

(20) en Jaume, na Maria, n’Antoni, n’Aina, n’Emília.

c. Languages which optionally use a definite marker with proper nouns, like spoken varieties of German and Italian. In Italian, the use of the definite article with a proper noun is optionally found in some dialectal varieties, and is considered as colloquial: as example (21.a.) shows, Maria is possible as well as La Maria; on the contrary, the definite article is obligatory with the surname of female celebrities (21.b.):

(21) a. Maria vs. La Maria. Maria the-DEF-F.SG Maria b. * Loren vs. La Loren. Loren the-DEF-F.SG Loren

Now, the question is: how can we account for the use of the definite article with proper nouns? This use has generally been viewed as a puzzling problem: if proper nouns are definite by themselves, as traditionally assumed in literature, what is the function performed by the definite article in such a case? I shall return to this point after having considered some examples from Ancient Greek.

5.2 Proper nouns in Ancient Greek

As already said (cf. 1.1., 3.2.), in Ancient Greek, the use of the definite article with proper nouns is far from being compulsory, even though it turns out to be quite widespread. The traditional interpretation may be summarized by a quotation from Gildersleeve (1911): “Proper names being in their nature particular do not require the explicit article, and when the article is used with them, it retains much of its original demonstrative force”. A similar assumption is found in Humbert (1960): “En principe, l’article n’est jamais nécessaire avec un nom propre, puisque celui-ci, en raison de son caractère singulier, ne peut être confondu avec un autre. Cependant l’article est fréquemment employé, pour des raisons subjectives fort va-riées. D’une façon générale, la présence de l’article indique que la personne (ou

© 2009. All rights reserved

590 Maria Napoli

la chose) désignée par le nom propre est familière à celui qui parle ou l’intéresse particulièrement [italics original]” (Humbert 1960: 46–47). Both these scholars noted that there is a remarkable variation depending on subjective and stylistic reasons.25 Clearly enough, authors vary: in Plato, for instance, the definite article is regularly employed with a proper noun in introducing the speaker of a direct speech, as in (22):

(22) pantápasin, éphē ho Kébēs, ADV say-3SG the-DEF-NOM.SG Cebes-NOM.SG hoútō phaínetaí moi ADV seem-IMPS me-DAT.SG ‘That — said Cebes — seems to me quite evident’. (Pl., Phd. 103 a.3)

As a matter of fact, there are some special cases in which the Greek article is fre-quently found with proper nouns, and in which it can be regarded as having a discourse-pragmatic function.

A typical instance of the use of the definite article with proper nouns, as point-ed out by Humbert (1960), is represented by those cases in which the notion of ‘familiarity’ is involved, as in (23), where the speaker is familiar to the listener:

(23) ton Íōna chaírein the-DEF-ACC.SG Ion-ACC.SG welcome-INF ‘Welcome, Ion’. (Pl., Ion 530 a.1)

There are some cases in which special emphasis is put on the individual named, because of his/her being generally known — examples (24.a.) and (24.b.) — or because of his/her prominence within the text (examples 25.a. and 25.b.):

(24) a. ho dè stólos noin esti/ the-DEF-NOM.SG PART mission-NOM.SG now be-3SG para ton Tēréa beside the-DEF-ACC.SG Tereus-ACC.SG tòn épopa […] the-DEF-ACC.SG Hoopoe-ACC.SG ‘Our mission now is to visit Tereus the Hoopoe’. (Ar., Av. 46–47) b. kinduneúei, o áriste, eikótōs ho Perikles may.be-3SG oh best-VOC reasonably the-DEF-NOM.SG Pericles-NOM.SG pántōn teleotatos eis ten every-GEN.PL the.most.perfect-SUP-NOM.SG in.regard.to the-DEF-ACC.SG

© 2009. All rights reserved

Aspects of definiteness in Greek 591

rhētoriken genésthai rhetoric-ACC.SG be-INF ‘I suppose, my friend, Pericles is the most perfect orator in existence’. (Pl., Phdr. 269 e.1–2)

(25) a. tís an phráseie poũ ’stin hē Lusistrátē? who MOD tell.can-3SG where be-3SG the-DEF-NOM.SG Lysistrata-NOM.SG ‘Who can tell us where Lysistrata is?’ (Ar., Lys. 1086) b. – […] ho gar Euripídēs the-DEF-NOM.SG for Euripides-NOM.SG kat’ épos basanieĩn phēsi tas tragōidías. PREP word-ACC.SG examine-INF say-3SG the-DEF-ACC.PL tragedy-ACC.PL – e pou baréōs oĩmai ton Aischúlon ADV ADV heavily guess-1SG the-DEF-ACC.SG Aeschylus-ACC.SG phérein bear-INF – ‘[…] because Euripides says he’s going to examine the tragedies word for word’. – ‘I’d guess that Aeschylus is pretty sore about that’. (Ar., Ra. 801–803)

In (25.a.), Lysistrata appears to be the key-character in order to arrive at the happy ending of the homonymous play; in (25.b.), the two poets (Euripides and Aeschy-lus) are rivals and involved in a competition concerning their poetic techniques (so, we can say the articles have a contrastive force).

Sometimes, the definite article is found with a deictic function also with prop-er nouns. An example is (26):

(26) - […] límnē ne Día lake-NOM.SG PART Zeus-VOC.SG haútē ‘stìn hen éphraze, kaì ploĩón g’ horo. this-NOM.SG be-3SG that-ACC.SG tell-3SG and boat-ACC.SG PART see-1SG – Ne tòn Poseido kásti/ PART the-DEF-ACC.SG Poseidon-ACC.SG and-be-3SG g’ ho Chárōn houtosí PART the-DEF-NOM.SG Charon-NOM.SG this-here – ‘It’s the very lake that he told us about, and I see a boat too’. – ‘Yes, by Poseidon, and that’s Charon himself ’. (Ar., Ra. 181–183)

© 2009. All rights reserved

592 Maria Napoli

In (26), the definite article occurs with the proper noun (ho Chárōn) plus the deic-tic form houtosí (‘this man there’), with the reference to an entity physically given in the surrounding perceptual space, as is proper of deictic elements.

However, there is another interesting case to which no sufficient attention has been paid: the anaphoric use of the article with proper nouns. When a proper noun is mentioned by a certain speaker for the first time, it is generally bare; how-ever, when the same speaker mentions it again, the definite article is often found;26 in subsequent mentions, the article can be omitted, especially if it occurs with a preposition. Let us consider a case from Aristophanes:

(27) a. hoútōs en neanískos Melaníōn tis ADV be-3SG young.man-NOM.SG Melanion-NOM.SG one-NOM.SG ‘In olden times lived a young man named Melanion’. (Ar., Lys. 784/785) b. hoútō tas gunaĩkas ebdelúchthē ‘keĩnos, hēmeĩs t’ ADV the-DEF-ACC.PL woman-ACC.PL loathe-3SG he we and oudèn hetton toũ Melaníōnos hoi sophrones NEG less the-DEF-GEN.SG Melanion-GEN.SG the-DEF-NOM.PL wise ‘That’s how much he loathed women. And, being wise, we loathe them just as much as Melanion did’. (Ar., Lys. 795–796)

In (27), the chorus tells Melanion’s story, and introduces his name by using the in-definite pronoun tis ‘one’ (27.a.), which “commonly signifies one named” (GREL, s.v.) when it occurs with proper names. The name is pronounced for the second time as in (27.b.), plus the definite article.

The anaphoric use of the Greek definite article with proper nouns is particu-larly interesting if we consider that, as pointed out by Diessel (1999), also in lan-guages with articles, demonstratives are often the preferred form when a referent is mentioned for the second time and is established as the topic of the discourse (see Abraham 2007a, 2007b, Leiss 2007, who explain this phenomenon in a con-vincing manner).27

In Ancient Greek, the use of the definite article with proper nouns is closely related to the second time in which the noun is mentioned, mainly if the referent of the proper noun is a central topic in the discourse; in all subsequent mentions, the article can be omitted (and actually it is, as a tendency). This could reflect the demonstrative force of the definite article in this stage of the history of the Greek language. Let me quote one more example from Aristophanes (28), and one ex-ample from Plato (29):

(28) a. oúkoun Oréstēs toũt’ epì toi túmbōi légei not.then Orestes-NOM.SG this at the-DEF-DAT.SG tomb-DAT.SG say-3SG

© 2009. All rights reserved

Aspects of definiteness in Greek 593

toi toũ patròs tethneotos? the-DEF-DAT.SG the-DEF-GEN.SG father-GEN.SG dead-PTCP-GEN.SG ‘Now doesn’t Orestes say this at the tomb of his dead father?’. (Ar., Ra. 1139–1140) b. oú phēmi ton Oréstēn kateltheĩn oíkade NEG say-1SG the-DEF-ACC.SG Orestes-ACC.SG come-INF house.to ‘I deny that Orestes was coming home’. (Ar., Ra. 1167)

(29) a. – o phíle Phaĩdre, poĩ de kaì póthen? oh dear-VOC.SG Phaedrus-VOC.SG whither PART and from.where – para Lusíou, o Sokrates, toũ Kephálou from Lysias-GEN.SG oh Socrates-VOC.SG the-DEF-GEN.SG Cephalus-

GEN.SG - ‘Dear Phaedrus, whither away, and where do you come from?’ - ‘From Lysias, Socrates, the son of Cephalus’. (Pl., Phdr. 227 a.1) b. gégraphe gar de ho Lusías represent-3SG for PART the-DEF-NOM.SG Lysias-NOM.SG peiromenón tina ton kalon tempt-PTCP-ACC.SG one-ACC.SG the-DEF-GEN.PL beauty-GEN.PL ‘For Lysias has represented one of the beauties being tempted’. (Pl., Phdr. 227 c.5–6)

The sentences in (28.a.) and (29.a.) represent the first mention of Orestes’ and Lysias’ names, respectively, whereas the sentences in (28.b.) and (29.b.) represent their second mention by the same speakers; only in the second mention does the definite article occur with these nouns. More in general, if the text has a dialogical form, as in (28) and (29), it is possible to note that the mention of a proper noun plus the definite article is located in the text at a certain distance from its first mention without the definite article. In this respect, the following case, taken from Plato, is quite interesting:

(30) a. kaì oĩmai kállista anthropōn légein perì and consider-1SG nice-SUP-ACC.PL man-GEN.PL say-INF on Homerou […], Homer-GEN.SG oúte állos oudeìs ton popote genoménōn NEG another nobody the-DEF-GEN.PL ever.yet exist-PTCP-GEN.PL éschen eipeĩn hoútō pollas kaì kalas dianoías have-3SG say-INF ADV many-ACC.PL and fine-ACC.PL comment-ACC.PL

© 2009. All rights reserved

594 Maria Napoli

perì Homerou, hósas ego on Homer-GEN.SG as.much.as-ACC.PL I ‘And I consider I speak about Homer better than anybody […], nor any one that the world has ever seen, had so many and such fine comments to offer on Homer as I have’. (Pl., Ion 530 c.8-d.3) b. kaì men áxión ge akoũsai, o Sokrates, and indeed worth PART hear-INF oh Socrates-VOC.SG hōs eũ kekósmēka ton Hómēron how well embellish-1SG the-DEF-ACC.SG Homer-ACC.SG ‘And indeed it is worth hearing, Socrates, how well I have embellished Homer’. (Pl., Ion 530 d.6–7)

In this dialogue, the speaker (Ion) mentions Homer’s name for the first time as in (30.a.), where it is pronounced without any determiner twice, at the beginning and at the end of his discourse; after a comment by Socrates, Ion cites Homer again as in (30.b.), with the article. The significant point is that in Ion, Homer’s name oc-curs 47 times, but (30.b.) represents its only occurrence with the definite article. Thus, what is relevant to the use of the definite article is not necessarily the ‘second mention’ of a proper noun (if this mention is close to the first), but the ‘second context’ in which such a name is uttered, mainly if this mention is important for establishing the referent as the topic of the discourse. To conclude, the fact that the article is ‘optional’ with proper nouns in Ancient Greek clearly reflects its not being a marker of logical definiteness, as opposed to Modern Greek, where the article actually is a marker of logical definiteness. In this stage of its diachronic development, the article simply covers pragmatic functions with proper nouns.

6. A tentative assessment of the contribution of Greek

At this point, we should ask: what is the contribution of Greek to the problem of the relationship between definiteness, proper nouns and generics? In my opinion, this is exactly the topic Ancient Greek could provide a new insight for, especially as regards proper nouns.

6.1 Which explanation for proper nouns and generics?

Before trying to explain why, I shall briefly examine two proposals, which are dif-ferent, but at the same time have something in common: both tried to provide a

© 2009. All rights reserved

Aspects of definiteness in Greek 595

unified explanation for the behaviour of proper nouns and generics in terms of definiteness.

The first proposal, which has a long tradition, consists in regarding the (pos-sible) occurrence of the definite article with both proper nouns and generics as ‘redundant’ or ‘expletive’, i.e., as semantically empty. I shall quote from the recent book by Alexiadou, Haegeman & Stavrou (2007: 186): “The important fact about the (definite) article accompanying proper (and generic) nouns is that the defi-niteness conveyed by this article can be said to be redundant. […] Proper names denote a single/unique individual — they form a priori and by definition single-member sets. Their use, in other words is ‘rigidly designating’. This makes the pres-ence of the definite article with a proper name redundant in a sense, since proper nouns by themselves seem to fulfill the function that the definite article serves with common nouns. Such a redundant use of the definite article in front of proper nouns has been called an ‘expletive’ use (Vergnaud & Zubizarreta 1992; Longo-bardi 1994) because it resembles the redundant occurrence of the expletive subject with certain verbs (weather subjects, such as it in it rains for instance). Like it in it rains, the Greek definite article o in o Janis (‘the John’) is devoid of any semantic content”. Here, I will not deal with the individual syntactic analysis of one author or the other. Nevertheless, it is worth mentioning that this topic has often been discussed in the generative perspective, and that scholars disagree in interpreting the distribution and the use of this ‘meaningless’ article.28

On the other hand, Lyons (1999) explicitly rejected the characterization of the definite article as ‘redundant’ or ‘expletive’ in some of its uses, on the basis of his view that articles are always a grammatical device for encoding definiteness. Consequently, he put forward a different hypothesis, interpreting proper nouns as a kind of generic:29 “Proper nouns are a kind of generic; they do denote ensembles (or better in this case, kinds), but always generically; because these ‘ensembles’ consist of only one entity. […] This analysis would lead us to expect proper nouns to have the same form as other generics. And this expectation is borne out for many languages, including English, where the typical generic is the bare noun or nominal (plural if count), and where proper nouns too are bare” (Lyons 1999: 194; 197). The author concluded that proper nouns and generics have to be considered as typically definite in languages like Catalan and Modern Greek, where they are both definite, and as typically indefinite in a language like English where they are both bare.

6.2 Proper nouns and generics as distinct classes

First of all, I will briefly comment on Lyons’ (1999) proposal cited above. In order to do that, let me introduce an important remark. As a matter of fact, there is an

© 2009. All rights reserved

596 Maria Napoli

‘ontological’ difference between proper and common nouns, which cannot be ig-nored, especially when examining definiteness; I quote from Prandi (2004: 129): “Both the distribution of determiners and their interaction with the different kinds of nouns prove that the distinction between countable individuals, classes and masses is both ontologically relevant and strictly formal”. The phenomenon of conversion of proper nouns into common nouns is possible only by adopting specific strategies: for instance, by adding an indefinite determiner or by inflecting them as plurals. A typical case which can be regarded as involving conversion of proper nouns into common nouns is the so-called metaphorical use: in this case, proper nouns are regarded “as prototypes of a specific property” (Bazzanella & Morra 2007: 67). This means that proper nouns are assumed to be the paradigm of a special class, as in the following example, from Aristophanes:

(31) ego gar oĩda taítion. Mían gar/ I for know the-reason-DEF-ACC.SG one-ACC.SG for ouk an eípois NEG MOD cite.can-2SG ton nũn gunaikon Pēnelópēn,/ the-DEF-GEN.PL now woman-GEN.PL Penelope-ACC.SG Phaídras d’ hapaxapásas Phaedra-ACC.PL PART all.together-ACC.PL ‘Well, I can tell you why: you can’t cite me a single Penelope among all the women now alive; absolutely all of us are Phaedras’. (Ar., Th. 549–550)

Penelope is the paradigm of the class of virtuous women, who are rare, Phaedra is the paradigm of the class of sinful women, who are supposed to be the majority: the first proper noun is accompanied by the numeral ‘one’, the second is inflected as a plural.

Thus, proper nouns are ontologically different from common nouns, even though they can be converted into common nouns themselves. This makes Lyons’ assumption that proper nouns are a kind of generic semantically unconvincing. Moreover, this proposal does not help us to explain the distribution of definite markers with these nouns. At a formal level, as said above (cf. 4), in each language genericity is normally conveyed by different noun phrase types, even though one of them tends to have an unmarked status. In English, bare plurals are the un-marked generics, according to Lyons (1999), but a singular definite noun phrase is also possible as kind-referring: the two types seem to differ in terms of frequency, but they are equally acceptable. In other words, generics can be definite or in-definite in English, whereas proper nouns are always bare. The only exception is represented by cases of conversion. More specifically, in a language like English,

© 2009. All rights reserved

Aspects of definiteness in Greek 597

this conversion is possible by adopting those determiners, like the definite article, which are not normally employed with proper nouns. I quote from Anderson (2007: 194): “personal names in English, prototypical names, reject the definite article unless in the particular context identification is not assured, […], or there is reference to (especially diachronic) instances or aspects of the entity whose name is being used”. The first case is exemplified by (32.a.), the second case by (32.b.), both taken from Anderson (2007: 194):

(32) a. The Bill with red hair. b. The young Byron, the France I’m fond of.

In instances like (32.a.) and (32.b.), as already noted by Lyons (1999), the proper noun is ‘re-categorized’ by adding the definite article plus some descriptive mate-rial.

In this respect, I will quote an example that I found in a recent novel by the American writer Paul Auster:

(33) a. – “And what name is that?” – “Joyce” – “Joyce?”. I paused for a moment in a kind of addled wonder. “Are you

telling me you’re married to a man named James Joyce?” – “Uh-huh. Just like the writer”. b. “I was introduced to her children (Devon and Sam); her mother Joyce;

and her Foley-walker husband, Jim, the James Joyce who was not James Joyce”

(Paul Auster, The Brooklyn Follies)

In the dialogue quoted above (33.a.), the first-person narrator talks with a woman married to a man who has the same name as the famous Irish writer James Joyce. When the narrator, later, speaks of his meeting with him, he mentions his nick-name Jim, adding what is given in (33.b.): the James Joyce who was not James Joyce. In this case, the article has a clearly identifying function; it serves to distinguish be-tween two men who have the same name: only one is James Joyce, the well-known and easily identifiable writer with whom everybody would associate that specific name and surname. Moreover, this is the reason why the identity of Jim cannot be simply established by mentioning his name and surname: his proper name is not a uniquely-referring expression, because he is not the famous Irish writer to whom this proper name is normally associated. The expression the James Joyce who was not James Joyce is a definite description which re-establishes Jim’s identity by specifying that he is not the famous Irish writer. Cases like this are the only ones in which the English definite article is admitted with proper nouns of individuals. Interestingly enough, when an English surname is used to denote the members of

© 2009. All rights reserved

598 Maria Napoli

a specific family, it is inflected as plural and, as opposed to generic noun-phrases, normally bare in the plural number, it requires the article (The Simpsons, to quote just an example). This confirms the different distribution of generics and proper nouns in English.

To conclude, my impression is that, in terms of definiteness, the case of ge-nerics has to be considered apart. The point about proper nouns is that they do require or do not require a definite marker in a single language. By contrast, the point about nouns used as generics concerns the types of determiners (normally, more than one type) that they admit with this value. Moreover, Lyons’ (1999) pro-posal does not hold from a cross-linguistic perspective: Lyons (1999) himself rec-ognizes that the formal correspondence between generics and proper nouns does not hold absolutely in some languages, as in French, where generics are typically definite but proper nouns are typically bare.

The expression of genericity, which is strictly language-specific, appears to depend on the extent to which an individual language uses its definite and/or indefinite articles. For instance, the usage of bare plurals with a generic value in English has to be analysed in the light of the fact that this language uses the bare plural form more extensively than a language like Italian: as is well-known, English completely lacks an indefinite marker in the plural, whereas Italian uses the partitive form, preserving the distinction between definite and indefinite. As seen in Section 4.1., an indefinite noun phrase is possible with a generic value in various languages, like English and Italian, but also French and German (cf. Lyons 1999: 185); on the contrary, Modern Greek does not admit this use. This could depend on the fact that this language has developed an indefinite article which is normally employed as a ‘referential indefinite’ form (Schroeder 2006: 598): if this article denotes a referent which is indefinite and specific, it is clearly incompatible with a generic meaning, non-referential and non-specific by definition. This point obviously deserves further research.

6.3 Points of view: Against ‘redundancy’

In this section, I shall concentrate on the notion of ‘redundancy’ as referred to the use of the definite article with proper nouns (cf. 6.1.). In my opinion, there are two main counter-arguments against this view.

First of all, this proposal is too rigid to capture the whole of the cross-linguistic situation. This is what the diachronic data from Ancient Greek show, and it is cor-roborated by evidence from other languages. It is worth quoting an example from the Tuscan dialect: in a variety spoken in Pisa, the definite article with feminine proper nouns is used if the referent is not simply ‘familiar’ to the speaker, but known to him/her from childhood.30 Another significant case concerns anaphora:

© 2009. All rights reserved

Aspects of definiteness in Greek 599

anaphoric uses of the definite article with proper nouns are not only attested in Greek. Frajzyngier (1991) quoted the case of Mupun, where a definite marker with an anaphoric function occurs after any noun, including proper nouns. What I mean is that the discourse-pragmatic function of definite articles with proper nouns can be taken as a counter-argument against the interpretation of the defi-nite article as devoid of meaning in its occurrence with such nouns.

On the other hand, considering, more specifically, the contribution of Modern Greek, an important and general point is in order here. The notion of the ‘redun-dancy’ of the definite article with proper nouns is generally invoked in order to explain the fact that some languages lack it with these nouns.31 Nevertheless, it is only from the point of view of a language like English that one would assume that proper nouns must occur without a definite article. If we analyse these data from the point of view of, for instance, Modern Greek, we should obviously expect them to be used with a definite article. In other words, in Modern Greek, the use of the definite article is obligatory with proper nouns, as well as in other contexts where a marker of definiteness is required.

So, it is probably nonsense, from both a diachronic and a synchronic point of view, to assert that the definite article is ‘redundant’ with proper nouns. From a diachronic perspective, Ancient Greek provides evidence for a definite article which is far from being semantically empty with proper nouns. The use of the article with these nouns, which in origin was linked to the expression of pragmatic functions, and limited to some specific contexts, is generalized and becomes a grammatical marker of definiteness in Modern Greek. As Friedrich Newmeyer has pointed out, “functional linguists and generative linguists with a functional bent have provided (to my mind) incontrovertible evidence that grammars are shaped in part by performance considerations” (Newmeyer 2003: 683). In my opinion, the development of the definite article from Ancient to Modern Greek, and its becom-ing an obligatory marker of proper nouns, can be taken as a significant example of fuctional factors shaping grammar, to use Newmeyer’s expression. In other words, what can be said about Greek is that the use of the definite article with proper nouns, which in Ancient Greek spread from contexts of pragmatic definiteness, becomes a matter of logical definiteness in Modern Greek, i.e. it becomes an obliga-tory marker of the inherent definiteness of such nouns. This is perfectly consistent with the well-known phenomenon that “grammatical conventions are explained by change” (Moravcsik 2006: 284). The fact that this development has not taken place in English, for instance, does not allow us to speak of a ‘redundant’ article, different from a meaningful one. It is not clear, in particular, why the definite arti-cle should be redundant with proper nouns and meaningful with nouns of unique entities. In some languages, proper nouns and nouns of unique entities have the same behaviour in terms of definiteness: for instance, in Bulgarian and Rumanian,

© 2009. All rights reserved

600 Maria Napoli

proper nouns do not take a definite article; however, “sometimes also another type of nonanaphoric reference, namely, unique reference, may be indicated by the non use of the article, most in prepositional phrases and typically with nouns which refer to unique institutions or natural phenomena” (Schroeder 2006: 598). This correspondence is true also from a diachronic point of view, as noted in the case of Greek and Ancient Egyptian (cf. 3.2.).

Before concluding, a remark is in order here: could we accept that Lyons (1999) is right in claiming that “definiteness marking is obviously not essential to communication” (Lyons 1999: 48, note 48)?32 In my opinion, the crucial point is that, usually, definite articles do not really add information independently of the semantics of nouns, even independently of the context (linguistic or extralinguis-tic). If ‘non-ambiguity’ of reference is the crucial feature of definiteness, as Chris-tophersen (1939) and Löbner (1995)33 believe, it must be taken into consideration that what determines the non-ambiguity of the referent is the noun with which the article occurs or the context in which it is uttered. The definite article is simply a linguistic device to make this ‘non-ambiguity’ explicit. I will try to make this point clearer by means of an example. If I am in a library with a friend of mine, and I say to him ‘give me the book, please’, I obviously mean ‘the book you know’, or ‘the book I previously mentioned’, or ‘the book that is on the desk, in front of us’; other-wise, this sentence would be absolutely nonsense. On the other hand, if I say ‘give me a book, please’, I obviously mean any book (a non-specific one). If I spoke a lan-guage without articles, in the same situation I could probably use a demonstrative in the first case (in order to make the referent non-ambiguous), an indefinite pro-noun in the second case. But if I meant a specific book, and I were sure the hearer could identify it (because we have just mentioned it, for instance), I could simply use the expression ‘*give me book’ (without any determiner). Similarly, provided that the definite article “acts as a form of instruction” (Hawkins 1978: 113), in a natural language with such an article, and in the most of the ‘natural’ conversa-tions we have, in principle this instruction could be absolutely unnecessary (with regard to cases of logical as well as pragmatic definiteness), because the evidence for identifying the referent is provided, as said above, by the semantic content of the noun, by the situation or by the linguistic context.

In such a sense, we could say that definiteness is not essential to communi-cation. Nevertheless, when it is grammaticalized in a specific language, and ex-pressed by means of articles, it is essential to choice between a definite and a non-definite marker, which could be a zero mark (as in Ancient Greek or, with regard to the plural number, in English). In other words, the fact that the article is present in a given language as a grammatical device, always leads us to a choice between definite or indefinite markers, even though the context is non-ambiguous by itself. In this sense, we always need definiteness.

© 2009. All rights reserved

Aspects of definiteness in Greek 601

7. Conclusions

It is time to sum up and provide some conclusions. In the present paper, I have attempted to sketch an outline of some problems related to the debate on definite-ness. One of my major aims was to show that we should probably give up the idea of a characterization of definiteness based on a single concept or property, such as uniqueness, inclusiveness, or identifiability. Indeed, this is what proper nouns and generics suggest: we cannot treat ‘definiteness’ as a unitary phenomenon, but we need to subdivide it into specific cases. In particular, we should start from a preliminary distinction between two types of definiteness, i.e. between those cases in which definiteness is pragmatically inferred from the context (because of ana-phora or deixis, for instance), and those cases in which definiteness is “logically” inferred, because of the particular semantics of nouns (as for proper nouns and nouns like the sun, the moon, denoting unique entities). This clear-cut distinction, which has already been noted, needs to be raised to a theoretical level: it can be formulated as a distinction between pragmatic definiteness and logical definite-ness. This distinction is consistent also diachronically: as already said, languages with an article system seem to mark pragmatic definiteness before logical definite-ness (as Greek or Egyptian did).

In a canonical system (on the basis of Corbett 2005), if a language has a dedi-cated marker for definiteness, we should expect this marker to be applied to all definite noun-phrases, regardless of whether their definiteness is a matter of se-mantics or pragmatics; likewise, if a language has markers for number, codify-ing the singular/plural opposition, we should expect it to apply such markers to all kinds of singular/plural noun phrases. Nevertheless, as is well-known, this is not the case: languages have their own idiosyncrasies and particularities in using grammatical markers. The next task could be to investigate more in depth why languages differ so greatly in the field of definiteness, and, in particular, of logical definiteness: why some languages use markers for definiteness to the maximum extent, including, for instance, proper nouns and nouns of unique entities, and others do not.

On this topic, many issues should be dealt with which have not been treated so far, and more research is obviously needed. However, in my opinion, new insights into the field of definiteness are possible if we analyze it from the perspectives I have tried to illustrate here.

© 2009. All rights reserved

602 Maria Napoli

Acknowledgements

My investigation on definiteness began during my stay at the Department of Linguistics and English Language, at the University of Lancaster (October-December 2006): I wish to thank Anna Siewierska, the head of the Department, for providing a very stimulating research envi-ronment. I presented some preliminary results of this research at the University of Oslo (Greek and Latin from an Indo-European Perspective 2, July 5–6/2007), and at the University of Cam-bridge, as an invited speaker at the Linguistics Seminar (November 21/2007, Faculty of Clas-sics): I am grateful to both audiences for their questions and observations, and, in particular, to James Clackson, Andrew Garrett, Dag Haug and Geoffrey Horrocks. I am indebted to Eystein Dahl, Romano Lazzeroni and Carlotta Viti, who read draft versions of this paper, and provided some useful remarks. Special thanks are due to two anonymous reviewers for their helpful and thoughtful comments.

Appendix: List of standard abbreviations

1 = first person M = masculine2 = second person MOD = modal particle marker3 = third person NEG = negationACC = accusative NOM = nominativeADV = adverb(ial) PART = particleDAT = dative PL = pluralDEF = definite PREP = prepositionF = feminine PTCP = participleGEN = genitive SG = singularIMPS = impersonal form SUP = superlativeINDF = indefinite VOC = vocativeINF = infinitive

Notes

1. In the present paper, the label ‘Ancient Greek’ is generically used (including Classical Greek), as opposed to the label ‘Modern Greek’. Greek texts and translations of the examples quoted here are taken from the editions cited in the bibliography. In general, the quotations from Ancient Greek are not literally translated. On the other hand, the meaning of individual words is given in interlinear glosses, on the basis of Leipzig Glossing Rules (http://eva.mpg.de/lingua/files/mor-pheme.html).

2. Cf. Gildersleeve (1890, 1911), Schwyzer & Debrunner (1950), Chantraine (1953), Humbert (1960), Morpurgo Davies (1968), and, more recently, Sansone (1993), Lombardi-Vallauri (2002), Basset (2006), Bauer (2007), Manolessou & Horrocks (2007).

© 2009. All rights reserved

Aspects of definiteness in Greek 603

3. It must be noted that “since the indefinite article hénas is identical to the numeral ‘one’, it is not always possible to decide in which function the word is being used. It is nevertheless true to say that the meaning ‘one’ is often presented even when the word is used as an article; and where the notion of singularity is not being emphasized the article is often omitted” (Mackridge 1985: 200).

4. “On a pu supposer qu’a l’époque d’Homère, la langue courante connaissait déjà l’article, mais que l’épopée conservait traditionnellement l’emploi démonstratif de l’article” (Chantraine 1953: 165). By contrast, Mycenean Greek does not document the use of the demonstrative pro-noun as a definite article (cf. Bartoněk 2003).

5. Among the most significant contributions, see Christophersen (1939), Greenberg (1978), Diessel (1999), Harris (1980), Lyons (1999), Himmelmann (1997, 2001), Putzu (2001), Heine (2003), Schroeder (2006). Cf., in particular, Schroeder (2006: 545): “it seems that in the process of the emergence of definite articles from demonstratives as well as in the anaphoric use of de-monstrative no clear-cut distinction can be drawn between deixis and anaphora. The emergence of articles from demonstratives starts with ‘pointing back within texts’, and the anaphoric use of demonstratives is best described as a continuum from deixis to anaphora”.

6. For a new approach to the anaphoric function of demonstratives and articles, the reader is referred to Abraham (2007a, 2007b).

7. It is worth mentioning Abraham’s (2007b: 250 f.) discussion on the emergence of the definite article from Old High and Middle High German up to Modern Standard German: the author identifies different diachronic stages, by an insightful analysis of the discourse-functional prop-erties of the demonstrative and the article.

8. On Latin see Pinkster (1990), Lehmann (1991), Rosén (1991), Bolkestein (1996, 1998), Bauer (2007). References on Russian can be found in Trenkić (2002). The narrative categories of fore-grounding vs. backgrounding can influence word order, as pointed out by Leiss (2000), and by Abraham & Leiss (2007) with regard to Old Scandinavian.

9. Cf. Lambrecht (1994: 90): “the common cognitive property which unites all instances of iden-tifiability and therefore justifies expression by a single grammatical category, is the existence of a cognitive SCHEMA or FRAME within which the referent can be identified”. He distinguishes between the cognitive category of ‘identifiability’ and its formal expression by means of arti-cles, and he recognizes that “there is no one-to-one correlation between identifiability or non-identifiability of a referent and grammatical definiteness or indefiniteness of the noun phrase designating that referent. Obvious evidence for this lack of correspondence is found in the fact that the use of the definite and the indefinite article varies widely from language to language, in idiosyncratic and sometimes quite subtle ways, while the mental ability to identify referents is presumably the same for speakers of all languages” (Lambrecht 1994: 79–80).

10. A detailed survey of the numerous analyses of definiteness in terms of logical and formal semantics would fall outside the scope of the present work. The reader is referred to Lyons (1999: 253 f.) for a review of the debate on definiteness within different theoretical frameworks and approaches. I completely agree with Abbott’s (2004) comment on such attempts to charac-terize definiteness “with varying degrees of formality”: what Abbott (2004: 147) writes is that “each has a foundation in intuition, as well as some degree of grammatical effect. However, it is not clear that any of them corresponds cleanly to formal categories. As so frequently seems to be the case, grammar is wilfully resistant to attempts at tidy categorization”.

© 2009. All rights reserved

604 Maria Napoli

11. “Roughly speaking, there are two kinds of uses for definites. In those cases which I want to call ‘semantic definites’ the referent of the definite is established independently of the immediate situation or context of utterance. […] ‘Pragmatically definite’ NPs, on the other hand, are essen-tially dependent on special situations and contexts for the non-ambiguity (and existence) of a referent. They include deictic, anaphoric, and endophoric uses of definites” (Löbner 1995: 298).

12. In particular, the label ‘semantic’ has been used as referring to the category of definiteness in those languages which do not encode it as a grammatical category. Cf. also Lyons’ (1999) dis-tinction between semantic and grammatical definiteness (Section 2.1. in this paper).

13. Obviously, Figure 1 does not include all possible cross-linguistic instances of definiteness, considering that, as already noted, such instances are partly language-specific. For instance, languages differ with respect to the use of definite articles with possessives (Lyons 1999: 22 f., 124 f.). A particular case is represented by languages showing two articles: I shall mention just one example, taken from Schroeder (2006). In Danish, there are two articles: the pospositive article (which is suffixed to the noun) and the preposed article. Only the preposed article (which is called ‘adjective article’) occurs before nouns with adjectives, as in de små kartofl-er ‘the small potatoes’ vs. kartofl-er-ne ‘the potatoes’. Nevertheless, the distribution of the two articles can be influenced by anaphora and by the expression of contrastiveness (Schroeder 2006: 562–563).

14. Cf. Hawkins (1978), who included all these cases into a broad category, labelled ‘larger situ-ation use based on specific knowledge or on general knowledge’: “When the definite referent is to be located in a larger situation the speaker and hearer must either share general knowledge of the existence of that entity in that situation (with additional specific knowledge being merely op-tional), or, where there is no such general knowledge basis of understanding, they must possess specific knowledge of this object as such in some shared situation” (Hawkins 1978: 119–120). For a discussion, see also Lyons (1999: 3 f.).

15. A referee wonders if the notion of a relative world does not appear to be an acknowledgment that the context is essential even to logical definiteness; he also points out that under the ‘logical approach’ what is missing is the notion that speakers employ articles to give hearers instructions for identifying the entities referred to. I consider the idea that, in principle, context is essential to definiteness in languages which encode it by articles as uncontroversial: in Italian one generally speaks of il sole (‘the sun’), but it is also possible to speak of un sole (‘a sun’) when attributing particular and temporary properties to it, i.e., in the context in which the entity referred to is singled out as a ‘specific occurrence’ of the sun (cf. also 6.3.). As for the Map of Definiteness, what is crucial is the definiteness status of nouns: the definiteness of entities unique in a relative world does not depend on a single, transitory situation shared by speaker and hearer (as in cases of pragmatic definiteness illustrated by examples 5 and 6), nor does it depend on the linguistic context. It is reality that determines the uniqueness of the referents presupposed by such nouns: the Prime Minister of Italy is only one because this is established by Italian government.

16. In Hawkins’ (1978) ‘immediate situation use’, one is informed of the existence of a specific object, and is also instructed to use the situation in order to identify which object is referred to, even though he doesn’t see it. To use Hawkins’ words, “the hearer is being instructed to ‘locate’ the referent in the immediate situation of utterance. In contrast to the demonstrative, he is not being instructed to actually perceive it, but only to assign it to the situation which he is in” (Hawkins 1978: 114).

© 2009. All rights reserved

Aspects of definiteness in Greek 605

17. I wish to tank Tania Kalapova, a Bulgarian native speaker, for providing me with these ex-amples.

18. Cf. Abraham (2007: 242): “neither generics nor uniques appear with the topical Det form, nor do proper names as in the spoken varieties of German (*(der) Hans ‘the John’, *(die) Mutter ‘the mother’, strongly stigmatised, however, in Modern Standard, written German)”. See also the quotation from Leiss (2007) in Section 1.

19. I shall not deal with mass and abstract nouns here.

20. Cf. Carlson & Pelletier (1995, eds.), in particular Krifka et al. (1995); see also Chierchia (1998), Lyons (1999), Behrens (2005), Schroeder (2006), Alexiadou, Haegeman & Stavrou (2007), Greenberg (2007).

21. Cf. Behrens (2005: 310): “English does not really have definite plural generics, as the use of definite plural phrases with kind-reference is allowed only in certain lexically and syntacti-cally restricted cases”. On generic bare plurals in English see Carlson (1977, 1979), Krifka et al. (1995: 81 f.), Lyons (1999: 189 f.), Guéron (2006), and the interesting contributions which are found in Vogeleer (2006, ed.). As is well known, the use of bare plurals is quite limited in Ro-mance languages: on this topic, cf. the book by Vogeleer & Tasmowski (2006, eds.).

22. The examples in (11) and (12) are taken from Alexiadou, Haegeman & Stavrou (2007: 176).

23. Some scholars interpreted this mark, which in Turkish occurs only in the accusative case, as a mark of ‘specificity’, rather than definiteness. Here, I will leave this point aside. The reader is referred to von Heusinger (2002) for a recent analysis of this problem.

24. “L’article personal s’aplica a qualsevol mena de noms referits a les persones (noms de pila, llinatges, malnoms). El paradigma complet d’article personal de la llengua antiga només es conserva a hores d’ara a les Illes Balears i Pitiüses. Presenta les formes en (davant noms mascu-lins començats per consonant), na (davant noms feminins començats per consonant) i n’ (da-vant noms masculins i feminins començats per vocal)” (Colomina 2002: 548). A referee informs me that definite articles can be used with proper nouns also in some varieties of Spanish (for instance, in colloquial Peruvian Spanish), if the referent is ‘familiar’ to the speaker and unique in the context: this use resembles the occurrence of the definite article with proper nouns in some Italian dialectal varities (cf. ex. 21.a.) and also in Classical Greek (cf. examples in Section 5.2.).

25. The variation can depend also on the type of text: cf. Federspiel (1995) on the definite/in-definite opposition in the language of mathematics.

26. A statistical study of this phenomenon in Ancient Greek has never been accomplished, to the best of my knowledge, and it is definitely desirable. However, the high frequency of the anaphoric use of the Greek definite article with proper nouns (also in the Egyptian papyri: cf. Eakin 1916: 336) is undisputed. Some interesting considerations are found in Parenti (1995), and Lombardi-Vallauri (2002: 26).

27. Cf., in particular, Abraham (2007b), who demonstrates how “the distribution of definite determiners in the early linguistic periods of Germanic converges with the distribution of de-monstrative” (Abraham 2007b: 252).

© 2009. All rights reserved

606 Maria Napoli

28. The reader is referred to the book by Alexiadou, Haegeman & Stavrou (2007), which con-tains an interesting survey of this debate: cf., in particular, the discussion on the proposals by Longobardi (1994) and Chierchia (1998). A referee objects that the redundancy of the definite article with proper nouns is ‘a matter of typological classification’. But the point of this discussion is that a classification based on the notion of redundancy may not be useful from a typological point of view, because it does not deal with those languages or with diachronic stages of languag-es in which the article is not redundant, but meaningful, with proper nouns (cf. 6.2. and 6.3.).

29. Carlson (1977, 1979) was the first who theorised an affinity between proper nouns and generics (at least in English); however, his view is the opposite of Lyons’ view. In particular, the main assumption in Carlson (1977, 1979) is that bare plurals in English correspond to proper names of kinds: “just as there are proper names for objects, such as ‘John’ or ‘Fido’, there are constructions in English which serve as proper names for kinds. This is the function of the English ‘bare plural’ construction” (Carlson 1979: 54). Cf. also Krifka et al. (1995), Vogeleer & Tasmowski.

30. Carlotta Viti, personal communication.

31. Cf., for instance, Löbner (1995): “the redundance of the definite article in semantic definites offers an explanation for the absence of any article in certain cases of definites. One general example are personal and other proper names” (Löbner 1995: 311).

32. This is what Lyons (1999) asserted in commenting on the existence of languages without articles: “Definiteness marking is obviously not essential to communication. Yet many languages which do not mark simple definiteness can be argued to compensate by having other distinc-tions with a similar function. Definiteness may be thought as one of a number of categories which serve to guide the hearer in working out how the discourse is structured and how entities referred to fit into it: markers of topic and focus come to mind here” (Lyons 1999: 48, note 48). See, however, Section 2.1. in this paper.

33. “Non-ambiguity is the property of an expression that allows for only one interpretation (possibly under additional constraints)” (Löbner 1995: 291).

References

Abbott, Barbara. 2004. Definiteness and indefiniteness. The handbook of pragmatics, Lawrence R. Horn & Gregory Ward (eds.), 122–149. Oxford: Blackwell.

Abraham, Werner. 2007a. Discourse binding. DP and pronouns in German, Dutch and English. Nominal Determination. Typology, context, constraints, and historical emergence; Elisabeth Stark; Elisabeth Leiss & Werner Abraham (eds.), 21–47. Amsterdam: John Benjamins.

Abraham, Werner. 2007b. The discourse-functional crystallization of the historically original demonstrative. Nominal determination. Typology, context, constraints, and historical emer-gence, Elisabeth Stark; Elisabeth Leiss & Werner Abraham (eds.), 241–256. Amsterdam: John Benjamins [Studies in Language Companion Series 89].

Abraham, Werner & Elisabeth Leiss. 2007. On the interfaces between (double) definiteness, aspect and word order in Old and Modern Scandinavian. Working Papers in Scandinavian Syntax 80: 17–44.

© 2009. All rights reserved

Aspects of definiteness in Greek 607

Alexiadou, Artemis; Liliane Haegeman & Melita Stavrou. 2007. Noun phrase in the generative perspective. Berlin: Mouton de Gruyter.

Anderson, John M. 2007. The grammar of names. Oxford: Oxford University Press.Bartoněk, Antonín. 2003. Handbuch des mykenischen Griechisch. Heidelberg: Winter.Basset, L. 2006. La préfiguration dans l’épopée homérique de l’article défini du grec classique.

Word classes and related topics in Ancient Greek, Emilio Crespo; Jesús De La Villa & Anto-nio R. Revuelta (eds.), 105–120. Louvain-La-Neuve: Peeters.

Bazzanella, Carla & Lucia Morra. 2007. On understanding metaphor. Lingue e Linguaggio 6(1): 65–84.

Bauer, Brigitte L.M. 2007. The definite article in Indo-European: Emergence of a new grammati-cal category? Nominal Determination. Typology, context, constraints, and historical emer-gence, Elisabeth Stark; Elisabeth Leiss & Werner Abraham (eds.), 103–139. Amsterdam: John Benjamins [Studies in Language Companion Series 89].

Behrens, Leila. 2005. Genericity from a cross-linguistic perspective. Linguistics 43(2): 275–344.Birner, Betty J. & Gregory Ward. 1998. Information status and noncanonical word order in Eng-

lish. Amsterdam: John Benjamins [Studies in Language Companion Series 40].Bolkestein, A. Machteld. 1996. Free but not arbitrary: ‘Emotive’ word order in Latin? On Latin.

Linguistic and literary studies in honour of Harm Pinkster, Rodie Risselada; Jan R. de Jong & A. Machteld Bolkestein (eds.), 7–24. Amsterdam: Gieben.

Bolkestein, A. Machteld. 1998. What to do with topic and focus? Evaluating pragmatic informa-tion. Functional grammar and verbal interaction, Mike Hannay & A. Machteld Bolkestein (eds.), 193–214. Amsterdam: John Benjamins [Studies in Language Companion Series 44].

Carlson, Gregory N. 1977. A unified analysis of English bare plural. Linguistics and Philosophy 1: 413–456.

Carlson, Gregory N. 1979. Generics and atemporal when. Linguistics and Philosophy 3: 49–98.Carlson, Gregory N. & Francis Jeffry Pelletier (eds.), 1995. The Generic Book. Chicago IL: Uni-

versity of Chicago Press.Chantraine, Pierre. 1953. Grammaire Homérique. Paris: Klincksieck.Chesterman, Andrew. 1991. On definiteness. A study with special reference to English and Finnish.

Cambridge: Cambridge University Press.Chierchia, Gennaro. 1998. Reference to kinds across languages. Natural Language Semantics 6:

339–405.Christophersen, Paul. 1939. The articles. A study of their theory and use in English. Copenhagen:

Munksgaard.Colomina i Castanyer, Jordi. 2002. Paradigmes flectius de les altres classes nominals. Gramatica

del catala contemporani, Vol. 1, Joan Solà; Marie-Rosa Lloret; Joan Mascaró & Manuel Pérez Saldanya (eds.), 535–582. Barcelona: Editorial Empúries.

Comrie, Bernard. [1981] 19892. Language universals and linguistic typology. Syntax and typology. Oxford: Blackwell.

Conte, Maria-Elisabeth. 1996. Anaphoric encapsulation. Coherence and anaphora, Walter de Mulder & Liliane Tasmowski (eds.). Belgian Journal of Linguistics 10: 1–10. Amsterdam: John Benjamins.

Corbett, Greville G. 2000. Number. Cambridge: Cambridge University Press.Corbett, Greville G. 2005. The canonical approach in typology. Linguistic diversity and language

theories; Zygmunt Frajzyngier; Adam Hodges & David S. Rood (eds.), 25–49. Amsterdam: John Benjamins [Studies in Language Companion Series 72].

© 2009. All rights reserved

608 Maria Napoli

Diessel, Holger. 1999. Demonstratives: Form, function, and grammaticalization. Amsterdam: John Benjamins [Typological Studies in Language 42].

Dryer, Matthew W.S. 2005. Definite articles. The world atlas of language structures; Martin Haspelmath; Matthew M.S. Dryer; David Gil & Bernard Comrie (eds.), 154–157. Oxford: Oxford University Press.

Eakin, Frank. 1916. The Greek article in first and second century papyri. The American Journal of Philology 37(3): 333–340.

Federspiel, Michel. 1995. Sur l’opposition défini/indéfini dans la langue des mathématiques grecques. Les Études Classique 63: 249–293.

Frajzyngier, Zygmunt. 1991. The de dicto domain in language. Approaches to grammaticaliza-tion, Vol. I, Elizabeth Closs Traugott & Bernd Heine, B. (eds.), 219–251. Amsterdam/Phila-delphia: John Benjamins [Typological Studies in Language 19:1].

Frege, Gottlob. 1892. Über Sinn und Bedeutung. Zeitschrift für Philosophie und Philosophische Kritik 100: 25–50. (English version translated as ‘On sense and reference’ by Peter Geach & Max Black 19803 (eds.), Translations from the Philosophical Writings of Gottlob Frege, 56–78. Oxford: Blackwell).

Gildersleeve, Basil Lanneau. 1890. On the article with proper names. The American Journal of Philology 11(4): 483–487.

Gildersleeve, Basil Lanneau 1911. Syntax of classical Greek. New York NY: American Book Com-pany.

Givón, Talmy. 1978. Definiteness and referentiality. Universals of human language; Joseph H. Greenberg (ed.), 293–330. Stanford CA: Stanford University Press.

Greenberg, Jospeh H. 1978. How does a language acquire gender markers? Universals of human language, Joseph H. Greenberg (ed.), 49–82. Stanford CA: Stanford University Press.

Greenberg, Yael. 2007. Exceptions to generics: Where vagueness, context dependence and mo-dality interact. Journal of Semantics 24: 131–167.

GREL: Liddell, Henry George & Scott, Robert (eds.). 1996. Greek-English lexicon. With a revised supplement. Oxford: Clarendon Press.

Guéron, Jacqueline. 2006. Generic sentences and bare plurals. Non-definiteness and plurality; Svetlana Vogeleer & Liliane Tasmowski (eds.), 219–234. Amsterdam: John Benjamins [Lin-guistik Aktuell/Linguistics Today 95].

Harris, Martin B. 1980. The marking of definiteness: A diachronic perspective. Papers from the 4th international conference on historical linguistics, Elizabeth Closs Traugott; Rebecca Labrum & Susan C. Shepherd (eds.), 75–86. Amsterdam: John Benjamins [Current Issues in Linguistics 14].

Hawkins, John A. 1978. Definiteness and indefiniteness: A study in reference and grammaticality prediction. London: Croom Helm.

Heine, Bernd. 2003. Grammaticalization. The handbook of historical linguistics, Brian D. Joseph & Richard D. Janda (eds.), 575–601. Oxford: Blackwell.

von Heusinger, Klaus. 2002. Specificity and definiteness in sentence and discourse structure. Journal of Semantics 19: 245–274.

Himmelmann, Nikolaus P. 1997. Deiktikon, Artikel, Nominalphrase. Zur Emergenz syntaktischer Struktur. Tübingen: Niemeyer.

Himmelmann, Nikolaus P. 2001. Articles. Universal units in the lexicon. Language typology and languages universals, Martin Haspelmath; Ekkehard König; Wulf Oesterreicher & Wolfgang Raible (eds.), 831–841. Berlin: de Gruyter.

Humbert, Jean. 1960. Syntaxe Grecque. Paris: Klincksieck.

© 2009. All rights reserved

Aspects of definiteness in Greek 609

Jakobson, Roman. 1971. Boas’ view of grammatical meaning. Selected writings II. Word and language, 489–496. The Hague: Mouton (already printed in The anthropology of Franz Boas: Essays on the centennial of his birth. Memoirs of the American Anthropological Association 89, Walter Goldschmidt (ed.), 1959, 139–145).

Krámský, Jirí. 1972. The article and the concept of definiteness in language. The Hague: Mouton.Krifka, Manfred; Francis Jeffry Pelletier; Gregg N. Carlson; Alice ter Meulen; Godehard Link &

Gennaro Chierchia. 1995. Genericity: An introduction. The generic book, Gregg N. Carlson & Francis Jeffry Pelletier (eds.), 1–124. Chicago IL: University of Chicago Press.

Lambrecht, Knud. 1994. Information structure and sentence form. Cambridge: Cambridge Uni-versity Press.

Lehmann, Christian. 1991. The Latin nominal group in a typological perspective. New studies in Latin linguistics. Selected papers from the 4th international colloquium on Latin linguis-tics, Cambridge, April 1987, Robert Coleman (ed.), 203–232. Amsterdam: John Benjamins [Studies in Language Companion Series 21].

Leiss, Elisabeth. 2000. Artikel und Aspekt. Die grammatischen Muster von Definitheit. Berlin: Mouton de Gruyter.

Leiss, Elisabeth. 2007. Covert patterns of definiteness/indefiniteness and aspectuality in Old Icelandic, Gothic, and Old High German. Nominal determination. Typology, context, con-straints, and historical emergence, Elisabeth Stark, Elisabeth Leiss & Werner Abraham (eds.), 73–102. Amsterdam: John Benjamins [Studies in Language Companion Series 89].

Lewis, Geoffrey L. 1967. Turkish grammar. Oxford: Oxford University Press.Löbner, Sebastian. 1985. Definites. Journal of Semantics 4: 279–326.Lombardi Vallauri, Edoardo. 2002. L’articolo greco fra identificabilità ed esclusività del referen-

te. Studi Italiani di Linguistica Teorica e Applicata 31(1): 7–33.Leonardi, Paolo. 2007. Meaning and context. Lingue e Linguaggio 6(1): 9–23.Longobardi, Giuseppe. 1994. Reference and proper names: A theory of N-movement in syntax

and logical form. Linguistic Inquiry 25(4): 609–665.Loprieno, Antonio. 1980. Osservazioni sullo sviluppo dell’articolo prepositivo in egiziano e nelle

lingue semitiche. Oriens Antiquus 19: 1–27.Lyons, Christopher. 1999. Definiteness. Cambridge: Cambridge University Press.Mackridge, Peter. 1985. The Modern Greek language. Oxford: Clarendon Press.Manolessou, Io & Geoffrey Horrocks. 2007. The development of the definite article in Greek.

Studies in Greek Linguistics 27. Proceedings of the annual meeting of the department of lin-guistics, school of philology, Aristotele University of Thessaloniki, May 6–7, 2006: 224–236.

Moravcsik, Edith A. 2006. Review of Newmeyer, Frederick. 2005. Possible and probable lan-guages. A generative perspective on linguistic typology (Oxford: Oxford University Press). Linguistic Typology 10(2): 277–286.

Morpurgo Davies, Anna. 1968. Article and demonstrative: A note. Glotta 46: 77–85.Newmeyer, Frederick. 2003. Grammar is grammar and usage is usage. Language 79: 682–707.Parenti, Allesandro. 1995. Aspetti diacronici della definitezza in greco e rumeno. PhD disserta-

tion, University of Perugia.Pinkster, Harm. 1990. Evidence for SVO in Latin? Latin and the Romance languages in the early

Middle Ages, Roger Wright (ed.), 69–82. London: Routledge.Plank, Franz. 2003. Double articulation. Noun phrase structure in the languages of Europe, Franz

Plank (ed.), 338–395. Berlin: de Gruyter.Prandi, Michle. 2004. The building blocks of meaning. Ideas for a philosophical grammar. Amster-

dam: John Benjamins [Human Cognitive Processing 13].

© 2009. All rights reserved

610 Maria Napoli

Putzu, Ignazio. 2001. Quantificazione totale/universale e determinatezza nelle lingue del Mediter-raneo. Pisa: Edizioni ETS.

Rosén, Hannah. 1991. The definite article in the making, nominal constituent order, and re-lated phenomena. Linguistic studies on Latin. Selected papers from the 6th international col-loquium on Latin linguistics, Jószef Herman (ed.), 129–150. Amsterdam: John Benjamins [Studies in Language Companion Series 28].

Russell, Bertrand. 1905. On denoting. Mind 14: 479–493.Sansone, David. 1993. Towards a new doctrine of the article in Greek: some observations on the

definite article in Plato. Classical Philology 88(3): 191–205.Schroeder, Christoph. 2006. Articles and article systems in some areas of Europe. Pragmatic

organization of discourse in the languages of Europe, Gioliano Bernini & Marcia L. Schwartz (eds.), 545–611. Berlin: Mouton de Gruyter.

Schwyzer, Eduard & Albert Debrunner. 1950. Griechische Grammatik. München: Beck.Sornicola, Rosanna. 2006. Interaction of syntactic and pragmatic factors on basic word order

in the languages of Europe. Pragmatic organization of discourse in the languages of Europe, Giuliano Bernini & Marcia L. Schwartz (eds.), 357–544. Berlin: Mouton de Gruyter.

Stark, Elisabeth; Elisabeth Leiss & Werner Abraham (eds.), 2007. Nominal determination. Typol-ogy, context, constraints, and historical emergence. Amsterdam: John Benjamins [Studies in Language Companion Series 89].

Stvan, Laurel Smith. 2007. The functional range of bare singular count nouns in English. Nomi-nal determination. Typology, context, constraints, and historical emergence, Elisabeth Stark; Elisabeth Leiss & Werner Abraham (eds.), 171–187. Amsterdam: John Benjamins [Studies in Language Companion Series 89].

Trenkić, Danjiela. 2002. Establishing the definiteness status of referents in dialogue (in languages with and without articles). Working Papers in English and Applied Linguistics 7: 107–131.

Vergnaud, Jean-Roger & Maria Luisa Zubizarreta. 1992. The definite determiner and the in-alienable constructions in French and in English. Linguistic Inquiry 23(4): 595–652.

Vogeleer, Svetlana (ed.). 2006. Bare plurals, indefinites, and weak–strong distinction. Amsterdam: John Benjamins [Belgian Journal of Linguistics 19].

Vogeleer, Svetlana. & Liliana Tasmowski. (eds.), 2006. Non-definiteness and plurality. Amster-dam: John Benjamins Linguistik Aktuell/Linguistics Today 95].

Texts

Aristophanes, Birds. Lysistrata. Women at the Thesmophoria. With an English translation by Henderson, Jeffrey. 2000. Cambridge: Harvard University Press.

Aristophanes, Clouds. With an English translation by Henderson, Jeffrey. 1998. Cambridge: Harvard University Press.

Aristophanes, Frogs. With an English translation by Henderson, Jeffrey. 2002. Cambridge: Har-vard University Press.

Plato, Ion. With an English translation by Lamb, Walter Rangeley Maitland 1952 (19251). Cam-bridge: Harvard University Press.

Plato, Phaedo. Phaedrus. With an English translation by Fowler, Harold North. 1977 (19141). Cambridge: Harvard University Press.

Plato, The Republic. With an English translation by Shorey, Paul. 1956 (19351). Cambridge: Har-vard University Press.

© 2009. All rights reserved

Aspects of definiteness in Greek 611

Plato, Theaetetus. With an English translation by Fowler, Harold North. 1952 (19211) Cam-bridge: Harvard University Press

Author’s address

Maria NapoliDipartimento di LinguisticaUniversity of PisaVia Santa Maria 3656100 Pisa

[email protected]