the spanish postnominal demonstrative, Study notes of Spanish

Download the spanish postnominal demonstrative and more Study notes Spanish in PDF only on Docsity! THE SPANISH POSTNOMINAL DEMONSTRATIVE IN SYNCHRONY AND DIACHRONY DISSERTATION Presented in Partial Fulfillment of the Requirements for The Degree Doctor of Philosophy in the Graduate School of the Ohio State University By David B. Alexander, M.A. The Ohio State University 2007 Dissertation Committee: Approved by Dieter Wanner, Adviser Fernando Martínez-Gil ___________________ Javier Gutiérrez-Rexach Adviser Spanish and Portuguese Graduate Program Brian Joseph iv VITA August 2, 1979………………………..Born Mayfield Heights, Ohio 2001…………………………………...B.A., Miami University 2003……………………………………M.A, The Ohio State University 2001-2007……………………………..Teaching Assistant, The Ohio State University FIELD OF STUDY Major Field: Spanish vi Page Abstract…………………………………………………………………………………………... ii Acknowledgments………………………………………………………………………………. iii Vita……………………………………………………………………………………………… iv Table of Contents……………………………………………………………………………….. vi List of Figures…………………………………………………………………………………... ix Introduction…………………………….………………………………………………………… 1 Chapters: 1: Hispanic Context………………………………………………………………………... 5 1.1 The Spanish Demonstrative in Context……………………………………….… 5 1.2 Diachronic Demonstrative Distribution………………………………………... 15 1.3 The Postnominal Demonstrative…………………………………….……….… 17 1.4 Prenominal versus Postnominal Demonstratives…………………..…………... 21 2: Syntactic Analysis……………………………………………………………………… 25 2.1 The Determiner Projection: Articles and Demonstratives………….………….. 25 2.2 Demonstratives, Articles, and Adjectives……………………………………… 43 2.3 DP and Phonetic Form within the Nominal Domain……………….………….. 48 2.4 A Synchronic Analysis of the Spanish DP……………………….………….… 52 2.5 The Determiner Phrase and the CP…………………………………………….. 54 2.6 The Article-less Post-nominal Demonstrative Construction……….………….. 56 2.7 Overview of Synchronic Syntax………………………………….………….… 58 3: Information Structure…………………………………………………………………... 59 3.1 The Semantics of Demonstratives…………………………………….……….. 59 4: Synchronic Analysis…………………………………………………………………… 82 4.1 Methodology…………………………………………………………………… 82 vii 4.2 …..Synchronic Informational Categories- este/ese………………………………... 89 4.2.1 Anaphoric …………………………………………………………... 91 4.2.2 Mutual Knowledge………………………………………………….. 96 4.2.3 Cataphoric ………………………………………………………….. 97 4.2.4 Discourse Topic………………………… ……………………… 99 4.2.5 Situationally Accessible……………………………………………. 101 4.2.6 Retrieval and Repair………………………………………………... 102 4.2.7 Affective……………………………………………………………. 106 4.2.8 Exclamation……………………………………………….………... 108 4.3 Synchronic Structural Categories- este/ese…………………………………… 109 4.3.1 [SPEC DP ART [D’ [NP N [DEMlo DEM]]]]……………………….…….. 109 4.3.2 [SPEC DP [D [NP N]]]]i…[DP DEM]i….…………………………………. 112 4.3.3 [CP COMP [NP N]]i [DP DEM]i ……………………………………... 114 4.3.4 [NP ADJ N]i…[DP DEM]i ………………………………………….... 114 4.4 Overview of Synchronic- este/ese……………………………………………... 115 4.5 Synchronic Informational Categories- aquel……………………………….… 117 4.5.1 Evoking Past Events………………………………………………... 118 4.5.2 Meta-Discourse…………………………………………………….. 118 4.5.3 Affective…………………………………………………………… 120 4.5.4 Exclamation………………………………………………………... 121 4.6 Synchronic Structural Categories- aquel………………………….…………... 122 4.6.1 [SPEC DP ART [D’ [NP N [DEMlo DEM]]]]……………………………... 123 4.6.2 [SPEC DP [D’ [NP N]]]i…[DP DEM]i…………………………………… 124 4.6.3 [NP ADJ N]i…[DP DEM]i.………….………………………………… 125 4.6.4 [CP COMP [NP N]]i…[DP DEM]i….………………………………….. 125 4.7 Overview of Synchronic aquel………………………………………………... 126 4.8 Synchronic este/ese vs aquel……………………………………………….…. 127 5 Diachronic Analysis……………………………………………………………….….. 130 5.1 Introduction………………………………………………………………….. . 130 5.2 Diachronic Informational Categories- este/ese………………………………. 140 5.2.1 Anaphoric……….……………….…..…………………………..….. 140 5.2.2 Anaphoric coreferential……..……………………………………..... 148 5.2.3 World Knowledge…………………………………………………... 149 5.2.4 Cataphoric Reference………………………………………………….……… 150 5.2.5 Situationally Accessible………………………………………….…. 151 5.2.6 Deixis ad oculos…………………………………………………….. 155 5.2.7 Discourse Topic…………...………………………………………… 155 5.2.8 Retrieval and Repair………………………………………………… 158 5.2.9 Affective……………………………………………………………. 158 5.2.10 Exclamation…………………………………………………………. 160 5.3 Diachronic Structural Categories- este/ese……………………………………. 161 5.3.1 [SPEC DP ART [D’ [NP N [DEMlo DEM]]]]…………………….………... 161 5.3.2 NP-shift……………………………………………………………... 170 5.3.3 NP-shift COP PP/CP……………………………………………...… 174 5.3.4 NP-shift-COP-w/o XP………………………………………………. 175 1 INTRODUCTION Aim of Study In this study I investigate the Spanish demonstrative and its use in discourse. In particular, the postnominal demonstrative construction and its development toward the synchronic situation where such placement has become conventionalized appearing alongside the more standard prenominal position. This study elucidates the relative function of this special demonstrative placement through the use of discourse frameworks that categorizes referring expressions vis-à-vis their information structure. Informational categories better explain the Spanish data and the development of the construction in question, which exhibits both syntactic and semantic stability from its origins in Late Latin to Modern Spanish usage. These frameworks elucidate not only information structure but also the concrete function of the construction in discourse where the speaker encodes the referent for the addressee as hearer-old and therefore as part of the common ground of discourse. From this core come the main communicative functions of the construction: Anaphoric, Cataphoric. Shared Knowledge, Situationally Accessible and Affective. 2 A specific aim of the present investigation is to clarify the role of the post-posed demonstrative in the “pejorative” or affective use frequently commented on in the literature (Ramírez Fernández 1951:316, Esbozo 1973:436, Macías 1997). By investigating both synchronic and diachronic data the relevant range of the meanings conveyed by the displaced demonstrative becomes clear, showing that this qualitative function is only the most salient part of a far larger pattern. The Affective use is a meta-discourse function meant to comment on the referent at hand and color the cognitive state of the addressee toward the entity being referred to (Lakoff 1974). In that this use communicates more than ‘what is said’ reveals that the inverted syntax of the postnominal construction is in itself a small portion of the communicative function of the expression within its full discourse context. Therefore in Chapter 1 Pan-Romance and Hispanic Context the postnominal construction is placed within the context of Spanish demonstratives este, ese, aquel and the system of Spanish deixis where the three term distinction is utilized for communicative purposes beyond that envisioned by traditional accounts. Also other Romance languages such as Rumanian and Catalan are investigated where the post-posed demonstrative occurs; Chapter 2: Syntactic Analysis investigates the derivation of the demonstrative within the Determiner Phrase in various frameworks (see Bernstein 1997, 2001 Brugé 1996, Panagiotidis 2000, Grohmann&Panagiotidis 2004, Roca 1996). In addition, the prenominal vs postnominal demonstrative is also investigated across languages where the prenominal is shown to be properly basic for Germanic and 3 Romance. The postnominal demonstrative is also compared with adjectival placement in Spanish and Rumanian, which exhibit a mirror image of the demonstrative. The postnominal is derived from the prenominal position, which is properly basic in the structure. Also the postnominal construction ART-N-DEM is distinguished from the article-less string N-DEM, which is more limited in its distribution, and CP-N-DEM used in exclamations. Chapter 3: Semantics and Information Structure investigates the demonstrative as a referring expression in a number of discourse frameworks (see Prince 1981, Givón 1983, Levinson 1987, Ariel 1990, Gundel et al. 1993, Blackwell 2003) from which a core number of informational categories are derived: Anaphoric, Shared Knowledge, Situationally Accessible and Affective. After these preliminary sections, which lay the groundwork for the understanding of the postnominal construction, comes the data analysis portion. In Chapter 4: Synchronic Analysis the informational categories derived in Chapter 3 are applied to synchronic data featuring the postnominal construction. Here the demonstratives este/ese and aquel are distinguished following the discussion of the Spanish demonstrative system in Chapter 1 where distal aquel has the specialized uses of Meta-discourse and Evocative. These informational categories are then associated with structural categories in order to clarify the syntax-information structure interface. It is important to distinguish genuine tokens of the postnominal construction from alternative constructions N-DEM and CP-N-DEM. In Chapter 5: Diachronic Analysis I apply the synchronic informational categories to diachronic data. Here the investigation necessarily relies on a direct knowledge of the synchronic language in order to interpret diachronic tokens. 6 (1) a. Este hombre DEM N b. This man DEM N Here the relevant deictic element este and this modify the noun ‘man’. In philosophical terms, deictic determiners, also known as demonstratives, are indexicals in terms of Peirce, which are necessarily context dependent. Yet they differ from other indexicals since they may include a pointing gesture, which is inherent to their deictic character. Demonstrative words are quintessential deictic expressions since they explicitly point out a referent in the deictic field of the discourse (Gutiérrez-Rexach 2002). These words are called demonstratives in that they show or point out the referent in question. “Demonstratives are place deictics. They indicate the relative distance of an object, location or person vis-à-vis the deictic center (also called the origo) which is usually the speaker” (Diessel 1999, p.36) In Spanish we note a tripartite demonstrative system este- ese- aquel which encodes features [+prox], [+medial] and [+distal] respectively. As traditionally understood the demonstrative este encodes a relative proximity to the speaker, ese refers to an area nearer to the addressee, and aquel points to regions removed from both participants meant to highlight the distance of the referent in deictic space. In turn, these demonstratives are seen as the counterparts of locative adverbs aquí- [+prox]- ‘here allí- [-prox, -dist]- ‘there’ allá- [+dist]‘over there’ respectively within the wider system of Spanish deixis (Hottenroth 1982). 7 Aquel allá Figure 1.1 Spanish System of Deixis A further tack taken by the grammars is a person-based account whereby proximal este corresponds to first person (yo), medial ese to second (tú), and distal aquel to third person (él, ella). In his review of demonstrative systems in languages of the world Diessel (1999) makes a point of contrasting such person-based systems with distance- based frameworks. However, these two accounts are often combined to yield a description of Spanish demonstratives as both personal and locative in nature (Eguren 1999). According to Eguren (1999) unlike terms of personal deixis such as yo, demonstratives are deictically opaque in that they require gesticular or contextual information. In term of their semantics these terms have an indeterminate content and are therefore incomplete deictic signs. A basic problem arises since this unique status allows for flexibility within the system that the traditional frameworks fail to reflect. In contrast to the Spanish system, English employs only two terms this [+prox] and that [+distal] for the same function. In this light, it has long been under debate why demonstrative Ese allí Este A aquí . Speaker 8 systems such as Spanish este- ese- aquel continue to maintain a tripartite distinction while many other Romance languages like French, Brazilian Portuguese and Modern Italian, show a tendency toward two-term systems as evidenced in the decline of medial terms aqueix and codesto in Catalan and Standard Italian respectively (Jüngbluth 2003, Meira 2003). These two- term systems consist of binary distinctions [+prox] vs [+dist] as in French ce livre-ci vs ce livre- là on par with colloquial English this book here, that book there (Bernstein 2001). Yet the existence of such differing systems in the same language families raises the question: is the spatial deixis properly basic to demonstratives in Spanish? Or rather do discourse factors determine usage? The flexibility of demonstratives allows them to express not only spatial but emotional content. For example, in Spanish the use of ese in place of este is seen as a distancing strategy that is in some sense pejorative (Ramírez 1951:316, Esbozo 1973:432, Eguren 1999). However, Gutiérrez-Rexach’s 2002 and Jüngbluth’s 2003 investigations of Spanish demonstratives in discourse reveal that ese is not the medial demonstrative but rather an unmarked form which encodes addressee reference. Due to its unmarked status, ese has the widest use across temporal, spatial, and discourse deixis (Gutiérrez-Rexach 2002). In this manner ese corresponds to the weak locative proform [áj] which refers to a non-specific region of space (Gutiérrez-Rexach 2002). (2) Por ahí [áj] PREP-by LOC[-prox,-dist] ‘Thereabouts’ 11 Postnominal (d.) el templo aquel ART-the N-temple DEM-that ‘That temple’ In (3a&c)) both the possessive adjective and the demonstrative precede and modify the noun. Whereas in (3b&d)) the definite article heads the construction with the possessive adjective and the demonstrative in postnominal position. Given the variation in the use of the article Bello posits that possessive and demonstrative pronouns are assumed to include articles within themselves when they precede the noun. (Bello 1848:250) Bello (1848) implies that the pre-posed demonstrative and possessive pronouns perform the function of the article. In other words, these prenominal elements do double duty as deictics and markers of definiteness. So when these same items follow the noun, an article is needed to mark definiteness in its stead. As previously mentioned, this postnominal use is often defined in traditional 20th century literature as emphatic and in some sense derogatory (Ramírez Fernandez 1951:316, Esbozo 1973:432, Macías 1997). However, Gutiérrez-Rexach (2002) defines post-nominal ese as a focalizer of NP’s as shown in the conversation below: (4) A: Dame la cartai. A: V-give CL-1st-s ART-the letter A: ‘Give me the letter.’ B: ¿Qué carta? B: INTER-what letter B: ‘What letter?’ C: La carta esai que te mencioné. C: ART-the letter DEM-that CP-that CL-2nd-s V-1st-s mentioned C: “That letter I mentioned to you.” (Gutiérrez-Rexach 2002) 12 Here in (4) post-nominal esa focuses the NP la carta, which has become a topic of discourse. The unique place of ese in the system of Spanish deixis is further illustrated by its status as a default focalizer as demonstrated below: (5) Estos/ esos/ aquellos tres coches esos DEM[+prox] /DEM[+/-prox] these DEM[+dist] those three Ncars DEM[+/-prox]those (Gutiérrez-Rexach 2002) Notice that while all three demonstratives in (5) are felicitous in prenominal position, only ese is felicitous post-nominally in all three combinations. It can be said that Spanish utilizes its tripartite system for expressive purposes in addition to a gradation of distance from the deictic origo (Bühler 1937), which is often neutralized in a discourse context. That is, spatial dimensions of demonstratives are reinterpreted in accordance with the discourse context. This neutralization is confirmed in discourse practice as shown by Jüngbluth’s 2003 study of speaker interaction in Toledo (Spain). Importantly, Jüngbluth (2003) envisions a conversational dyad which takes into account the shared space between speaker and addressee in contrast to Bühler (1937) who only envisions a single deictic origo centered on the speaker. In speaker interviews Jüngbluth (2003) finds that the only situation where the tripartite system is consistently maintained is in Side-by-Side communication between speaker and addressee where the three-part gradation este- ese- aquel is used to encode referents and their relative distance from the discourse agents as shown in Figure 1.2. 13 este ese aquel [+prox] [-prox, -dist) [+dist] Figure 1.2 Side-by-Side Communication (Jüngbluth 2003: 24-27) In contrast, a Face-to-Face context features a binary opposition between este and aquel to encode shared versus non-shared space respectively between speaker and addressee in Figure 1.3: aquel este aquel (not-shared) (shared space) (not shared) Figure 1.3 Face-to-Face Communication (Jüngbluth 2003) While a Face-to-Back configuration encodes ese as being within the addressee’s sphere as opposed to that of the speaker in Figure 1.4: Addressee Speaker Speaker Addressee 16 languages such as French and Italian (see (13) below), features aqueste -aquesse -aquel. Diachronic corpora studies (Figure 1.5 below) reveal a steady presence of este with an ever- increasing use of ese beginning in the late 19th century, both prenominal: ese hombre (1886) and postnominal las operas esas (1873). Consequently, ese begins as a minority pattern in the 1200’s at just .4% eventually rising to 37.6% in the 20th century. The later predominance of ese can be partly explained via its association with more colloquial usage coupled with the nature of the 20th century corpora, which feature spoken exchanges in written discourse However, Salvá 1837’s observations on the divergence between the uses este and ese in the Modern language is still pertinent in that ese is seen as tied to a discourse function. All of the aforementioned changes are to the detriment of aquel, which encodes referents outside the discourse. Therefore, while aquel constitutes a dominant form at 57.6% in the 1200’s the same form drops off rapidly in the 20th with a meager 8% as the usage of medial ese expands to its present range leaving aquel as a minority option. This limited use is confirmed by oral discourse where aquel is largely absent, being confined to written or stylized uses. 17 0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 1200 1300 1400 1500 1600 1700 1800 1900 Este Ese Aquel Figure 1.5: Diachronic Distribution (Corpus del español) In the present study I am interested in the communicative use of demonstratives in discourse. Such uses are related to the fact that language is “responsive to the fundamental need of speakers to convey and assess feelings, moods, dispositions, and attitudes. This need is as critical and as human as that of describing events. Interlocutors need to know not only what predication a speaker is making [but also] the affective orientation the speaker is presenting with regard to that particular predication” (Ochs and Schieffelin 1989:9 cited in Finnegan 1995 p.5). 18 1.3 The Postnominal Demonstrative In particular, one of these expressive uses involves special syntax such as the post-posed demonstrative shown in (8b). (8) a) Este hombre b) El hombre este DEM N ART N DEM [+def] [+def] [+loc] [+loc] [+deic] [+deic] The main contrast between (8a) and (8b) is that (8a) features the demonstrative in prenominal position: [DEM-N], while (8b) has a postnominal placement for the same element introduced by the definite article: [ART-N-DEM]. At a more basic level (8b) spells out the semantic features of the referring expression:[+def], [+loc], [+deic] independently. The definite article in (8b) identifies the referent via a feature of definiteness [+def] while the post-posed demonstrative expresses a locative [+loc] and deictic features [+deic]. In contrast, prenominal este expresses all three features in itself. More importantly, the prenominal demonstrative este adds a measure of deictic import that the noun hombre does not possess by itself. Prenominal este, with its concentration of features in a single term, reinforces the referent introduced by the demonstrative and brings the deictic-locative features to the fore (Eguren 1999) whereas the postnominal construction separates the deictic and locative features from the noun. Eguren (1999) discusses parallel uses in English (10), which utilize the demonstrative coupled with a locative along with their Spanish counterparts (9): 21 (12) near(sg) near(pl) far(sg) far(pl) remote(sg) remote(pl) Northern this thir that tho yon/thon yon/thon Southwest thease theys that they thik thik (Milroy 993: 214-238) In a broader context, comparative data reveals the prenominal demonstrative as the default form across Romance and Germanic languages in a sequence of type [DEM + N] exhibited in (13) for the NP “this man”: (13) a. Romance Spanish este hombre French cet homme DEM N DEM N Portuguese este home Catalan aquest hom DEM N DEM N Italian quest’uomo Rumanian acest om DEM N DEM N b. Germanic English this man German dieser Mann DEM N DEM N (Bernstein 2001) However, the pattern in (13) above is in contrast languages like Indonesian, which while featuring the same basic SVO word order as in English and Spanish it has a different demonstrative distribution as shown in (14) 22 Indonesian Demonstratives (14) a. Rumah ini N house DEM this ‘This house’ b. Rumah itu N house DEM that ‘That house’ (Johns 1995) Thus in (14) the noun rumah ‘house’ is modified via demonstrative ini ‘this’ and itu ‘that’ appear in postnominal position. This placement is significant in that it reveals how Non- Indo-European languages such as Indonesian feature a constituent order different at variance with Indo-European languages like English and Spanish (13). 1.4 Prenominal versus Postnominal Demonstratives It is essential to note that not only do languages feature a default position for functional items as exhibited for demonstratives in (13) vs. (14) but that said placement has an impact on the grammar. To this end, some Romance languages, such as Spanish, Rumanian and Catalan exhibit a distribution of unmarked prenominal versus marked post-nominal position for demonstratives: (15) Spanish a. este hombre b. el hombre este DEM N ART N DEM Rumanian c. acest om d. om-ul acesta DEM N N-ART DEM 23 Catalan: e. aquest hom f. el hom aquest DEM N ART N DEM Traditionally, if they are not stand-alone pronouns, demonstratives have been defined as adjectives since they necessarily modify a noun. Similar to demonstratives Spanish adjectives also have default and marked distribution, which has an impact on the interpretation of the Noun Phrase (NP). However, in contrast to the demonstrative the unmarked position is post-nominal as opposed to prenominal: (16) a. El hombre pobre ART-the N-man ADJ-poor b. El pobre hombre ART-the ADJ-poor N-man ‘The poor man.’ Here the emphasis of the adjective is affected, with the prenominal position giving a more qualitative reading beyond the truth-value content of the adjective at hand. This distribution is similar to that of demonstratives; however, here the positions are reversed. That is to say, both the prenominal adjective and the postnominal demonstrative are marked positions that have specialized functions within discourse. The contrast between adjective and demonstrative positions exhibited in (17a&b) and (16a) respectively seems to indicate that these terms cannot be identical since their distribution in the syntactic string illustrates a different underlying place in the hierarchy. 26 Notice that the article is infelicitous when appearing on its own in (18a). Articles are prosodically weak elements that seek support on the following noun as evidenced by the morphology of articles in Italian (ex. l’ before vowel, lo before consonant clusters, gli before masculine nouns beginning in vowel (18g)) and to a lesser extent in Spanish el, an allomorph of the singular feminine article appearing before nouns beginning in a vowel with initial stress (ex. el agua (17e)). In addition, Portuguese has a special phonological variation of the article similar to Portuguese clitic pronoun variation(ex. fazer+lo>faz-elo, fazem+lo> fazem-no). Here the syntactic connection is to the following nominal, but the phonetic variant is determined by the preceding word. Klavans (1985) demonstrates the difference between syntactic and phonological attachment for clitics: (18) a. b. X’ X’ Y’ X clitic X clitic=Y Syntactic host Phonological host /fazer o/ [faze- lo] V-inf-do CL-3rd-s ‘To do it.’ /fazem o/ [fazem- no] V-3rd-s-do CL-3rd-s ‘They do it.’ /por lo/ [po- lo] PREP-for CL-3rd-s ‘For it’ (Uriagareka 1995) 27 In (18) a clitic element attaches at two levels: the syntactic and the phonological, which do not directly correspond. Therefore in Portuguese the object clitic o when contiguous to various hosts produces allomorphs lo and no in faze-lo, fazem-no and po-lo respectively. Note that the syntactic and phonological boundaries between words do not directly correspond. In another example (17h) the English indefinite article has a different morphology depending on context. A is the default form of the definite article has an allomorph An appearing before nouns beginning in a vowel (17h). This morphological variation in the definite article depends on the phonology of particular dialects. For example, the noun herb has a glottal fricative “h” in British English a herb whereas American English lacks this initial sound and therefore features the alternate form: an herb. While articles and demonstratives differ in semantics and exhibit a slightly more varied morphology and prosody, said elements are both related terms in that they are determiners. Furthermore, as evidenced by child language, both articles and demonstratives are often confused with one another and thus used interchangeably (Giusti 1997). In addition these elements are also diachronically and morphologically related in that demonstratives often provide the lexical material for articles (Diessel 1999). An example is the oft-cited case of Latin distal demonstrative ILLE, which gives the article in Romance languages and Spanish el, la, los, las. However, the formation of articles also utilized alternative sources such as the identifier IPSE in an earlier stage of Romance development followed by a later wave of distal ILLE. Therefore languages like Sardinian utilize articles su, sa from IPSE as in sa limba ‘the language’. in addition to Balearic Catalan which features the article salat from IPSE along with the more standard ILLE as shown in (19) 28 (19) a. s’esglèsia b. L’Esglèsia ART-N ART-N ‘the church’ ‘The Church (at large)’ c. sa sala d. La Sala ART N ART N ‘the room’ ‘City Hall’ (Vincent 1997) Here the article from IPSE su, sa (19a&c) serves as a marker of specificity in which a specific church or hall is being pointed out. In contrast, the article from Latin ILLE (19b&c)is used for more generalized uses whereby the church and the hall are identified with the Church, and the City Hall respectively. This development is paralleled by many other languages such as the Modern Finnish demonstrative system: -tama ‘this’-se ‘that’ –tuo ‘that’ where the demonstrative se is presently taking on article-like functions. The Finnish demonstrative se, which encode referents in the speaker and addressee sphere have developed what is termed an ‘articloid’ function (Aebischer 1948), which involves the use of demonstratives as markers of definiteness (Laury 1997). According to Diessel (1996) this change involves a change whereby a term featuring a pointing gesture (deictic) becomes used for textual reference (anaphoric) and then to encode a definite referent as identifiable by the addressee (definite). In the case of Spanish and other Romance languages, the articloid phase surfaces in Late Latin texts. 31 (22) CP C I C V (23) a. DP b. DP DP N’ D N DET N DET N That e This book Este Este libro (Abney 1987, Roca 1996) Another consideration is the fact that functional elements such as determiners are closed classes requiring a single complement. DP’s are projections of determiners with an NP as complement. In other words, the NP supplies the descriptive content which a determiner lacks thereby creating a definite referring expression. However, the relationship between articles and demonstratives within the DP remains to be seen. In order to illuminate this state of affairs it is helpful to focus on a construction that features both of these elements, namely an article as well as a demonstrative. Spanish like English features a prenominal demonstrative system (ex. This book, Este libro). However, a further option is available in Spanish where the demonstrative appears in post-nominal position: (24) El libro este ART-the book DEM-this 32 Given that both elements, the definite article and the demonstrative in the above example appear together, it is unlikely that they occupy the same position in the syntactic structure (Cornilescu 1994). Yet in order to elucidate the structure of the DP it must be determined where said elements are generated within the projection, in addition to what derivation or movement is needed to constitute the present string. Along these lines it is interesting to note a parallel construction to (24) which involves the possessive (25) a. Mi/tu/su libro POSS N-book b. El libro mío/tuyo/suyo ART N-book ADJ-poss Here in (26b) the article head construction involves a noun modified by a postnominal possessive adjective. In contrast the article does not appear in the prenominal placement in (25a). The history of Spanish as well as other Romance languages exhibits a construction wherein the article and possessive adjective are contiguous in a string: ART-POSS-N. (26) a. Old Spanish b. Italian c. Portuguese La su casa la sua casa a sua casa ART POSS N ART POSS N ART POSS N This type of construction illustrates the position of elements within the DP where the article heads the projection. In an earlier stage of Spanish in addition to many Modern Romance language such as Italian and Portuguese allows for more elements within the DP than Modern 33 Spanish which has more constraints whereby an intervening element cannot appear between the article and the noun. This same phenomenon applies to the Spanish postnominal demonstrative construction where the relationship between the article and demonstrative in addition to the site where these elements are generated in the structure needs to be clarified. Literature dealing with the Spanish demonstrative and similar constructions has arrived at varying solutions. The theoretical frameworks used to analyze the post-nominal construction can be divided roughly into two groups: noun raising and demonstrative moving models. Within this second group demonstratives are generated either nominally or post-nominally and then moved accordingly. Brugé 1996 The first of these accounts, Brugé (1996) derives post-nominal demonstratives via noun- raising. For example, Brugé (1996) generates the demonstrative in situ and later moves the noun to the left to give the post-nominal construction: (27) a. Este libro b. El libroi este ____i c. El libroi interesante ____i Importantly, Brugé 1996’s analysis treats the demonstrative on a par with the adjective in that both these elements remain in situ while the noun is generated in post-demonstrative and adjectival position and then can be moved into a higher position. In addition, the author argues that demonstratives are encoded with both [+ref] and [+deic] features. Those with a strong [+ref] feature check in Spec, DP before spell-out, moving overtly. In contrast, demonstratives with a 36 In addition, Bernstein (2001) posits that the article in the post-nominal construction is an expletive that licenses the DP projection itself. A problem with Bernstein 1997&2000’s account is that while structurally parallel to ce livre-ci the Spanish el libro ese the postnominal demonstrative is prosodically weak (contra Bernstein 2001) as opposed to the locative portion of the French and Italian constructions in (29). Panagiotidis 2000 Panagiotidis (2000) bases himself on Modern Greek, which features co-occurring articles and demonstratives: (30) a. Afta ta nea fenomena DEM-this ART-the new phenomenon b. Ta nea afta fenomena ART-the new DEM-this phenomenon c. Ta nea fenomena afta ART-the new phenomenon DEM-this ‘This new phenomenon.’ While not strictly parallel to Spanish, Panagiotidis points out that in Greek the article is also obligatory with the post-nominal demonstrative. Therefore he base-generates the demonstrative in post-nominal position, which is then moved into SPEC DP: (31) [SPEC DP aftosi [D’ to [NP andras ti]]] 37 An analysis, which he then extends to Spanish in 14): (32) a. [SPEC DP e [D’ el [NP hombre este]]] b. [SPEC DP estei [D’ e [NP hombrei ]]] In said analysis the post-nominal position is properly basic whereas the default pronominal placement is derived via raising the demonstrative. Following Brugé (1996) he posits that features [+deic] and [+ref] are involved in the raising process. To this end, Panagiotidis (2000) notes that only the pronominal construction is felicitous in an expressively deictic context. (33) a. Thelo afto to apaho butaki V-1sts-want this the lean joint b. *Thelo to apaho afto butaki V-1sts-want the lean this joint c. *Thelo to apaho butaki afto V-1sts-want the lean joint this “I want this lean joint.” The example he gives in (34) takes place in a butcher shop where the customer literally points to a particular cut of meat he or she wishes to purchase with a gesture. In this context only the pre-adjectival form of the demonstrative is felicitous. This difference lies in the deictic import communicated by the expressions at hand, where the pre- adjectival placement in (34a) is the only constructions encoding a strong deixis [+deic], which involves a pointing gesture. On the other hand, the post-adjectival and post-nominal positions have features of weak deixis used for referential purposes and identifying discourse antecedents where no overt pointing is needed. 38 This distinction is important in that the same analysis can then applied to the prenominal vs. postnominal alternation of the demonstrative in Spanish. Where the prenominal variant features a deictic feature in first position [+deic] as shown in (31) in contrast to postnominal where this content appears later. Thus the deictic import [+deic] is emphasized in the prenominal position as opposed to the prenominal, similar to the Greek example in (33). According to Panagiotidis 2000, in Spanish when a strong deictic feature [+deic] is present it attracts the demonstrative to SPEC of DP. Otherwise an operator licenses the D head as an article which remains phonologically null if the demonstrative is moved to SPEC of DP: (35) a. OPexp el hombre este b. Estei e hombre ti [+deic] In the spirit of syntactic parallelism of Abney (1987), Panagiotidis (2000) likens demonstrative movement within the DP to wh-movement within the CP. In Panagiotidis 2000’s account the demonstrative is originally generated in post-nominal position and then moved: Estei e hombre ti. Panagiotidis 2000’s account that the demonstrative is the element that undergoes movement seems the most probable option (contra Brugé 1996). In addition, the use of deictic and referential features to explain demonstrative placement. However, assuming that the post-nominal position is marked it would not be a natural place for the derivation to begin (contra Panagiotidis 2000). 41 suggested by Cornilescu (1994). In this regard, both authors call for the article to appear in SPEC of DP (39a). (39) a. [SPEC DP el [D’ [NP libro]]] b. [SPEC DP [D’ este [NP libro]]] In this schema the definite article and demonstrative appear in different layers of the DP, the article as a licenser in SPEC DP and the demonstrative lower in the structure in D’. Roca (1996) generates the demonstrative prenominally contra Panagiotidis (2000). In this regard the prenominal is the default position for demonstratives not only in Spanish, but in a number of Romance and Germanic languages. (see (13), Chapter 1) Romance: Spanish French Este hombre Cet homme DEM N DEM N Portuguese Catalan Este home Aquest hom DEM N DEM N Italian Rumanian Quest’uomo Acest om DEM-N DEM N Germanic: English German This man Dieser Mann DEM N DEM N (Bernstein 2001) 42 To further illustrate this point, other non-Indo-European languages such as Indonesian exhibit the opposite pattern where post-nominal is the unmarked position as mentioned in Chapter 1: (see (17), Chapter 1 ) Indonesian Demonstratives a. Rumah ini N-house DEM[+prox]-this ‘This house’ b. Rumah itu N-house DEM[+dist]-that ‘That house’ (Johns 1995) The case of Romance and Germanic in (39) contrasted with Indonesian in (17) demonstrate how the default demonstrative position plays an integral role in the language system in which it appears. Given the state of affairs present in the Romance and Germanic data in (39) where the demonstrative appears prenominally it is more plausible to assume that the demonstrative originates in unmarked prenominal position in Spanish along the lines posited by Roca (1996). However, a peculiar Romance case of default post-nominal is that of Rumanian where the definite article is cliticized onto the noun: (40) Lupu- l N-wolf- ART-the ‘The wolf’ 43 Rumanian is different from other Romance languages in that it has the article appear categorically in postnominal position similar to Balkan languages such as Bulgarian and Macedonian as shown in (40). In this case it makes the most sense to generate the article clitic in-situ after the noun rather than move it into position from another generation site. Similarly, the fact that the Spanish demonstrative has its own default position within the language lends credence to the fact that the marked post-nominal must be derived from overt movement of the demonstrative from pre to post-nominal position rather than base generated and later moved into SPEC DP. Therefore, the language specific distribution of functional terms as evidenced in (13) vs. (14) is related to the derivation of these same terms in syntactic structure. Here the normative position is properly basic in that it provides the underlying generation site from whence the special position is derived via movement. 2.2 Demonstratives, Articles, and Adjectives Not only do elements in the DP have a default position within a given language said elements interact in different ways to form a syntactic chain as is exemplified in the contrast between Spanish and Modern Greek. (41) a. Afta ta nea fenomena e. El hombre este DEM ART ADJ N ART N DEM b. *Este el hombre f. Ta nea afta fenomena DEM ART N ART ADJ DEM N c. *El este hombre g. Este hombre ART DEM N DEM N d. El hombre h. Ta nea fenomena afta ART N ART ADJ N DEM 46 unmarked adjective options for pragmatic purposes. Furthermore, adjectives in the marked pronominal position are parallel to post-nominal demonstratives in that they are pragmatically salient and encode qualitative and affective content. These are part of a discourse technique whereby the speaker encodes his/her attitude toward a referent via special syntax. The interaction between elements in the DP varies across languages. For instance, an analysis of Rumanian reveals similar fluctuations: (47) a. Omul acela biet Man-ART DEM poor b. Bietul om acela. poor-ART man DEM c. Acel om biet. DEM man poor ‘That poor man.’ Here the article in (47b), which normally attaches as a nominal suffix, can cliticize onto the adjective when the ADJP is moved into SPEC position. This adjective position is the equivalent of the marked prenominal position in Spanish which has qualitative overtones as shown in (46)&(47) (Agard 1958, Cornilescu 1995). Therefore biet meaning ‘poor’ has two readings depending on where it appears. The adjective gives a literal reading in postnominal positioning and a qualitative interpretation in prenominal position as in the Spanish examples ((45)&(47)). Another parallel with Spanish involves the placement of the demonstrative, which can appear in prenominal and post-nominal position (48). In Spanish post-nominal placement is related to the complementary distribution of AP’s, PP’s along with the demonstrative. 47 (48) Postnominal Demonstrative a. El libro (ese) de sintaxis (ese) ART-the book (DEM-that) of syntax (DEM-that) b. el (nuevo) libro (nuevo) ART-the (ADJ-new) book (ADJ-new) c. el libro (de sintaxis) nuevo (de sintaxis) ART-the book (PREP-of syntax) new (PREP-of syntax) d. el (nuevo) libro (nuevo) (de sintaxis) ese (de sintaxis) Prenominal Demonstrative e. ese (nuevo) libro (nuevo) (de sintaxis) ((nuevo)) (Gutiérrez-Rexach 2002) Following Roca (1996), prenominal adjectives are derived from the default post nominal position within the DP. In contrast PP’s such as de sintaxis are adjuncts that join to the DP, which can then move to a higher position within the structure as shown in (49). (49) [SPEC DP [D’ (AP)i [NP N (AP)i (XP)j DEM [D’ (XP)j … ]]]]] The structure in (49) above is headed by a SPEC DP with an NP followed by DEM D’, which is equivalent to Grohman&Panagiotidis 2004’s DEMlo. The structures in parentheses parallel the optional adjectives (AP) (i.e. nuevo) or adjuncts (XP) (i.e. de sintaxis) as shown in (48). An XP is generated as an adjunct below the DEM D position which can then move into a site above this position. This is in contrast to AP (ADJP), which is generated higher up in the structure than the adjunct XP (i.e. above DEM D’) and can move into prenominal position just 48 below the DP. Both DP and AP are extended projections while PP’s are XP adjuncts. Thus the demonstrative is generated in prenominal position as opposed to the adjective, which is generated post-nominally. While the demonstrative can move into a DEMlo position (Grohman&Panagiotidis 2004) an adjective must move into prenominal position in order to be marked and elicit a qualitative reading. 2.3 DP and Phonetic Form within the Nominal Domain Also of note is the morphological variation between prenominal and post-nominal position in demonstratives such as Rumanian acel vs. acela (47). On par with Italian, Rumanian features special prenominal forms (Quest’uomo). Yet Italian goes further: (50) a. Quegli amici DEM-those friends b. Quei ragazzi DEM-those boys c. Quelle macchine DEM-those cars d. Quello sport DEM-that sport Here the distal demonstrative quello undergoes alterations similar in form to the article, which is diachronically related. To wit, Latin ILLE results in Italian quello via a compound 51 demonstrative. On the other hand, the postnominal demonstrative tends to refer to information already familiar to both speaker and addressing without explicitly pointing it out. It makes the most sense to posit a prenominal generation for the demonstrative, which can undergo movement to the right periphery, thus creating the postnominal construction. Since the demonstrative is a functional term the difference is less striking than in the adjective. Nevertheless, the pre- and post-nominal contrast is not in free variation but has undergone conventionalization in the grammar of the language whereby the postnominal demonstrative is used for previous mention, and common ground. In addition, this postposition has a qualitative content similar to that seen with adjectives such as pobre. The reading of the demonstrative can be pejorative depending on the context as found in both Spanish and Rumanian in (51). (51) a. Omul ǎla e de necrezut. Man-ART-the DEM-that COP PREP-of unbelievable ‘That guy [Bush] is unbelievable.’ (Calude 2002) b. Es increíble el hombre ese. COP incredible ART-the man DEM-that (52) a. El hombre ese ART-the man DEM-that b. Omul acela Man-ART-the DEM-that (Manoliu 1999, Calude 2004, Gutiérrez-Rexach 2002) 52 Here the speaker is talking about Bush in the context of his administration’s foreign policy and the Iraq War. The speaker is commenting on a known entity (president Bush) with a disparaging tone via the placement of the demonstrative after the noun in the syntactic string. The adjective increíble also serves to reinforce this pejorative connotation. 2.4 A Synchronic Analysis of the Spanish DP Along the lines of (Roca 1996) I posit that the article in expressions of type el hombre ese can be analyzed as a licenser for DP projections that spell-out features of definiteness when the SPEC of DP is empty via the movement of the demonstrative to post-nominal position. Following Grohman&Panagiotidis (2004) the node D plays the role of C in the DP. In this framework the article in el hombre este is only an apparent article, which results from the copy- spell out of the demonstrative operator in D, inserted for PF (Phonetic Form) reasons. Structurally the article serves as a marker of definiteness and identifiability that licenses the whole DP projection. In addition, incorporating the observations of Brugé (1996), Panagiotidis (2000), Grohman&Panagiotidis (2004) on the difference between prenominal and post-nominal constructions the features I posit that the feature [deic] has a role in the derivation of the demonstrative position. Therefore in referring expressions with demonstratives a strong [+high deic] feature attracts the demonstrative to prenominal position into SPEC of DP. A lack of such a feature signals a weak deixis [-high deic], which is referential in character [+ref] allowing the demonstrative into marked post-nominal position. The trace of this movement is linked in a chain with SPEC of DP which in order to license the projection spells out the number and gender of the referent as an article. 53 The post-nominal demonstrative is prosodically weak and therefore does not receive stress (contra Bernstein 2001) and in this way it is distinct from constructions such as French ce livre-ci which otherwise exhibit structural parallels. (53) a. Ese hombre b. [SPEC DP _(el)__i [D’ ti [NP N hombre [DEMlo DEM esei]]]]] DP DP SPEC D’ SPEC D’ (DEMhigh) Esei +DEF > (el)i [+high deic] D NP D NP ti N ti hombre N DEMlo hombre esei [+low deic] ([+ref]) The postnominal use is strictly referential in origin, meaning that it refers back to entities within a discourse. Therefore it does not specifically point out a referent with a gesture as in (33) but relies on the addresee to identify the entity being referred to. Thus this referential use is marked by a weak deixis [+low deic] as opposed to strong deixis [+high deic]. The lack of a strong [+high deic] feature (here represented with [+ref] or [+low deic]) in the structure above (53a&b) is significant in that the post-nominal demonstrative is not emphatic contra Fernández Ramírez (1951): 316, Esbozo (1973): 436, Macías (1997) but deictally weaker than the prenominal position. Yet this use lend itself to qualitative functions whereby the speaker signals his or her attitude toward a referent to the addressee. 56 2.6. The Article-less Post-nominal Demonstrative Construction This last CP construction appears, in turn, to be related to a post-nominal construction, restricted to certain registers, which lacks the definite article all together (see Chapter 5 for the diachronic relationship with CP): (57) el corresponsal comenzó diciendo que: "Es difícil hacer ART-the correspondent V-began GER-saying REL-that COP difficult V-inf-make un cálculo, cosa esta a la que ART-a calculation thing DEM-that PREP-to the REL-that debíamos estar más que entrenados V-1st-pl-should V-inf-be more than prepared. ‘The correspondent began by saying: “It’s difficult to make a calculation, a thing which we should be more than prepared for.’ (Noticia: Cuba:CubaNet:Febrero 4, 1998) (CREA) According to Roca (1996), in contrast to the post-nominal demonstrative headed by an article this other construction can only appear as an appositive or a complement. So without a licenser in SPEC of DP this construction cannot properly be a Determiner Phrase and therefore it receives a very limited distribution. A DP must necessarily be licensed in order for it to function as a definite referring expression in the grammar. Therefore the more canonical post-nominal construction the article and the article-less variants are not structurally parallel. (58) a. [SPEC DP(la)i[D’_i [NP televisión [DEMlo esai [CP que] b. [SPEC DP [D’__i [NP cosa ]]]i…[DP esta]i [PP a] 57 The definite article within SPEC of DP in (58a) is a member of a chain involving D’ and the post-nominal demonstrative, which spells out the number and agreement features within the DP. In addition, as a specifier of the DP, the article licenses the DP projection as a whole. In contrast, the article-less variant has no licenser in SPEC DP. The article-less construction (58b), heretofore described structurally as N-DEM, is parallel to a deictic expression in British English which features a post-posed demonstrative preceded by an adjective and a noun: (59) a. A very impudent fellow this. ART-a ADV ADJ N DEM (Goldsmith, Oliver. She Stoops to Conquer, 1773) b. Careless fellow that. ADJ N DEM (Waugh, Evelyn Brideshead Revisted, 1945: p.117) The above construction (59a&b) involves syntactic inversion which consists of a prosodically weak element following the verb: Careless fellow . . . that. However, the English right-lying demonstrative is set off by a pause and is akin to an afterthought not fully integrated into the syntactic regime. Similarly, the construction of the Spanish article-less construction is distinct from the full-fledged postnominal construction with a two-tiered structure (58b) instead of a single unbroken string (58a). Consequently, both articles and demonstratives are fundamental elements within the DP, a projection encoding features of definiteness and deixis that serves to introduce referring expression like NP’s as per Abney (1987). 58 However, such a framework does not take into account the complex interaction between default versus marked positions of articles and demonstratives within the grammar as evidenced across languages such as Rumanian and Indonesian. The question remains whether the DP generates determiners (ex. articles and demonstratives) uniformly across languages, which then are moved into position via feature checking and how language specific constraints control such outcomes. 2.7 Overview of Synchronic Syntax For Spanish I recognize the following constraints: 1) Demonstratives are generated prenominally in D’, thus the marked post-nominal construction is derived via movement to a lower DEM position. Agreement features are then realized through copy spell-out in SPEC DP as a definite article via a chain involving SPEC DP, D’, and DEMlo positions. A minority pattern also exists in which there is no operator occupying SPEC DP but rather a two-tiered structure [SPEC DP [D’ [NP N]]]i…[DP DEM]i. The default prenominal position results from a strong [+deic] feature wherein the demonstrative moves from D’ into SPEC DP to license the projection. 2) Adjectives are generated post-nominally in AP which then can move into marked prenominal position (but below D’) in a direct opposite to the demonstrative. 3) PP’s are adjuncts to the DP, which can then be moved higher up the structure above DEMlo. 4)CP’s are parallel to DP’s in structure. In addition, CP’s can appear as adjuncts. In the Exclamative construction (see 67, Chapter 5) a CP subordinated NP with a postnominal demonstrative has neither an overt or covert licenser in SPEC DP. The minority article-less pattern N-DEM syntactically mirrors CP constructions such as the Exclamative CP- N-DEM and not the full postnominal paradigm, a relationship, which is discussed in Chapter 5. 61 The above formulas in (60) illustrate the intersection between the noun argument A; the context set C and the predicate argument B, which allows for the demonstrative to associate with definiteness. Here the feature proximal relates the demonstratives to participants in a discourse. In the case of este it relates the speaker to the context set. Whereas ese associates with the addressee. In contrast the demonstrative aquel is defined negatively as non-proximal and involves something or someone other than the speaker or addressee. Yet deictics differ from other indexicals since they commonly include a pointing gesture, which is inherent to their deictic character (Gutiérrez-Rexach 2002). However, in contrast to personal pronouns tú and yo demonstratives have an indeterminate content and are therefore incomplete deictic signs (Eguren 1999). A demonstrative contains definite, locative and deictic features. That is to say, a demonstrative refers to a definite referent, which it points out in an area within the deictic sphere. (see (1), Chapter 1) a Este hombre DEM-this N-man Therefore in (1a) the demonstrative este encodes the noun hombre for definiteness [+def], placing said referent as proximal within the speaker’s sphere [+prox], and deictic [+deic] in that the expression points the entity out to the addressee. The placement of the demonstrative stems from discourse techniques meant to encode the cognitive status of a referent. This involves not only a syntax-semantic interface but the relationship between syntax and Information Structure as defined by Lambrecht (1994) in (62) below: 62 (62) Information Structure: That component of sentence grammar in which propositions as conceptual representations of states of affairs are paired with lexicogrammatical structures in accordance with mental states of interlocutors who use and interpret these structures as units of information in discourse contexts (Lambrecht 1994:5) Thus, information structure can be regarded as the manner in which interlocutors package their cognitive representations within linguistic structures. One of the basic elements of language is the structural or syntactical relationship between the constituents that make up an utterance. The interface between information structure and syntax is significant that Spanish, Catalan, and Rumanian feature both prenominal and post-nominal positions for the demonstrative. (see (15) Chapter 1) Spanish: a. Este hombre b. el hombre este DEM N ART N DEM Rumanian: c. Acest om d. omul acesta DEM N DEM-ART DEM Catalan: e. Aquest hom f. el hom aquest DEM N ART N DEM As has been previously discussed on the syntactic portion of my analysis, although it varies across languages generally speaking the prenominal is the unmarked position of the demonstrative Romance languages. Therefore the post-nominal demonstrative construction is a marked construction as defined by Horn (1984): 63 (63) Horn’s Division of Pragmatic Labor The use of a marked (relatively, complex and or prolix) expression when a corresponding unmarked (simpler, less effortful) alternate expression is available tends to be interpreted as conveying a marked message (one which the unmarked alternative would not or could not have conveyed). (Horn 1984:22 cited in Blackwell 2003: 23) The postnominal construction [SPEC DP ART [D’ [NP N [DEMlo DEM]]] is relatively more complex than it prenominal counterpart [SPEC DP DEM [D’ [NP N]]]. Thus the underlying status of a referent encoded in Spanish post-nominal constructions is marked. Moreover, such markedness involves the interactions between different parts of the grammar, which are motivated within the language. The placement of the demonstrative is determined by the model the speaker has of the addressee’s access to possible referents. Yet according to Chafe (1994) there is a limited amount of active referents in a discourse. This situation is due in part to the limits of short-term memory. So in the speaker’s point of view some referents are more active than others. But in order to remain active a referent must be constantly refreshed and actualized. Again following Chafe (1994) activeness measures the focus of attention in terms of the speaker’s discourse model. This involves what the speaker believes the status of the referent to be in the addressee’s mind within the immediate linguistic and extra-linguistic context. Referents range from more active to less active (semi-active), which are termed given and accessible respectively. Therefore the cognitive status of a referent can be encoded via a referring expression that serves as a processing cue to the addressee. 66 cognitive accessibility. In the same way, a first name introduced by a demonstrative: Esa Lola (DEM-N) and a variant with a post-nominal demonstrative La Lola esa (ART-N-DEM) are more salient referents than unmodified Lola (N). A later framework, Gundel et al. (1993) derives a taxonomy for the cognitive status of referring expressions, which fall out due to Grice’s Maxim of Quantity (Grice 1975). The forms that make up this Givenness Hierarchy serve as processing signals to the addressee to retrieve the referent. Also of note is that the statuses of Givenness indicated below are not separate but related via a Horn scale in which each member is implicationally related to the other (Horn 1984). That is to say each status entails all lower statuses from most restricted (in-focus) to less restricted (type identifiable): Type identifiable a N Referential indefinite this N Uniquely identifiable the N Familiar that N Active that, this, this N In-Focus it Figure 3.3: Givenness Hierarchy (Gundel et al. 1993) According to the Givenness Hierarchy (67) demonstratives are active and familiar as well as encoding various degrees of identifiability. That is to say demonstrative expressions encode referents that constitute shared information within a discourse. Said discourse entities are represented in current short-term memory. In terms of Gundel et. al (1993) English “that” means 67 ‘you are familiar with this, and therefore can identify it’. In a related case, Lakoff (1974) investigates the use of English demonstratives, especially that, in a use she terms Emotional deixis. (see (6) Chapter 1) a. How’s that throat? b. That Henry Kissinger really knows his way around Hollywood. Here in (6b) that encodes a referent, which forms a part of world knowledge. Therefore the referent, Henry Kissinger, is familiar to both speaker and addressee given the discourse and temporal context (circa 1974). The demonstrative serves as a marker of solidarity and shared knowledge whereby an opinion is shared between speaker and addressee. Being able to distinguish between knowers and non-knowers of a piece of information is a precondition to being able to obey the maxim “Don’t tell people what they already know,” a special subcase of which is the rule “Don’t tell people what you’ve already told them.” (Sacks 1971, Oct. 19:9 cited in Levinson 1987:82). This observation is important in that it pertains to referring expression produced in the course of a discourse whereby demonstratives are used to encode entities already introduced into the discourse via previous mention or common knowledge. This use is mirrored in Spanish with the post-nominal demonstrative. Therefore la pasión esa in (65) encodes the identifiable commonplace of the passion of newly married couples. (65) La pasión esa del principio te dura muy poco. ART-the passion DEM-that of-ART-the beginning CL-2ndS-you V-3rdS-lasts very little 68 ‘That beginning passion lasts a very short while.’ (Informe Semanal 27/05/95 TVE1) (CREA) The nature of the ‘passion’ alluded to is never specified given that the speaker assumes the hearer already has access to this information via previous mention or common knowledge. Blackwell (2003) bases her study on Neo-Gricean pragmatic principles following Levinson (1987, 2000) and Huang (2000). In Logic and Conversation, Grice (1975 had stipulated guidelines for any conversation. The most important of these premises are the Maxims of Quantity and of Manner: (66) Quantity: Make your contribution as informative as is required for the current purposes of the exchange, but no more informative than is required. Manner: Be perspicuous, specifically, avoid obscurity and ambiguity, be brief (avoid prolixity), and be orderly. (Grice 1975, Blackwell 2003: 13) Participants in a discourse must therefore balance informativeness with efficiency. If a speaker is not explicit enough the addressee cannot identify a referent. On the other hand, if the speaker produces too extravagantly then the discourse loses its coherence. The tension between these two premises of conversation allows for what Grice (1975) calls conversational implicatures in which the speaker communicates “more than what is literally expressed by the conventional sense of the linguistic utterance” they produce. Blackwell (2003) argues that 71 involving a hat with an object on it. Here the demonstrative pronoun indicates reference with a less salient referent pirulo, earlier in the discourse. Significantly, the referent is referred to again with a post-nominal demonstrative. So the referent is upgraded to a full NP to make sure the addressee can identify the referent. The speaker switches from an application of the I-principle whose aim is to say as little as possible (eso) to Q, which calls for a more explicit reference (el pirulín ese). This is an example of backward anaphora (Blackwell 2003) or cataphora whereby a speaker uses a pronominal expression to refer ahead textually to an entity appearing later on in the discourse. Shifts in referring expressions as found in (68) are characteristic of oral registers where the speaker changes the type of referring expression in an attempt to accommodate the assumed cognitive status of the entity in the addressee’s mind. While the prenominal demonstrative in (69) signals to the addressee that both speakers are already familiar with the friends they are talking about. So one could paraphrase as follows: Those friends of ours. You know the ones who were here yesterday. That is to say the discourse participants have mutual knowledge of the referent in question. (69) Precisamente en casa de estos amigos de los Precisely in house of DEM-those friends of ART-the que estuvieron ayer. CONJ-that V-were here ‘Precisely at the house of those friends who were here yesterday.’ (Blackwell 2003: 131) 72 Another use of the marked post-nominal expression is analyzed in conjunction with the Pear Stories in (70): (70) Entonces aparece un hombre por un camino con su cabra. Then V-3rds-appear a man by a road with his goat ¿Le-le dice algo el hombre ese o? CL-3rds CL-3rds V-3rds-say anything ART-the man DEM-that or? ‘Then a man appears on the road with his goat. Does the man say anything to him?’ (Blackwell 2003: 228) Here the referent hombre is introduced as an indefinite expression un hombre. According to Levinson (1987) one would expect a more minimal unmarked expression for subsequent reference as per the I-principle. Instead the interlocutor produces el hombre ese which is co- referential with the previous referent un hombre. Blackwell (2003) suggests that this marked expression is due to the thematic prominence of the referent hombre as well as a tracking use. Thereby hombre is contrasted with another referent encoded with the clitic le. So in the text there is a man picking pears who is referred to with le as opposed to the man with the goat encoded with el hombre ese. In addition, Himmelmann (1996) refers to such tracking use as immediate anaphora after first mention. Whereby thematically salient referents are introduced into discourse and then referred to again via a full lexical NP. This discourse technique is often used at the beginning of stories. For example: Once upon a time there was a king. This king had . . (Himmelmann 1996). Here the referent a king is introduced and immediately referred to again via the demonstrative 73 NP this king. In this light the Spanish the post-nominal demonstrative has a recognitional use, which encodes a prominent theme in the discourse context proper. While she does explore the discourse use of demonstrative expressions Blackwell (2003) does not concentrate on the main difference between prenominal and postnominal demonstrative use. In her analysis both estos amigos and el hombre ese denote referents that are part of mutual knowledge. Perhaps the only distinction would be that the post-nominal demonstrative signals more prominence in the discourse. Brizuela (1999), another study examining Spanish referring expressions, finds that post- nominal determiner (el coche ese) is a marker of high accessibility in comparison to pronominal demonstratives (ese) and noun with articles alone (el coche). In this light, post-nominal demonstrative placement is used as a means of grounding discourse referents with the addressee in mind (Traugott 1995). Though not exhaustive these hierarchies are importance since they demonstrate how the form of a referring expression depends on the assumed cognitive status of a referent, which the speaker wants to communicate to the addressee (Gundel et al 1993). I argue that the post-posed demonstrative is a marker of high accessibility and shared knowledge in addition to speaker attitude and involvement. This use is part and parcel with discourse factors such as speaker-hearer solidarity and shared knowledge (Lakoff 1974’s emotional deixis), which go well beyond the taxonomy presented by Ariel (1990), Gundel et al. (1993), and Blackwell (2003). That is to say, it is important to take into account not only conversational implicatures in the traditional sense but the affective content of a referring expression as such. In fact, much of the literature on Spanish sees the post-nominal demonstrative construction as pejorative in nature (Ramírez 1951:316 Esbozo 1973:436, Macías 1997). 76 Referential > Affective. Thereby a referential use encoding old information then develops accordingly as a way to communicate meta-referential information about the discourse entity at hand. This is a discourse strategy whereby a speaker communicates a qualitative judgment about a referent via a marked position as further evidenced by adjectival placement in (46). A postnominal adjective is the default position whereby a post-posed adjective communicates a more nuanced subjective comment about the noun it modifies. (see (46), Chapter 2) a. El pobre hombre. ART-the ADJ-poorN-man ‘The poor man.’ b. El hombre pobre. ART-the N-man ADJ-poor ‘The poor man.’ Prenominal pobre denotes poor’s canonical meaning as lacking in wealth and resources while post-nominal pobre encodes the qualitative sense of unfortunate and miserable. Significantly the second reading is implicated by the first since certain attributes are associated in the mind with a condition of poverty. In English these two meanings are disambiguated via context whereas Spanish has conventionalized pre vs. post-nominal use for some adjectives. Similarly, by post-posing a demonstrative the speaker implicates a certain cognitive status of a referent as well as their affective stance vis-à-vis the object in question. Taxonomies dealing with information status of referring expressions need to take into account the dyadic structure of a discourse including speaker and addressee along the lines of the Neo-Gricean 77 Framework. However, care should be taken to investigate phenomena such as the post-posed demonstrative in Spanish, which has a parallel use in Rumanian, and Catalan. As evidenced by the Spanish data, post-nominal demonstrative placement does not have a clear-cut status but tends to encode accessible, salient entities within the discourse, which interact with the semantics of the referent, the speaker’s model of the addressee’s sphere, as well as the discourse context proper. Therefore both marked and unmarked positions in the syntax interact to encode discourse referents and the speaker’s attitude towards them. These type of phenomena need to be taken into account in future frameworks in order to capture the dynamic nature of referring expression within the discourse dyad of speaker and addressee. One of the first proposed frameworks, Prince (1981), has a direct bearing on the Spanish data: Evoked Textually > unused > inferable > brand new Evoked Situationally Figure 3.4 (Prince 1981) Here she draws a crucial distinction between textually and situationally evoked status. That is to say a referent is introduced in the discourse and is either brand new or inferable from previous mention. In contrast, the discourse situation also includes referents, which are directly accessible in a physical sense termed deixis ad oculos by Bühler (1937). In addition, for Prince (1981) Givenness is important in that it is based on the relaying of information that the speaker estimates the hearer can interpret felicitously; that is: retrievable from previous context, in consciousness of hearer, and as a part of shared knowledge. 78 Along these lines, Clark and Marshall (1981) elaborate on Prince 1981’s Givenness framework to give a taxonomy of types of mutual knowledge: (73) Mutual Knowledge a. Community membership mutual knowledge Assumed knowledge plus a system of inference. b. Physical co-presence mutual knowledge Knowledge speakers have by attending to physical situation they are in. c. Linguistic co-presence mutual knowledge Previous mention in discourse. Givón (1983) further refines types of shared information of shared information available to speaker and addressee. (74) Givón 1983’s Definition of Discourse Context a. Generally shared knowledge coded in the culturally shared lexicon and known semantic likelihoods. b. Specifically shared knowledge of the particular discourse, what was said earlier and various inferences thereof including verbal and non-verbal feedback. c. Specifically shared knowledge of the particular speaker and hearer, what they know or tend to assume about each other, their respective knowledge, motivation, and propensities, not excluding possible telepathy, however unlikely on general grounds. (Blackwell 2003: 71) Lambrecht (1994) subsumes these referents under the header of Identifiability. Under this are those referents, which are identifiable. Identifiable referents can said to be inactive, active, or merely accessible. In which case they are accessible textually, situationally, and inferentially. 81 Figure 3.6 (continued) Situationally Accessible- Clark & Marshall (1981) b), Prince (1981)- Evoked situationally, Lambrecht 1994- Situationally accessible Affective- Lakoff (1974)- Emotive deixis In turn these categories can serve as the basis of a taxonomy covering the uses of the post-posed demonstrative construction. A taxonomy along these lines will be elaborated on the basis of synchronic data in the following section. 82 CHAPTER 4 SYNCHRONIC ANALYSIS 4.1 Methodology I carry out the investigation via CREA an online corpus of synchronic Spanish data from Spain and Latin America accessible via the Real Academia Española (The Spanish Language Academy) website www.rae.es. The 20th to the 21st century texts found in CREA range from novels (see (88)), newspapers, television (see (65)), radio (see (90)), in addition to speaker interviews (see (92c)) and transcriptions of conversations (see (82)). As such CREA consists of a wide birth of both oral and written material. Another source of synchronic data is the much smaller database www.corpusdelespanol.org by Mark Davies, which includes data up to the year 2003. In addition, diachronic tokens are provided via CORDE a diachronic counterpart of CREA also available from the Real Academia website: www.rae.es that features texts from the 13th to the 19th century. The alternative corpus www.corpusdelespanol.org also includes range of texts from the same period. The character and range of these texts is discussed in Chapter 5 in Diachronic Analysis portion of the dissertation In terms of their use interface the corpora CORDE and CREA allow for search options determined by: country of origin, text genre, and century. However, in order to search for a particular construction one must enter the query exactly (i.e. el señor este.) 83 The drawback to this search engine is that one cannot search for a construction in more general terms via grammatical categories. Also the absence or presence of an accent can yield completely different results. Take, for example, este vs. éste, which are two forms of the demonstrative. Here the more sophisticated search engine in www.corpusdelespanol.org offers an advantage in that it allows for the user to input wild-card codes for parts of speech: * N- noun, *.V- verb etc. In conducting a search for tokens of el señor este one has the option of inputting: *.DET_ART_DEF *.N *.DET_DEM. A drawback in Corpus is that it provides very little textual context in data results. Therefore one gets the following result after conducting a search: (77) o no me acuerdo ya bien cómo estaba el asunto or NEG REFL-1st-s V-remember already well how V-was ART-the situation ‘I don’t remember how the situation was el caso fue que ella mató a sus hijos para ART-the case V-was COMP-that PRO-3rd-s V-killed PREP-to POSS-3rd-s children PREP-of It turned out that she killed her children que no le estorbaran digamos, para sus relaciones COMP-that NEG CL-3rd-s V-thwart V-say PREP-for POSS-3rd-s relations So that they wouldn’t get involved with her relationship con [el señor este] CONJ-with ART-the man DEM-that With this man. Enc. El caso de Medea. Enc. ART-the case PREP-of Medea Surveyer: Like Medea. Inf. --Y entonces, los mató y los echó a Inf. CONJ-and then CL-3rd-pl V-killed CONJ-and CL-3rd-pl V-threw PREP-to Informant: And then she killed them and threw them una canasta. ART-indef basket 86 V-was ART-the matter ART-the case V-was COMP-that PRO-3rd-s V-killed PREP-to POSS-3rd-pl Exactly how the situation was. The case was that she killed hijos para que no le estorbaran digamos, para sus children PREP-for COMP-that NEG CL-3rd-s V-thwart V-say PREP-for POSS-3rd-pl her children so they couldn’t get in the way, let’ say, relaciones con [el señor este]i El caso de Medea. Relations CONJ-with ART-the man DEM-this. ART-the case PREP-of Medea With her relations with this man. Just like the story of Medea Y entonces,los mató y los echó a una canasta. CONJ-and then CL-3rd-pl V-killed CONJ-and CL-3rd-PL V-threw PREP-to ART-indef basket And then she killed them and threw them into a basket. El mismo caso de Medea, que los echó en una fuente. ART-the same case PREP-of Medea COMP-that CL-3rd-pl V-threw ART-indef fountain The same thing as in Medea who threw them into a fountain. Bueno, el caso fue que ella se deshizo de sus Well ART-the case V-was COMP-that PRO-3rd-s REFL-3rd-s V-unmade PREP-of POSS-3rd-pl It turned out that she got rid of her children hijos, ¿no?, los mató, y total el señor este la dejó children neg CL-3rd-pl V-killed CONJ-and total ART-the man DEM-this CL-3rd-s V-left right? She killed them and to top it off the man left her. o no se volvió a aparecer, y ella se volvió or NEG REFL-3rd-s V-returned PREP-to V-inf-appear CONJ-and PRO-3rd-s REFL-3rd-s V-turned or disappeared and she went loca, y dicen que desde entonces, de remordimiento 87 mad CONJ-and V-say COMP-that from then PREP-of regret crazy. They say that ever since then que el alma de esta señora anda penando hasta COMP-that ART-the soul PREP-of DEM-this lady V-go V-GER-suffering PREP-until This woman’s soul goes about suffering from regret la fecha, y que grita: "¡mis hijos!" Ahora, ¿por qué ART-the date CONJ-and COMP-that V-yells POSS-1st-s children Now PREP-for WH Up to our time. And she yells: My children! Now why le dicen a las mujeres las lloronas en los cantos? CL-3rd-s V-say PREP-to ART-the women ART-the crying PREP-in ART-the chants Why do they call women “the one that crying’ in songs? ¿Las lloronas? Sí, porque hablan :“¡llorona, lloroncita!” ART-the crying-ones yes because V-say crying- crying-one-DIM Crying ones?. Well it’s because they say: Crying woman, crying woman.” (ORAL, México, Mujer de 25 años. Trabaja en una agencia de viajes. Estudios: High School y Artes Plásticas) (CREA) The additional context helps to identify el señor este- ‘this man’ with the anaphor el señor – ‘the man’ as well as provide the subject matter of the text where the individual in question forms part of a Mexican legend La Llorona- about a ghostly apparition of a crying woman. A weakness that both corpora and search engines share are the misleading results elicited by even the most detailed of searches. Therefore: 88 (79) & por onrra & fiesta deste su njeto fezo CONJ PREP-for honor CONJ feast PREP-of-DEM-this POSS-his grandson V-3rdS-made ‘And to honor and celebrate this grandson of his the lord had a feast on this day el señor este dia esta fiesta aqui se ajunto ART-the lord DEM-this day DEM-this feast here REF-3rdS gathered mucha vianda segund su costunbre much meat according-to their custom Here a lot of meat was gathered according to their custom.’ (Gónzalez de Clavijo, Ruy. Historia del gran Tamorlán. 14th) (CORDE) In contrast to the search result in bold, the demonstrative este in (85)is not co-referential with el señor but modifies día. Here mere surface linearization is not sufficient for an accurate culling of data. The relevant data can only be isolated via manual revision of search engine results to weed out random sequences from legitimate ones. For the Synchronic investigation I investigate the demonstratives in groups Este/ese and aquel. This grouping stems from the deictic import of the demonstratives in question. Proximal este and medial ese are characterized as referring to elements within the sphere of discourse while aquel encodes information outside this sphere (see Figure 1.1 Chapter 1). This contrast in turn affects the interpretation of the demonstrative expression in discourse. In this regard este/ese share a similar distribution, which is studied via Informational Categories in 4.2 followed by Structural Categories in 4.3.