Bare indefinites and distributivity in basque

Argument Selectors, 2019

Basque is a language without known surviving relatives spoken by some 700,000 people in the Basque Country (Araba, Biscay, Gipuzkoa and Navarre in northeastern Spain, and Labourd, Low Navarre and Soule in southeastern France), in addition to some small Basquespeaking communities found in the Americas. It is used by bilingual speakers of all ages, but the highest percentages and/or numbers of speakers are found mainly in non-metropolitan areas of Biscay, Gipuzkoa, and Navarre. There are several regional varieties and a standardized form (euskara batua), which is the one addressed in this study. Basque morphology is largely agglutinative, i.e., it is predominantly concatenative and of separative exponence (except in the person-number inflection of verbs), with some flexivity (i.e., the allomorphy found in inflectional phenomena is not purely phonological) in both the verbal and nominal domains. Basque clauses show double-marking patterns and pragmatically conditioned deviations from the default SOV constituent order. The present paper surveys the argument selectors found in different areas of Basque grammar. Overt coding selectors are presented in Section 2 (with 2.1 covering the relatively straightforward dependent-marking patterns and 2.2 addressing the comparatively convoluted head-marking ones). Section 3 surveys behavioral selectors involving coreference (adverbial clauses, as well as control and raising phenomena), while Section 4 deals with other behavioral selectors (relativization sites, focus constructions, and subject of imperatives, with 1 The second author has conducted her part of this work thanks to support by the following funding schemes / research projects: Bas&Be FF2011-26906, from the Ministry of Economy and Competitiveness (MINECO), AThEME-Advancing the European Multilingual Experience (EC FP7/SSH-2013-1 AThEME (613465)), from the European Commission, and The Bilingual Mind (IT665-13), from the Basque Government. The authors are indebted to Jon Ortiz de Urbina for his help with some of the data, for the references to several studies mentioned here, and for his valuable comments on a previous version of this paper. The usual disclaimers apply. 2 some comments on Basque voice). The conclusions presented in Section 5 characterize Basque as a showing a complex picture regarding grammatical relations. From the perspective of Chomskyan linguistics, unproblematic notions like subject, direct object, and indirect object are instantiated in quite intricate ways, especially by the verb morphology, which makes for an interesting kaleidoscope of coding details related to both the lexicon and the grammar of the language. From the perspective of functionalist-typological linguistics, highly problematic notions like subject and object are shown to be particularly tricky in Basque morphology, but much less so in Basque syntax. 2. Overt coding selectors 2.1 Dependent marking When overt, the NP expressing the single argument of a monovalent verb of simple clauses (S) is most often found in the unmarked absolutive case. In (1), for instance, the S of the verb hil 'die', i.e. gizona 'the man', appears unmarked: (1) Gizon-a hil d-a. man-DET[ABS] die.PFV TAM-TAM 'The man has died.' This is certainly so with patientive monovalent verbs ("unaccusatives"), which align with motion and posture verbs. Agentive monovalent verbs ("unergatives"), by contrast, show some variation regarding how the nonstandard varieties treat them-roughly, their S tends to appear in the absolutive in the east and in the ergative in the west. 2 In Standard Basque, some monovalent verbs like irakin 'boil' have their S appear in the ergative (an "extended ergative" according to Dixon 1979; cf. Ortiz de Urbina 1989): (2) Ura-k irakin d-u. water-ERG boil.PFV TAM-have 'The water has boiled.' Bivalent predicates come in three guises. The ergative-absolutive pattern can be illustrated with ikusi 'see'; the agentive argument (A) takes ergative case marking (-k) while the patientive argument (P) appears in the unmarked absolutive: (3) (Ni-k) (zu) ikusi z-a-it-u-t. 1SG-ERG 2SG[ABS] see.PFV 2.I-TAM-PL-have-1SG.II 'I have seen you (SG).' The absolutive-dative pattern, on the other hand, can be illustrated with gustatu 'please'; the experiencer A bears dative case and the stimulus P is unmarked: (4) (Ni-ri) ardo-a gustatzen z-a-it .

Anuario Del Seminario De Filologia Vasca Julio De Urquijo, 2013

The generative tradition has built up various arguments to enrich the inventory of functional categories inside the noun phrase! beyond the NP-DP basic structure proposed by Brame (1982), Abney (1987) and others; just to mention a few: the existence of clause-like agreement data in Hungarian (Szabolsci 1984 and subsequent work), the co-occurrence of demonstratives and noun-raising to a clitic-like article in languages like Rumanian (Giusti 1997), the interaction of the construct state with the genitive in Hebrew (Ritter 1991), the different ordering of the head noun with respect to the NP-internal adjectives (Cinque 1994), the noun-adjective concord (Carstens 2000). All these proposals have yielded a more articulate structure of the noun phrase and all have in common that some functional category beyond D, be it Number and/or some kind of nominal Agreement, is proposed to capture the relevant languageparticular and cross-linguistic generalizations. With respect to Basque, the progress towards understanding the internal structure of noun phrases has been more modest, probably because Basque generativists have paid more attention to other, more salient, aspects of the language such as the pluripersonal verbal agreement, the explanation of ergativity, the topic-focus field, and so on. Recent works that tackle the structure of the Basque noun phrase (e.g. Eguzkitza 1993, Laka 1993 and Artiagoitia 1997, 1998) stop short of providing any languageinternal evidence for other functional category beyond D itself In this article, I intend to make a small contribution to the understanding of the Basque DP: I shall argue for the existence of an intermediate functional projection lying between N and D. The argument is based on Longobardi's (1994, 2000a) work; I will show that some Basque DPs headed by the article must have their D position • Ohore eta atsegin handia da Rudolf-en omen aid ian parte hartzea eta ekarpen xume hau hari eskaintzea. This research is supported by grant PI-1999-18 from the Basque Department of Education, Universities and Research and by grant 9/UPV 00027.130-13587/2001 from the University of the Basque Country. I "noun phrase" is a purely descriptive term; NP stands for the maximal projection of a head noun, which need not correspond to a noun phrase. Abbreviations in this article include: ART = article; AUX = auxiliary; DAT = dative; ERG = ergative; GEN = genitive; Num = number; sg = singular; pi = plural.

A Grammar of Basque, 2003

Leaving aside a small set of verbs (see 3.6.3.), Basque finite verbs are composed of an morphologically independent lexical verb carrying aspectual information, and a clitic auxiliary bearing Tense, Agreement and Modal affixes. The choice of auxiliaries in Basque seems to be largely dependent on the valency of the predicate. Intransitive, transitive, and ditransitive auxiliaries typically correspond to monovalent, bivalent and trivalent predicates. The arguments of the verb (participants in the event, such as agents, themes or patients, and beneficiaries) are mapped systematically by person and number morphology corresponding to grammatical functions such as subject, object and indirect object. Changes in the argument structure of the verb (as in the causative/inchoative alternation) are also signaled in the choice of the auxiliary. However, in some cases the correlation between valency and choice of auxiliary does not, or does not seem to obtain. The mismatch between valency and morphology is due in these cases to the contribution of an aspectual dimension (Grimshaw,90) (see section 1.3). In order to maintain the two domains clear (lexical structure and morphology), I will refer to the valency of the verb with categories such as monovalent, bivalent or trivalent (see Hualde, section 3.6.3 in this book), and to its morphological expression with categories such as intransitive and transitive. The latter are familiar from the structural analysis of basic verbal paradigms (Hualde, in section 3.6.3.) and are based on the presence/absence of ergative morphology. When ergative morphology is present in the paradigm corresponding to a finite form, I will refer to that form as transitive. Otherwise, I will refer to that form as intransitive. 1.1. Case and Agreement patterns Basque is an ergative language in both its case marking system and in its verbal morphology (with splits depending on Tense, see Hualde, section 3.6.3). That is, a language where subjects of intransitive verbs and objects of transitive ones are Casemarked and cross-referenced in the agreement elements of the verb identically and differently from subjects of transitive verbs. The Case marking of transitive subjects is called ergative, that of intransitive subjects and objects of transitives, absolutive. Table 1 shows the absolutive and ergative case endings in the three numbers (singular, plural and indefinite) 1.3. Basque as an "extended" ergative language Ergative languages divide into two different Case patterns. One, exemplified by Dyirbal or Samoan, marks all subjects of intransitive verbs identically, irrespective of the aspectual or agentive properties of the verb or the agentive properties of the single argument. Another one, exemplified by Basque and Georgian, among other ergative languages, marks some subjects of intransitives as objects of transitives (that is, absolutive), and some others as subjects of transitive verbs (with the ergative Case). A pattern that Dixon (1979,1994) calls "extended ergative". This split case marking pattern depends apparently on the aspectual properties of the verb, particularly on whether it is telic or not (see section 4.1. for a discussion on Basque; and Lyell, 1995, p.120-123 for Georgian). In this, Basque seems to express morphologically a distinction that has been noted syntactically in other languages between unaccusative and unergative predicates (Perlmutter, 1978; Burzio, 1981). Subjects of unergative predicates take the ergative Case, subjects of unaccusative predicates take the absolutive Case. In accord with Case marking, unergative predicates take the transitive auxiliary, and unaccusative predicates the intransitive one: (iii) a. Jonek ardoa ekarri du Jon-erg wine-abs bring-partc Aux-T 'Jon brought the wine' b. Jonek saltatu du Jon-erg jump-partc Aux-T 'Jon jumped' c. Jon etorri da 'Jon came' Heldu also means "to ripen, to mature". See next section.

2010

Basque is well known for its complex verbal morphology, where the in-flected verb, i.e. the predicate core, shows agreement with up to three ar-guments (ergative, absolutive, and indirect object). This entails that the syntactic relations between the central participants and the situation core are perfectly reflected in verbal and nominal morphology, by both agree-ment and case. The Basque agreement pattern illustrates the syntactic hier-archy between the central participants in different argument constellations. The argument slots on the top allow of most variation. For those at the end, possible candidates are restricted to third person. Auxiliaries correspond-ing to uncommon argument-constellations fall gradually out of use. As a consequence of the obsolescence of verbal morphology, new strategies for the expression of syntactic relations (and semantic content) emerge. The evolutions illustrated in this paper are symptoms of the typological shift from agglutinative to fusional morphology, and even beyond, as parts of the grammar have entered what may be called “the post-fusional stage”. The influence of Romance contributes to the obsolescence as well as the preservation of morphologically complex forms.