Syntactic variation of object arguments in Lime{u00F1}o contact varieties

Abstract

Generative linguists have mostly been concerned with either idealised data sets that nicely fit their theory, or an idealised relationship between speakers and their homogeneous speech community. Lexical-Functional Grammar is different from other formal theories. In allowing for a less rigid theory, structures and constraints can be linked to incorporate linear order and information structure. This constraint-based theory accommodates variation and change, it allows for my descriptive analysis of nonstandardised variation. The goal of this dissertation is to explore the complex relationship between differential object marking and clitic doubling in nonstandardised variation data from Lime{u00F1}o Spanish contact varieties (LSCV). The main focus of this study is on the microvariation of the person three clitic paradigm cooccurring with extended differential object marking in nonstandardised linguistic phenomena. In particular I focus on 'strange lo', a featureless and invariate form, which is part of the feature specifying third person direct object clitic paradigm, used to optionally crossreference animate and inanimate objects in LSCV. Contact speakers use a double object marking strategy that allows them to use either marked or unmarked forms. Dynamic competition of these morphosyntactic forms leads to competing grammars, a process that contributes to language change. Both mark essentially the same grammatical relation, however motivated by a difference in semantic and pragmatic strategies. In this work I argue that nonreferential grammaticalised direct object clitics are a vital part of a pragmatic marking strategy to mark the direct object in a monotransitive clause as the secondary topic. The motivation for these grammaticalisation processes is analysed as information packaging taking also into account the correlation between contact and inherent changes as trigger factors. Ultimately the syntactic strategy is twofold, firstly to mark the secondary topic of a monotransitive clause and secondly, to compete with the indirect object for the primary object space. The latter may lead to a new accusative case marker which would imply a typological change in monotransitive constructions only. This proposal is supported by similar findings from typologically related and related languages such as Hindi/Urdu, Persian, Basque, Ibizan Catalan and Mexican Spanish. The major linguistic elements for the historical development of the third person object clitics and object marking are multiple grammaticalisation processes of clitics and the differential object marker a, inherent diachronic instability of the clitic paradigm and prolonged contact mediated by Andean Spanish. All factors combined provide the setting for the variation found that may ultimately lead to change. An example for completed change would be the single clitic paradigm found in Ecuador and Paraguay. My hypothesis is based on two competing but not mutually exclusive theories. Alsina (1996b) proposes a binary distinction of [DAT+] and [DAT-] for Romance languages such Catalan and Spanish, where both objects are internal arguments, direct functions and distinct from the external arguments subject and oblique. In that case, the unmarked object is the semantically restricted theta role bearing object. In a new theory of differential object marking (Dalrymple & Nikolaeva, 2007), the differences between marked and unmarked objects are analysed as different information structures expressed in syntactic terms representing two different grammatical functions. Different from semantically-based theories, this new approach takes the communicative context into account. In these terms, marked direct objects can be analysed as the primary object and the secondary topic in a monotransitive clause. Both analyses are not mutually exclusive but complementary, signaling different stages of language change.