On the Development of Differential Object Marking in Romance

Abstract

Through an investigation of novel diachronic and synchronic data, this dissertation explores and analyses the emergence and development of Differential Object Marking (henceforth, DOM) in Romance. Although DOM has been posited as the output of semantico-pragmatic parameters such as animacy, this study follows in the vein of research that has emerged in recent decades showing DOM to be the output of a cluster of variety-specific structural (viz. nominal, clausal, verbal), semantico-pragmatic (viz. topicality, saliency, referentiality, animacy), and lexical parameters. In chapter 1, I briefly summarise the state-of-the art of the literature regarding DOM in Romance and establish several paucities within said literature that form the basis of the research questions of the present dissertation. Namely, I identify gaps within the literature that fail to account for: (i) the diachronic emergence and development of DOM in Romance at a general level, (ii) the advancement of DOM with nominal parameters along an implicational hierarchy of highly referential, topical, and salient nominals, (iii) the diachronic distribution of DOM in relation to structural, semantico-pragmatic, and lexical parameters, (iv) the diachronic interplay between these parameters and how this affects the realisation (or lack thereof) of DOM, (v) the interplay and possible cause-and-effect relationship between DOM and other syntactic parameters contemporaneous to it such as the Verb Second parameter in old Romance and low lexical V(erb)-movement in modern Romance, and, finally, (vi) the structural and non- structural parameters that underpin the absence of DOM. In chapter 2, I discuss and analyse the distribution of DOM in modern Romance and the extensive literature regarding this phenomenon from the earliest functionalist and prescriptivist accounts to modern minimalist accounts. I identify a D vs. DP split in the licensing of DOM with nominal types and establish strong parallels between non-standard Romance and old Romance in which DOM with salient phrasal nominals (i.e., DPs) is highly restricted in its diatopic and diachronic distribution. Furthermore, I discuss and analyse the importance of Ledgeway’s (2020) North-South Divide that establishes a typology of syntactic behaviours in Romance, deviations from which are investigated in chapter 7 of this thesis. Finally, I establish a correlation between high N(oun)-movement in the nominal core and the availability of DOM in modern Romance that is highly indicative of DOM being unavailable in varieties in which the noun does not target the higher D-area within the DP. In chapter 3, I investigate the emergence of DOM in old Romance with semantico- pragmatic and nominal parameters through an analysis of novel quantitative and qualitative data from four old Romance varieties: old Spanish, old Sardinian, old Portuguese, and old Neapolitan. In particular, I establish that the licensing of DOM with salient nominals indeed follows an implicational hierarchy à la Aissen (2003) and that the semantico-pragmatic parameters to which DOM is sensitive in old Romance —whilst subject to notable diatopic and diachronic variation— can be generalised as those of topicality and syntactic specificity in the nominal. Additionally, I report key empirical data plotting the height of N-movement in the respective diachronies of the Romance varieties under investigation. By means of a comparison of the data regarding N-movement and those regarding the distribution of DOM in old Romance, I identify old Spanish as a striking outlier in which N-movement is systematically high and in which DOM is readily licensed with both salient Ds and salient DPs. In contrast, in old Portuguese, old Sardinian, and old Neapolitan, N-movement is low-to-medial and DOM is not readily licensed with salient DPs. In chapter 4, by means of the novel data presented in chapter 3, I propose that DOM in Romance emerged through the probing of highly referential, topical, (animate), and salient nominals by a TopP functional head in a Giusti-style nominal left periphery (or, KP) (cf. Giusti 1992, 2002; Löbel 1994). Specifically, I argue that diatopic variation in the height of N- movement in Romance in synchrony and diachrony within the DP is responsible for significant diatopic and diachronic variation in the availability of DOM in relation to salient Ds vs. salient DPs (cf. Longobardi 1994; Ledgeway 2015). In other words, I propose that the D vs. DP split observed in modern and old Romance follows from the of the height of N-movement and the probing by the KP of DP1 and/or DP2 within a split-DP (cf. Bernstein, Ordóñez&Roca 2018, 2020, 2021). In chapter 5 of this dissertation, I explore the role of lexical parameters in the emergence of DOM in Romance. In particular, by means of novel quantitative and qualitative data and analyses, I propose a distinction between: (i) lexically-determined DOM as a consequence of the lexical features of highly transitive predicates à la Hopper and Thompson (1980) in Spanish and Portuguese, and (ii) lexically-determined dative licensing (i.e., the lexically determined licensing of indirect objects) in old Neapolitan and old Sardinian which I interpret as the continuation of a Classical and late Latin phenomenon. In chapter 6, I investigate the link between DOM and the proposed V2 parameter in old Romance. I propose that the availability (or absence) of DOM in old Romance follows from the height of V-to-C. That is, I propose that the salient KP’s associated general [+T]opic feature that is licensed upon a goal being provided for TopP can be checked by movement to a specialised and dedicated TopP in V-to-Fin varieties (viz. old Neapolitan and earlier old Portuguese). However, I argue that this [+T]opic feature cannot be checked in V-to-Force grammars (viz. old Spanish, old Sardinian, and later old Portuguese) in which movement to dedicated and specialised Topic projections is unavailable, necessitating the spell-out of this feature at PF and, thus, the differential marking of the salient KP. Finally, in chapter 7, I explore the link between low V-movement and DOM in modern Romance and, in particular, I examine the outlying cases of Brazilian Portuguese and Modenese. In the case of Brazilian Portuguese, I propose that a period of high V-movement along clausal spine in the 18th and 19th centuries is responsible for engendering the loss of DOM and the obligatory spell-out of subjects in [Spec, TP], as highlighted by diachronic data which identify a crossover between the two phenomena (viz. Duarte 1995; de Araújo Adriano 2023). I posit that non-structural factors, such as extensive superstrate contact with Bantu varieties during the Atlantic Slave Trade, further disfavoured the general acquisition of a-marking, a conclusion further confirmed by the increasing loss of a-marking in the marking of indirect objects in Brazilian and African varieties of Portuguese. In the case of Modenese, I argue that DOM is a morphological and paradigmatic residue in the accusative pronoun series following a previous stage of low V-movement between the 15th and 16th centuries in a proposed divergent grammar of early modern Emilian. All in all, this dissertation enriches an exiguous literature on the emergence of DOM at a pan-Romance level and, at the same time, explores and provides new data and analyses in relation to unexpected cases of DOM (or the absence thereof), especially in relation to contrasts between low and high V-movement grammars.