Journal of Historical Linguistics - Volume 14, Issue 1, 2024

While previous studies have analyzed the changing nature of polarity items (PIs) in Latin (see Gianollo 2018, 2020) and the licensing conditions of PIs in modern languages (see Homer 2021), less research has analyzed the diachronic behavior of PIs in the development of the Spanish language. The present study takes a quantitative approach to historical corpus data in showing that in older varieties of Spanish, there was an increased degree of competition between items of the alg- series (i.e., alguno ‘some’) and items of the n- series (i.e., ninguno ‘none’) in negated clauses which later decreased as the language entered its modern age. We find that the competition between these items in negated clauses is influenced by factors such as register, the syntactic role of the PI, and activation status (following Larrivée 2012, 2017). These data provide quantitative support for Martins (2000), who suggested that earlier forms of Spanish exhibited more versatile licensing conditions of PIs, and that this variation gradually decreased over time due to a greater salience of the n- series in negated clauses. In total, the present work aims to use corpus data to connect historical linguistic research to theoretical approaches regarding the contemporary usage of PIs.

Persian has a polyfunctional case marker, ‘rā’, which diachronically varies greatly in the range of functions it covers. In this paper, we give an account of different case functions of ‘rā’, in New Persian, an era from the 8th century (C.E.) to present. To analyze the functionality of ‘rā’ in different texts, we selected 78 books from the New and Contemporary Persian eras and studied one thousand tokens of ‘rā’ from each century. The data show that ‘rā’ has been a polyfunctional case marker in New Persian, marking about 13 different case roles. Its main role was to mark direct objects, and gradually it has become its sole function in Contemporary Persian. However, during the time span, some of the ‘rā’-marked roles remained constant and some of them replaced ‘rā’ with other adpositions. We follow a historical semantic map approach as a typological grid to examine our data. The findings show that ‘rā’ has shifted from animate to inanimate concepts gradually. While in the 12th century about 750 out of 1000 (about 75%) roles marked with ‘rā’ were animate, it has decreased to about 400 out of 1000 (about 42%) in the 19th century and less than 30% in the 20th century. Our data show that ‘rā’ has not gone further to mark inanimate relations, and it has gone toward core case roles, specifically direct object.

This study sets out to discuss the evolution from oblique to core case as a manifestation of overtly-marked nominative-accusative alignment in Sinitic languages. This is due to the emergence of a type of ‘optional’ marking on preverbal direct objects in a construction type that has become widespread in Sinitic (Chappell & Verstraete 2019). In particular, I examine spoken discourse data from Taiwanese Southern Min whose comitative preposition, ka⁷, has grammaticalized into an optional object marker. It is argued that this marker is undergoing morphologization into a direct object index (doi) on the main verb in the predicate, subsequent to the omission of the resumptive pronoun it governs. The new index takes over this function of cross-referencing the lexical direct object, typically located in the immediately preceding discourse, if not in clause-initial position.

In an epilogue, I also briefly treat the evolution of local cases such as the allative and the perlative to optional object markers in the Southern Min languages of Shantou and Jieyang, situated in Guangdong Province, China. Both of these are extremely rare sources in the Sinitic family, yet common in Tibeto-Burman and Romance languages. The approach adopted is in harmony with recent diachronic studies which target source morphosyntax in order to explain the emergence of a variety of synchronic patterns, all bearing similar discourse and grammatical functions (Cristofaro & Zúñiga 2018).

Besides highly grammaticalized analytic verb forms that constitute the system’s main tense, aspect, and mood forms, Basque has a handful of less grammaticalized periphrases for secondary aspectual and modal meanings. In both older and more recent texts, some of these periphrases have been reanalyzed as monoclausal and readjusted in accordance with the auxiliated verb’s argument structure. This readjustment or actualization process involves changes in two respects: case-marking and indexation through auxiliary change. The reanalysis and actualization of theses periphrases seem to be driven by analogy with highly extended and frequent analytic verb forms. With regard to their actualization, it seems to depend on three factors: word order, valency of the auxiliated verb, and plural patient agreement.

Case marking alignment has been assumed to reflect principles of optimization: dedicated case marking is limited to arguments more in need of disambiguation, and semantically or pragmatically similar arguments are encoded by the same case forms. This view is based on the synchronic properties of the relevant alignment patterns and the cross-linguistic rarity of other logically possible ones, not diachronic phenomena involved in their emergence or cross-linguistic distribution. This paper explores several developmental processes that recurrently give rise to accusative, ergative, and active case marking alignment cross-linguistically, including reanalysis of argument structure, the development of case forms through grammaticalization or phonological reduction, and the extension of an existing case form to novel contexts. These processes appear to be driven by inherent or contextual properties of particular source constructions, independent of principles of optimization in the use of case marking. The synchronic properties of the resulting alignment patterns cannot be taken as evidence for such principles either, because they are due to inheritance (a case form inherits the distribution of particular source elements or developmental processes, which is unrelated to the assumed optimization principles) or residue (a case form becomes restricted to particular arguments as a new form develops for the other arguments, also independently of these principles). These facts call for a source-oriented approach to case marking alignment and recurrent cross-linguistic patterns in general, one where the focus shifts from the synchronic properties of individual patterns to unraveling the effects of several different diachronic phenomena that give rise to individual patterns and shape their cross-linguistic distribution over time.