Final Report Summary - EVALISA (The Evolution of Case, Alignment and Argument Structure in Indo-European)
Alignment and argument structure lies at the heart of all current theoretical models in linguistics, both syntactic models and research within typology. This project, EVALISA, represents the first large-scale comprehensive study of the historical development of case marking and argument structure carried out in modern times, using modern linguistic approaches and frameworks, and covering the earliest layers of all 11 branches of Indo-European. More specifically, the project has investigated case marking and argument structure from a historical perspective, in particular non-nominative case marking of subjects, focusing on its development from prehistory to the earliest documented phases of the Indo-European languages.
One of the products generated by this project is an electronically searchable database of verbs and compositional predicates occurring with non-nominative subject marking, a database available to the research community at large, for further research on the topic. Each lexical entry has been coded for case marking of the first and the second argument, as well as for meaning, etymon, language and language branch. Each entry also contains some examples, which are all glossed, translated, and provided with information about context and textual source.
Another product emerging from this project is a methodology for reconstructing syntax and grammar, based on the tools of Construction Grammar. The theoretical framework of Construction Grammar is easily extendible to syntactic reconstruction, due to the basic status of form–meaning pairings in that model, and hence the more lexicon-like status of the grammar. This creates a natural leap for Construction Grammar from synchronic form–meaning pairings to historical reconstruction, based on form–meaning pairings. Within this project, verb-specific constructions have been reconstructed for Proto-Germanic and Proto-Indo-European, verb-class specific constructions for Proto-Germanic and Proto-Indo-European, the Dative-of-Agent construction for Proto-Indo-European, different word order constructions for Proto-Indo-European, involving both basic word order and a clause-initial focus position. A conceptual metaphor has been reconstructed for Proto-Indo-European and its interaction with an argument structure construction existing in the grammar of Proto-Indo-European speakers. Further, the constructional formalism has been shown to be extendible to reconstructions of different levels of schematicity and even different levels of granularity within a construction’s given schematicity level.
In particular, the project has established different Proto-Indo-European sources for non-nominative subjects, one arising through an antipassive-like construction in the development from a semantic alignment system to the well-documented accusative system found in most of the earliest stages of the Indo-European daughter languages. The other has arisen through a process of anticausativization, involving a reduction in valency, as well as the retaining of the object case of the causative on the subject of the anticausative. This type of dependent-marking anticausativization with the anticausative marker found on the subject is more or less unknown in the typological literature on anticausatives, where only the head-marking type, in which the anticausative marker is found on the verb, is discussed. Both of these developmental paths of non-nominative subjects originate in prehistory, yielding the non-nominative subject construction reconstructable for Proto-Indo-European.
A second major achievement of this project relates to attested examples showing that potential non-nominative subjects behave syntactically as subjects not only in the early Germanic languages, as already documented prior to this project, but also in Old Irish, Latin and Ancient Greek. One of the most conclusive evidence for a subject analysis of potential non-nominative subjects involves control infinitives where the non-nominative is left unexpressed, a syntactic behavior generally confined to subjects and not found with objects. The project’s documentation of such examples is a major breakthrough, suggesting that the subject behavior of potential non-nominative subjects in these early Indo-European languages may indeed be reconstructable for Proto-Indo-European.
A third major achievement of this project relates to the interaction between verbal semantics and argument structure in language history, in that argument structure constructions turn out to be pivotal for explaining changes in verbal semantics, as opposed to changes in verbal semantics leading to changes in argument structure which is the general assumption in the field. Subsequently, several obscure verbal etymologizations might benefit greatly from taking argument structure into account. The project has also established a certain degree of directionality in semantic change, on the basis of the interaction between conceptual metaphors and changes in argument structure.
One of the products generated by this project is an electronically searchable database of verbs and compositional predicates occurring with non-nominative subject marking, a database available to the research community at large, for further research on the topic. Each lexical entry has been coded for case marking of the first and the second argument, as well as for meaning, etymon, language and language branch. Each entry also contains some examples, which are all glossed, translated, and provided with information about context and textual source.
Another product emerging from this project is a methodology for reconstructing syntax and grammar, based on the tools of Construction Grammar. The theoretical framework of Construction Grammar is easily extendible to syntactic reconstruction, due to the basic status of form–meaning pairings in that model, and hence the more lexicon-like status of the grammar. This creates a natural leap for Construction Grammar from synchronic form–meaning pairings to historical reconstruction, based on form–meaning pairings. Within this project, verb-specific constructions have been reconstructed for Proto-Germanic and Proto-Indo-European, verb-class specific constructions for Proto-Germanic and Proto-Indo-European, the Dative-of-Agent construction for Proto-Indo-European, different word order constructions for Proto-Indo-European, involving both basic word order and a clause-initial focus position. A conceptual metaphor has been reconstructed for Proto-Indo-European and its interaction with an argument structure construction existing in the grammar of Proto-Indo-European speakers. Further, the constructional formalism has been shown to be extendible to reconstructions of different levels of schematicity and even different levels of granularity within a construction’s given schematicity level.
In particular, the project has established different Proto-Indo-European sources for non-nominative subjects, one arising through an antipassive-like construction in the development from a semantic alignment system to the well-documented accusative system found in most of the earliest stages of the Indo-European daughter languages. The other has arisen through a process of anticausativization, involving a reduction in valency, as well as the retaining of the object case of the causative on the subject of the anticausative. This type of dependent-marking anticausativization with the anticausative marker found on the subject is more or less unknown in the typological literature on anticausatives, where only the head-marking type, in which the anticausative marker is found on the verb, is discussed. Both of these developmental paths of non-nominative subjects originate in prehistory, yielding the non-nominative subject construction reconstructable for Proto-Indo-European.
A second major achievement of this project relates to attested examples showing that potential non-nominative subjects behave syntactically as subjects not only in the early Germanic languages, as already documented prior to this project, but also in Old Irish, Latin and Ancient Greek. One of the most conclusive evidence for a subject analysis of potential non-nominative subjects involves control infinitives where the non-nominative is left unexpressed, a syntactic behavior generally confined to subjects and not found with objects. The project’s documentation of such examples is a major breakthrough, suggesting that the subject behavior of potential non-nominative subjects in these early Indo-European languages may indeed be reconstructable for Proto-Indo-European.
A third major achievement of this project relates to the interaction between verbal semantics and argument structure in language history, in that argument structure constructions turn out to be pivotal for explaining changes in verbal semantics, as opposed to changes in verbal semantics leading to changes in argument structure which is the general assumption in the field. Subsequently, several obscure verbal etymologizations might benefit greatly from taking argument structure into account. The project has also established a certain degree of directionality in semantic change, on the basis of the interaction between conceptual metaphors and changes in argument structure.