The first stated goal of the project consists in a systematic diachronic study of a number of grammatical features of Basque as we know it today. The choice of the specific research topics is guided by three types of considerations: on the one hand we want the aspects under study to correspond to central aspects of the grammar of Basque, so that it will be possible to examine all the associated changes in the system and compare them with wider typological generalizations; on the other hand, we will privilege those aspects of the grammar of Basque that are less common in the context of the well studied European sprachbound; finally, the object of study should be one that allows testing the impact of the new research methodologies
in the studied topics.
The second stated goal of the proposal is to create a comprehensive annotated historical corpus of Basque that will allow for systematic searches in diachronic research and will set a model for future works in historical corpus construction in Basque. This corpus will comprise both part-of-speech and syntactic annotation, and a rich set of metadata structure. Our database will allow us to search the annotated corpus by words, lemmas, grammatical categories, by sequences of grammatical categories, and by specific structural configurations that we will also annotate to ease the use of the database (that is, searches based on constructional notions such as relatives, correlatives, finite subordination and others). The searchable database will be based on a 750.000 word reference corpus, expanding from the XVth century to the mid XVIIIth century, the period encompassing Archaic and Old Basque. We intend the highly complementary research and technical agendas of this proposal to be a showcase of the potential advantages derived from the use of searchable databases and quantifiable data in the domain of Basque historical linguistics.
In order to achieve those goals, we have set the following workflow chart:
As regards the interface for searching examples of the analysed grammatical phenomena, results will be displayed in the transcribed versions of the corpus but those transcriptions will be linked to their original facsimiles, thanks to the application developed in the Testu historikoak (Historical Texts) project, as shown in the example below: