The purpose of this Project is to develop a macroanalysis and distant reading of Spanish Golden Age Sonnets, from the Renaissance Era (Garcilaso de la Vega) to the poetry from the end of the Baroque period (Sor Juana Inés de la Cruz). Computational methods will be used in order to perform the analysis, so that the main recurrent traits —both metrical and semantic— can be detected and singled out.
Unlike previous studies, our aim is not to analyze the limited number of sonnets that have been standardized as ‘canonical’ poetry, but rather to achieve the identification and characterization of those literary traits, both metrical and semantic, that all the sonnets from the 16th and 17th centuries share (Navarro Colorado, 2015, 2016), using computational techniques.
The latest advances in natural language processing and text mining are giving way to new approaches to the literary text. The most significant ones are the so-called distant reading (Moretti, 2007, 2013) and macroanalysis (Jockers, 2013). Contrary to more traditional literary criticism methods, which focus on a detailed analysis of a very small amount of texts, distant-reading and macroanalytic methods propose the analysis of very large literary texts corpora, with the purpose of detecting and defining the general literary traits that characterize all the works written in a certain period.
Far from being opposite models, both approaches are perfectly compatible and complementary. Only by knowing the conventions from a literary period will we be able to determine which the specific features that characterize a particular author from that period are. In general terms, it is necessary to be familiar with the cultural and literary features of both the production and the reception contexts to correctly interpret and study a literary work within its right coordinates.
Project Goals and Objectives
- To study and analyze the basis of the distant-reading method, or computational-based macroanalysis, and determine its relations with other literary criticism methods.
- To compile a large corpus of Golden Age Sonnets that is representative of all the production for the whole period, and to annotate all the metrical information for each verse in a semi-automatic way.
- To develop and optimize an automatic scansion system for Castilian Spanish that is able to extract the metrical pattern of each verse in the corpus.
- To detect significant literary and metrical traits shared by the sonnets in the corpus, previously annotated, using natural language processing techniques and also text mining procedures, both basic (frequency values) and advanced (topic modelling (Blei, 2012)).
- To analyze and compare results with other Golden Age poetry studies, and establish thus a list of all the regular and recurrent characteristics regarding metrics, semantics and topics, shared by all the Golden Age sonnets.