The Computational Aesthetics of Kojiki : the Machine Learning Approach for Regenerating the Primitive Japanese Text, and the Relationship between Machine and Authorship.

Literature have influenced and shaped humanities since human’s imagination have broaden the horizon of language beyond a communication platform. Literary text has a metaphoric layer which are based on human’s conceptual mapping, visual mapping, or the combination of both (Nehaniv). This complex network of values and thoughts within literature reflects its critical role, which is to capture the societal phenomenon in the artistic way. As the society accelerated into the world compiled by digital information and machines, the way of crafting literature has evolved. The digital humanities scholars have explored and reinvented the new aesthetics at the intersection computational power and human values by creating machine-human hybrid literary pieces. In this paper, the author aims to use one of the most profound Asian literature and the oldest manuscript in Japan that aesthetically capture the hierarchical and societal landscape of the 8th century Japan, Kojiki to examine the relationship between author and computer. There is a similarity when a computer creates a new piece literature based on its reference text and the way that human writes story based on his/her prior readings. Therefore, the only distinction between text generated by human and computer are the source and amount of references. Thus, the idea of authorship is not determined by the type of agent, but the role that he/she/it/they play in crafting the literary piece.