“Reuse” of Biblical Quotes in Swedish 19th Century Fiction

Dimitrios Kokkinakis and Mats Malm Introduction Multifaceted relations between texts can be complex, abstract, diverse or subtle. Digital humanists are interested in identifying pairs of text passages likely to contain substantial overlap and empirically supporting (hopefully) new interpretations of historical texts. For instance, Cordell [3] discusses how digital interpretive tools can help make better sense of enlarged bibliographies, and the continuous digging into digital archives promises to effect exciting revisions to our literary history. Intertextual similarities [12] between historical texts embrace a larger set of morphological, linguistic, syntactic, semantic and copying variations, thus adding a complication to text-reuse detection. Recycled text chunks are frequently only small portions of a document and may be significantly modified [4,5]. Older language variants and dialects are less standardized; their evolution spanning centuries [1]; unlike today, e.g. verbatim quotes in older texts were not visually enclosed in quotation marks, making it hard for us to discern reuse from ‘original’ text; some authors quote other authors we know nothing about or whose works do not survive. Moreover, spelling and orthographic variations as well as OCR-errors can be problematic for the identification of (historical) text reuse [1,2]. Therefore, the task of detecting text re-use is challenging with NLP having a major role to play in this process. NLP techniques to discover intertextual similarities between historical texts is a major topic of considerable interest among scholars from both a theoretical and practical point of view. Biological sequence alignment is one available method also used for detection of similar passages in text collections [2,5,7,9,10]. We use the Pairwise Alignment for Intertextual Relations (PAIR[11]), a simple implementation of sequence alignment for text analysis which supports one-against-many comparisons.