Text Processing Procedures for Analysing a Corpus with MedievalMarian Miracle Tales in Old Swedish

Sep 28, 2026

Upload by Sarah

A text corpus of 101 Marian miracle tales in Old Swedish, written between c.1272 and 1430, compiled digitally from three 19th-century transcriptions. Interpreting the medieval language requires specialized knowledge because vocabulary, spelling, grammar, and rich morphology differ markedly from modern Swedish. The paper preliminarily tests automated text-processing strategies including frequency-list analysis and spelling-variation methods to generate stop-word lists and reveal key terms, supporting word-form grouping, lexicon lookup, and lemmatization.

LanguageEnglish

CategoryResearch & Report

FormatPDF

Pages7

Reading Time18 min

AccessFree to download

1 INTRODUCTION

2 THE DATA MATERIAL

2.1 The Old Swedish Legendary

2.2 Book of Miracles

2.3 Solance for the

# Text Processing Procedures for Analysing a Corpus with MedievalMarian Miracle Tales in Old Swedish Bengt Dahlqvist Department of Linguistics and Philology,Uppsala University,P.O.Box 635,75126 Uppsala,Sweden Keywords:Text Mining,Medieval Texts,Miracle Stories,Old Swedish,Stop Words,Word Similarity,Spelling Variations,Key Words. Abstract:A text corpus of one hundred and one Marian Miracle stories in Old Swedish written between c.1272 and1430 has been digitally compiled from three transcribed sources from the 19th Century.Highly specializedknowledge is needed to interpret these texts,since the medieval variant of Swedish differs significantly fromthe modern form of the language.Both the vocabulary and spelling as well as the grammar show substantialvariances compared to modern Swedish.To advance the understanding of these texts,automated tools fortextual processing are needed.This paper preliminary investigates a number of strategies,such as frequencylist analysis and methods for identifying spelling variations for producing stop word lists and exposing thekey words of the texts.This can be a helpto understand the texts,identifying different word forms of the sameword,to ease a lexicon lookup and be a starting point for lemmatisation. developing more automated text processing tools foranalysing texts written in Old Swedish,primarilywith the aim to mine and deconstruct texts intoconstituents and group similar forms together. ## 1 INTRODUCTION To make computer analyses of texts in Old Swedish,which was in use between 1225 and 1534,is still notan easy task.The standard lexicon for Old Swedishwas prepared in the late 19h Century by the Swedishphilologist(Soderwall,1884).An electronic versionof this exist,but gives no support for spelling variantsnor word inflections,which makes it hard to usepractically for unknown word forms.Very littlesupport exists for automated tasks.For instance,anefficient part-of-speech tagger or a parser is not to befound,even if some work in this area has been donein the last years(Adesam,2016).Neither has notmuch been done regarding the essential problems forthis type of text,which shows many features that areproblematic to handle,foremost in the area of the richmorphology and abundance of word forms.A non-specialist user,wishing to understand or eventranslate a given text,faces many problems.In thispaper,a number of text processing strategies will bediscussed and applied to a small corpus of medievalmiracle tales. Aside from this,the texts themselves areinteresting as witnesses of religious thinking at thetime and as evidence of the interchange andtranslation of textual material within the Europeanmedieval culture.More and better tools for theanalysis of Old Swedish as such may in this way pavethe way for more serious studies of the content ofthese types of texts,and help to facilate both literaryunderstanding and analysis. ## 2 THE DATA MATERIAL The data material used in this study consists of 101medieval miracle tales in Old Swedish where theVirgin Mary figures as a saint and wonder worker.This text collection is believed to consist of allsurviving complete tales of this kind. Miracle tales constitute a specific subgenre inmedieval religious writing,aside from hagiographies,visions and moral tales.The contents of miracle taleson the whole are for the most part purely apocryphal,in that they do not originate from the Bible.They alsooften take place in later times,after the death of the Especially,focus here will be given to the inherentproblems of word analysis,the elimination of wordvariants and text normalisation to be able to identifyword content with rich lexical meaning.This can beseen as a first step in an ongoing research aiming at 452 protagonist saint.Stories of this type were commonin many European languages at the time,from Latinand Greek to French and German.Translations wereproduced into several other languages,including OldNorwegian,Danish and Swedish. In the fol

FAQs

What is the corpus used in the study, and where does it come from?

The study uses a corpus of 101 Marian miracle tales in Old Swedish, digitized from three transcribed sources published in the 19th century.

Why are Old Swedish texts difficult to analyze computationally?

The texts show substantial differences from modern Swedish, including spelling variation, inflectional word forms, and rich morphology that make automated tasks difficult.

Which text-processing strategies are investigated to support analysis?

The paper investigates strategies such as frequency-list analysis and methods for identifying spelling variations to build stop-word lists and expose key words, facilitating word-form grouping and lexicon lookup.

Sep 28, 2026

Upload by Sarah