The omnipresence of electronic texts in our time warrants questions about the power to define what text really is, now held by technology — the textuality and representational qualities of digitized texts are undoubtedly influenced by the technological frameworks that decide their possible content and form, making the issue a point of interest for LIS and the digital humanities. This thesis explores the effects of commonly applied technology and practices in electronic publishing of Japanese language texts. Special attention has been paid to the possible issues concerning gaiji characters, ruby annotations and kana orthography. The predominant approach taken is a comparative analysis of two literary texts extracted from the corpora of the digital libraries Aozora Bunko and Japanese Text Initiative, accomplished by a comparison of the electronic editions with print. The findings include issues and weak points related to character encoding and the handling of annotations as well as other aspects of textuality being affected by the digitization process. The resulting discussion explores these findings as well as possible workarounds and solutions offered by developments in technology and methodology. The main conclusions of this thesis are that the textuality and the bibliographic codes of Japanese language texts may be hampered by the representational limitations of technology and digitization practices, but that the application of text encoding using contextual markup, as well as developments in character representation, may prove to be beneficial for the preservation of Japanese language textuality.

Studien genomfördes för att undersöka om associationer mellan utformningen av typsnitt i latinska alfabetet och i japanska skriftsystem och icke-språkliga ljudstimuli. Ytterligare syfte var att undersöka om associationerna påverkas av modersmål och kultur hos respondenter. Teorin som denna studie byggs på inkluderar ljudsymbolismen och BoubaKiki-fenomenet som visar att vissa ljud associeras med vissa former, samt studier om typsnittets personlighet. Två online-enkäter skapades, en engelskspråkig och en japanskspråkig. Enkäterna innehöll samma ljud och samma frågor. 10 typsnitt valdes för latinska alfabetet och 10 för japanska skriftsystem. Ljuden skapades med ett datorprogram och innehöll huvudsakligen toner, rytmer och brus. Respondenterna ombads att lyssna på ett icke-språkligt ljud och sedan välja ett typsnitt från de befintliga svarsalternativen som de upplevde likna ljudet mest. Resultatet visade att det finns ett svagt mönster som kan tyda på att människor associerar snabbare rytm med tunnare typsnitt och långsammare rytm med tyngre typsnitt. Brunt brus associerades starkt med kantiga tunna typsnitt. Toner som förändrades i tonhöjd visade på en svag association med rundare typsnitt. Höga toner associerades med luftiga tunna typsnitt. Det verkar som att japanska och västerländska respondenter inte hade några större skillnader i korssensoriska associationsmönster. Mindre skillnader kunde dock ses, som kan bero på olika faktorer och kan delvis förklaras med sättet studien genomfördes på. / The aim of this study was to investigate associations between typeface design of the Latin alphabet and the Japanese writing system and nonspeech sound stimulus. Moreover, the goal was to study if the associations are influenced by the native language and the culture of a person. Theories that this study was based on included the sound symbolism and the BoubaKiki phenomenon, which shows that certain sounds are associated with certain shapes, and the theory on the typeface personality. Two onlinesurveys were created, one in English and one in Japanese. The surveys were distributed to two groups, the Latin alphabet native users and the Japanese native speakers.  The questions of the surveys were the same and were split in two parts depending on the language of typefaces. 10 typefaces were chosen for the Latin alphabet and 10 for the Japanese writing system. Sound files included tones, rhythms, and the Brownian noise. Respondents were asked to listen to a sound and then choose a typeface that felt “closest” to it. Result shows a somewhat weak pattern that people associate quicker rhythm with thinner typefaces and slower rhythm with heavier typefaces. Brownian noise was strongly associated with angular and thin typefaces. Tones that changed the pitch were somewhat weakly associated with rounder typefaces. High pitched tones were associated with thin and light typefaces. The differences in cross-sensory perception between the Japanese and the western respondents were not major. However, some minor differences could be noticed which might depend on the way the study was conducted or the other factors.