Understanding Understandability: Evaluating Understandability as a Part of the Linguistic Accessibility of a Text and Its Translation

In this thesis, I attempt to answer the question of what makes a text understandable by creating a set of criteria for the evaluation of linguistic accessibility and understandability. The aim is to create a language-independent set of criteria that can be used both to evaluate a text and as a set of guidelines to follow when writing accessible content. The main resources I used in the creation of the criteria are techniques for making a text easier to read from the Web Content Accessibility Guidelines 2.1 and the Selkomittari, a tool meant to evaluate whether a text qualifies as simplified Finnish or not. Simplified language shares some features with accessible language, and this thesis focuses on the Finnish–English language pair, making the Selkomittari a valuable resource. I tested the criteria by evaluating the accessibility of a text found on the website of the Finnish Digital and Population Data Services Agency. It is a public sector website published on January 1, 2020 and should thus be following the EU Web Accessibility Directive. The analyzed text focuses on the registration of foreigners in the Population Information System, which means that the main audience of the text might not be fluent in any of the languages the text is available in. This increases the need for accessibility and understandability. I analyzed both the original Finnish text and its English translation. I found no major differences in accessibility between the two languages. Instead, they both seem to share the same accessibility issues – most significantly, long sentences with too much information and nondescriptive headings. Both versions had few issues related to grammar, vocabulary, or the overall structure of the text. Overall, both texts met the criteria to some extent. However, this is not sufficient, as it does not take those with cognitive issues or poor language skills into account. The criteria I created seemed usable as a tool to evaluate the accessibility of a text, at least in the data I used. The criteria were relevant, and I found all criteria applicable to the texts