Language forecasting, i.e., predicting the future state of a language, has long been regarded with a fair amount of skepticism. This is partly due to language change often being considered sudden, random, unpredictable, and viewed as the result of complex interacting factors that are not well understood (e.g., Keller 1994:72; Bauer 1994:25; Labov 1994:10; Croft 2000:3; discussion in Sanchez-Stockhammer 2015). Some have gone as far as to claim that "[d]iachronic linguistics is not a predictive science" (Bauer 1994:25). Nevertheless, more positive views on the possibility of language forecasting have emerged in recent years (Soskuthy 2015; Sanchez-Stockhammer 2015; Van de Velde 2017). In this dissertation I present arguments in favor of language forecasting, claiming that it can and should be practiced. I argue that forecasting can be used to test various expectations toward language change, including the understanding of the propagation of new linguistic variants through a language community. Using historical data in the form of regular time series, I produce short- to mid-range forecasts for two selected changes in Icelandic. The first change concerns the complex prepositions "a bak við" 'behind' and "við hliðina a" 'next to' which are occasionally encountered in a simplified form as "bakvið" and "hliðina," respectively. Although the change has been briefly mentioned before (Friðjonsson 2004, 2007 and Rognvaldsson 2021), it has not been systematically documented. The second change involves subject case marking with the predicate "hlakka til" 'look forward to', where an oblique case (accusative or dative) replaces an earlier nominative case with subjects. This change has been extensively studied (e.g., Svavarsdottir 1982; Jonsson & Eythorsson 2003; Nowenstein 2023), but the present work offers an original conception of its essence. The time series analysis and forecasting presented in the dissertation provide a novel type of documentation and a fresh insight into both types of changes. Since language forecasting is argued to require ample context to be comprehensible, efforts have been made to contextualize the changes under discussion to the extent possible. [The dissertation citations contained here are published with the permission of ProQuest LLC. Further reproduction is prohibited without permission. Copies of dissertations may be obtained by Telephone (800) 1-800-521-0600. Web page: http://bibliotheek.ehb.be:2222/en-US/products/dissertations/individuals.shtml.]