Odenman Holmberg , Mårten2025-06-262025-06-262025-06-26https://hdl.handle.net/2077/88426The theoretical focus of this thesis is the effect that performance in parsing syntactic structure (Hawkins, 1994) has on language change. The empirical focus is the OV→VO-change in Swedish and how it relates to a) the heaviness of the object b) ambiguity/syncretism in case marking on nominal arguments. Integrating Hawkins’ Performance Grammar Correspondence Hypothesis (PGCH) (1994; 2004; 2014) into Ritt’s evolutionary-memetic approach to language change (2004) I propose that performance in parsing syntactic structure constitutes a stable selection pressure on linguistic form-replication. The effect that object heaviness has had on the OV→VO-change in Swedish is tested using a logistic regression model with object length (in words) serving as the predictor of OV/VO in transitive clauses with a finite auxiliary (AuxO-clauses) in four historical Swedish texts. For the same AuxO-clauses, ambiguity in case marking is measured calculating the Shannon entropy on how well a certain form indicates its’ Grandmother node G (S/VP). This measure is correlated with the aggregate IC-to-word ratios for the same AuxO-clauses using Spearman’s rank order correlation. The findings are these: 1) there is a clear positive correlation between the heaviness of the object and VO-preference in the history of Swedish. 2) Ambiguity in case forms is positively correlated with increased rapidity in constituent recognition through word order (represented by the IC-to-word ratio measure). 3) A performancememetic approach to the OV→VO-change is a better account than prior ones (see Delsing, 1999) of why certain objects and object types persevere in OVconfigurations.engThe Performance Grammar Correspondence HypothesisEvolutionary Linguisticslanguage changeOV→VOobject lengthcase lossGermanicSwedishHistorical linguisticsHawkinsRittmemeadaptationText