Abstract
Current theories of language propose an innate (Baker 2001; Chomsky 1981) ora functional (Greenberg 1963; Dryer 2007; Hawkins 2014) origin for the surfacestructures (i.e. word order) that we observe in languages of the world, whileevolutionary modeling (Dunn et al. 2011) suggests that descent is the primaryfactor influencing such patterns. Although there are hypotheses for word orderchange from both innate and usage-based perspectives for specific languages andfamilies, there are key disagreements between the two major proposals formechanisms that drive the evolution of language more broadly (Wasow 2002; Levy2008). This paper proposes a universal underlying mechanism for word orderchange based on a large tagged parallel dataset of over 1,500 languagesrepresenting 133 language families and 111 isolates. Results indicate that wordclass length is significantly correlated with word order crosslinguistically,but not in a straightforward manner, partially supporting opposing theories ofprocessing, while at the same time predicting historical word order change intwo different phylogenetic lines and explaining more variance than descent orlanguage area in regression models. Such findings suggest an integrated"Min-Max" theory of language evolution driven by competing pressures ofprocessing and information structure, aligning with recent efficiency-oriented(Levshina 2023) and information-theoretic proposals (Zaslavsky 2020; Tucker etal. 2025).