Abstract With morphological and syntactic argument properties, some arguments behave alike (i.e., they align with each other) while others do not. Such alignment patterns have received significant attention in the literature, but claims as to their origin and development are sometimes difficult to assess, due to scant actual data. This paper surveys the main hypotheses proposed in early and recent work on the topic, focusing on alignment type change and on major alignment types (ergativity, accusativity, and split intransitivity) of morphological properties, with some remarks on syntactic properties. The survey shows that alignment type change may often occur when clauses denoting low transitivity are reanalyzed as clauses of either higher or lower syntactic valency, sometimes even introducing a partition in the verbal lexicon (occasionally being conditioned by semantic, pragmatic, or structural factors), or are extended from low‐transitivity predicates to most bivalent predicates. Lastly, alignment type change can be either functionally motivated or not.