“Reconsidering animacy hierarchy effects in Mayan: Experimental evidence from Ch’ol” 

(presenting joint work with Stefan Keine (UCLA), Juan Vázquez Álvarez (CIMSUR-UNAM), and Michael Wagner (McGill))

Like many other Mayan languages, Ch’ol has been described as restricting the combination of 3rd person arguments in a transitive clause according to their relative animacy (Zavala 2007, Vázquez Álvarez 2011; see Deal & Royer 2025 for an overview), as in (1):

(1) Ch’ol animacy restriction:
Third person subjects must be at least as high as third person objects on the scale HUMAN >> ANIMATE >> INANIMATE.

As in many other Mayan languages, the restriction holds only over 3>3 transitives; 1st/2nd person human objects are possible regardless of the subject’s animacy. Passivization is commonly described as a rescue for expressing hierarchy-violating 3>3 constructions.

Aissen (1997) provides an Optimality Theoretic account for these patterns in related Tsotsil, using constraints which enforce alignment between participant hierarchies and grammatical roles. More recently, Deal & Royer (2025) argue for an Agree-based approach to Mayan animacy effects based on Deal’s (2024) Interaction/Satisfaction model of Agree. While different in their formal mechanisms, both accounts (i) derive uniform ungrammaticality of all hierarchy violating transitives; (ii) invoke comparison of the relative animacy of subject and object; (iii) stipulate the immunity of 1st/2nd person pronouns.

We show that animacy effects in Ch’ol are more complex than previously described. We discuss results of a series of three experiments we conducted with 52 speakers of Ch’ol in Chiapas, Mexico: (i) a production task; (ii) a forced choice task; and (iii) a rating task. We do not find a binary opposition between hierarchy-obeying and hierarchy-violating constructions, contra expectations of (1), but rather more gradient and task-specific effects. We propose an account that attributes the animacy restriction to alignment constraints that demand subjects and external arguments to be high in animacy, combined with task-specific competition between syntactic structures. Unlike previous analyses, this account does not involve a comparison of the animacy of subject and object. Furthermore, it derives the fact that animacy restrictions arise only in configurations in which the verb form does not uniquely determine the mapping between nominals and their grammatical roles, correctly capturing the immunity of 1st/2nd persons and the rescuing effects of passivization.

Location: Fowler Museum A139