An analysis and reconstruction of transitive nominalization in Ch’olan languages

This paper reconstructs the transitive nominalizing suffix *-yaj (IPA */-jax/) in the Ch’olan branch of Mayan languages. I consider data from modern Chol, Chontal, and Ch’orti’ as well as colonial Ch’olti’ to reconstruct the phonological form and syntactic function of this morpheme. This suffix has been called nominalizing antipassive (e.g., Robertson et al. 2010:186-7), although it does not eliminate the object in all cases. Rather, I analyze it as a more general valency-reducing suffix. Each of the languages has undergone small phonological changes, and all of them allow truncation of the suffix to -aj in certain phonological contexts and in fast speech. This paper argues that the glide is underlying, rather than epenthetic, and that the final consonant reconstructs to the velar fricative /x/ rather than the glottal /h/. It also considers the distribution of these nominalizations in each of the languages, and the additional morphology that can appear on them. In particular, there has been a shift between colonial Ch’olti’ and modern Ch’orti’ in the preferred method for marking the thematic roles of the nominalized verb. Ch’olti’ requires a prepositional phrase to reference the patient or stimulus of the verb if it has been derived into an agentive, while Ch’orti’ uses the Set A possessor for the same function. When there is no agentive prefix in Ch'olti', the Set A proclitic can appear before the nominalization, as in Ch’orti’. Chol and Chontal use the *-yaj suffix very similarly to each other. Although there is some debate about the role of nominalizations in split-ergative languages like these, these particular forms act as syntactic nouns, taking nominal morphology including possessors and being incorporated into verbs like any other noun. Further fieldwork on the distribution of the allomorphs in these languages would be particularly useful, as would a closer study focused on the syntactic distribution