Published online by Cambridge University Press: 01 January 2026
Whether spoken language verbal classifiers and sign language classifier handshapes are comparable enough to be treated similarly is a subject of debate in the literature. In this article, I first show that both spoken language verbal classifiers and sign language classifier handshapes cross-reference the internal argument in intransitive and transitive clauses. Despite differences in the modality of expression (visual-gestural vs. auditory-oral), verbal classifiers end up accomplishing this same kind of work in the grammar, which falls under absolutive alignment. From a morphosyntactic point of view, however, there is more to the story, as data from body-part classifiers reveal. I show that Turkish Sign Language (TİD) is similar to Manam, Diegueño, and Cherokee with regard to classifiers cross-referencing the external or internal argument's body part. While some of this falls outside of absolutive alignment because cross-reference is to the external argument, I show that the syntactic distributions of clauses with body-part classifiers in both modalities can be accounted for with a few modifications to recent morphosyntactic proposals originally offered for sign languages. This supports the conclusion that verbal classifiers are comparable across modalities. Along the way, I refine Benedicto and Brentari's (2004) account and propose that there are building blocks (selected fingers and hand-parts) in the morphophonology of TİD that combine to yield the range of classifiers that researchers hitherto have tended to describe with holistic labels. Namely, all classifier types in sign languages (whole entity, handling, and body part) employ selected fingers that cross-reference the internal argument in some way, similarly to how many spoken language verbal classifiers cross-reference internal arguments. Furthermore, handling and body-part classifiers make use of hand-parts that can cross-reference the body part of the external argument. Similarities between the spoken languages Manam, Diegueño, and Cherokee and the sign language TİD in cross-referencing the body part of an argument in syntax become clear, and the morphophonological patterns of hand-parts also reveal handling and body-part classifiers in sign languages to be more similar than previously thought.
I am grateful to the Deaf individuals in Turkey for sharing their time, language, and intuitions with me. This work would not have been possible without them. I am especially grateful to Elvan Tamyürek Özparlak for discussions and for modeling sign language handshapes. This research was supported by Boğaziçi University Research Fund, Grant Number 14458 (PI: Kadir Gökgöz), and the SIGNHUB project, Horizon 2020, European Union funding for Research & Innovation, Grant Number 693349 (Coordinator: Josep Quer; PI in Turkey: Meltem Kelepir). I am also very grateful to the editors of Language, John Beavers, Diane Brentari, and Lisa Cheng, and to the anonymous referees for all of their comments, questions, and suggestions, which immensely improved the paper. I would like to extend my gratitude to A. Sumru Özsoy, Ronnie B. Wilbur, Diane Lillo-Martin, Elena Benedicto, Hande Sevgi, Demet Kayabaşı, and Onur Keleş for several discussions that helped me shape the ideas in this article. Versions of this work were presented at LingDay 2018 at Boğaziçi University and at TU+7—the Seventh Workshop on Turkic and Languages in Contact with Turkic—at the University of Connecticut. I would like to thank the audiences for invaluable questions and comments during those events. I dearly thank my wife, Vincenza Iadevaia, for listening to the ideas in this paper and for her endless encouragement. All remaining errors are my own.