The Causative-Inchoative Alternation and Age-of-Acquisition Effects on Multi-predicate Constructions in Turkish Sign Language

ABSTRACT

We discuss the causative-inchoative alternation in Turkish Sign Language (Türk İşaret Dili – TİD), and the age of acquisition effects on multi-predicate, complex constructions that are observed in both causative and inchoative events. We present a picture-description task performed by 24 adult signers, half of which were exposed to TİD from birth (native adult signers) and the other half after the age of 4 (late adult signers). The results show that (i) the causative-inchoative alternation in TİD is attested both as labile alternation and classifier alternation; (ii) there is no significant age of acquisition effects on the choice of alternation types; (iii) both causative and inchoative events have complex event structures that can be overtly expressed through multi-predicate, complex constructions; and (iv) late signers are less likely to produce multi-predicate, complex constructions than native signers. These results support the significance of deaf individuals’ early exposure to a sign language.

Acknowledgments

We are grateful to all the participants of this study. Without them, this study would not be possible. We want to thank Furkan Atmaca, Onur Keleş and Dr. Pavel Logačev for their help in statistical analysis and Muharrem Ayar for doing annotations for interrator reliability check. Special thanks to Elvan Tamyürek Özparlak for the pilot study and her help in data collection and to Gül Yiğit for her help in data collection. We are also grateful to the two anonymous reviewers for their detailed comments, questions and suggestions, which improved the paper vastly. All remaining errors are ours.

Disclosure statement

No potential conflict of interest was reported by the author(s).

Notes

¹ The sign glossed as ix or index is a pointing sign, which has multiple functions. When it is used within a noun phrase accompanying a noun, it is a modifier as in (4b) from ASL and (8a) from TİD. When the pointing sign has a letter subscript attached to it, it means that the pointing sign is directed to a certain location in space, which can be shared by predicates, modifiers, and pronouns to indicate co-reference. For instance, man and ix share the –a subscript in (8a) to show that the modifier ix establishes a referential locus for the head noun. The two indexes, ix_1SG and ix_3SG, are used as pronouns to refer to the first person and a third person respectively in (9).

² As one of the reviewers pointed out, the effects of this delayed exposure to a first language has not disappeared among deaf children despite the increasing use of hearing technology (Carrigan & Coppola, 2020).

³ Deaf children of hearing parents create language-like gestural systems, called homesign to communicate with hearing family members (Parts II–III in Goldin-Meadow, 2005). Although these systems include language-like properties such as a gestural lexicon composed of identifiable and differential items like actions, objects, and attributes (Goldin-Meadow et al., 1984; Chapter 12 in Goldin-Meadow, 2005), some combinatorial rules (Hunsicker & Goldin-Meadow, 2012), grammatical functions such as Subject (Coppola, 2002) and clause types (Franklin et al., 2011), homesign systems are not fully fledged conventional linguistic systems. Furthermore, the gestural input from parents of homesigners do not determine the emerging structures in these systems (Goldin-Meadow & Mylander, 1983) and homesign systems do not suffice to fully bridge the communication needs between deaf individuals and their family members (Carrigan & Coppola, 2012).

⁴ BPCLs have been reported to behave as unergative predicates in ASL (Benedicto & Brentari, 2004). However, our ongoing research about TİD has shown that BPCLs are used not only in unergative clauses but they are also used in causative clauses (see, Gökgöz & Sevgi, 2020; Gökgöz, under revision; Sevgi and Gökgöz, submitted).

⁵ Given that the event in (12a) is also of a causal nature, the only factor that affects the difference in the classifier type of the causing event is the [±human] feature that the causer carries. Since the animate non-human causer in this event does not have the same means of performing a grabbing/contact action as a human does, the initiation it performs is expressed with a BPCL, cross-referencing the “paw” of the dog in this case. BPCLs could be treated as a sub-type of contact HDCLs, differentiating based on [±human] feature. We note here that (12a) is contra Benedicto and Brentari’s (2004) observation that BPCLs can only be licensed in unergatives (see, Gökgöz (under revision) for a detailed account).

⁶ See the Appendix for the production data that are not used in statistical analysis for alternation strategies. They also show a similar pattern for native and late signers.

⁷ Thanks to one of the reviewers for raising this point.

Additional information

Funding

This work was supported by the Boğaziçi University Research Fund Grant Number 14458.