|Description||The Linguistic Typology of Sign Language Instrumentals |
The linguistic encoding of classifier predicates found in instrumental sentences displays the well-known distinction between hand-as-hand and hand-as-object iconicities (Padden et al., 2015). While the agentive nature of an instrumental utterance (Benedicto & Brentari, 2004) requires the use of a Handling classifier (H-CL; hand-as-hand iconicity), certain factors such as Iconic Handshape Preference, Instrument Sensitivity (Brentari et al., 2015) and Instrument Typicality (Brentari et al. 2016) encourage the use of an Object classifier (O-CL; hand-as-object iconicity), which foregrounds the instrument information in the classifier predicate and demotes the agent information. Preliminary observations on TiD (Hakguder, 2018) has shown that Argument Drop is another factor that affects classifier decision. While instrument drop encourages O-CL use, agent drop does not have a comparable effect on H-CL use. This research project aims to study the variation among languages regarding the distance of licensing between a classifier and its noun antecedent(s). While bringing novel dimensions into the equation, the study will also tackle factors that were previously identified. The ultimate goal of the study is to create a linguistic typology of sign languages built on statistical methods by analyzing data collected from 5 genealogically unrelated sign languages (10 signers each): ASL, TiD, HKSL, LIS and BSL. Each signer of these 5 languages are presented with standardized experiments that aim to lay out the linguistic typology of sign language instrumentals. The probabilistic nature of classifier decision in sign languages will be analyzed using a Maximum Entropy model.