Optimality Theory for Hong Kong Sign Language loan words

Emily Koenders

Keywords: cross-modal language contact, Hong Kong Sign Language, loan words, Optimality Theory, Sign Linguistics

TL;DR: An OT approach unveils the systematicity with which cross-modal loans are rendered into HKSL.

Abstract:

Optimality Theory for Hong Kong Sign Language loan words

Introduction
As any natural language, Hong Kong Sign Language (HKSL) borrows words from its contact languages – sign or spoken – to express concepts for which no lexical signs yet exist. Sign languages lend words from across modalities, a phenomenon which, unfortunately, hitherto has not yet been attested for the world’s spoken languages. HKSL’s loan words can be divided into categories based on their source language and borrowing strategy. Source languages include American Sign Language (ASL), Japanese Sign Language (JSL), Chinese Sign Language (CSL), written Chinese, Cantonese, English, and Korean. Borrowing strategies are direct loans, (partial) iconically motivated signs (e.g., salad, Figure 1), (partial) Signed Chinese/English/Korean (e.g., toast, Figure 2), character signs (e.g., introduce, Figure 3), fingerspelling, and initialisation (e.g., cappuccino, Figure 4).
[see PDF for figures]

What is interesting about the intra-modal direct loans is that HKSL adopts mouthing of the Cantonese translation equivalents. For example, concept (Figure 5) is a loan from ASL; HKSL adds mouthing of the Cantonese 概念 koi nihm ‘concept.’ The added mouthing facilitates the understanding of the borrowed sign. For intra-modally borrowed verbs (e.g., the ASL loan communicate, Figure 6), the accompanying non-manuals are native HKSL aspectual markers (i.e., protruding lips).
[see PDF for figures]

However, there is only a negligible correlation between source language and adaptation strategy. Broadly speaking, Chinese loans are adapted as Signed Chinese or character signs, and English loans make more use of fingerspelling and initialisation. These patterns are unsurprising and uninteresting: Deaf people rely on visual information, i.e., the written form of a language. What is more important is that there is no discernible consistent relationship between the source language of a loan word and the borrowing strategy applied. In other words, there is no direct relationship between a loan word’s source language and the cross-modal adaptation strategy used to render the loan word in HKSL.

Optimality Theory
It is therefore that I propose an Optimality Theory approach to capture and analyse the patterns in the dataset in a systematic and rule-based manner. These are the faithfulness and markedness constraints:

Faithfulness Constraints
[see PDF for constraint list]

Markedness Constraints
[see PDF for constraint list]

In the following tableaux, I use the examples of watsons and toast to show how the constraint ranking renders a cross-modal loan into HKSL. The constraint ranking of the tableau shows that faithfulness constraints are ranked higher than markedness constraints. This means that HKSL attempts to render loan words as faithful as possible.
[see PDF for tableaux]

watsons: There are no sign equivalents for the Chinese 屈, 臣, or 氏, which is why Signed Chinese candidates are disqualified. None of the three characters can be visually represented using a character sign without violating sign-phonological constraints. Fingerspelling all segments of Watsons would violate constraint 6. An iconic sign would violate constraint 4 by merging the three segments present in the input. The fingerspelling and initialisation candidates do not violate constraint 7 because Watsons is a name that forms a direct semantic association with the referent.

toast: full Signed Chinese candidates violate constraint 5 because there is no HKSL sign equivalent for the character 士. The two candidates that incorporate iconic, new signs violate constraint 4 by not fully representing all input elements in the output. The winning candidate, partial Signed Chinese, violates this same constraint but to a lesser extent because all input elements are still represented through the mouthing of source language dō sih.

Conclusion
In conclusion, the OT approach shows that there are clear rules (or constraints) that govern cross-modal loan word adaptation in HKSL. The pattern the OT analysis exposes is one that favours source-faithful renditions over the usually favoured modality-specific iconic word-formation strategies common in sign languages.

Submission Number: 6