This study examines handshape distribution in Taiwan Sign Language (TSL) naturalistic discourse across two metrics: type frequency using a corpus of signs compiled across six pedagogical sources and Ann’s (2005) ease of articulation categories. Ann posits a correlation between ease of articulation of a given handshape and its distributional frequency in naturalistic discourse, with more easily produced forms occurring more frequently. This study expands on her original analysis by building a larger corpus of conversational data and by widening the scope of the lexicon under investigation with an expanded data set across more sources, totaling 2000 lemmas. By increasing the corpus size, as well as diversifying discussion topics beyond a single conversation, this study is able to conduct statistical analyses more capable of achieving valid results, as well as mitigate confounds of overrepresentation of certain handshapes due to their morphosemantic content. Moreover, this study assesses the role of ease of articulation in understanding asymmetries in changes between handshapes, finding that starting from type frequency may offer additional insights to understanding how ease of articulation functions in the signed modality. 

Shoulda, coulda and woulda are emergent auxiliary verbs understood as contracted forms from should have, could have and would have, respectively. Having as hypothesis that these verbs express more than, and are used to mark tense, aspect and modality (TAM), this research aims to analyze the uses of shoulda, coulda and woulda in Present-day English (PDE) under a constructionist approach. For this, I rely on Goldberg (1995), Croft (2001), Traugott and Trousdale (2013) and Furtado da Cunha, Bispo and Silva (2013) with respect to the formal and functional properties of a construction, to the properties of schematicity, productivity and compositionality and to the characterization of constructional network; Heine (1993) and Bybee (2010) regarding the notion of auxiliary verbs and the development of modern auxiliaries and modals; and Lyons (1977), Palmer (1986) and Croft (2012) to the comprehension of the categories of modality and aspect. In order to establish an analysis within real communication situations, I use the Corpus of Contemporary American English (COCA), formed by eight text groups, namely, oral, fiction, magazines, newspapers, academic, blogs, webpages and TV/movies from 1990 to 2019. To process the data, I use WordSmith Tools 7.0 program. The idea is to measure token and type frequency, and verify the contexts of occurrences and their functions, and with that, to describe the place of shoulda, coulda and woulda, as auxiliary verbs, on the constructional network.