Quadros, Ronice M. de & Josep Quer. 2008. Back to the back (wards) and further: On agreement, auxiliaries and verb classes. In Ronice M. de Quadros (note.M.), Sign languages: spiders and unravellings of the past, present and future. TISLR9, forty-five documents and three posters from the 9th Theoretical Questions at the Sign Languages Research Conference, Florianopolis, Brazil, December 2006. Petrépolis/RJ, Brazil: Editora Arara Azul. In this article, we presented arguments for an analysis of the indication of verbs that rely on Liddell (2000) as a typologically unique and unimodal fusion of morphemes and dot fingers that function as a construction used for reference tracking. We explained how some patterns observed when displaying verbs coincide with an increasing volume of research in co-speech needle gestures and multimodal vocal/gesture constructions.
The similarities between gesture and verbal system display lead to one of the main ways in which the display of the verbage system does not resemble compliance marking – that is, the way they use space for deiztic references does not always lead to the systematic covariance normally associated with tuning systems (Corbett 2006). We have shown how this commonality corresponds to what we know in a number of sources, from the latest corpus studies to denakquise, grammation, sociolinguistic variations and language changes. We also showed how our analysis of indicator construction offers a unique way to account for the relationships between modification, gaze and staging in a way that is not taken into account by agreement analyses. This building grammar account also removes the need for rules-based explanations to explain why only a subset of sign language verbs indicates verbs, the optionality of the leadership obligation and the problem of backward verbs. The adoption of a design analysis is also an asset, as this commonality can be taken into account between the gesture of co-speech and the reference systems in sign languages and in multimodal communication with spoken languages. To better understand these patterns, more detailed comparisons with multimodal descriptions of language use are needed. With regard to the licensing of null arguments, research on the expression of variable subjects suggests that forecasts of agreement analysis are not generated by the data available from some of the sign languages we are working on. While studies by McKee et al. (2011) indicate that concordance verbs are actually slightly nil for sign language in Auslan and New Zealand (57% and 57% respectively). 54% of all tokens who do not have an undivided subject argument), as proposed by Lillo-Martin (1986) and others, also shows that spatial verbs (73% for Australia and 66% for New Zealand sign language) and simple verbs (60% and 53%) most often without a clearly formulated subject argument.
It is therefore not clear how the analysis of the agreement takes these models into account, and the data indicate that other factors, such as co-reference and structural primer (see McKee et al. 2011 for more details) are also important to influence the variable expressions of the subject (structural predictions about the prediction of the need to predict a use-based construction model, such as the one we are proposing here). In summary, our indication of the construction proposal has advantages over rule-based models, as it does not need additional mechanisms to explain why only a subset of verbs indicates verbs and why there are differences in the way the system is designed (with individual verbs and classes of verbs, such as verbs with regular and upside down verbs).