Theoretical Syntax at the Crossroads: Big Data, Citizen Science and Crowdsourcing

Based on the novel data pertaining to five syntactic phenomena (the position of the finite verb in embedded clauses, in sentences with a modal verb, negative concord, the position of focused light/heavy objects in main clauses with a complex tense and scrambling) in the heritage language Mòcheno collected via original fieldwork, we show that there are two populations – one exhibiting intra-speaker variation between German and Italian word orders, and one lacking it; and these two populations are the result of diatopic variation and, to a lesser extent, of diastratic variation. The results achieved using quantitative statistical analysis are partially convergent with those arrived at via the traditional theoretical syntax for Mòcheno, but our analysis has allowed us to shed new light on a series of phenomena that have been neglected or poorly understood thus far. More specifically, and for the first time, we discovered that there is a micro-variation resulting from diastratic (age) variation within the Roveda variety, which represents the only case in Mòcheno in which age is a relevant factor in determining variation. We also show that the traditional claim that the Palù variety is ‘more German’ than is the other Mòcheno variety is to be confirmed; however, we could refine it by showing that German word orders are also accepted by speakers of other varieties and that the acceptability of these word orders in competition with the Italian syntax is not due to their age (no diastratic variation). Finally, we show that the acceptance of German word orders across speakers varies according to the phenomenon investigated: German word orders are more likely to be accepted in sentences featuring a negation, whereas German word orders are more likely to be rejected in embedded clauses. Based on this fine-grained description of the distribution of OV/VO word orders across different contexts and groups and the available theoretical account for the derivation of OV word order given by Cognola (2013b), we propose that the observed variation can be parametrized along the lines of recent developments of Parameter Theory (Roberts, 2012; Biberauer et al., 2014 a.o.). More specifically, we propose that the movement of the non-finite verb form to lowForce°, which is responsible for OV in Mòcheno, can be captured in terms of a parametric hierarchy. When verb movement takes place in all syntactic conditions, including with all non-finite verb forms and when the auxiliary has not moved out of v° to Spec, CP, a macroparametric effect obtains which corresponds to the system instantiated by the Palù variety. The mesoparameter corresponds to a system in which the movement to the non-finite verb form can only be found when v° is empty, i.e., in main declarative clauses. The fact that for a subgroup of speakers from Fierozzo and Roveda OV word order is accepted with modal verbs follows from a microparamter: the movement of the non-finite verb form to lowFocus° can only take place with non-finite verbs. Finally, the fact that OV is obligatory for nearly all groups is captured in terms of a nanoparamenter associated with negative constituents.