About the lecture
"Linguistics in the second half of 20th century was changed enormously by Chomsky's and Montague's turn to formalization and mathematization, these changes are most dramatically visible in the syntax (Chomsky (1957), Chomsky (1993) among many others and in semantics (Montague (1973), Montague (1970), Winter (2001)).
After 40 years of the main parts of linguistics being formalized, from 90s we can observe next step in the history of modern linguistics which can be called experimental turn: more and more typologically diffuse languages are subsumed under the roof of formal methods founded in the decades before. The diverse data lead to widening and sharpening of the tools of modern linguistics naturally (Horn (1989), Gajewski (2005)) but as well to a new search for empirically more adequate frameworks where much more attention is payed to data gathering and their statistical interpretation than it was in the years before.
In sum: beginning with the nineties of the last century a lot of prominent linguists turned their attention to experimental methods, statistical approaches to linguistic data and generally to a more robust connection between the existing theories and realistically gathered and sampled linguistic data (see Baayen (2008) a.o.)
In my talk I will focus on:
a) short introduction of the modern formal linguistic history,
b) one concrete example (series of experiments analyzing Neg-Raising, published as Dočekal and Dotlačil (2016).
I will show the connection between the formal and experimental methods in current linguistics and I will demonstrate:
- predictions of formal theories and the interactions between theories and their verification/modification,
- construction of an experiment - design, fillers, latin-square design,
- statistical modeling of the experimental data - linear models, reaction times, working with the language R,
- final summary of the statistically modeled data with respect to their consequences for current formal theories.