The limits of freedom: A permutational approach to word order in South Slavic languages (J6-4615)

Typological and grammatical investigations of the sentence structure in the so called ‘free word order’ languages typically focus on a small set of rearrangements of the basic word order attributed either to syntactic displacement operations (e.g. ‘left-branch extraction’) or some other factor, e.g. information structure. The usual line of empirical inquiry, which we may term the approach 'from below', starts with the canonical word order and progressively explores various rearrangements of this basic order as they are brought to researchers' attention. At the same time, for a given sentence size n, where n is the number of words, there exists a natural upper bound of their possible rearrangements, namely, the set of permutations whose cardinality is n! An empirical assessment of sentence acceptability patterns over the full set of permutations has rarely been undertaken on a non-trivial scale because of the intractability problem connected with the infinitude of sentence sizes. Consequently, the extent of actual word order flexibility in a given language has so far been difficult to estimate.

The present project develops a new empirical approach to word order in South Slavic languages within the framework of experimental syntax. It sets up a methodology of empirical investigation of word order that we may term an approach 'from above'. The essence of this approach is: i) given a sentence in the canonical word order of a specific, fixed size (as a local solution to the intractability problem) and syntactic structure, consider the entire set of respective word order permutations as generated by a small number of well-defined displacement types or factors; ii) collect speakers' acceptability and response time data for all permutational variants and iii) model these data to predict acceptability of specific permutational variants as a function of 'weight' or impact of individual displacement types. This approach organically complements the traditional approach 'from below' in utilizing the naturally available upper bound of possible word order variations, but entails important empirical and theoretical advantages which significantly advance the inquiry into the free word order phenomenon. Empirically, it yields an acceptability 'heatmap' regarding word order variation within a particular sentence size and structure, which can serve as a basis for developing local cumulative measures of word order flexibility. It also makes possible multiple comparative studies of word order variation within different sub-classes of the tested permutational variants. Theoretically, this enables a deeper and more balanced perspective regarding natural classes of word order-changing operations applying within a particular type of syntactic structure, a re-evaluation of commonly postulated syntactic drivers of non-canonical word order (e.g. ‘scrambling’ and ‘left-branch extraction’), but also a possibility to explore the psychological reality of constituency in the free word order languages and the role of non-syntactic factors affecting word order such as information structure and processing considerations, among other things. These are the principal tasks that will be undertaken in this study.

The empirical portion of the proposed study includes a series of sentence acceptability experiments in three South Slavic languages: Slovenian, Serbo-Croatian and Bulgarian within the framework of experimental syntax. The theoretical portion of the research includes predictive analyses of acceptability profiles in different syntactic structures based on the collected sentence acceptability data and developing a new model of word order in South Slavic that incorporates the permutational approach 'from above'. Our investigation paves the way for future studies of the free word order phenomenon which bolster the robust experimental basis, on the one hand, and are solidly grounded in speakers' acceptability intuitions, on the other.

Project group:

Arthur Stepanov (PI) – University of Nova Gorica (sicris) (2022-)
Penka Stateva – University of Nova Gorica (sicris) (2022-)
Rok Žaucer – University of Nova Gorica (sicris) (2022-)
Danil Khristov – University of Nova Gorica (sicris) (2022-)

Duration, type, size, funding:

Project duration: 1.10.2022–30.9.2025
Project type: standard basic research project; category C
Funding agency: Slovenian Research Agency
Size: 2210 hours per year

In Sicris database: Link