The nun is in the saucepan: word-order harmony in silent gesture improvisation of verb phrases and adpositional phrases
- Rafaela Alford, University of Edinburgh, Ailsa Anderson, University of Edinburgh, Rea Caballero, University of Edinburgh, and Cliodhna Hughes, University of Edinburgh
- Teams Room 1, University of Aberdeen
The majority of languages represented in WALS follow one of two patterns: (1) the verb precedes the object, and the adposition precedes its complement, or (2) the verb follows the object, and the adposition follows its complement (Dryer, 2013). This is an example of word-order harmony: a phenomenon whereby the head of the phrase consistently either follows or precedes its complement(s), across phrase types within a language. Though there is debate regarding what causes harmony, experimental evidence shows that the mechanism behind it is particularly active in learning: adults and children are better at learning harmonic noun phrase orders, and children are more likely than adults to shift to a harmonic pattern when given non-harmonic input, regardless of whether their native language is harmonic (Culbertson, Smolensky, & Legendre, 2012) (Culbertson & Newport, 2015) (Culbertson, Franck, Braquet, Barrera Navarro, & Arnon, 2020). Although there is much research on harmony in learning, there seems to have been little work done to investigate whether or not the mechanism that causes harmony is also active in improvisation.
We utilised the silent gesture improvisation paradigm (Goldin-Meadow et al., 2008) to investigate whether the relative order in which participants gesture the object and verb in verb phrases primes their subsequent relative ordering of adpositions and their complements in adpositional phrases, in silent gesture improvisation. In their silent gesture improvisation experiment, Schouwstra & de Swart (2014) found that participants given stimuli depicting extensional events were more likely to produce gestures in a subject-object-verb (SOV) order, and those given stimuli depicting intensional events were more likely to produce gestures in a subject-verb-object (SVO) order. We utilised these findings to encourage one group of participants to produce SOV-ordered gestures, and the other SVO. Both groups were then given a set of stimuli depicting adpositional phrases (eg. the nun is in the saucepan). We then analysed whether the extent to which participants produced VO order for the events stimuli predicted the extent to which they produced prepositional orders for the adpositional stimuli.
Our results do not provide support for the hypothesis that verb phrase head-directionality has an effect on adpositional phrase head-directionality in silent gesture improvisation. We discuss a number of possible reasons for this. Firstly, methodological issues with our experiment resulted in the exclusion of much of our data, and thus our results are based on a small number of participants. Secondly, perhaps the mechanism that causes harmony is only weakly active in improvisation, and our experimental design was insufficient to demonstrate such a weak effect. Finally, maybe there is no effect to be found: the mechanism that causes harmony with regards to verb phrases and adpositional phrases may not be present in adult improvisation. This could indicate that the mechanisms causing harmony are not active from the beginning of language emergence, suggesting harmony develops later, perhaps as subsequent generations of children learn the language and their tendency to switch to a harmonic pattern when given a non-harmonic input, as found by Culbertson & Newport (2015), drives a change in the language.