Kennard (2013)

De Arbres
  • Kennard, Holly Jane. 2013. Breton morphosyntax in two generations of speakers: evidence from word order and mutation, DPhil. University of Oxford.


 abstract:
 "Following a decline over the twentieth century, Breton has seen an increase in revival efforts, including Breton-medium education. This study investigates the effect of the language transmission gap on the morphosyntax of verbs. Fieldwork was undertaken with three distinct age groups: older native speakers (aged over 65), and two groups which make up a younger generation of speakers: children in Breton-medium education, and young adults who have been schooled in Breton. The question of word order and the placement of verbs in Breton has been controversial, largely because it is complex and variable, making the identification of basic word order difficult.
 
 The data show that usage across the older generation is fairly consistent, with V2 word order in matrix clauses. Verbal mutation is also maintained. Despite the transmission gap, younger adults from French-speaking homes do not systematically replace Breton patterns with French SVO. Rather, they avoid SVO in some contexts, and indeed use it less than the senior adults. The amount of input speakers receive is crucial: children in bilingual schooling, with only half of their classes in Breton, tend to oversimplify word order patterns and show French influence. In contrast, those with additional Breton input from a family member are more proficient. Children have difficulty acquiring mutation rules, and do not seem to have grasped the system of verbal mutation, but young adults use mutation proficiently, like the older speakers.
 
 Consequently, despite strong French influence, Breton word order has remained consistent. The fact that verbal mutation is variable in children reflects late acquisition, since the young adults rarely diverge from the expected usage. Thus, the changes in Breton morphosyntax are subtler than expected in light of the unusual transmission pattern and close proximity to French. The crucial factor appears to be sustained input in the language."


protocole et données

Les données sont des données d'élicitation venant de deux protocoles. Le premier montre des images ou de petits films qui sont ensuite décrits. Pour obtenir une négation, une image est présentée rayée d'une croix rouge. L'autre protocole, destiné à éviter les tournures progressives, consiste en un jeu de cartes de mots non-fléchis et non-mutés qu'il est demandé de composer en une phrase. Ce dernier, utilisant l'écrit, est moins bien passé avec les plus vieilles générations ou les plus jeunes.

Il est parfois difficile de comprendre dans le texte quel exemple a été obtenu avec quel protocole, et relativement peu de données brutes sont fournies au moment de leur analyse. Malgré la formule choisie des élicitations, il n'y a pas de données négatives (données confirmées agrammaticales pour un locuteur donné). L'échantillon de locuteurs est forcément réduit, mais les résultats sont en pourcentages d'occurrences.