Retour vers liste

Détail de la contribution

Auteur: Richard INGHAM

The diachrony of a child L2 variety: later Anglo-Norman

Abstract/Résumé: An L2 variety, in its systemic aspects, can be seen as the product of Universal Grammar set to the values of an L1, supplied with L2 input after early childhood (Herschensohn 2007, Meisel 2011a,b). Over time, if the L1 UG values change, and if the L2 input also changes, an L2 variety should show diachronic change. Establishing how this operates over historical time, however, is not straightforward. Written L2 production from earlier periods (other than with classical languages such as Latin or Sanskrit) is usually not plentiful enough to allow detailed study of the question, and where it exists has generally been subject to normative pressures. An exception, in both respects, is Anglo-Norman, the variety of French used in England between the Norman conquest and the beginning of the 15th century. Large quantities of Anglo-Norman literary and non-literary materials are available (Trotter 2007); its spelling forms and morphology noticeably resisted continental French literary norms (Pope 1934). Anglo-Norman showed a clear evolution across time, c. 1100-c. 1400, but in two very different ways. Its segmental phonology changed so as to align itself on L1 English, but its syntax changed in accordance with the evolution of continental French. As demonstrated by Tanquerey (1915), phonological neutralisations in later Anglo-Norman interfered with French morphological distinctions of noun gender agreement and conjugation assignment. In addition, English exerted increasing influence on Anglo-Norman metrical phonology. A series of syntactic studies (Ingham 2006, 2010, 2011, 2012), however, has found that in numerous respects Anglo-Norman clausal syntax not only showed little or no interference from English, but conformed with ongoing syntactic changes in medieval French. Parameterised syntactic features of medieval French were transmitted successfully, but consonant and vowel oppositions not in the phonemic repertoire of English failed to be transmitted. L1 effects therefore relate to traits acquired in the earliest years of acquisition, the phonemic inventory and suprasegmentals. By contrast, freedom from L1 influence in the domain of clausal syntax implicates a window of nativelike acquisition that remained open for somewhat longer, into middle childhood. It is shown that these outcomes accord with contemporary testimony of the acquisitional setting of Anglo-Norman in the relevant period (Orme 1973), and offer no support to earlier notions of Anglo-Norman as an instructed L2 (Rothwell 1975).