Back to list

Detail of contribution

Auteur: Olivier BONAMI

Co-Auteur(s): Ana R. LUÍS, U. of Coimbra

A morphologists perspective on creole complexity

Abstract/Résumé: The complexity of creoles has generated a great deal of discussion during the last decade (McWhorter, 2001; de Graff, 2003), and inflectional morphology has been perceived as a defining domain for determining whether creole languages are simple or complex (Siegel, 2004; Plag, 2006). However, while the concept of complexity is taken to be of central importance for our understanding of language structure, it remains a rather ill-defined concept and, hence, no widely accepted and objective quantification exists for either linguistic or morphological complexity (Moscoso del Prado Martín, 2011). In this paper, we survey some of the different measures that have been adopted to define morphological complexity. Such measures include i) counting cells (paradigm size), ii) counting the number of morphs per word (word size), iii) counting the number of distinct bound morphs (affix inventory size), iv) detecting the presence or absence of certain morphological phenomena or structures that are chosen as representative of complexity, v) calculate implicative relations between paradigm cells (Nichols, 2010; Shosted, 2006; Malvern et al., 2004; Xanthos & Gillis, 2010, Ackerman, Blevins & Malouf, 2009; Bonami, Boyé & Henri 2011). For historical reasons, criteria i) to iii) have been at the forefront of criteria for measuring morphological complexity in general and creole complexity in particular (Sapir 1921). Our survey will reveal, however, that by themselves they give a skewed view of the overall complexity of the systems. We argue in favour of a notion of morphological complexity that is meaningful both for creole and non-creole languages, by examining the synchronic system of a creole language with respect to other languages, rather than focusing on what it has lost from its lexifier. In addition, we defend the view that morphological complexity is multidimensional and that it should be analysed along different dimensions, such as inflection classes, syncretism, suppletion, prevalence of fusion, allomorphy, morphotactic size and overabundance. Given our present understanding of morphological systems, designing separate measures for each of these dimensions constitutes a more promising research agenda than attempting to rank languages using one single measure. Ultimately, we will claim that examining different aspects of complexity constitutes a more purposeful goal than deciding which language is more complex. One further goal of our talk will be to examine whether each one of these dimensions is triggered by phonological or syntactic factors or whether they are purely morphological, in which case they will lend support to the autonomy of morphology.