Relabeling and Word Order: A Construction Grammar Perspective

Relabeling and Word Order: A Construction Grammar Perspective
of 31
All materials on our website are shared by users. If you have any questions about copyright issues, please report us to resolve them. We are always happy to assist you.
Related Documents
  To appear as Chapter 5 in C. Lefebvre, The Theory of Relabeling in Creole Genesis: The Hong Kong Lectures . Oxford: Oxford University Press. Chapter 5 Relabeling and Word Order: A Construction Grammar Perspective *  Claire Lefebvre and Renée Lambert-Brétière Abstract Word order in creoles does not systematically reflect that of either of their contributing languages. This puzzle has generated a significant amount of research from different  perspectives. On the basis of a sample of Caribbean creoles, this chapter addresses the question of how word order is established in a relabeling-based account of creole genesis within the framework of Radical Construction Grammar framework. In this framework, word order is not specified as part of atomic constructions. Rather, it is specified as part of the constructions in which individual words appear (e.g., [ Q DEF ADJ    N ]). We show that this model allows for a straightforward and principled account of how the different word orders are established in creole genesis. 1.   Introduction   1.1. The puzzle of word order in creoles and in creole genesis Word order in creoles and in creole genesis has long been, and still is, a puzzle for researchers. While the vast majority of creoles are subject-verb-object (SVO) languages (e.g., Caribbean creoles, see Muysken 1988a: 85), some are subject-object-verb (SOV) (e.g., Hiri Motu), and some even have both SOV and object-subject-verb (OSV) order (e.g., Pigin Yimas (-Arafundi)). While the vast majority of creoles derive their word order from their superstrate languages (e.g., Plag 2008; Siegel 2008), some appear to constitute counterexamples to this generalization. For example, Berbice Dutch is SVO in spite of the fact that both of its contributing sources, Dutch   2 and Ijo, are SOV languages (e.g., Kouwenberg 1996; Muysken 1988a). While Saramaccan is generally an SVO language, like its superstrate languages, English and Portuguese, it has  postpositions, a feature of OV languages, like its substrate Gbe languages (e.g., Essegbey 2005). In some cases, it may look as if the respective contributions of the source languages to a creole’s word order are split between lexical and functional categories. i  For example, the structures below provide an overview of constituent and word order in the nominal structures of Fongbe, Haitian, and French ( OBJP =objective phrase, GENP =genitive phrase, RC=relative clause). (1)   OBJP    N   ADJ   GENP   ADJ    NUM   DEM   DEF   PL   Q   RC   F ONGBE  RC INDEF  (Lefebvre & Brousseau 2002: 56) (2)   Q    NUM   ADJ    N   OBJP   GENP   ADJ   DEM   DEF   PL   RC   H AITIAN   INDEF  RC (3)   Q   DEF    NUM   ADJ    N   OBJP   OBJP   ADJ   F RENCH   DEM  RC POSS  As can be seen from these structures, the constituent and word orders of Haitian correspond  partially to those of Fongbe, a substrate language, and partially to those of French, the lexifier language. Apart from relative clauses, which appear to the right of the head noun in all three languages, it looks as though major category lexical items  —  adjectives, numerals, and quantifiers  —  follow the word order of French, whereas functional category lexical items  —  demonstrative terms, definite determiners, and plural markers  —  follow the word order of Fongbe. For example, while adjectives, numerals, and quantifiers follow the noun in Fongbe, they all  precede it in Haitian, on the model of French. Furthermore, in Haitian, adjectives may either   3  precede or follow the noun, as in French. By contrast, Haitian definite determiners ( DEM DEF PL ) all follow the noun, as in Fongbe, and unlike in French where these items all precede the noun. The generalization that lexical categories follow the word order of the superstrate language, and functional categories that of the substrate languages, is only partially correct, however, as the word order of some functional categories in creoles departs from that of their substrate languages. For example, unlike Haitian and Fongbe, where the plural marker follows the noun, in Martinican Creole it precedes it. This is illustrated in (4). (4)   a. liv la yo H AITIAN               F ONGBE    book DEF   PL   ‘these books’   (Lefebvre’s field notes)   b.  se liv la  M ARTINICAN   PL    book DEF   ‘these books’   (Lefebvre’s field notes)  A similar example is provided by Saramaccan where, in contrast to Martinican, Haitian, and Fongbe, the definite determiner precedes the noun, as shown in (5). (5)   a. liv la H AITIAN &   M ARTINICAN   wèmá     F ONGBE    book DEF   ‘the book’ (Lefebvre’s field notes)   b. di búku  S ARAMACCAN   DEF  book ‘the book’   (Lefebvre’s field notes)  Other similar examples can be found in Lefebvre (1998). Thus, the hypothesis that the contributions of the source languages to a creole’s  word order are divided between lexical and functional categories falls short in view of counterexamples of the type in (5) b) (see also the discussion in Plag 2000). Another avenue, explored in Lefebvre (2007), is that word order in creoles is established on the basis of the position of the form that provides the label for a lexical entry in the creole in   4 question. For lexical categories, this is illustrated by the surface position, in Haitian, of adjectives, numerals, and quantifiers (see (2)), with respect to that of the same categories in Haitian’s source languages (see  (1) and (3)) (see also Aboh 2006). For functional categories, this is illustrated by the position of the definite determiners in (5). The form of the Haitian and Martinican definite determiners is derived from the French postnominal adverbial deictic form là ; the definite determiner is thus postnominal in these creoles (see (5)a)) (for details, see Lefebvre 1998: 78  –  9, and the references cited therein). The form of the Saramaccan definite determiner di  (< di(si) ) is derived from the English prenominal form this ; the definite determiner is thus prenominal in this creole (see (5) b)) (for details, see Lefebvre 2012). In some cases, lexical items have been retained from the substrate languages. For example, in Saramaccan, the Fongbe focus marker   has been retained as such, including its low tone (e.g., Smith 1996). In this case, there is also a link between the label and the word order. As is shown in (6) ,  the  position of the creole lexical entry is the same as that of the substrate language form (see also Smith 1996: 126). (6)   a.   ví       ,     F ONGBE  Massè child PL arrive ‘It is the people of Massè who have arrived.’  (from Hounkpatin 1985: 218)  b.  Andí   i bói.  S ARAMACCAN  what FOC  you cook ‘What did you cook?’  (=(11b) in Smith 1996: 117) In this chapter, we will argue in support of the generalization that word order in creole genesis is mainly determined by that of the form (from either the superstrate or the substrate language) that provides the label for the creole lexical item. We will show, however, that this  proposal is not sufficient to account for all cases of word order. For example, while both the definite determiner and the plural marker in Haitian are individually derived from French   5  postnominal forms  —  là  and eux , respectively  —  as we saw in Chapter 3, there is nothing in the French nominal structure that could account for the surface order of these morphemes with respect to one another in the creole (see (4)). Furthermore, some data appear to constitute counterexamples to the above generalization. For example, the fact that there are postpositions in Saramaccan such as báka   ‘behind’ from English back   and déndu   ‘in’ from Portuguese dentro   ‘in’ constitutes a counterexample, as back   and dentro  in English and Portuguese are not  postpositional but prepositional. At first glance, then, it looks as if there were no principled way in which word order would be established in creoles. In addition to the general problem posed by word order in creole genesis, as outlined above, there is the specific problem posed by the choice of a theoretical framework. For example, the relabeling-based account of creole genesis advocated in Lefebvre (1998, and related literature) was formulated within a theoretical model in which directionality properties are specified as part of individual lexical entries (e.g., the notation V specifies that the verb takes its complement to the right). Since a creole’s word order does not necessarily reproduce that of its substrate languages, as we saw above, word order phenomena constitute systematic counterexamples to a relabeling-based account of creole genesis in such a model (see, e.g., the discussion in DeGraff 2002: 355  –  67). Since relabeling was otherwise shown to account for a vast amount of creole data, we would not want to falsify a relabeling-based account of creole genesis on the basis of word order alone. Instead, we conclude that word order should not be included in individual lexical entries. The account of word order in creole genesis proposed in this chapter is thus set within a theoretical framework in which word order is not specified as  part of individual lexical entries, namely Croft’s (2001) Radical Construction Grammar (RCxG).  
Related Search
We Need Your Support
Thank you for visiting our website and your interest in our free products and services. We are nonprofit website to share and download documents. To the running of this website, we need your help to support us.

Thanks to everyone for your continued support.

No, Thanks