Although it was long believed that creole languages lack productive morphology, recent work shows this to be untrue. Careful investigations by numerous researchers have demonstrated that creole languages have both derivational and inflectional morphology (see, e.g., DeGraff 2001; Lefebvre 2003; Plag 2003, 2006). However, it is also clear that the creole morphemes are generally not the same as those in the lexifier languages, in terms of their phonological forms, meaning, conditions of use, or some combination of these, which raises interesting questions with regard to the development of creoles, the answers to which are important not just for our understanding of these languages but also for our understanding of the human capacity for language learning and creation more broadly. Questions related to the nature and source of the morphological forms are of natural interest to linguists. For instance, did the morphemes arise through borrowing or grammaticalization within the creole? Are creole morphemes more semantically transparent than those in non-creole languages? (See Plag 2006 for some discussion of these themes.)
Other issues remain, however, even after these questions have been answered; for instance, how does a form that is likely variable in use become a stable part of a grammar, shared by all speakers, and why that form and not another (particularly when a substrate form is borrowed)? While these might seem to be psycholinguistic or social-psychological rather than strictly linguistic questions, understanding these other processes is necessary to understanding the full dynamics of language emergence, formation, and change (Mufwene 2010).
Here we explore one such process: how a form comes to have a shared meaning in an emerging linguistic community—that is, the emergence of conventionality [End Page 109] (Hockett 1960). Our discussion is inspired by the development of spatial morphology in Nicaraguan Sign Language (NSL); however, the issue is relevant for grammatical forms in any new contact language. Indeed, one might assume that it is less of a problem in a signed language where some degree of iconicity can get around the problem of discerning meaning for novel forms, since the meaning of signs (iconic ones at least) is often quite obviously related to their form.1 However, as we will show, even here it remains a problem.
We begin by describing spatial morphology in NSL, a signed language that developed in Nicaragua in the 1980s. Although not a typical creole, in that it lacks a lexifier or superstrate language and substrates shared by multiple participants involved in the language creation process (Kegl and McWhorter 1997), NSL has nevertheless been described as a creole by some researchers (see, e.g., Senghas 1995). Whatever its status, NSL presents an interesting case for investigating the emergence of conventionality in a morphological system. After briefly describing how space is used in the language at present, we discuss why it is surprising that the forms in question took time to emerge and/or stabilize in NSL and present results from an experiment with adult hearing speakers that explores why this might be the case. Finally, we discuss the problem of conventionalization with respect to emerging languages more broadly and speculate on how grammatical forms might come to have shared meanings in a newly emerging language.
1.1 Background on Nicaraguan Sign Language
The initial social conditions enabling the emergence of Nicaraguan Sign Language occurred in the late 1970s when an elementary school for deaf children was established. Although the school was oralist, the children were allowed to gesture with each other in the playground. Many of these children continued on to a vocational school established in 1980 and, importantly, began to socialize outside school hours, using the emerging gestural system that they were actively involved in creating. The community, and with it the language, continued to grow over time as more children and adolescents had the opportunity to attend the two schools. Eventually, a Deaf Association was formed, providing a locus for the Deaf community post-school (see Senghas et al. 2005, for a more complete description of the history of the language).2
Like many signed languages (Supalla...