Paradigmatic complexity in pidgins and creoles

The last decade has seen increasing attention paid to questions of grammatical complexity, in particular regarding the extent to which some languages can be said to be more ‘complex’ than others, whether globally or with respect to particular subsystems. Creoles have featured prominently in these debates, with various authors arguing that they are particularly simple when set against non-creoles, with an apparent lack of overt morphology in creoles often cited as one of the ways in which their grammars are especially simplified. This paper makes two contributions to this discussion. First, it develops metrics of grammatical complexity that derive directly from a well-known model of creole development, thus providing an explicit link between the sociohistorical circumstances in which creoles formed and grammatical outcomes. Second, it applies these metrics to the newly published dataset from the Atlas of Pidgin and Creole Structures, setting this data against that from the well-known World Atlas of Language Structures, allowing for a more comprehensive and rigorous quantitative comparison of complexity in contact and non-contact languages than has previously been possible. It will be seen that there is good evidence that contact languages are simplified overall with respect to a class of complexities labelled PARADIGMATIC here but that this general conclusion nevertheless masks significant underlying variation among them.