In earlier work Takia, an Oceanic Austronesian language, has been compared with its Papuan neighbour Waskia to show that much of the grammar of Takia can be explained as the outcome of metatypy on the model of Waskia or a language typologically like it. Takia belongs to the Western subgroup of the small Bel family, the members of which have all undergone metatypy to some degree. In this paper data from the Bel languages are used to reconstruct the history of certain Takia grammatical features, especially clause chaining and postverbal enclitics, in order to confirm or refute earlier inferences about their history. These inferences are largely confirmed, and the origins of clause chaining and postverbal enclitics are situated as a sequence of changes within a history of the Bel languages. The investigation shows that, as expected, metatypy is a gradual process. It also confirms that the innovation of clause-chaining was a normal piece of grammaticisation, but one that was triggered by speakers' bilingualism in a clause-chaining language or languages.The investigation has yielded an unexpected spin-off by showing that metatypy was an ongoing process during the development of the Bel languages. Reconstructing the genesis of Western Bel clause-chaining and enclitic sequences forces us to reconstruct ongoing contact, whilst Matugar, an Eastern Bel language, reflects metatypy probably on the model of its Western Bel relatives.
If the inline PDF is not rendering correctly, you can download the PDF file here.