The inclusion of the poorly known Karas is best argued in Cowan, H. K. J. 1953: 33-36 , with systematic correspondences in pronominals and a few items of basic vocabulary. Evidence for inclusion in Trans New Guinea is weak C. L. Voorhoeve 1975: 432-437 , both lexically and pronominally, cf. Andrew Pawley 2005: 94-95 . Likewise, the East Timor/Alor comparisons in Hull, Geoffrey 2004 are flimsy.
C. L. Voorhoeve 1975: 432-437