Berawan, an Austronesian language spoken in northern Sarawak, Malaysian Borneo, is one of several languages in central and western Borneo that have unusuallyinnovative phonologies. Not only are these phonologies rich in number of changes,and the effect they sometimes have on concealing cognation (e.g., Malay bǝruaŋand Long Terawan Berawan kǝbiŋ ‘the Malayan sun bear: Ursus malayanus’ arecognate), but the search for theoretically-supported motivations for some changesleads nowhere. One of these changes in all dialects of Berawan is *b > k in intervocalic position, a change that is abundantly attested, and therefore not in question asa valid transition from an earlier to a later state of the language. A basic question iswhether this was a one-step change, or a telescoping of several phonetically more‘natural’ changes, and while it can be resolved into a two-step change, this hardlyrelieves our sense of theoretical angst, since so far as the evidence allows us toinfer, these changes were *b > g in intervocalic position, followed by intervocalicdevoicing of g from two historically distinct sources. Attempts so far to show thatthis change involved other intermediate steps have yet to be successful.