>In a message dated 9/26/2001 8:25:29 PM Pacific Daylight Time, >[log in to unmask] writes: > >> Yeah, I've noticed that /k/ and /g/ are a much harder pair to >> distinguish in Japanese than /p/ and /b/ or /t/ and /d/. Is there >> something about the velar POA that makes them harder to distinguish? [snip] >> > Yes, in a word. Voicing is harder to keep up the further you go back in >stops and in fricatives. In fact, voicing is hard to keep up for >non-sibilants, so they often turn into stops, such as the [D] in the English >"the". Our professor showed us most of the time that [D] is really [d_d] >(dental stop--is that the way to do it?). But, yeah, there was some formula >he showed us for how, on stops, voicing becomes more and more difficult the >further back you go, and this can be shown where even in languages with a k/g >contrast, [k] appears more than [g] (a given), and all other voiced >consonants appear far more than [g] as well. The reason has to do with >surface area and the aerodynamic voicing constraint... We're learning this >right now, so I'm no expert--hopefully by the time the midterm rolls around I >will be. :) > Actually, this is kind of cool, since it gives me a reason of sorts to drop [g] out of Simafira, but not the other voiced stops, [b] and [d]. Still unusual, but not without reason. I had something like that in mind just from observation of how I personally pronounced them - but didn't have the justification. -- Steve Kramer -=oOo=- scooter at buser dot net Thought for today: "You! What PLANET is this?" Dr. McCoy, "The City on the Edge of Forever"