[YG Conlang Archives] > [engelang group] > messages [Date Index] [Thread Index] >
selpa'i, On 17/09/2012 08:46:
Am 17.09.2012 03:27, schrieb And Rosta:selpa'i, On 17/09/2012 00:11:Really? [x] is hard to distinguish from [h] for English speakers?No, e.g. /ihi/ and /ixi/ would be hard to distinguish for human beings in general.So am I an exception? I couldn't really believe that, so I opened Audacity and recorded [ihi] and [ixi] in random order. The difference is like night and day to me, so I'm not sure what this is all about.
I'd like to hear what you're doing. If you say, for example, [ipi], the tongue is going to remain in i-position during the [p], and the same goes even for, say, [ili]. For [ihi] to even be phonetically possible, i.e. with the intended [h] remaining as [h] rather than becoming [c,], you'd have to move the tongue out of i-position into a position as open as [@]. That's a very unnatural and unnatural-sounding thing to do. Perhaps -- I am speculating -- you made the x variant with closer stricture, which causes saliva to vibrate, giving rise to 'scrape'. If so, that's not a very robust contrast, and isn't a contrast between [h] and [x] (but rather is a contrast between an wider-stricture [c,] and a narrower-stricture [c,]).
I think as long as<'> doesn't become [?], I won't compain.There had been a consensus for<'> being [?]! Xorban<'>, that is. There's no consensus on whether Lojban<'> has any reflex in Xorban.Oh! Uhm... okay :) . I had expected<'> to be [h]. Don't you then have two [?], once as<'> and once as<q> ?
<'> would be an optional allograph of <q>, used optionally when /q/ has its name-stem-delimiting function. --And.