> That's true but isn't exactly the same thing. "Irregularly" pronounced > combinations are still valid combinations. I'd say the universal example > for languages that are written in Latin alphabet or a variation thereof > would be the (notorious) 'fgsfds.' It's an invalid combination because > there is _no_ pronunciation at all--except 'figgis-fiddis' which is a > really recent, and ground-breaking, invention ;-)
by this definition, one could devise a valid input method with which it would be impossible to type "xyzzy". > no written language stands independent of its pronunciation rules. > alphabets need a somewhat larger set of rules than syllabaries, but > that's true independent of language. i'm not sure they are fully dependent. consider acronyms. or even variable names. (sometimes these need to be referred to in speech.) there are special hacks for making these pronouncable. in mathematics the same symbol can have many pronunciations that depend entirely on the context. i'm not a linguist, but the linguists i know subscribe to the viewpoint that the written and spoken language are separate. and evolve separately. i would derive from this that writability is independent of pronouncability. trying to think as a linguist, i would consider spoken acronyms to be cognates from the written language. as an homage to j. arthur seebach i'd say, "english is *neat*". - erik
