/N/ vs /Ng/ (was: Re: English notation)

From: "David Peterson" <DigitalScream@...>
> In a message dated 6/28/01 9:43:36 AM, nsampat@IX.NETCOM.COM writes:
> << This makes me wonder: Is there a need for a distinction between /N/ and
> /Ng/?  It seems to me that that's a difference in rendering as opposed to
> phonemics. >>
>
>     I think there is.  I say /fiNgr=/, but never /fiNr=/.  And I say
/riN/,
> but never /riNg/.  Yet, they're not in complimentary distribution, for I
say
> /lONr=/, never /lONgr=/.
I read somewhere (either this list or a different one) that the general rule
for /N/ vs. /Ng/ is that the /g/ drops morpheme-finally but not
morpheme-medially (hence /&Ngr=/ 'anger' but /h&Nr=/ 'hang-er').  But for
some reason this doesn't generally apply before -er of comparison while it
does for -er of agency: you may have /lONr=/ but I have /lANgr=/ "longer:
more long", while "longer: one who longs" would be /lANr=/.

This doesn't seem to be an unbreakable rule, though, as there's (for me,
anyway) /N/ in monomorphemic "hangar" and "dinghy".


    *Muke!