Re: [TIED] from /m/ to /t/

----- Original Message -----

From: Glen Gordon

To: cybalist@egroups.com

Sent: Tuesday, May 23, 2000 8:13 AM

Subject: Re: [TIED] from /m/ to /t/

Gerry asks a fun question:
>Piotr, is there any evidence of a /m/ to /t/ linguistic
>switch in English.  For example, a switch from idiom to idiot?

Glem wrote (on *m > *t):

I can imagine it, Piotr :) But it does require more than one step and the
plausibility depends on the environment of the word. If it's final, one
might suppose the following: *-m > *-n > *-t. Not too impossible, I dare
say.

True, but it would still be rare. Nasals are rarely denasalised except through dissimilation (e.g. n...m... > d...m...) or gradual lenition, and as for the symetrical change, e.g. d > n (spontaneous nasalisation), it happens even more rarely, except through assimilation (such as dn > nn). Word-finally, as in your hypothetical example, phonological contrasts are typically lost and odd things may sometise happen. However, Gerry obviously meant direct, context-free developments, not restricted to auslaut or assimilatory environments.

 

For Gerry's information: idiom and idiot ARE related, but not as ancestor and descendant via phonological change. They just share the same morphological base. Like _idiosyncrasy_ or _idiolect_ they both come from Greek _idio-_ 'private, peculiar' with different suffixes added.

 

In initial position, it would require a good deal more than a generation...
Maybe we could suppose the following: *m- > *n- > *d- > *t-. Unlike the
first example, this latter recipe requires a good 20,000 years to bake :)

Let me tell ya, one has a better chance being struck by lightning than for
*m to regularly and directly, or even indirectly, become *t in a language
anyway. Even if it were to do so, it's practically impossible for the
phoneme to undergo a sweeping change like this in all positions Initial,
medial AND final?! Forget it!

Precisely.

Piotr