CONtract vs. ConTRACT

[Rudolph C Troike]

(Interesting how we can adapt the graphic system to indicate prosodic/
suprasegmental features.)

        Re Arnold Zwicky's (as usual) useful and thorough contribution, a
couple of minor observations:

        ICE(d) cream  --  ice(d) TEA

Many of these things are lexically idiosyncratic, but time or frequency
may be a factor. Natalie's (and others') "ice CREAM" may be a retention of
the earlier phrasal condition, as Arnold suggests. Is "ice cream" older
than "ice tea" (Barry?).

        Chomsky and Halle worried about the application of the Alternating
Stress Rule to "TennessEE", since "TENNessee" would have been predicted.
(Has SPE been so soon forgotten?) As dInIs recently attested, the latter
is in fact the natural pronunciation of many natives of the territory.

        Rudy

NB: re "Hallowe'en/Hollowe'en", some of the [ae/a/)] variation goes back
to Old English, whence, inter alia, we get semantically differentiated
alternate pronunciations in BAND:BOND and STAMP:STOMP, and in tense
vowels, HALE:HEAL(TH):WHOLE:HOLY:HALLOW ["make holy"]. (BTW, my MWCD-10
gives /a/ as a second pronunciation for Hallowe'en.)