Patio : The world from South Texas

[Arnold Zwicky]

first, my own linguistic history re PATIO: born 1940, raised in
eastern pennsylvania, heard and used only [ae] in the word, except
from one family who used [a], and were widely mocked for their
pretentiousness in this (and some other linguistic matters).

now, there are several countervailing tendencies here.  on the one
hand, in the early stages of borrowing, when the use of a word is
associated with its language of origin, there's a tendency to preserve
as much of the phonetics of the original as possible.  even later,
some speakers will want to preserve this pronunciation, on the grounds
that it's the "original", hence the "correct", one.

on the other hand, once the borrowing becomes felt to be merely just
another word in the borrowing language, there's a tendency to nativize
it fully; if english is the borrowing language, nativizing often
includes using the default spelling-to-sound relationships.  in
addition, some speakers will want to avoid the non-nativized
pronunciation, on the grounds that it's "foreign"; this tendency is
especially strong for speakers who want to dissociate themselves from
the culture of origin (many anglos in nevada have invariable [ae] in
their state's name, specifically because [a] would sound "mexican"),
or to mark solidarity with their own culture (the english are famous
for reproducing french loan words inaccurately; actually sounding
french would be, well, un-english), or to protect themselves from
accusations of pretentiousness (i assume that the latinate [a]
pronunciation for DATA has failed to spread, while the nativized [ae]
and [e] compete with one another, for just this reason).

the result is often variation.  not always: as far as i know, english
speakers invariably pronounce SAN and SANTA, in place names, with
nativized [ae], never with the spanish original [a].  but sometimes.

different speakers will use different variants; they'll use the one
they first heard, from people they identify with, or they'll shift
completely to another variant, for one or another of the reasons
above.  sometimes there's variation within individuals; i've heard
young coloradans shift back and forth, in a single conversation,
between [a] and [ae] pronunciations of their state's name.  as i do
between [i] and [E] versions of ECONOMICS.  such shifts might be
triggered by the choices of other participants in a conversation
(which a speaker might accommodate to, or resist), by the vowels in
other words in the context of the word COLORADO, by subtle shifts in
the speaker's sociocultural identifications, by changes in the topic,
or of course by sunspots.

i'd imagine most of the folks on this mailing list know all this
already, but nobody actually said it, so i thought it might be
useful for me to make it explicit.

arnold (zwicky at csli.stanford.edu)