Re: Need some help with terms: was "rhotic miscellany"

On Fri, Nov 05, 2004 at 12:14:53PM -0500, Sally Caves wrote:
> What I need help with is understanding "rhotic" and especially
> "approximant."
Well, "rhotic" is basically used to describe the family of sounds which
have been represented by the letter R in Latin scripts, or its
equivalents in other scripts.  It is a "family" relationship because
there is no set of isolated traits you can point to and say "that
constitutes rhoticity".  Any two rhotics will have some features which
overlap, but there will be other rhotics which completely lack the
feature shared by the first two (overlapping elsewhere).

Fundamentally, what makes them rhotics is that people think they sound
similar to other rhotics.  And, in fact, there may be some *acousttic*
properties that are shared by all rhotics, but there don't seem to be
any *articulatory* ones.

> Is rhotacized the same as rhotic?
No.  The term "rhotacized" (or "rhoticized"; I've seen both spellings
and tend to use the latter) refers to the influence of a rhotic on an
adjacent non-rhotic sound.    For example, most Americans have [a] in
both the words "car" and "father", but the vowel in "car" sounds
different because of the following [r\].  Basically, you start to
pronounce the [r\] before you're done pronouncing the [a], and the
result is a mixture of the two sounds, spelled [a`].  (In IPA, there is a
little hook that looks like a small r used instead of the `.)

Vowel rhotacization tends to happen with *sonorant* rhotics - the ones
that can double as vowels, because they can be pronounced at the same
time as a vowel.  It's easy to see that [r\] is such a sound; in fact,
the sequence schwa + r decomposes into just [r\=] ("=" means
"syllabic") for many speakers in words like "speaker" [spikr\=].

> Somebody else told me that my American "r" (in "American" and "car")
> was probably an approximant, and he distinguished it from a
> "retroflex."  Have I misunderstood him? (Can't remember who it was;
> I'm trying to consolidate my responses into one big one here.)
That was me.  I didn't intend to contrast "approximant" with
"retroflex"; "approximant" is manner of articulation, while "retroflex"
is place of articulation.  I certainly don't have a retroflexion
in [r\], and would be surprised to find that most Americans do.

> What's an approximant?  A sound made where
> the point of articulation is almost reached but isn't?
Precisely.  The sounds [l], [r\], [w] and [j] (the English consonantal Y
sound) are all approximants; there's no friction involved in producing them,
because the closure isn't tight enough to produce any.  Nevertheless, the
partial closure affects the sound of the air stream, so you can make out
the sounds.

Approximants are sometimes called "glides".  The sounds [w] and [j] are
also called "semi-vowels" because they are basically shortened versions
of [u] and [i] (or [U] and [I], or some other vowels in that area).  The
sounds [l] and [r] aren't generally called semivowels because they don't
correspond to any vowels, although as I said [l=] and [r\=] do serve *as*
vowels in some languages.

-Marcos

Reply