Stops
A complete stop involves three "phases":
- approach
- closure
- release
An extreme idealized picture of segment slicing pretends that the stop
is its closure phase.
Release
Normally, once the active and passive articulators have cut off the
airflow by touching, they have to move apart again and let the airflow
resume.
Unreleased stops
Many languages require their stops to be released. English
doesn't. It's possible to release the final stop in an utterance, e.g.,
That's my cat, but it's not necessary. Because it's not necessary,
it usually doesn't happen.
IPA marks an unreleased stop or a stop whose release can't be heard with
the "corner" diacritic:
The absence of a release is even more obvious in a compound word like
catnip. Here, the tongue tip touches the alveolar ridge at the
beginning of the [t] and stays there all the way through to the end of the
[n]:
The same situation occurs in words like button and widen.
In normal speech, there is no vowel in the second syllable of these words
-- the airflow remains blocked at the alveolar ridge throughout the entire
syllable.
In situations like this, it is often (unfortunately) said that the stop is
nasally released. There are also lateral releases: in
a word like bottle, the tongue tip doesn't completely leave the
alveolar ridge after the [t], only the sides of the tongue drop down to
form the following [l].
Overlapping
In an ideal world (for listeners), every stop would be released before the
next segment began. In English clusters of stops, the first stop is
usually not released till after the second stop is in its closure phase.
This means:
- There's a brief period of time where there are two closures.
- You can't hear the first release -- by the time the first stop is
ready to let the airflow move again, there's already a blockage somewhere
else in the mouth.
Situations like this are typically diagrammed as follows:
Next: Timing in stops
Previous: Making sense of allophones
Up: table of contents