Stops: timing

Timing

As well as varying one of the parameters (e.g., passive articulator), there can be variations in timing:

of one consonant relative the other sounds
between constriction gestures within a consonant
between a constriction gesture and glottal activity or nasality

Overlapping

One example is the kind of overlapping we have already seen between the closure phases of [p] and [t] in apt or [k] and [t] in doctor.

Taps

The second consonant in words like city or butter is quite different from an ordinary [t].

It is extremely fast.
The closure phase is so brief that the vocal cords usually don't have enough time to stop vibrating, then start again for the following vowel. (Often they don't even try.)

This kind of constriction is called a tap. It is usually treated as a minor type of constriction degree (alongside stop, fricative, etc.) and given a separate row in consonant charts. The IPA symbol for an alveolar tap is a lower-case [r] without the upper serif.

For example: city: tap-r

Rogers contrasts a tapped-t with an ordinary stop [d], but it is common for d's to be tapped in English as well. There is often no difference in the alveolar constriction between latter and ladder, or futile and feudal. Even when the vocal cords fail to stop vibrating during the "voiceless" tap, this usually doesn't make the words identical in Canadian English: the secondary cues to consonant voicing are still there, e.g., effects on the pitch and length of the preceding vowel.

Aspiration

The vocal cords need not turn on and off at exactly the same instant that a stop closure is created or released. There can be different timing relationships between vocal cord vibration and stop closure:

vibration occurs throughout the closure (fully voiced)
vibration begins during the closure (partially voiced)
vibration begins simultaneously with the release (voiceless unaspirated)
vibration begins after the release (voiceless aspirated)

The last three are illustrated in Rogers' Figure 3.3:
Rogers 3.3

These possibilities are referred to as Voice Onset Time (VOT). Stops transcribed as "voiced" and "voiceless" in a broad transcription can have different VOTs in different contexts and in different languages.

Voiced stops in French are typically fully voiced. English voiced stops are often fully voiced between vowels, but usually only partially voiced (and sometimes even voiceless unaspirated).

Voiceless stops in French are typically unaspirated (i.e., no delay between the release and the onset of voicing). Voiceless stops in English are unaspirated in some contexts (e.g., after an [s]) and aspirated in others (e.g., at the beginning of a stressed syllable).

In English, the delay in voice onset after an aspirated stop will occur whether the following sound is a vowel or a consonant. If the following sound is a consonant, it will become (at least partially) voiceless:
plum [p voiceless-l ^ m]

Next: Vowels
Previous: Releases
Up: table of contents