Is voicing a gradient scale?

Question

In one online linguistics community, I read the statement that "voicing is not all-or-nothing and that it is a gradient scale." This got me thinking: is this statement true or false? I guess it may be true, but I have no idea how.

If true, what about the "relative degree of voicing" between the consonants [z] and [d͡z]? Considering the definition of voicing, I would say that [z] is more voiced than [dz] because the vocal chords vibrate more in case of [z] than for [d͡z]. Is my opinion true? And also for what consonants do we see this phenomenon?

Answer 1

Voicing is gradient in the sense that in some languages, voicing onset time begins sooner than in others. For example, Spanish voiced stops are more voiced than their English counterparts (also see Lisker and Abramson 1964). What English counts as a /b/, with little or no prevoicing of the actual stop segment, would be phonetically equivalent, on the basis of VOT, to a Spanish /p/.

Is it also gradient that particular sounds are more voiced than others? I do not know for sure, but since the level of difficulty varies in voicing different segments, I'd say possibly. For example, it's very easy -- in fact the default -- to voice vowels, and relatively difficult to maintain voicing through the complete oral closure during a stop. If this intuition is true, it will be easier and thus more common for [z] to be fully prevoiced than [dz] or [dd].

Answer 2

To add to Knitter's answer, many Phonologists have considered the relative voicing of sounds, especially when trying to make sense of the syllabification of segments. The precise meaning of "voicing" is important here: are we simply asking whether the vocal chords are vibrating? If we are, then voicing is a binary (yes/no) distinction. For this reason, Phonologists usually use the term sonority to refer to the relative 'strength' of voicing.

A great place to start reading about this is pp. 38-64 in Ferdinand de Saussure's "Course in General Linguistics". We have the sense that voicing is not an "all or nothing" phenomenon, and de Saussure explains this nicely.

The so-called sonority hierarchy is a well-established method of ranking the relative sonority of sounds, which is closely related to voicing. This is central to modern phonology, check out John Goldsmith's "The Syllable" for a great overview of modern thought on this question.