Saturday 28 September 2013

approximant, sonorant, resonant or what?

In my blog entry on open approximation - see here - I wrote that academics call sounds which are produced with open stricture either approximants, sonorants, resonants, glides, semi-vowels or even frictionless continuants.

A helluva lot of terms for this sound group, innit? Which one should be preferred?
Here's a brief random survey first.

Ashby, P. (2011), Understanding phonetics
Sounds which are produced with wide approximation are called approximants by the author. She points to the fact that phonologists distinguish sonorants (= vowels, approximants, nasals) from obstruents (= plosives, fricatives, affricates. The English sounds belonging to the group of approximants are [ɹ, w, j, l].

Ashby, M., Maidment, J. (2005), Introducing phonetic science
In their glossary of the book an approximant is described as a "consonant sound made with a constriction between two articulators which is not narrow enough to cause air turbulence" (190). On p. 57 approximants are equated with sonorants - "[...] approximants are sonorants [...]" - meaning that approximants are a subgroup of sonorants. Reference is made to "older terms" (59) for approximants: glides or frictionless continuants. English approximants are [ɹ, w, j, l].

Collins, B., Mees, I. (2013), Practical phonetics and phonology
The authors present a clear-cut, unambiguous division of the English consonant system on p. 52:

credit: Routledge



Cruttenden, A. (2008), Gimson's pronunciation of English
Cruttenden's classification of sounds according to their noise component due to the degree of constriction in the vocal tract leads to two classes of sound - obstruents (comprising plosives, fricatives and affricates) and sonorants (nasals, approximants and vowels).

Knight, R.-A. (2012), Phonetics - a coursebook
When the articulators are positioned in wide approximation we produce the English approximants /w, j, l, r/. Approximants together with nasals and vowels belong to the larger class of sonorants. Plosives, fricatives and affricates are subsumed under the category of obstruents.

Laver, J. (1994), Principles of phonetics
Sounds produced with open approximation during their medial phase are called resonants by Laver. They are divided into central resonants - comprising the approximants /j, w, r/ and the vowels (which Laver calls vocoids) - and the lateral resonant /l/.

Lodge, K. (2009), A critical introduction to phonetics
Sounds produced with the articulators too far apart for friction to occur are called approximants (Lodge also makes mention of the equivalent terms frictionless continuants and semi-vowels). This class of sounds encompasses [ɹ, w, j, l].

Roach, P. (2009), English phonetics and phonology
The sounds /m, n, ŋ, l, r, w, j/ are called continuants (p. 46), and are subdivided into the nasals /m, n, ŋ/ and the approximants /l, r, w, j/.

Windsor Lewis, J. (1969), A guide to English pronunciation
Here is Windsor Lewis's classification in diagrammatic form based on paragraph I, section 8 of the book:

What are we to make of all this?
The majority of phoneticians mentioned above prefer to use the term approximant for sounds produced with wide/open approximation. So I'm going to call [ɹ, w, j, l] approximants. (Kraut locutus, causa finita)

15 comments:

  1. A wise decision.
    BTW, may I quote a short passage from J. C. Catford’s A Practical Introduction to Phonetics:
    “… the articulatory channel for a fricative is so narrow that the airflow through it is always turbulent, and hence noisy, whether it is voiced or voiceless. The articulatory channel for an approximant, however, is a little wider than that of a fricative, just to the extent that airflow through the channel is non-turbulent (hence no hiss-sound) when it is voiced, but it IS turbulent and hence noisy (though not strongly so) when it is voiceless.
    … [i] is a typical approximant, but [æ] is clearly something different. We call this latter type of sound, which never has noisy local turbulence, a resonant. We shall, in fact, have little occasion to use the term resonant because the only sounds with this very wide type of stricture are vowels like [ɛ] and [æ], and we normally use a different terminology in the description of vowels.”

    ReplyDelete
    Replies
    1. Thanks, Emilio, for reminding me of Catford's book. I have it (= 1st ed.) on my shelf but somehow had forgotten about its existence.

      Delete
    2. Just bought the 2nd ed of Catford's book

      Delete
    3. I mean, Catford -of course!

      Delete
    4. Of course, you are a sweet guy too!

      Delete
  2. You've turned up what's essentially a set of synonyms here, just one example of many in phonetic terminology. Let's just hope some bright author doesn't start inventing distinctions within this this set (imagine a PhD thesis proposing non-approximant sonorants vs. approximant sonorants to solve some obscure problem that has eluded us all for centuries). These synonyms probably arise from national traditions, theoretical schools and individual influential authors.

    Advice to students? Be pragmatic, stick to the terminology used by the person setting your exam, switch as necessary whenever you get a new teacher using a different terminology (maximizing your exam performance). Advice to teachers? Make your mind up (as you did) and adopt a course book accordingly (and if your course book is imposed on you by someone else, take a deep breath and pretend you like it, maximizing your career possibilities).

    Sooner or later we'll have to consider the status of vowels and consonants. I haven't yet heard a watertight definition, except by simply listing them (the vowels are /i e æ .../ and consonants /p t k .../ etc), which is what most grammars do. And then I reached the end and found the example taken from JWL. Well done Jack.

    ReplyDelete
    Replies
    1. You're in good company. I just checked the latest IPA definitions, they call them all approximants too. There's still work to be done on approximant rhotics, there's still too much focus on the tongue blade and too little on the pharynx, I see people mentioning "bunched tongue" more often, and I've needed a uvular approximant for Inuit. Up to now I've used ʀ and redefined it.

      Delete
  3. Off topic, sorry. But, does anyone know whether Mr Jack Windsor Lewis is well? He hasn't written anything in his "Phonetiblog" for more than two months. Thanks.

    ReplyDelete
    Replies
    1. He IS well. He's just extremely busy at the moment.

      Delete
  4. Thank you. I'm very glad to hear that.

    ReplyDelete
  5. Hating the term 'frictionless continuant', I was delighted with Peter Ladefoged's excellent coinage, introduced with no fanfare, recorded in the OED thus:
    '1964 P. Ladefoged Phonetic Study W. Afr. Lang. vi. 25 The term approximant is used here to describe a sound which belongs to the phonetic class vocoid or central resonant oral, and simultaneously to the phonological class consonant in that it occurs in the same phonotactic patterns as stops, fricatives and nasals.
    I shd thank Sidney for his too kind pat on the back in regard to a book publisht so long ago (1969) that it's about to be re-publisht in a series of historic texts to my simultaneous gratification and embarrassment.
    My thanks also for the kind enquiry in respect of the long gap in my series of 'phonetiblogs' which I expect to resume shortly.

    ReplyDelete
  6. This comment has been removed by the author.

    ReplyDelete
    Replies
    1. This comment has been removed by the author.

      Delete
    2. True enough, then the other day I was reading something about phonotaxis, that reminded me you would need to subdivide approximants. In syllable onsets, semivowels come closest to the vowel. Depending on the language of course. In English you have words like squire, /skwV/. English doesn't seem to have so many general examples for /j/. There's miaow or piano, depending on your pronunciation, I'd say /mjaw/, with the semivowel located later than the nasals.

      Anyone working with Latin during the past 2000 years or more will also have needed to identify semivowels, in order to handle alternations between [i, u] and [j, w]. Again, if you find a corner dealing with processes exclusively affecting [l, r] in the phonology of some language, you will need to isolate the liquids.

      Delete