Monday, 4 June 2012

Halfway through the Knight

I've dug myself through 10 of the 20 units in Dr Rachael-Anne Knight's book with the short and succinct title Phonetics.

Unit 1 is a good intro to the relations between sounds and letters and the CV-structures of words.

Unit 2 introduces the concepts of voicing in general and of voiced consonants and the transcription symbols for GB consonants. On p. 18 we read the somewhat infelicitously formulated claim "that the speaker does not use muscles to open and close the folds for every cycle of the vibration." If you know what makes the folds open and close regularly to produce phonation, then the statement is okay. Without such knowledge, the reader might be a bit at a loss. The author continues her explanation: "Instead, muscles are used to narrow them by just the right amount, so that aerodynamic contraints take over and they vibrate in the airstream of the lungs [...]" (18).
On p. 20 in exercise 2.4 the reader is asked to

"[l]ook at the bold consonants in each word below.  
many feather check bat need "
What is set in bold type is letters representing consonants, not the consonants themselves. Granted - it's only a minor point.

Unit 3 deals with place of articulation and introduces both passive and active articulators. Some of the exercises, e.g. 3.6, are not intended for NNSs of English.

Unit 4 familiarises us with manner of articulation as another feature of consonant description.

Unit 5 shows how consonants are represented by the symbols of the IPA, and midsagittal sections as a means to illustrate articulator configurations for consonants are introduced.

Unit 6 deals with vowels, their similarities and differences to consonants are identified and the cardinal vowels are introduced. The FLEECE vowel is transcribed as /i/ and the KIT vowel as /ɪ/. Dr Knight justifies her choice like this: 

[...] as we have seen, /i/ and /ɪ/ are rather similar, but do differ in the height and frontness of the tongue. Since they are represented by different symbols (/i/ and /ɪ/), it seems unnecessary to distinguish them further by adding a length mark to the longer vowel (that is, /iː/). [...] The only time this can really cause us problems is when we transcribe vowels in unstressed syllables, like that at the end of 'happy'. [...] In this book we will transcribe both vowels as /i/, while remembering that there is in fact a difference in duration between the two.
This may be fine with NSs who grew up with the shortness of the final /i/-sound in words like city or happy. But when it comes to teaching English pronunciation and phonetics to NNSs, I beg to differ with the author. What may cause further confusion with the reader is a set of three transcriptions to be found on p. 76:

Complying with Ms Knight's rule/description/explanation that we ought to transcribe a short i-vowel in unstressed syllables as /i/, shouldn't anything in a), in in b) and believe, eleven and units in c) be transcribed with an /i/, i.e.  shouldn't it be /biliv/? (Why should we transcribe <to write> as /tu rɑɪt/ rather than /tə rɑɪt/, why is the diphthong in <write> /aɪ/ on p. 74 but /ɑɪ/ here and why is there no weakform plus linking-r indicated in the phrase <are on> in sentence b)? Just typos?)

In general the text is typeset flawlessly with the exception of fig. 6.2, where some of the vowel symbols are taken from different fonts (three, at least).

On p. 71 we are informed that /ə/ "only ever occurs in unstressed syllables". Really? Look up <because> in LPD3 or CPD18 and you find a pronunciation variant with schwa in the second, stressed syllable. And what about <just>? (See also this blog post by Jack Windsor Lewis).

In fig. 6.8 the /ɑ/ symbol is the only one which is inside the quadrilateral, whereas in fig. 6.10 it's printed on the outside as all the other primary cardinal vowels.

Unit 7 focuses on airstream mechanisms. Pulmonic egressive and ingressive, glottalic egressive (= ejectives) and ingressive (= implosives) and velaric ingressive (= clicks) sounds are introduced. Mention is even made of the velaric egressive airstream in section 7.5.4, although we could easily dispense with it as "no such sounds are found in human language" (96). 

Unit 8 discusses syllable division, ambisyllabicity, syllabic consonants and some effects of lexical and sentence stress. 

Unit 9 treats allophonic variation, the concept of the minimal pair, devoicing and aspiration while Unit 10 deals with variations in the place of articulation of consonants. Here the author shows that the place of articulation (= POA) of a consonant can be more advanced or more retracted if pronounced in a certain phrase or sentence than what the POA is in the canonical articulation of a consonant. 

More on the book in a future blog post.


  1. I shd like to say that I completely agree with Dr Knight's treatment of redundant colon length marks which I consider a waste of time and space in phonemic transcriptions for current mainstream General British etc pronunciation. She apparently needs to make it clear that weak /i/ occurs at the ends of words (and morphemes) as for example in 'soberly' /`səʊbəli/ which may be compared with, to choose a rather topical word, 'jubilee' /`ʤubəˌli/ (also commonly pronounced /ʤubə`li/) where stress
    marks may be used if it's felt necessary to clarify the choice of strong versus weak /i/.

    1. I agree with both R-A Knight and JWL on not marking length of vowels iff [sic] we restrict this practice to NSs of English who know if a vowel is long or short. But what about NNSs (and, maybe, even the odd NS as well) who are faced with transcriptions like /kraɪsiz/, /beɪsiz/ or /sItiz/. I think quite a few will be rather at a loss.

  2. `kraɪˌsiz `beɪˌsiz `sɪtiz

    1. laɪmi: Not YOU! You will most definitely not be at a loss!

  3. Some care about colon length, others don't.