At the end of March, I met with other City Lit language tutors to discuss how we teach pronunciation to our learners. Pronunciation was described as the elephant in the language classroom. It’s a problem from the outset for every single learner and yet it’s difficult to sort it out. Worst of all is trying to sort it out with advanced learners who might speak with pristine grammar, but with pronunciation that is at times hard to understand. Pronunciation is difficult for a number of reasons. First of all, it’s comprised of many different elements: the way you say individual sounds, intonation (variations in the pitch of voice), rhythm (the variation in stress over syllables and words) to name just a few. Whilst there’s the issue of what element of pronunciation we ought to focus on in class, there’s also the added problem of learners having to produce sounds they don’t have in their own language(s).
Drilling is the most common phenomenon. The teacher stands at the front and blasts you with a load of sounds which are either completely new or if they aren’t new then they’re in combinations that your own language doesn’t allow (e.g. <gwl> in Welsh or strč prst skrz krk in Czech/Slovak). There’s also the additional problem of trying to hear the difference between sounds which your own language doesn’t count as different. If you’ve ever tried to learn Mandarin you’ll immediately think of those seemingly impossible sounds that are represented in Pinyin as: z, zh, j, c, ch and q. They’re all affricates, i.e. scratchy sounds produced with turbulence in the mouth. But for an English speaker, the types of phonetic differences that are so obvious to a Mandarin speaker just can’t be heard. Drilling isn’t going to help you understand the differences between these two sounds. You need some metalinguistic knowledge- you need to be introduced to the phonetic underpinnings of sounds in order to understand the differences. Doing so will enable you to hear the differences and then to start producing the sounds yourself.
Phonetics is the scientific study of the sounds of human language. Trained phoneticians can produce any sound that occurs in any language by consulting their nifty chart. So how can we use phonetics to give our language learners insight about the sounds they are trying to produce. I’m not talking about blinding them with phonetic labels and terms for the sake of it, but using phonetics in such a way that it has a practical application in a language learning environment. With my beginner learners, I’m currently trying vowel quadrilaterals as a way of improving the pronunciation of vowels.
Vowels are incredibly slippery things and always so difficult for learners to get their heads around. The problem is worse for speakers of lots of British varieties of English because over 70% of vowel sounds in English end up being pronounced as the vowel schwa /ə/. This is the vowel at the start of the word ago or at the end of the name Adam. Vowels in unstressed syllables in English become this vowel. This is “good English” and students on TEFL and English as second languages courses are introduced to this rule as a way of improving their pronunciation.
ɑ ɵ ɛ ɪ œ ɶ ø
Daniel Jones was a 20th-century phonetician who came up with the idea of “cardinal vowels”. These are a bit like reference vowels with which we can compare and contrast vowels of the languages of the world. The cardinal vowels are arranged on a grid formation shown below. Vowels on the right of a pair are rounded (i.e. lips pursed into a circle- think French and Swedish vowels or how someone from Cardiff might say “ear”), whilst vowels on the left of the pairs are unrounded (i.e. pronounced with spread lips). There are two axes: height (running vertically) and frontness/backness (running laterally).
Height is referring to jaw height, i.e. how open the mouth is when the vowel is pronounced. So /ɑ/ is said with a gaping at-the-dentist style mouth, whereas /u/ is made with a very narrow mouth opening. Fronted means that the tongue is further forward (or at least that’s one definition- really there’s also just something about /i, e/ and the rest of the front ones that just sounds more “front” than the ones at the back. Whilst the quality of the vowel is determined by height and frontness, there’s also the dimension of roundness. This means we can have two vowels with exactly the same height and frontness, but which differ as one is rounded and one is unrounded. For example, if you were struggling to say the Swedish vowel written as ö in the word nöt (nut), you could aim for the unrounded cardinal vowel /e/ and then round your lips to make the rounded /ø/ sound.
Welsh Vowel Quad
A work in progress! Southern Welsh vowels (N.b. these aren’t IPA symbols)
This quadrilateral uses the actual letters of Welsh instead of phonetic symbols, but this doesn’t matter so much for teaching a phonemic (“phonetic”) language like Welsh where there is a(n almost) one-to-one correspondence between a letter or letter chain (called a grapheme) and a sound (called a phoneme).
A graph like this can be useful to help learners visualize the differences between the sounds they are trying to produce. It also goes beyond simple pronunciation instructions such as “if you see a circumflex over a vowel then it’s just a long version of that vowel”. This rule doesn’t quite cut it. w and ŵ aren’t really just long and short versions of each other. w, e.g. in the word cwm (valley), sounds like the vowel in the word cup when said by a northern English speaker so that it rhymes with put, wheres the ŵ has a different quality. ŵ is produced slightly higher, with the mouth more closed and the tongue slightly further back.
One colleague described how the problem of pronunciation is amplified when teaching British learners as opposed to learners from other countries. My colleague pointed out the eager readiness of French people to correct learners’ pronunciation, however, she felt that Brits never deign to correct a non-native speaker’s pronunciation. The situation is perhaps amplified in London where every interaction with a person is in a different accent. Perhaps we also just need to be more direct with our learners who are often extremely anxious about improving their pronunciation and sounding comfortably intelligible. This is a problem learners want us to help them solve. I think by employing more phonetic methods we can expunge the embarrassment element from the process by equipping learners with the words they need to describe and analyse the sounds they are themselves aiming to produce. There’s a different between “I just can’t make that sound- I never get it right!” and a learner instead thinking “I need to say that vowel as rounded and not unrounded” or “I need to increase the height of my i sound”.