10 January 2011

Teaching English /r/ and /l/ to Asian EFL learners: a lexical approach (Part II)

This is the second part of a series on teaching English /l/ and /r/. The piece published below incorporates the first installment and then discusses possible sequences for teaching English /l/ and /r/.

Teaching English /r/ and /l/ to Asian EFL learners: a lexical approach (Part I-II) 
Charles Jannuzi
University of Fukui, Japan


English /r/, /l/ and contrasts between these two categories of sounds are often cited as pronunciation and listening perception problems for a variety of EFL learners, most from E. Asia. The language backgrounds most often associated with these problems are Japanese, Korean, Chinese and some languages of SE Asia (e.g. Thai but also Cantonese Chinese). Other language speakers have also expressed an interest in improving their pronunciation of English /r/ and /l/, including Russian and German EFL learners.

Perhaps the most well-known group to have a problem with the two categories of sounds is Japanese EFL learners. This could be because their native language background creates the most difficult problems to overcome. It could also be because Japan attained affluence before most of the rest of Asia and hired native speakers of English to help teach and model the language. So a lot of information based on knowledge and experience of Japanese and Japanese learners of EFL has been exchanged and discussed in 'global ELT'.

What is the issue for Japanese learners of English?

In the case of Japanese learners of English, just what is the issue? The most common account is based on a simple 'contrastive analysis'. Japanese is said to have one categorical sound (or phoneme) whereas English has two. The Japanese sound is often referred to as a type of [r] that is tapped, flapped or trilled.  The Japanese sound never closes a syllable and has a very limited distribution in Japanese. One form of the Japanese /r/ helps to form the syllables used in grammatical inflections (such as verb forms which are suffixes). Also, word-initial Japanese /r/ is limited to words of foreign origin (e.g., ramen, the type of Chinese noodles now a national dish in Japan, or ramune, a soda the name for which seems to be derived from the English word lemonade).

English-speaker descriptions of the Japanese sound or of the Japanese learner of English's sound represent the Japanese sound as variably resembling English /l/, /r/, or /d/ (especially [d] in the middle of a word, like in the word 'middle'). Phonetic descriptions have also said that the American medial voiced [t] of words such as little  are quite like the Japanese /r/.

However, it is not really clear how useful a cross-linguistic, contrastive analysis of phoneme inventories is in diagnosing the problems or in helping Japanese learners of English to overcome them. For one thing, the often-read argument that Japanese has only ONE phoneme, Japanese /r/, is arguably wrong. That is because, using structuralist criteria for determining what is and what is not a phoneme, we can isolate at least two Japanese [r] sounds that are distinct: initial [r] in the word rou  (candle) from palatalized intial [r] in ryou (dormitory).

It is also misleading to teach EFL learners that there is one English /r/ and one English /l/. That is because they will hear native and fluent speakers of English make a wide array of both sounds in actual speech. In terms of articulation, there is a wide variety within both categories of sounds. Interestingly, the distribution in the lexicon of English [r] sounds strongly parallels English [l] sounds: word-initial (right vs. light), word-initial cluster unvoiced (crime vs. climb), word-initial cluster voiced (grow vs. glow), post-vocalic (fear vs. feel), medial (correct vs. collect), and unstressed syllabic (batter vs. battle).

There is some complementary distribution if we consider clusters: [tr-] as in true but no [tl-], [sl] as in slide but no [sr-], [shr-] as in shred but no [shl] (except some loan words), and [l] can cluster with [r] post-vocalically, as in girl or world, but not vice versa. Moreover, since both of these sound categories tend toward 'vowel-like', it is not surprising that in some cases they might reduce to a vowel or vowel lengthening in some accents, dialects and word contexts (such as post-vocalic [r] in London, Boston and NY Englishes, or the lost [l] of the word chalk).

Given the variety of English /r/ and /l/ sounds and how they parallel each other in the lexicon of English, it is little wonder that EFL learners, even after they have practiced making an English /r/ vs. /l/ distinction, lose the ability when actually communicating orally. Therefore, it is best to teach--over a period of time and through a variety of activities--the full parallel variety of English /r/ and /l/ sounds as found in the most frequent words of the lexicon. A proposed sequence is this: first the variety of English /l/s, then the variety of English /r/s, then /l/ vs. /r/ contrasts in common words, then a follow up on the variety of post-vocalic [r]s in rhotic accents, such as US and Canadian Englishes.

What sequence should be used to teach English /l/ and /r/?

One line of reasoning might say that, since English /r/ is typically the last consonant acquired by native speakers, it might be best to teach English /l/ first to EFL learners because of the inherent difficulties with the English /r/ group of related sounds. A different line of reasoning that might support this first approach might be based on interlanguage analysis. If Japanese has its own /r/ sound, the argument goes, then Japanese learners of EFL would find it easier to differentiate and master English /l/ first. However, as noted above, Japanese /r/ and Japanese learners' interlanguage /r/ and /l/ sounds are often described as sounding more like English /l/ (or /d/ or medial voiced tapped [t]) than English /r/. If all these arguments are considered together, the issue of which sequence to follow starts to look rather difficult to decide.

One compromise might be to conclude that the Japanese /r/ and palatal /r/ are not acceptable substitutes for either English /l/ or /r/ and to start an instructional sequence with English /l/ first. The strongest justification for English /l/ coming first is that even English native speakers acquire /r/ rather late in their language development and this complex set of sounds actually often requires remedial practice.

So one possible proposed sequence of instruction (as a pronunciation segment of a longer class, such as 20 minutes out of a 90 minute class period) is as follows:

First class: English /l/
Second class: Review of English /l/, English /r/
Third class: Review of English /l/, /r/, contrast of English /l/ vs. /r/
Fourth class: Review and revise as necessary
Subsequent classes: Work on problem contrasts within and across the categories (e.g., 'farm' vs. 'firm', 'walk' vs. 'work', etc.). 

English /l/ should be taught using the most frequent and useful words of English as possible (and teachers should be prepared to provide an L1 translation in the case of less common words) and should cover the following types of /l/:

Types of English /l/

Word-initial [l]:  like, lake, let, lot, low, lamp, leap, last, etc. 
Unvoiced cluster [l]:  clean, close, clock, place, play, please, slide, slow, slip, etc.     Voiced cluster [l]:  blue, blow, blood, blame, glad, glue, glow, glass, etc.  
Post-vocalic [l] (dark l):  feel, fall, fail, call, sale, all, deal, tall, etc.
Medial [l]:  follow, hollow, yellow, jello, hello, pillow, filling, collect, etc. 
Unstressed syllabic [l]:  settle, battle, riddle, middle, puddle, little, tunnel, etc. 
[-rl] cluster:  girl, world, whirl, hurl, curl, twirl, swirl, unfurl, etc.  

Then, English /r/ should be taught using the most frequent and useful words of English as possible and should cover the following types of /r/:

 Types of English /r/

Word-initial [r]:  right, raise, rise, risk, rose, run, red, road, etc. 
Unvoiced cluster [r]:  cry, cream, crazy, tree, true, try, pray, praise, prize, three, throw, etc.  
Voiced cluster [r]:  broom, bring, British, grow, grass, great, dry, draw, dream, etc. 
Post-vocalic [r] (reduces to or alternates with 'schwa' in many dialects and accents):  car, fear, far, tear, fair, form, farm, dear, hear, more, war, etc.  
Medial [r]:  correct, Korea, porous, preferring, occurring, recurring, referring, transferring, etc. 
Unstressed syllabic [r]:  ladder, litter, batter, motor, runner, sadder, madder, heater, etc.  
[-rl] cluster:  girl, world, whirl, hurl, curl, twirl, swirl, unfurl, etc.  

In the next part of this series, we will look at actual classroom activities that might be used to teach and practice English /l/ and /r/, both in terms of listening perception and language production.

No comments:

Post a Comment