all about phonology

Testing Pronunciation (by Augustin Simo)
Home
Testing Listening (by Kitao)
Testing Pronunciation
Testing Listening
Testing Pronunciation (by Augustin Simo)

Testing Pronunciation

 

by Augustin Simo Bobda


This article is motivated by the author’s observation that the evaluation of students’ pronunciation is not given the place it deserves in many EFL and ESL classes in Cameroon and probably elsewhere. It is supported by the fact that the purpose of testing pronunciation is not only to evaluate knowledge and award grades, but also, and probably more importantly, to motivate students to be sensitive to this aspect of English. Given that the motivation of many students for learning English is instrumental rather than integrative, pronunciation tends to be neglected by many learners as long as they know they will not be tested on it.

Obviously, pronunciation is tested globally in different types of conversational exchange, interview, reading aloud, etc., that go on in the classroom. What seems to be insufficient is the testing of accuracy-that is, testing to assess the learner’s management of specific features, segmental or suprasegmental. This insufficiency is due to two main causes. First, many teachers do not consider it useful to test specific features. This attitude is based on the belief that the mastery of specific features, taken individually, does not matter much in real-life situations where the context always (?) provides the cue for the learner to interpret what he hears (Heaton 1988:64) or to make himself understood even if the ideal quality of phonemes is not reached.

It is possible for people to produce practically all the correct sounds but still be unable to communicate their ideas appropriately and effectively. On the other hand, people can make numerous errors in both phonology and syntax and yet succeed in expressing themselves fairly clearly. (Heaton 1988:88)

The second, and surely more important, cause is the particular difficulties involved in testing oral skills. One of the greatest problems in oral testing is administration. It is often impossible to manage the large number of students to be tested. Testing equipment, like laboratories or tape recorders, is scarce in many Third World countries where English is taught. Even when such material is available, testing may be rendered impossible by the lack of even more basic facilities like electricity.

A further difficulty in oral testing arises when English is part of a school-leaving or promotion examination for an entire country. In most countries offering such examinations (e.g., Cameroon), candidates over a large area have to respond, often in writing, to the same paper. This exacerbates the problem of logistics.

Taking segmental phonemes and word stress as illustrations, this article explores some ways of testing specific features of English pronunciation, both as a teaching activity and as part of an examination. The ideal way of testing pronunciation is to actually listen to the learner. But since this is not always possible or suitable, the alternatives discussed below can be used for testing segments and word stress. Throughout the discussion, the illustrations are based on pronunciation problems of Cameroonians.

Dictation

Given that speaking and listening skills are interrelated, dictation, an old exercise, remains one of the ways of testing the learner’s pronunciation. This testing method is based on the assumption that, most often, if the learner has a deviant pronunciation of a word, he will not understand it when it is read with a different pronunciation. For example, if a student’s pronunciation of sword is [swOd], he/she will not understand and therefore not spell it correctly if it is read (RP) [sOd].

A dictation exercise may appear in different forms. First, it may consist of a whole passage incorporating target words to be tested. It may also consist of a set of individual words incorporating the segmental or stress features being tested. A third interesting type of dictation consists in a cloze test: the testee is given a text from which target words have been removed and replaced by blanks; the examiner reads the full passage and the testee fills in the blanks with the words he has heard. One precaution to take here is that the context should be as neutral as possible; a context that is too supportive will elicit the correct word even if the student’s pronunciation of it is faulty.

This type of cloze test has recently been used by Talom (1990) with Upper Six Anglophone students of the Government Bilingual High School, Yaounde. Using a British reader and a Cameroonian, for comparison, the author obtained fascinating results. For example, the British reader’s pronunciation of climbing as [klaImI˜] was often understood and spelt as *climate, [fjU@l] (fuel) was understood as *few, [|pEz@nt] (peasant) as *persons, *patients, *prisons, [|lEp@ds] (leopards) as *lepers, *labourers, *letters, *left but, *locust, [mE@] (mayor) as *man, [pEsl] (pestle) as *pencil, *parcel, *person, [|b{rI@] (barrier) as *barrack, *garage, [|kvItId] (coveted) as *cavity, *carvity, *quality, and so on. The fact that students’ spelling problems were caused by their deviant pronunciation was confirmed by the fact that parallel groups of students submitted to a Cameroonian reader encountered virtually no problems.

Test Segments

In addition to the various forms of dictation analysed above, there are many listening activities (and others that could involve the interpretation of gestures and pictures) designed to test the learner’s ability to discriminate phonemes or groups of phonemes. Following are a few examples:

1. Same or Different? The testees listen to a pair of words or pairs of sentences and indicate whether they are the same or different; e.g.,

a. suck - sock

b. but - bought

c. seat - seat

d. hut - hurt

e. Is that my pen? Is that my pan?

f. He was severely beaten by his wife. He was severely bitten by his wife.

The exercise can also be done by showing the testees a set of pictures corresponding to words that elicit contrasting sounds; one of the words is spoken by the examiner or played on tape.

2. a or b (or c)? A multitude of sound-discrimination tests can be grouped under what can be broadly termed an a or b (or c) test. For example, the testees are shown pictures eliciting the following words:

1. a. sock     b.     sack     c. suck

2. a. cat     b. cut     c. cart

3. a. court     b. caught     c. cart

The examiner says, for example:

1. sack

2. cat

3. court

The testee writes the letter corresponding to the most appropriate word; i.e., 1. b; 2. a; 3. a.

The exercise, in which the list can be reduced to minimal pairs, can be done without pictures. But pictures are useful because they make the class more lively. This type of exercise is probably one of the simplest sound-discrimination tests.

3. Which Definition? A word is read twice, and several different definitions, including one that is correct for the word, are given; the testees are asked to select the correct definition for the word heard; e.g.,

1. bought - bought

a. a vehicle that moves in the sea
b. past participle of buy
c. coordinating conjunction

2. hid - hid
a. not to like [hate?]
b. placed where it cannot be seen 
c. knock

This type of exercise has the extra advantage that it tests vocabulary at the same time.

4. Which Ones Are the Same? The testees listen to a list of words and mark the ones that are the same.

1. a. pot     b. pot     c. port

2. a. bid     b. bit     c. bid

5. Fill the Gap. The testees listen to a sentence and select from a set of words the one they hear; e.g.,

1. Did you see the --- you were looking for?

a. people     b. pupi1     c. purple

2. He died at the age of --- .

a. forty     b. fourteen     c. thirty

Using colour cards

Colour cards are particularly useful in testing the many phonological alternations that exist in English; e.g., [s, z, Iz]; [t, d, Id]; [˜g, ˜], [aIn, aIt; In, It], [S, Z], [ks, gz], etc. After giving each testee a set of cards of different colours corresponding to the various alternates, the tester pronounces or writes forms and asks the students to show the corresponding card. He may start by pronouncing some forms, a fairly simple exercise if the lesson has been properly taught; e.g.,

[s], [z] or [Iz]?
books, schools, cats, churches, students’, plays, James’s

[t], [d] or [Id]?
wanted, added, jumped, robbed, increased, showed, carved

[ng] or [n]?
finger, singer, hanging, prolongation, prolonging

[aIn, aIt], or [In, It]?
Catherine, Muscovite, acolyte, finite,
masculine, infinite, valentine

[S] or [Z]?
version, invasion, conclusion, Persian, division, coercion, tension, casual, measure, mansion

[ks] or [gz]?
maximum, taxi, exist, Texas, exhaust, taxonomy


Tests other than listening comprehension

Turning now from listening exercises to those specifically testing the learner’s ability to perceive and identify segments, several writing tests not combining listening are available. They include the controversial phonetic transcription exercises, finding odd members out of a set, regrouping, matching, and many other miscellaneous types.

1. Phonetic Transcription. It may be useful to observe that many EFL and ESL textbooks on the syllabus in Cameroon use phonetic transcriptions; e.g., Grant et al.’s (1977) Secondary English Project, Atanga et al.’s (1987) Intensive English, and even more systematically, Cripwell and Linsel’s (1990) Go for English.

Opinion is divided among teachers as to the relevance and/or possibility of teaching or testing phonetic transcription in secondary school. The reluctance of many teachers is due to the fact that they themselves cannot cope with the exercise.

I have personally noticed that when secondary-school pupils have a good introduction to phonetic transcriptions, many of them acquire the skill quickly and even enjoy the exercise in the long run. My suggestion is that, for want of a wholesale introduction to phonetic transcription, students at least be taught to attempt partial transcription.

For example, students can be gradually introduced to the sounds of English and eventually be made to transcribe specific sounds in a word. They can, then, attempt such tasks as “Transcribe the sounds represented by the underlined letters”: journey, peasant, favourite, penal, southern.

2. Finding an Odd Member. The testees are given sets of words in which one word has a sound that differs from the others. The question can be put thus: In each of the following sets of words, three words have the same sound and one does not. Write down the number and the letter of the one that does not.

1. a. dull     b. bull     c. wool     d. pull

2. a. warn     b. dawn     c. scorn     d. barn

3. a. pour     b. poor     c. sure     d. tour

3. Regrouping. The testees are given a list of words and asked to regroup the words that have the same sound (it can be specified whether the common sound is a vowel or a consonant):

let, say, gene, quay, meat, rate, maid, says, said

4. Matching. The testees are asked to find words that have the same sound as a given word. This exercise is similar to the one above. But here, the words illustrating the key sounds are suggested and the testee is asked to find from the list words that have the same sound; e.g., Find from the list below words that have the same sound as the following:

cut, pot, push

tin, than, thatch

List of words:

swamp, buffalo, cook, one, swallow, bosom, country, squander, bush

thing, Thames, either, although, three, Mathilda, clothes, Theresa, cloth

Matching can also be done by finding the words that rhyme with a given word; e.g., Pick out from the following list words that rhyme with cat, lone, tore, poor, here, pair, respectively:

Joan, plait, bat, mere, tour, clear, rare, pour, chair, share, roar, known, sure

5. Miscellaneous Ways of Testing Pronunciation. These include asking testees to circle silent letters (silent letters abound in English words, and the importance of such an exercise cannot be overemphasised); e.g., Circle letters that are not pronounced in the following words:

aren’t, weren’t, sword, debt, bombing Greenwich, Parliament, evening

Note that it is better to have students circle rather than underline the letters; this avoids hesitant students cheating by drawing a line under two letters.

It should be noted that the foregoing exercises require a thorough knowledge of English pronunciation, yet they do not necessitate the use of phonetic symbols.

Testing Word Stress

It does not seem relevant here to dwell on listening comprehension as a method of testing word stress; the method can be used in about the same way as for testing segments. There are other methods of evaluation that are particularly useful for testing knowledge of stress.

One possible method derives from the very nature of English stress. In native English speech, stress is so strong that it is generally accompanied by a movement of some part of the body (head, eye, hand, etc.). To exploit this characteristic of English stress, one first teaching exercise may consist in asking the testees to identify the stress of a word by a bodily movement, like tapping, shaking one’s fist, moving one’s head, etc.

|teacher A|merican Senega|lese
|pardon lieu|tenant kanga|roo
|challenge a|rena appoin|tee

In writing, the testees may be asked to use one of the conventional ways of marking stress; e.g., Put the stress symbol before the stressed syllable or on the stressed vowel.

|salad or sálad

suc|cess or succéss

pre|paratory or prepáratory

The testees can also be asked to use less conventional methods, like underlining, making a circle below the stressed syllable (see Hubbard et al. 1983:215), circling the stressed syllable, capitalising, etc.

The disadvantage of capitalisation is, first of all, that it is likely to cause bad spelling habits in the learner. It is also inadequate when the syllable receiving stress consists of a capital vowel, as in |Agatha. The shortcoming of underlining is that, as seen above, in a serious test situation, a learner who does not want to commit himself will put the line between two syllables if he is not quite sure of the answers; e.g., develop, Agatha, etc. To forestall cheating, therefore, circling remains the most appropriate unconventional method of testing word stress.

Conclusion

The above discussion hopefully convinces the EFL/ESL teacher that pronunciation can be tested in various ways. It can even be tested in various end-of-term, promotion, and public examinations organised in Cameroon and elsewhere, like the BEPC (Brevet d’Etudes du Premier Cycle), Baccalauréat, and GCE Ordinary Level, where the required uniformity among all the testees can be reached without special logistic support. I suggest that to the usual sections on Grammar and Vocabulary, Reading Comprehension, Essay, etc., be added a section called Pronunciation, using the hints suggested above. Some of the hints are quite simple; e.g., for segments-matching, finding an odd member, regrouping, circling silent letters, etc.; for stress-circling, underlining the stressed syllable, etc.

REFERENCES

Atanga, A. S., B. O. Oluikpe, T. Y. Obah and M. K. Okole. 1987. Intensive English for secondary schools, Books 1, 2 and 3. Bamenda: Africana.

Cripwell, K. and J. Linsel. 1990. Go for English. London: Macmillan.

Grant, N., D. O. Olagoke and K. R. Southern. 1977. Secondary English project. Essex: Longman.

Heaton, J. B. 1988. Writing English language tests. London: Longman.

Hubbard, P., H. Jones, B. Thornton and R. Wheeler. 1983. A training course for TEFL. Oxford: Oxford University Press.

Talom, P. 1990. The intelligibility of some RP forms in Cameroon. Unpublished post graduate dissertation, Ecole Normale Supérieure, University of Yaounde.

 

ALL ABOUT PHONOLOGY - ALL RIGHTS RESERVED