Adapting and Writing Language Lessons/Chapter 6

by Earl W. Stevick

CHAPTER 6

CUMMINGS DEVICES

HISTORY

In the year of our Lord 1219, and the thirteenth year of his conversion, Brother Francis held a general chapter at Santa Maria della Porziuncola, and sent brethren to France, Germany, Hungary, Spain, and those provinces of Italy which the brethern had not yet reached.... The German mission was led by Brother John of Parma with some sixty or more brethren. When they were come into Germany, not knowing the language, and when men asked whether they desired lodging or meat or any such thing, they answered Ja, and thus received kindly welcome from some folk. Seeing therefore that this word procured them humane treatment, they resolved to answer Ja to all questions whatsoever. Wherefore, being once asked whether they were heretics, come now to infect Germany after the same fashion wherewith they had already perverted Lombardy, they answered Ja so that some were cast into prison, and others were stripped of their raiment and led to the common dancing-place where they were held up for a laughing-stock to the inhabitants. The brethren therefore, seeing that they could make no fruit in Germany, came home again; and this deed gave the brethren so cruel a report of Germany, that none dared return thither but such as aspired to martydom....

Jordan of Giano (in Ross, 1949)

Thomas Cummings taught languages of India to missionaries in the early part of this century, some seven hundred years after the Franciscans' disagreeable experiences with German. He published some of the fruits of his work in 1916, in a little book titled How to Learn a Foreign Language. one of his most characteristic emphases was that his students should know not only some answers, but also some questions to which each answer is appropriate, and not only some questions, but also a number of useful answers for each question. Cummings saw that each language has only a small number of question-words, and realized what power those few questions give to a student who wants to elicit new vocabulary. Furthermore, the same questions can be applied to one center of interest after another, in accordance with the student's changing needs. The answers to such a set of questions can readily be combined into meaningful and interesting texts, whether those texts be written exposition or genuine unrehearsed conversation. Having more than one answer to each question insures that the student does not merely memorize a fixed sequence, but that he is always aware of the choices without which discourse cannot qualify as communication.

Half a century after Cummings, this writer was attempting to solve problems of materials development for Swahili and Hausa at the Foreign Service Institute, and for Chinyanja (now called Chichewa) in the Peace Corps. This work led to three observations:

1. The shorter a dialog, the less unexplained, confusing clutter it contains. The shortest possible dialog consists of two lines.

2. Differences in progress were less between trainees of low aptitude and trainees of high aptitude when material was true, important and, if possible, autobiographical, and greater when material was general, fictitious and of no immediate use.

3. Students seemed to retain material better when they have used it for communication of some kind.

Cummings' use of questions and answers made sense in all these respects. It thus became the historical source of what was called the 'microwave' format for writing language lessons. This unfortunate label was selected as part of an elaborate electronic metaphor; its meaning was that the length of one ‘cycle’ (defined as the length of time from first introduction of a new item until its use in communication) was extremely short.

The microwave format itself, in what we may a little wryly call its 'classical form,' contained a basic utterance (usually but not always a question) and from four to eight potential answers or other appropriate rejoinders. If the basic utterance and the rejoinders are well chosen, they can lead to almost immediate real or realistic (Chapter 2, p. 28-29) conversation in class, and are also likely to find use in real life outside of class. At the same time, new structures and new vocabulary can be kept to a minimum.

A microwave 'cycle' was divided into an M-phase and a C-phase. M stood for mimicry, manipulation, mechanics and memorization, and C for communication, conversation, and continuity. Within the M-phase, the first section usually introduced the answers or rejoinders, often in the form of a substitution drill with a separate column for cue words. The second section contained the question(s) or other basic utterance(s). The C-phase combined the elements of the M-phase with each other and, ideally, with material from earlier lessons, to form a short sample conversation.^[1]

At least in the early stages, all sentences were kept fairly short (very few with as many as 12 syllables). They also were simple in their structure, and in most cycles all of the rejoinders exemplified a single surface structure. Cycles were therefore relatively light and transparent, in the sense of Chapter 3 (p. 47f). In the C-phase, parentheses ( ) were placed around those nouns, verbs, or adjectives that were subject to replacement, and users were urged to 'relexicalize' the cycle by adding their own vocabulary at those points.

From this brief description, it should be obvious that microwave cycles have potentially high ratings for usability (Chapter 2, Assumption I), responsiveness (Assumption III), and responsibility (Assumption IV). Just how much of this potential is realized for any one textbook or any one program depends on three factors:

1. The internal structure of individual cycles.
2. The relationship of the cycles to one another.
3. The degree to which the content is pertinent to the needs of the students—‘strength’ in the sense of Chapter 3.

The same three factors of course affect the success of non-microwave lessons. But while an inappropriate microwave lesson is no more unmotivating than an inappropriate course of any other kind, it was felt that an appropriate microwave lesson could go far beyond most other formats, at least for young American adults who were about to go abroad.

These ideas took shape in 1964, and were first discussed publicly at a conference in Bloomington, Indiana, in the spring of the following year (Stevick, 1956). This was a period in which the Peace Corps need for new materials in new languages was at its peak, and so it happened that the micro-wave format was adopted for use in dozens of courses, written under extreme pressure for time, by materials developers with highly miscellaneous backgroundsfor the job. Results were sometimes surprisingly good, and in many cases were probably better than what the same writers would have produced in other formats, but the experience of the next five years also proved instructive in some negative ways:

1. Microwave is not a theory, nor a method, but only a format.

2. There are certain pitfalls in writing individual cycles.

3. A course that consists of nothing but cycles violates Assumption V ('Plura1ism,' p. 36), and is also unsatisfactory in other ways.

The remainder of this chapter will deal with the implications of these three statements.

MICROWAVES AND CUMMINGS DEVICES

First, on microwave as a format. A distressingly large number of people have talked and even occasionally written about microwave as a theory or as a method. This may be due to the ease with which the term enters as the first member into noun—noun constructions: 'the microwave method,' and so forth. The only such construct that is justifiable is 'microwave cycle.' If that phrase is replaced by 'Cummings device,' then the originator will receive credit for his idea, and at the same time unwarranted collocations will be blocked by the very mechanics of the English language: people will not easily slip into talking about 'Cummings device theory,' or 'the Cummings device method.' On the contrary, the term 'device' is intended to suggest a small part of a total method, which may or may not be consonant with one or another theory.

WRITING CUMMINGS DEVICES

Problems in writing Cummings devices arise in connection with both the manipulative and the communicative phases.

Example 1 (French)

Bonjour Monsieur.	Good morning. (said to a man)
Bonjour Mademoiselle. Good morning.	(said to an unmarried woman)
Bonjour Madame. Good morning.	(said to a married woman)
Comment allez—vous?	How are you?
Bien, merci. Et vous?	Well, thanks. And you?
Tres bien, merci. Et vous?	Very well, thanks. And you?
Pas mal, merci. Et vous?	Not bad, thanks. And you?
Ca va bien, merci. Et	Fine, thanks. And you? vous?
Je vais bien, merci. Et vous?	I'm fine, thanks. And you?
Au revoir.	Good bye.
Au revoir, à bientôt. Good bye, until later.

Summary:	A.	Bonjour (Monsieur).
	B.	Bonjour (Mademoiselle). Comment allez-vous?
	A.	(Très bien), merci. Et vous?
	B.	(Ça va bien), merci.
	A.	Au revoir.
	B.	Au revoir, à bientôt.

Comments on Example 1

The sentences are of suitable length. The subject matter is appropriate for most groups, and the sentences are all idiomatic. As shown in the summary, ( ) can be filled in various ways so as to make several different conversations.

Example 1 however departs from the format of a classical Cummings device in three ways: (1) It is actually a composite of three such devices, whose basic utterances are respectively (a) Bonjour, (monsieur) (b) Comment allez-vous?, and (c) Au revoir. (2) All of the rejoinders to Comment allez-vous? are practically synonymous with one another, (3) There is only one rejoinder to the last basic utterance. Meaningful choice, and hence communication, are thus impossible in two-thirds of this particular device.

Example 2 (English)

Basic utterance:	What is your name?
Potential rejoinders:	My name is Bill Williams.
	My name is Clyde Bonney.
	My name is Ethel Redd.
	My name is Carol Singer.
	My name is Fletcher Arrowsmith.
Summary:	what is your name?		My name is (Bill Williams).

Comments on Example 2

The subject matter--getting people's names-—is well-chosen if the cycle is used by trainees who are still getting acquainted with each other. It is also appropriate, but less so, for groups that have passed that stage, since any trainee can look forward to having to get people's names at some time in the future.

The length of the sentences (4-7 syllables) is ideal. The names that are selected for use in presenting the lesson may be chosen either for their phonetic problems or for their lack thereof. After initial presentation of the Cummings device, names of real people should be used at the point indicated by ( ).

This Cummings device has a serious flaw, and it is the kind of flaw that writers of language lessons most easily overlook. The short, uncomplicated sentences and the usefulness of the subject matter should not blind us to the fact that the question simply is not idiomatic. 'What is your name?' is used only to children and to inferiors. If I want to know the name of another adult, I must find it out in some other way. For example, I may volunteer the information that 'My name is _____________ and expect him to reciprocate.

Example 3 (Swahili)

Basic utterance:

Unatoka mji gani?	What city are you from?
Unatoka jimbo gani?	What state are you from?
Unatoka nchi gani?	What country are you from?

Potential rejoinders:

Ninatoka mji wa Topeka.	I'm from (the city of) Topeka.
Ninatoka jimbo la Kansas.	I'm from (the state of) Kansas.
Ninatoka nchi ya Amerika.	I'm from (the country of) America.

Summary:

Unatoka (mji) gani?

Ninatoka (mji) (w)a (Topeka).

Comments on Example 3

Suitability of subject matter is as for Example 2. Length and complexity of sentences are still within the ability of beginners. This Cummings device provides a fairly realistic way of learning to choose among wa,la, ya in agreement with mji, jimbo, nchi.

Example 4 (French)

Basic utterances:

Qu'est—ce que vous faites à 6 heures du matin?	What do you do at 6 a.m.?
Et après, qu'est—ce que vous faites?	And then what do you do?

Potential rejoinders:

Je me réveille.	I wake up.
Après je me léve.	Then I get up.
Après je me lave.	Then I wash.
Après je m'habille.	Then I get dressed.
Après je vais au réfectoire.	Then I go to the dining hall.
Après je prends un casse-croûte.	Then I have a bit to eat.
Après j'etudie le français.	Then I study French.

Comments on Example 4

Experienced language teachers will recognize their ancient and trusty friend the action chain, disguised here as a Cummings device. The subject matter is appropriate for almost any group, although the questions themselves are seldom asked outside of a language classroom. The questions would have been a bit long for absolute beginners, but this was No. 68 in a series of Cummings devices.

Example 5 (Bini)

See pages 320-321.

Comments on Example 5

This is an excellent example of a Cummings device set out in standard microwave format. The list of key words in the left—hand column of M-l makes that section of the lesson into a simple substitution drill. C-1 and C-2 could have been combined, by putting the noun into parentheses. As a good C—phase should, C-3 takes the user beyond mere mechanical combination of what was in the M-phase.

CYCLE 26

M-l

èpɛnì	èpɛnì nàkhin	pen	This is a pen.
èpsnso	èpensò nàkhin	pencil	This is a pencil.
èbe	èbe nàkhin	book	This is a book.
ugbɛ̀khùn	ugbɛ̀kùn nàkhin	belt	This is a belt.
aga	aga nàkhin	chair	This is a chair.
èteburù	èteburù nàkhin	table	This is a table.
ibâtà	ibâtà nàkhin	shoes	These are shoes.
ɛwû	ɛwû	nàkhin	dress/shirt	This is a shirt.
èsìga	èsìga nàkhin	citarette.	This is a cigarette.
M-2
	bh' ònà a-khin?	What is this?

C-l
A:	Bh' ònà a-khin?	A:	What is this?
B:	èsìga nàkhin.	B:	This is a cigarette.
C-2
A:	Bh' ònà a-khin?	A:	What is this?
B:	ɛwù nàkhin.	B:	This is a shirt.

C-3

Continue this cycle using actual objects in the room. Additional vocabulary should be given for objects present for which the Bini equivalent is unknown. Have the students try both asking and responding.

To The Student:

The prefix /a—/ attached to the verb occurs when the question word /bhe/ introduces a question. There are times when /a—/ becomes fused with e preceding vowel, or is elited since retention of the preceding vowel is required.

CYCLE 19

M-l
èmwan	èmwan ɔ ghi dì'à	here	He lives here now.
ghi	ɛ̀bhò nâ ɔ̀ ghi ye	now	He lives in this city now.
ebha	ebha ɔ ghi dì'à	there	He lives there now.
diya	òwa nâ ɔ ghi dì'à	live in	He lives in this house now.
ye	idùmwun nâ ɔ̀ ghi ye	to be	He now lives (is) on this street.
M-2
k'èkè	d'èkè n' èjonì ghi di'ǎ?	What place	Where does John now live?
kere	diɛ̀bhò n' ɔ kerě?	come from	Where does he come from?

C-1
A:	d-ɛ̀b?ò ? ejɔnì kerě?	A:	where is John from?
B:	èshìkagò ɔ kerè.	B:	He is from Chicago.
A:	èbha ɔ ghi dì'à?	A:	Does he live there now?
B:	ɛò, ɛ̀r̄ ebha 3 ghi dì'à	B:	No, he does not live there now
	émwan ɔ̀ ghi ye.		He is here now.

Comments on Example 6

Again, the C-phase is relatively strong. Notice that it brings in yes-no questions, which had been covered in a previous lesson, and also that it recombines elements from the M-phase more boldly than C-1, C-2 of Example 5.

The sentences of the M-phase of Example 6 all exemplify very much the same surface structure, but the cue words in the left-hand column are chosen from three different parts of their sentences: verb, place expression, and time expression. The reason for this is not apparent, since it will probably make M-I more difficult to use as a drill. In any case, the relationship between cue words and expected responses is one of the problems that writers of Cummings devices must recognize and deal with.

Example 7 (Swahili)

See pp. 323-324.

This lesson seemed to have much to recommend it. It was about air transportation, which all of the trainees expected to be using shortly after the end of their language study; it was illustrated with a reproduction of an authentic airline schedule; and it was obvious how the content of this lesson could be replaced by up-to-date information on actual flights that the trainees expected to take. Nevertheless, this lesson was heavily criticized by nearly everyone who tried to use it, and eventually had to be dropped. It may therefore stand as a warning to other developers of language materials. Its chief flaws seem to have been the following:

(1) The individual sentences are rather heavy, in the sense of Chapter 3 (P. 47). The heaviness results not only from their length, but also from the fact that in translating time into Swahili, one must add or subtract six hours: 7 o'clock is literally 'one o'clock' and 1 o'clock is literally 'seven o'clock.'

CYCLE 43

[Refer to the timetable which appears below. ]

M-l

New York	Ndege namba 35 huondoka New York saa 5 na dakika 30 asubuhi.	New York	Flight 35 leaves New York at 11:30 a.m.
Chicago	Ndege namba 35 huondoka Chicago saa 7 na dakika 25 mchana.	Chicago	Flight 35 leaves Chicago at 1:25 p.m.
Kansas City	Ndege namba 35 huondoka Kansas City saa 9 na dakika 20 mchana.	Kansas City	Flight 35 leaves Kansas City at 3:20 p.m.

M-2

Chicago	Ndege namba 35 hufika Chicago saa 6 na dakika 48 mchana.	Chicago	Flight 35 arrives in Chicago at 12:48 p.m.
Kansas City	Ndege namba 35 hufika Kansas City saa 8 na dakika 42 mchana.	Kansas City	Flight 35 arrives in Kansas City at 2:42 p.m.
Albuquerque	Ndege namba 35 hufika Albuquerque saa 10 na dakika 10 mchana.	Albuquerque	Flight 35 arrives in Albuquerque at 4:10 p.m.

C-1

Ndege namba 23		What time does Flight 23
	(fika/ondoka)			(arrive/leave)
	saa ngapi.		?
(jina la mji)		(name of city)
Hu		saa		.	It		at		.
	(fika/ondoka)		(wakati)			(arrives/leaves)		(time)

C-2

[Ask and answer the same questions about Flights 27, 107, 137, etc. The students should of course have the timetable before them for this cycle.]

(2) The upcoming air trip to East Africa, though a dramatic event, was one that would not be part of daily life.

(3) Arrangements for any air travel that they might undertake within East Africa could best be made in English anyway.

Example 8 (Lao) See pp. 326-327.

This example consists of the M-l segments of Cycles 38 and 73 in Lao Basic Course. In their format, they are identical. The problem of interest here is again choice of content. Each contains a miscellaneous list of predicate expressions. In the C-phase of Cycle 38, it is easy to see how the content of the lesson could be brought to bear on one member of the class after another, so that they would get better acquainted with one another at the same time that they were practicing their Lao. No comparable focus is obvious for the material of Cycle 73. Cycle 38 talks about what one individual or another can do, while Cycle 73 is a list of unconnected bits of information illustrating a grammatical pattern.

One more question that arises in the writing of Cummings devices is the extent to which they should depend on the use of translation. The same problem of course comes up in connection with lesson material of other kinds, such as dialogs and drills. Obviously, if the intent is to do with little or no translation, then there is a correspondingly greater premium on the quality of transparency (Chapter 3, p. 48).

CHAPTER 6

CYCLE 38

M-1

tiicák	to type
láaw tiicák dàj bɔɔ?	Can she type?
aāk hùup	to take pictures
láaw sāk hùup dàj bɔɔ?	Can she take pictures?
sỳ khɔ̌ɔŋ	to shop
láaw sỳ khɔ̌ɔŋ dàj bɔɔ?	Can he shop?
tát phǒm	iáaw tát phǒm dàj bɔɔ?	Can she cut hair?
púk hýan	to build a house
léaw púk hýan dàj bɔɔ?	Can she build a house?
khúakin	láaw kháakin dàj bɔɔ?

M-2

khúakin
vɔ̄ɔ̄ dàj,phɔ̄wāā láaw khúakin bɔ̄ɔ̄ pen	No, because she doesn't know how to cook.
púk hýan	build a house
bɔ̄ɔ̄ dàj,phɔ̄wāa láaw púk hýɔ̄ɔ̄ pen	No, because he doesn't know how to build a house.

CYCLE 73

M-l

dan-fáj	to make a fire
càw si sàj njǎŋ daŋ-fáj?	What will you use for making a fire?
hɔ́ɔŋ, nāŋ	to underlay, place beneath; sit
càw si sàj njǎŋ hɔ́ɔŋ naŋ?
lābaj, nàm	to control the flow, water
càw si sàj njǎŋ lābaaj nàm?	What will you use to control the water flow?
sǎmûlat, bə̌eŋ	survey, inspect, look at, see
càw si sàj njǎn sǎmlûat bə̌eŋ?	What will you be inspecting?
lỳak,mâak màj	select, choose; fruit
càw si sàj phǎj lỳak mâak maj?	Who will you use to select the fruit?
kinkhâw	to have ones meal
càw si sàj njǎnŋ kinkhâw?	What will use to eat with?
phán,mýy	to wrap around,hand
càw si sàj njǎn phǎn mýy?	What will you use to wrap around your hand?

RELATIONSHIPS BETWEEN CUMMINGS DEVICES AND OTHER COMPONENTS

We have already said (p. 314) that a number of sets of materials have appeared which consisted but nothing but series of Cummings devices. The advantages in writing text-books in this way are that each individual device provides a sharp focus for one class hour, and that the absence of any further printed helps leaves no doubt in the teacher's mind that he must be creative, but in general this kind of textbook is inadvisable. Cummings devices have, however, entered into a number of other courses either as the central elements of the lessons, or in peripheral and supporting roles some of the possibilities that can be documented are the following:

1. Cummings device central in each lesson, supported by usual types of drill, glossary, grammar notes, and a few suggestions for use. (Stevick, Indakwa, et al., An Active Introduction to Swahili)

2. Cummings device followed by suggestions for a related 'cross—cultural experience' and 'routine language experience,' and information of interest to learners who also expect to teach English in their host country. (Kim, Lee, Crowley, Lessons in the Korean Language and Culture)

3. Cummings device paired with experience in singing folk songs and enjoying poetry. Often only a distant connection between the Cummings device and the poem. (Bailey, Jamaican Creole Lanquage Course)

4. Cummings device followed by explanatory notes, followed by a series of self-testing frames related to something in the first two parts of the same lesson (Yates and Sayasithsena, Lao Basic Course).

5. Heavily modified Cummings device format as the 'principal stage,' preceded by self-instructional frames which present other matters, and followed by detailed suggestions for other activities both in and outside of class. (Blair etal., Cakichiguel Basic Course)

6. Lessons that are built around Cummings devices are interspersed among other types of lesson. (Kamoga and Stevick, Luqanda Basic Course)

7. Cummings devices as one of several components which lead to performance of well—defined objectives for use in and outside of class (Rehg and Sohl, Kitail Lokaiahn Pohnpei see Appendix R, p. 346. See also Appendices E (Mauritian Creole) and G (Thai), which show the use of Cummings devices as subordinate parts of new materials, and all the appendices to Chapter 3, in which Cummings devices appear as one means for adapting existing lessons.)

SUMMARY

This chapter has described an ancient device, and to some extent has chronicled a recent flurry of attention to it. But to look only at the format itself would be a mistake. Mueller (1968) has said that 'as soon as [a] pattern has been mastered, the student must progress further to the creative stage, where he learns to use what he has mastered.' [emphasis added] Even though it may be the case that in this book we mean more by ‘use’ than Mueller did, still his words may serve as a statement of what we can call 'the Cummings principle.' Experiences of the past six or seven years suggest that the Cummings principle deserves general recommendation and wider application. The Cummings device, with its shortness, clearly defined scope, and goal of immediate use, is certainly an excellent way of realizing the principle. Nevertheless, the device itself should play a supporting, or at most a co-starring role in the design of published lessons.

1. The terms 'M-phrase' and 'C-phrase' were applied by Garner and Schutz (1969) in much the same sense but on a quite different scale.