Dr Klaus Bung
68 Brantfell Road
Blackburn BB1-8DL

Klaus Bung:
Some thoughts on programming
languages for the tape recorder

First publication in print:
Klaus Bung: "Some thoughts on programming languages for the tape recorder", in: International Conference 'Modern Foreign Language Teaching', Berlin, Germany, 1964. Preprints Part I, p 425-439. Berlin, Germany
Published on the Internet on 2012-04-12
© 1964 and 2012 Klaus Bung
The beginnings / endings of each page have been marked in such a way that the printed original version can be quoted from this html version

Introduction (2012)

This ancient paper is re-published here for historical reasons. It contains some of the seeds of the IDYLL® METHOD™and its learning algorithms which have slowly developed on the coal face of learning and teaching over the decades that followed and led to the fairly elaborate system which we have now. The references to tape recorders and language laboratories are now out-of-date. In their place which have computers and mp3-players which combine all the functions which at the time had to be handled by a variety of different presentation devices.
The incipient theory presented in this paper, however, is not out of date and has gradually been further developed in the publications which followed. Some modern readers may find it easier to find their way into the modern system if they start digesting the simple proposals made in this old paper. It shows already the determination to guarantee success (and test and modify procedures until success can be predicted), which is characteristic of Cartesian Language Learning and was typical for programmed instruction. Some of this determination has sadly been abandoned in subsequent decades by the teaching profession as it moved from one fashionable approach to the next (Decoo 2001) and often lost the focus inherent in programmed instruction that learning *** success *** is of paramount importance.
We now have all the presentation devices that were lacking, or were being invented, at the time, but we have lost the focus on *** good *** theory. Some modern course material created for use with computers looks pretty and entertaining, teachers and pupils enthuse about it, but that does not mean that it is effective (as programmed teaching materials are meant to be): it is sub-standard from the theoretical point of view. The rigour of the programmed instruction movement has gone and its insistence on effectiveness has gone. Unlike Descartes, today's educators do not insist that the problems (teaching and learning tasks) be successfully be solved. They accept failure as an option.
The prevailing attitude in schools at present (2012) is that some students succeed and others fail and are not even encouraged, or permitted, to attempt learning a foreign language. Success in language learning is random, failure is considered to be the student's fault, the teacher does his spiel in the classroom (as he has been "trained"), but whether the student performs well in the language or not does not concern the teacher.
Advocates of the programmed instruction movement would consider this attitude scandalous. In the programmed learning movement (which is Cartesian in nature) every teaching and learning step is carefully considered, tested and sequenced. If the student fails to perform as desired, we do not blame the student (e.g. by saying that he has no aptitude or talent or whatever) but we modify the teaching/learning steps which have failed (or their sequence) and continue experimenting with them until they produce the desired results in *** all *** students. If the student fails, it is the programme's (= the teacher's) fault, or in the case of the IDYLL® METHOD™, it may be the fault of the learning algorithm employed by the student. In that case, the learning algorithm is adjusted, which we have been doing over decades until we reached the most recent version presented elsewhere in this book. This is the approach of an engineer who looks at each component (e.g. of a car) and tests and modifies it until it performs as required. There are no untalented cars, but only good engineers and sloppy or romantic engineers.
To make the contrast quite clear: if an application/implementation of programmed instruction (Cartesian Language Learning) fails, we do not change the "approach", as is customary in our fad-ridden education system, but we minutely and systematically change the teaching steps, until they work.
It is the purpose of this book to revive the old rigour and improve teaching and learning procedures based on it and make language learning more successful.
"There is nothing more practical than a good theory." (Ludwig Bolzmann)


Here starts the original paper of 1964
-------------------------------------------- page 425 -------------------------------------------

Single-Skill Programmes

When learning a foreign language, most students have to acquire a large number of basic skills, such as understanding, oral expression and written expression. These basic skills can be broken down into a large variety of more narrowly defined skills. (Note 1)
Many language programmes define their criterion behaviour as a combination of all or some of the above skills. These I will call 'complex-skill programmes'. I have some reservations about such programmes:
1 The optimum teaching (and programming) method and the best medium of presentation can vary widely from skill to skill. For instance, striving for accuracy obstructs fluency. If both are pursued at the same time, neither skill may be taught effectively.
2 The desirable arrangement of linguistic materials to proceed from simple to complex behaviour, and from known to unknown facts, can vary widely from skill to skill. For instance, some words that are easy to articulate are not easy to spell, or may be used as part of a complex structure only, or express a concept difficult for the learner.
3 We do not know much about the precise effect that the possession of one skill has on the acquisition of another.
-------------------------------------------- page 426 -------------------------------------------
4 There is no general agreement about the relative importance of these skills in a non-specialised language course and about the amount of time that should be given to each. The relative importance of the skills changes not only with the teachers' and examiners' whim, but also with the specific purpose for which the language is being studied.
If all the skills are firmly combined in one programme, this program will be the result of a large number of arbitrary decisions. A conscientious programmer will be constantly worried and inhibited by this knowledge, the resulting programme will suit few teachers and may turn others against the ideas of programmed learning altogether, and it cannot suit all purposes of study. (Note 2)
As programming is an expensive and time-consuming job, programmes should be so constructed that they can be used as widely and for as many different purposes as possible. As programmes have, within their own limits, a great degree of finality, they should be so organised that they do not unnecessarily bind the teacher's hands but leave him as much freedom as possible to assert his own personality when working with them.
I therefore suggest the construction of 'single-skill programmes'. The aim of each programme should be as narrowly defined as possible. Each programme starts from scratch and does not presuppose that the student has acquired the skills taught in any other single-skill programme (or, in some cases, relies as little as possible on any other programme). All programmes together can be arranged as 'parallel streams' of one language course. The teacher decides for each student which programmes are to be used, how much weekly time is to be spent on each and at what point each programme is started. (Note 3)
-------------------------------------------- page 427 -------------------------------------------

Correct oral responses to organised stimuli

I now want to discuss in more detail two types of programme for the tape recorder, teaching the student to make correct oral responses to organised stimuli. One programme can teach the student to respond to questions, using abbreviated structures only, thus
'Where did you go yesterday?'
-- 'To school.',
rather than
'I went to school yesterday.'
I will call this type a 'short-response programme'.
The other type is a 'sentence-structure programme', in which the student learns to say things like 'The man went home at 8 o'clock.' Since the majority of questions do not require a complete sentence as an answer, this programme will make comparatively little use of questions as stimuli but use drawings and translation instead. (Note 4)
Good and unequivocal drawings can be described by the student in complete statements, and a series of such drawings can elicit a narrative. To cover cases where one picture cannot unambiguously elicit the description of a complex situation, I am now trying, in collaboration with two typographers, to develop a set of graphic symbols. This can be used by programmers for the construction of pictorial stimuli. Thus, the sentence 'Der Mann geht nach Hause' (= The man is going home.) would be elicited by three picture frames, the first indicating 'man', the second showing a pair of feet to indicate 'goes' and the third showing a house indicating 'nach Hause' (home). (Note 5)

Grammatical selectiveness

The programmer, in order to be effective, should not only separate the various linguistic skills but also limit seriously the grammatical features and discriminations being taught. In a German programme, for instance, gender should not interfere as long as case concepts are being taught. I therefore suggest
-------------------------------------------- page 428 -------------------------------------------
teaching nothing but masculine nouns (which show the case discrimination most clearly) until all cases, all tenses and all basic structures have been mastered. The same material can then be covered again for the two other genders, and this revision, which teaches only forms but no concepts, can proceed at a much faster rate. Similarly the third person singular alone could initially be used for all structures and tenses of the sentence-structure programme, as it is probably the commonest. In those sections of a German programme for English students which rely heavily on translations, or in programmes which are based exclusively on teaching by translation, the German imperfect tense could be introduced before the present tense and be used exclusively for a long time because it can be more easily matched with the English simple past that can the German present with the English present.
In the short-response programme, on the other hand, it is more natural to use the first and second persons.

Responses in unistructural programmes

Both the sentence-structure programme and the short-response programme are unistructural, i.e. each basic structure will be practised exhaustively before the next one is tackled. The preparatory work for both types of programme co-incides. The vocabulary to be taught has to be collected and marked according to difficulty in various respects, according to frequency, usefulness and to whether it can be pictorially presented. Then basic 'grammatical situations' are selected and sequenced, e.g. 'direct object situation', 'indirect object situations', 'relative order of adverbs', 'subordinate clauses' (e.g. in German), etc.
Next basic patterns from each 'grammatical situation', are selected for teaching. From the direct object situation in German for instance I would select the following for teaching:
-------------------------------------------- page 429 -------------------------------------------

1 Subject-Verb-Object:
Der Mann kauft einen Hund.
(The man buys a dog.)

2 Subject-Verb-Adverb-Object:
Der Mann kauft heute einen Hund.
(Today the man buys a dog.)

3 Subject-Verb-Object-Adverb:
Der Mann kauft den Hund heute.
(The man buys the dog today.)

4 Verb-Subject-Object. -- Imperative:
Kaufen Sie den Hund!
(Buy the dog.)

5 Verb-Subject-Adverb-Object. -- Imperative:
Kaufen Sie heute einen Hund!
(Today buy a dog.)

6 Verb-Subject-Object-Adverb. -- Imperative:
Kaufen Sie den Hund heute!
(Buy the dog today.)

7 Verb-Subject-Object. -- Question:
Kauft der Mann einen Hund?
(Is the man buying a dog?)

8 Verb-Subject-Adverb-Object. -- Question:
Kauft der Mann heute einen Hund?
(Is the man buying a dog today?)

9 Verb-Subject-Object-Adverb. -- Question:
Kauft der Mann den Hund heute?
(Does the man buy the dog today?)

Each of these patterns in turn will be the basis of a big programme section. We take first 'Der Mann kauft einen Hund' (= The man buys a dog) and develop it further.
(The same procedure can be applied to all the other patterns.)
Two patterns of the qualified negative form are taught, namely 'nicht' (= not) after the object ('Der Mann kauft den Hund nicht.' / The man does not buy the dog.) and 'nicht' before the object ('Der Mann kauft nicht einen Hund, sondern eine Katze.' / The man does not buy a dog, but a cat.)
The non-qualified form includes both 'Der Mann kauft einen Hund' (= The man buys a dog.) and 'Der Mann kauft keinen Hund' (The man buys no dog. = The man does not buy a dog.). We have thus obtained three varieties of 'Der Mann kauft einen Hund'. To get a list of all responses we want from the student, we subject each of these varieties to the process of substitution, expansion and contraction.
Substitution requires little comment: for each word in a starting pattern all words are substituted that could take its place.
Expansion can occur as multiplication and as qualification.
  • Multiplication could turn 'der Mann' (the man) into 'der Mann und der Junge' (the man and the boy) or 'kauft' (buys) into 'kauft oder verkauft' (buys or sells).
  • Qualification could lead from 'einen Hund' (a dog) via 'einen großen Hund' (a big dog) via 'einen sehr großen Hund' (a very big dog) into 'einen nicht sehr großen Hund' (a not very big dog).
Next substitution can be applied to the expanded patterns.
Contraction can also be applied to both the starting patterns and the expanded patterns.
-------------------------------------------- page 430 -------------------------------------------
It is not possible to list sensible contractions without considering immediately which stimuli would complement and elicit these contracted responses. Here is a list of some stimuli and contracted responses (they will not necessarily be programmed in this sequence):
1 Haben Sie einen Hund? - Ja.
2 Haben Sie eine Katze? - Nein.
3 Wieviele Hunde haben Sie? - Einen.
4 Haben Sie zwei Hunde? - Nein, einen.
5 Welchen Hund kaufen Sie? - Den jungen.
6 Was für einen Hund haben Sie? - Einen großen.
7 Haben Sie einen kleinen Hund? - Nein, einen großen.
8 Was haben Sie da? - Einen großen Hund.
9 Kaufen Sie den Hund? - Nein, ich verkaufe ihn.
10 Wohin geht der Junge? - Nach Hause. or: In die Schule.
Translation for readers who do now speak German. Note that the German structures sometimes are much simpler than the English ones.
1 Do you have a dog? - Yes.
2 Do you have a cat? - No.
3 How many dogs do you have? - One.
4 Do you have two dogs? - No, one.
5 Which dog will you buy? - The young one.
6 What kind of dog do you have? - A big one.
7 Do you have a small dog? - No, a big one.
8 What have you got there? - A big dog.
9 Are you buying the dog? - No, I am selling it.
10 Where is the boy going? - Home. or: To school.
The introduction of pronouns is part of the contraction process. Some substitution can be applied to the contracted responses.
Thus a complete list of wanted responses is developed. The responses are sequenced. Most responses occur only in one of the two unistructural programmes.

Construction of stimuli

After having listed all wanted responses and arranged them in a suitable sequence, the programme constructs a stimulus for each response. In the 'repetitive-unit programme' to be described later, stimuli and responses are grouped into units of 10 uniform pairs of items. Each unit begins with a general instruction or 'unit stimulus' applicable to all 10 items - such as 'Answer the following questions' or 'Name the objects drawn on page 5' or 'You want the objects on page 5. Ask your friend if he has them.' Or 'Translate' or 'State the opposite' or 'Tell a friend to perform the actions depicted on page 5', etc etc. Thus statements, questions and commands can be elicited. Unit stimuli should, of course, try to refer to life situations rather than to grammar.
-------------------------------------------- page 431 -------------------------------------------
The presence of unit stimuli sometimes allows the 'item stimuli' to be less elaborate. If the unit is headed 'Name the objects on page 5', the item stimulus consists only of the number of the object to be named.
The pictures have the function of 'context-prompts'. Instead of pictorial context-prompts, verbal context-prompts can be used, particularly in the more advanced stages of the programme. For example:
  • Context-prompt:
    Es klingelte. Sie machten die Tür auf.
    (The bell rang. You opened the door.)
  • Stimulus:
    Was machten Sie, als es klingelte?
    (What did you do when the bell rang?)
  • Response:
    Ich machte die Tür auf.
    (I opened the door.)
With the same context-prompt as above but the changed stimulus 'Wann machten Sie die Tür auf?' (When did you open the door?), the response would be 'Als es klingelte.' (When the bell rang.) (Note 6)
The stimulus-response relationship should be sensible, interesting and correct. Here is a negative example. In a programme draft on English for foreigners setting out to teach the uses of 'some' and 'any', the following type of item was suggested:
Verbal context-prompt: I haven't any cigarettes.
Stimulus: Have you?
Response: No, I haven't any.
This item has at least two faults. Firstly, it is uninteresting: The context-prompt 'I haven't any cigarettes' and the response 'No, I haven't any' are almost identical. Secondly, the response, although in itself correct English, is not likely to occur after the question asked. The proper response to the question 'Have you (got any cigarettes)?' Is 'No, I haven't.' In fact it appears that 'any' is never used in response to a question. It *** is *** used in response to an imperative. I therefore suggested items of the following type:
Pictorial context-prompt: drawing of cigarettes.
Student: Give me some cigarettes, please.
Friend: I can't. I haven't got any.
Two sets of units with appropriate unit stimuli induce the student to take both roles in this setup in turn.
-------------------------------------------- page 432 -------------------------------------------


Many language textbooks, particularly those produced outside the country whose language they teach, are full of mistakes, even if the author was a native speaker of the language. Such books are comparatively harmless because students tend to forget the incorrect along with the correct. But if such mistakes get into programmes, they become dangerous because programmes are validated as to their effectiveness in shaping the student's linguistic behaviour. While the mistakes in textbooks normally consist of undiplomatic, wrong or senseless utterances, language programmes built on a stimulus-response play can also contain and teach wrong stimulus-response relationships. These relationships are subtle and experts often find it difficult to agree on them.
In the absence of extensive and detailed research providing, for instance, a catalogue of stimulus-response relationships and other linguistic facts to which the programmer can refer, I suggest that all language programmes before being validated for their teaching efficiency undergo the procedure of what I will call 'subvalidation': all model responses and all reinforcements should be deleted from the programme, and the programme should then be worked through by native speakers of the target language and by non-native speakers who know the language extremely well. Their responses should be noted and analysed.
The items that produce a variety of correct responses or unwanted responses should be reconsidered. Items which a native speaker cannot solve at a first attempt should be considered too difficult for a foreigner. Such a procedure will try to ensure that stimulus and model response present a natural relationship, that model responses are in themselves correct and meaningful, that stimuli, including pictorial stimuli, are unequivocal and that the student does not learn to do things the native speaker cannot do or does not do.
Subvalidation experiments provide plenty of //snip//
(* Some lines are missing here in the original publication, bottom of p 432, and cannot be recovered.)
-------------------------------------------- page 433 -------------------------------------------

Repetitive-unit programmes

Apart from linguistic content and individual frame construction, the programmer has to consider the learning procedure that the students will use. I shall describe a programme pattern and learning system that I call 'repetitive-unit programme ' and 'repetitive-unit learning system'.
All items are presented in units of 10. As a rule, all items in a unit demonstrate, or demand application of, the same 'rule', that is, they all have the same structure and the same stimulus-response relationship. Each item has a number which is called out before the stimulus. The student has a printed scoring book which has numbers corresponding to the item numbers in a vertical column. Depending on the teacher's preference, the student either listens to each unit once or twice before trying to respond or he tries to respond at first presentation even though the unit may contain entirely fresh material. This first attempt at responding is called 'test run'. The student ticks off all items he has answered correctly. He counts the correct responses at the end of each run through the unit and enters the total at the bottom of the current column. If during a test run he has made 10 correct responses, i.e. 100 per cent, he proceeds to the next unit.
If the student makes too many mistakes or if his mistakes occur in a certain pattern, the principle of 'intensive runs' comes into operation. An 'intensive run' is concentrated practice of one item or a smaller number of consecutive items without scoring. For the teacher's benefit, the student indicates that he has done an intensive run by writing, instead of the normal scoring marks, the word 'intensive' into the current column of his scoring book. For example, if, in the test run, he counts three or less correct responses, he has to make an intensive run. In this case, he works through the following parts of the unit repeatedly in turn:
-------------------------------------------- page 434 -------------------------------------------
  • items 1 to 3,
  • 4 to 6,
  • 1 to 6,
  • 6 to 10.
Another run from beginning to end follows, and the progress is scored. The student continues working on the unit until he has reached the 'passing score', defined as 'two consecutive runs with no errors'. If the score shows the same item wrong in two consecutive runs, the student makes an intensive run of that item because apparently it gives him trouble. If the score shows two or more consecutive items wrong in the same run, the student makes an intensive run of the wrong items, the reason being that individual items are difficult to find on tape and consecutive errors are therefore a conveniently large target for attack.
This learning method places on the student a responsibility for functions which the early 'Skinner disk' for graphic programmes fulfilled automatically but which today's tape recorders cannot fulfil.
The learning method influences the actual programme written for it, and I suggest that there are, as far as language programmes are concerned, the following advantages over the conventional linear method with small frames:
  1. Since correct responses at first attempt are not expected, the programme can be more concise, and more challenging for the intelligent student. Each student hears one item only as often as he needs to hear it. This makes the repetitive-unit programmes particularly suitable for brushing-up work.
  2. As the units are small and an unlimited number of repetitions is possible, weak students are not discouraged.
  3. Concentration is aided by the aim of mastering each unit with as few 'through runs' as possible. The system of intensive runs makes sure that the student does not repeat the whole unit indefinitely without ever coming to grips with the 'difficult' items. The student is never in doubt what to do next. The challenge is always before him.

  4. -------------------------------------------- page 435 -------------------------------------------

  5. Mastering one unit by reaching the passing score gives the student a feeling of well-deserved success. In a conventional linear programme, the mastery of one frame occurs so often and demands so little effort from the student that the feeling of triumph quickly wears off.
  6. The units give the student a sense of progress. The progress from unit 40 to unit 41 is a definite step forward, whereas a linear programme has so many frames that it makes little difference to a student if he has reached frame 540 or 560.

Media of presentation

As as long as machines that will synchronise acoustic and visual presentation and be capable of branching are not generally available in language laboratories, I suggest the use of a programmed book in combination with the tape recorder. A clear system of numbering the units on tape, the textual frames in the book and the pictorial context-prompts in the book makes this quite workable. Spoken instructions on tape tell the student what pictures to look at while making his responses, or what frame to turn to in the book when he has finished a unit. If the frame numbers in the book are suitably printed, any form of programming can be employed depending on the specific problem just being taught: simple linear, simple branching, complex branching, short or long frames, overt or covert responses, or straight non-programmed writing. Switching from one technique to another is possible at any time.

Final remarks

The suggestions contained in this paper are not intended to point the way forward or discuss 'ideal programming'. They seek to help the programmer to deal with *** today's *** teaching problems and to make the best of the equipment available in language laboratories *** today ***.
Acknowledgements are due to the Audio-Visual Language Association, London, on whose behalf this paper is being contributed to the Congress, and to Woolwich Polytechnic, London, and Holborn College of Law, Languages and Commerce, London, where most of the work described has been done.
-------------------------------------------- page 436 -------------------------------------------


In this paper, I have suggested the construction of 'single-skill programmes'. The tape recorder is particularly suitable for programmes teaching the student to make correct oral responses to organised stimuli. These programmes may try to elicit responses either in the form of short phrases or of complete sentence-structures. Depending on the type of programme and the skill being taught, different grammatical features should be selected for teaching. The building up of responses in unistructural programmes and the construction of stimuli has been discussed. The term 'subvalidation' was introduced for a procedure that tests the 'truth' of a programme rather than its effectiveness. The frame-work of 'repetitive-unit programmes' and the corresponding learning system has been described. The tape recorder and programmed books containing text and pictures have been suggested for presentation of language programmes.
-------------------------------------------- page 437 -------------------------------------------


Note 1: Here is a list of some of the skills most language students have to learn:
1 Understanding
  a Accurate hearing and discrimination of sounds
  b Comprehension, with the help of pictures, of language spoken at natural speed
  c Comprehension without 'crutches' of language spoken at natural speed
  d Comprehension of written language
  e Ability to derive the meaning of unknown words by systematic consideration of context
2 Oral expression
  a Correct reproduction of sounds heard
  b Correct articulation of sounds presented in writing
  c Correct responses to organised stimuli.
These can be
    i Abbreviated structures as commonly used in reply to questions or
    ii Complete sentence-structures, as used, for instance, in narratives
  d Fluent speech
3 Written expression
  a Correct spelling of sounds heard (dictation)
  b Correct written responses to organised stimuli (e.g. filling-in exercises; translation)
  c Fluent self-expression in writing (letters, essays)

Note 2: Programmed instruction is particularly suited to students who have very narrowly defined study aims: a complex-skill programme is particularly unsuitable just for these students who are predestined for programmed instruction. One should therefore consider if, perhaps, a complex-skill programme violates a basic principle of programming.
Note 3: I also suggest the introduction of a variety of tape recorders and other machines into one and the same language laboratory, each type of machine catering for the attainment of some particular skill. For instance, some recording machines, excellently suited for imitation and frequent repetition of short phrases, are difficult or impossible to use for exercises with long verbal context-prompts
-------------------------------------------- page 438 -------------------------------------------
or for exercises aiming at fluency in speech. Other machines can synchronise pictures and sound or text and sound, etc, but are unsuitable for frequent repetition. It is of course possible to build a 'super language teaching machine' which carries out all functions with equal ease. However, in a class working with programmes all students will do different types of work and a student will never utilise all functions of his 'super machine' at the same time. It may be more economical to let him work on one single-skill programme and one single-skill machine for 30 minutes or so and then let him switch places, machines and programmes with another student.
The change of machine may also help to break the monotony of work and thus enable students to spend longer periods in the language laboratory.
Note 4: Translation can be very useful when structures of native language and target language are to be compared and when the student is to be immunised against interference from his native tongue.
Note 5: This third frame could, of course, not be used for the English word 'home'. The German word 'fährt' (= goes/travels on a vehicle) may be indicated by a wheel. To distinguish between 'goes', 'walks' and 'comes' (provided all three imply 'on foot'), a pair of feet and the initial letters g, w or c may have to be shown. When the English word 'goes' implies travel by means of some vehicle, the wheels would be used to elicit it.
Note 6: If questions were found suitable for a sentence-structure programme, we might find the following stimulus-response relationship in such a programme:
  • Was taten Sie, als es klingelte? (What did you do when the bell rang?)
    Als es klingelte, machte ich die Tür auf. (When the bell rang, I opened the door.)
  • Wann machten Sie die Tür auf? (When did you open the door?)
    Ich machte die Tür auf, als es klingelte. (I opened the door when the bell rang.)
Subvalidation procedures would have to establish if the latter is a natural stimulus-response relationship.
-------------------------------------------- page 439 -------------------------------------------


Note: It is not normally necessary to devote 14 runs to one unit. This score has been designed to combine all the features that might occur in one score or another. End of this page

Note added in 2012

The scoring sheet was inspired by Robins and Harris 1959. The difference between the original and Klaus Bung's development is that, in the original, the student always had to tackle all ten items in succession until reaching the passing score, i.e. the intervals from between each repetition (revision) of items was always the same and comparatively long. It was therefore possible (but undesirable) that the student continued practising an exercise ad infinitum without ever reaching the passing score.
Klaus Bung introduced the intensive runs which dynamically shortened the intervals between repetitions when and where required by a particular student for a particular item. This brought the control of intervals down to the micro-level of seconds and increased them as and when possible. It therefore anticipated the proposals made by Pimsleur 1967 for short-term retention (micro-level in terms of seconds). This work was transferred from the acoustic medium to the graphic medium (and from short-term to long-term retention, from micro-level control to macro-level control), in successive publications, and led to the development of the theory of adaptive learning algorithms as realised in what is now called PAPA (pen and paper algorithm), REV (revision algorithm), ENFA (enforcer algorithm) and to the general memory model most elegantly and succinctly described in Bung 1991. The cheerful (uninhibited, even though ostracised) use of the mother tongue for explanations of grammar and vocabulary and for exercises in language (which brings so many benefits and has no drawbacks if intelligently handled) has more recently been confirmed by the successful practice of teachers such as Michel Thomas (see Solity 2008) which agrees in many other respects with the principles of the IDYLL® METHOD™, except for its rejection of writing in elementary language teaching and some of its formulations about the respective responsibilities of teachers and learners, which seem to contradict each other, but would largely coincide if both were formulated within the same theoretical framework, e.g. that of Helmar Frank's didactic variables (Bung and Rouse 1970)


  • Bung, Klaus 1991: "Dynamic Learning Algorithms ", Expanded version of a paper read at the Congress of Educational Cybernetics at Charles University, Prague, 26-29 August 1991. Reprinted in Klaus Bung: "Cartesian Language Learning", Vol 3, Institute for Dynamic Language Learning, Blackburn, England, 2012.
    -- German translation in: Lobin, Günter, Heinz Lohse, Siegfried Piotrowski and Eva Poláková (eds) 1998: 'Europäische Kommunikationskybernetik heute und morgen' (European communications cybernetics today and tomorrow). In honour of Professor Helmar Frank on the occasion of his retirement. 180 pp. Published by Kava-Pech, Prague, Czech Republic, ISBN 80-85853-38-8, and by KoPäd, Munich, Germany, ISBN 3-929061-83-5, p 1-18
    (English version reprinted in this volume)
  • Bung, Klaus, and Kenneth Rouse 1970: 'Introduction to Helmar Frank's concept of didactics'. In: RECALL, Vol 1, p 174-196. London, England (reprinted in this volume)
  • Decoo, Wilfried 2001: "On the mortality of language teaching methods"
  • Pimsleur, Paul 1967: "A memory schedule". In: MODERN LANGUAGE JOURNAL, Vol 51, p 73-75
  • Robins, Lewis, and Reed Harris 1959: " 'Instant' French". Pickwick Sales Corporation, The Keel Manufacturing Corporation, Freeport, New York, USA
  • Solity, Jonathan 2008: "The learning revolution", 248 pp. Hodder Education, London, England.
© 1964 and 2012 Klaus Bung