Wikipedia:WikiProject Ballet/redlinks

An Analytical Enquiry into the Functions and Value of Proper Grammar in the Philosophy of English, with special reference to the Teaching of acceptable Written English to Children aged sixteen to eighteen

By

Agida Michael

Thesis submitted in fulfillment of the requirements for the degree of DPhil in

English of the University of Oxford June 2017

The thanks of the writer are due to Professor Austin Chukwu the Dean of humanities and social sciences of Federal University Ndufu Alike Ikwo, Ebonyi State, Nigeria  who inspire him to study further on Grammar, Dr. Ben. O. Nweke the Head of Department of Languages/Linguistics/Literary Studies/Visual Arts of  Federal University Ndufu Alike Ikwo, Ebonyi State, Nigeria and in particular to the Senior English Teachers and English Staff of the five schools concerned in the research work, to his Academic Advisor Professor Charlse Paine, for their courteousness and help.

ABSTRACT

In this work, the worth of the traditional English grammar class in serving children to write correctly was tested. The grammar lesson was found to be certainly not grander, and in most instances was substandard, to direct

training in writing skills. The progress of five forms having no grammar lesson was measured on eleven counts against that of five similar forms following the same English course but taking one lesson a week of English grammar. At the end of two academic years, of the 55 resultant scores, 25 proved highly reliable. Ten of these showed a significant advantage (where t = more than 3) to the non-grammar forms, none to the grammar. Of a further 20 fairly reliable measures, one significantly favoured the non-grammar course, and none the grammar. Where t equaled or exceeded 1.5 in calculating the significance of a difference between the mean scores of the grammar and the non-grammar scores, that is, in thirty-four scores, thirty went to the non-grammar forms, as against only four to the grammar – none of these four reaching a level of significance in which t = 3. Significant gains were made by the non-grammar forms in the following measures: the number of words per common error (three forms out of five); the variety of sentence patterns used (two forms); the number of complex sentences currently used (four forms); the number of correct sentences (one form); and the total number of words written (one form). After a preliminary experiment of three months, a pair of forms from each of the five secondary schools took part for the first two years of their secondary course. There were 119 pupils in the grammar forms, and 109 in the non-grammar. Two of the schools were boys’ Grammar schools, one located in Oxford and the other in Central London, two were Technical branches of Comprehensive schools both located in Oxford and the only girls were from a Secondary Modern school in Oxford. All schools were in United Kingdom: two to the east, two to the west, and one central. Each child attempted a formal grammar test at the beginning and end of the sequence, and each wrote at an interval of two years, an essay on the same subject. The grammar forms wrote a total of 71, 130 words, and the non-grammar forms 60,313. From the breakdown of these essays, the eleven counts were taken by the subsequent measurements:

a) Typical length of the correct simple sentences

b) Instances of the inadvertence of the full stop

c) Numeral of words per common error

d) Amount of dissimilar sentence forms

e) Sum of non-simple sentences minus the number of simple, correct and incorrect contained within

f) Total of subordinate clauses

g) Overall number of words

h) Figure of correct multifaceted sentences

i) Number of correct simple sentences containing two or more  transforming phrases

j) Number of correct sentences exceeding the number of incorrect

k) Number of adjectival clauses and phrases.

Four of the pairs of forms were each taught by one teacher; the period of teaching, time allowed for the essays, and the content of the grammar and non-grammar courses were as far as possible the same for each child. The cogency of the eleven measuring instruments was proven by their abstraction from an assessment of the written work of ten-year-old and of fifteen-year old children. The measures were tested for dependability by being used on a pair of essays written at a week’s interval by 27 children aged sixteen to eighteen. The measuring instruments have two virtues: they test available skill and not mere recognition; and they are to some extent measures of maturation and not simply of error. Methods of assessing the importance of common errors are discussed, as is the lack of correlation between relatively successful grammatical study and improvement in writing-skill. The work includes various appendices, among them one remarking on the use of the Error Quotient recommended by Stormzand and others for relating errors in importance, another making suggestions for a course in language in the secondary school and intended to replace the traditional grammar course, and a third giving a summary of the grammatical contents of a number of text-books published since 1944. Despite various investigations of the value of formal grammar in teaching English, there has been little fundamental change in the teacher’s approach to the subject in Great Britain. This conservatism may be due to the unconvincingly short period of experiments previously undertaken. The present work covering a period of two academic years is an attempt to overcome this objection.

GENERAL CONTENTS

Chapter I

a)      Objective of the work, report of problem.

Chapter II

a)      Climate of opinion 1890-1960.

b)      Subjects of previous research.

Chapter III

Methodology of experiment, and its limitations.

Chapter IV

A pilot experiment.

Chapter V

Material of the experiment

I. a)   Population.

b)      Types of school; work undertaken.

II. The search for a measuring instrument.

III. Reliability of the measuring instrument.

IV. Results of a grammar test.

V.      Scoring of the main essays.

VI. Results of the error count.

Chapter VI

Conclusion and discussion

Appendices

Bibliography DETAILED CONTENTS

Chapter I – Aims and Problem

-Discussion of term ‘formal grammar’ and its limits as used.

-The crucial issue – does formal grammar help in writing.

-A note on some defences of formal grammar.

-Does formal grammar help older children to write correctly – some G.C.E. marks as evidence.

-Discussion of the term ‘correctness’ – need to include maturing style, as well as mere absence of error.

-That formal grammar still occupies much time in schools and in the textbooks.

Chapter II – Climate of Opinion 1890-1960 and Previous Research

A. Climate of Opinion 1890-1960

1. Early changes in opinion of the value of grammar in teaching English.

a) Grammar as a mental discipline;

b) As a pure science.

2. The general trend – the teachers’ desire for grammar as a tool of expression.

3. Official opinion and its general support of the position of formal grammar.

a) The inspectorate;

b) Professional groups.

4. Conflicting individual judgments on grammar’s value.

5. Orthodoxy of textbooks.

6. Some attempts by modern linguists to revise the grammatical approach

B. Themes of Previous Research

Theme 1. Can grammar be successfully taught?

a) popularity or otherwise of grammar lesson;

b) expected content of learning and age of acquiring it;

c) results of enquiries by Macauley, Cawley.

Theme II. Does grammar teaching fulfil its intended functions?

a) some implications of Cawley’s work;

b) inconclusiveness of research in this theme so far;

c) is grammar a useful mental discipline?

d) the relationship between knowledge of grammar and other subjects and

skills;

e) a possible explanation for lack of transfer of grammatical to other skills;

f) the value of grammar in correcting common errors;

g) neglect in Britain of results of American and other research, and reasons for

this.

h) is direct practice more effective than grammatical study in gaining writing

skill? Some limitations of the enquiry into this;

i) attempts to find new ways of teaching grammar; limitations of these.

j) grammatical knowledge and literary criticism.

k) attempts to reformulate grammatical terms (Fries).

Theme III. What are the valuable elements in formal grammar?

What are the major errors and difficulties children experience in writing?

a) the small number of grammatical items agreed to be important;

b) the main difficulties of children as writers;

c) important characteristics of an error to the practicing teacher – massiveness

and generality;

d) agreed signs of maturation in writing-skill.

Chapter III – Approach to the Experiment

a) Language changes slow to become established, therefore:

1) long-term work needed;

2) ordinary continuous written work used to test available not merely

recognitive skill.

b) Difficulties arising from long-term work –

1) limits of population, numbers and types of school;

2) lack of control of variables;

-- unequated groups, and changes in course of work;

-- teachers’ attitude and skill;

-- method of and speed of grammatical instruction;

-- influence of foreign languages.

Chapter IV – A Pilot Experiment

a) description of a three month trial course.

b) results of grammar test and of structure and error counts.

c) need for a more refined method of measuring changes in writing ability.

Chapter V – Material of the Main Experiment

Section I - Population

a) pattern employed;

b) population and amount of written material;

c) schools, teachers, content and procedure of course; texts.

Section II - Measuring Instruments

a) Search for a measuring instrument:

1) Fallibility of essay-marking; need for a measure of maturity as well

as of correctness;

2) Ineffectiveness of recognition tests based on common errors;

3) Steps in the analysis of essays.

b) A comparison of essays written at ten and at fifteen years of age, giving

criteria for assessing growth, and methods of scoring.

c) A list of common errors, and method of scoring.

d) The validity of the scoring criteria.

e) Sample essays, scored.

Section III - Reliability

Reliability of the measuring instruments – five very good, four fairly good,

two unreliable but retained.

Section IV - The Formal Grammar Test

Scores of the formal grammar test – not very high even by grammar forms, but sufficient, and with significant gains by grammar groups.

Section V – Scoring of the main essays.

a) Scores on eleven measures by the experimental groups after nine months, showing some advantage to the non-grammar groups but scores as yet unstable.

b) Similar scores at conclusion of two years. A much more pronounced advantage to the non-grammar forms, but still not decisive.

Section VI – Results of the Error Count

a) - d) showing importance of errors judged by commonness of occurrence and widespread among the children; and showing also the similarity of the types of error in the various schools.

e) The error-count in the comparison of methods.

f) Opportunity for common errors.

Chapter VI – Conclusion and Discussion

1. Some measurable changes in children’s writing.

2. a) the limited degree to which formal grammar is learnt.

b) The lack of advantage gained from grammatical study in writing-skills.

c) Is it advisable to teach grammar in the English lesson?

3. Some possible replacements for the formal grammar lesson.

4. The contents of the textbooks.

5. Nature and importance of common errors.

Appendices

1. List of common errors.

2. Contents of some courses of English published since 1944.

3. Preliminary recognition tests A and B, and Grammar test C.

4. Suggestions, with a summary, for a scheme of language study (A Grammar of Situation).

5. Examples and criticism of the ‘error-quotient’.

6. Headings of essay analyses.

7. Some results of the analysis of variance.

8. a) mean IQs of pairs of forms.

b) Mean scores of pairs of forms on eleven measures in first essays.

Bibliography LIST OF TABLES

Table 1 Pilot experiment

Comparison of scores in grammar test C by non-grammar and grammar forms

Table 2 Pilot experiment

Comparison of scores in error count by non-grammar and grammar forms

Table 3 Pilot experiment

Comparison of non-grammar/ grammar forms in structure count

Table 4

Population of main experiment – non-grammar forms

Table 5

Population of main experiment – grammar forms

Table 6

Average scores by 11-14 year old children in usage tests A and B

Table 7

Average scores by 11-14 year old children in grammar test C

Table 8

Scored differences between writings of children aged ten and fifteen years

Table 9

Scores in items a) to k) in the ten/fifteen comparison

Table 10

Reliability of the measures a) to k)

Table 11

Scores of main schools in Test C (formal grammar)

Table 12

Scores in measures a) to k) after nine months for schools A, B, C

Table 13

Final scores in items a) to k) after two academic years for schools A, B, C, D,E

Table 14

Correlation of the top third of scores made in the final grammar test with changes in score between the first and final essays

Table 15

Error count of first essays

Table 16

Error count for final essays

Table 17

Total error count for combined schools, first and final essays, grammar/ non-grammar groups

Table 18

Combined total error count in all essays 184 &

Table 19

Significance of the differences of mean changes between first and final essays by all non-grammar as compared with grammar pupils in common errors

Table 20-24

Significance of the differences of mean changes between first and final essays by non-grammar as compared with grammar pupils in common errors within each school

Table 25

Results of comparisons of errors 1b, 1c, 3d, 3f, 4 within schools

Table 26

Numbers of instances of scores in first and final essays where the critical ratio exceeds 1-5

Table 27

Grammatical terms used in fifteen courses of English

Table 28

Detailed scores in preliminary tests A, B and C

Table 29

Error quotient of separate errors for five children

Table 30

Analysis of variance, measure (c)

Table 31

Analysis of variance, measure (d)

Table 32

Analysis of variance, measure (h)

Table 33

Analysis of variance, F ratio for within classes variance, measure (j)

Table 34

Analysis of variance, F ratio for within classes variance, error (3d)

Table 35

a) Mean IQs of pairs of forms at beginning of course.

b) Mean scores for pairs of forms on the eleven main measures at beginning, in first essays

CHAPTER 1

AIM OF WORK AND DESCRIPTION OF PROBLEM

The aim of the present work is to decide how far there is any value in teaching Traditional Formal English Grammar to children in their early years at the secondary  school, and in particular whether the use of grammatical terminology in instruction  and in correction helps to foster correct writing.

The term “formal grammar” has an ill-defined denotation by this time, and the descriptive “traditional” may seem to blur its outline still further. Indeed, it is intended to serve this purpose, for it implies a grammar based well back in time, but not incapable of flexible development. ‘Nobody in our present days, in the present writer’s experience, is found to teach English grammar in the highly Latinized form in which the term would have been understood’ by, say, Dr Johnson. On the other hand, much that a more popular contemporary approach would call ‘functional grammar’ is still in  practice largely dependent on modes of thought and on categories worked out and  inherited from Latin days. Before 1850, language instruction was dominated by grammar, and although the drift from disciplinary to social values in education resulted in changing grammar   from an end in itself to a means to an end, the main result seems to have been a falling to disorder of the general attitude of teachers so that for example a wide disparity came to exist between one teacher’s judgment and another’s of the proper level at Which to teach certain items. Recently, in 2013 a committee of the National Council of Teachers of English in America sent out a questionnaire to which 271 replies were received, showing that “grammar in most schools seemed to occupy one half to three fifths of the school time devoted to English in the last three years of the elementary course”. In Britain the proportion now would seem to vary between one quarter and one fifth.

There has intruded into the formality a large reference to meaning, which while at first appearing to add a measure of concrete sense to the many abstractions and generalizations of the formal categories, has in fact on occasion merely multiplied these unnecessarily, and added confusion to complexity. Even this “functional” categorization is not at all new, as may be seen in for example Ben Jonson’s apportionment of gender in accordance with the extensional meaning of a word: Of the genders there are six. First, the masculine, which comprehendeth all males, or what is understood under a masculine species … Second the feminine, which compriseth women, and female species … Third, the neuter, or feined gender: whose notion conceives neither sex; under which are compris’d all inanimate things… Fourth, the promiscuous, or epicene, which understands both kinds, especially when we cannot make the difference … Fifth, the common, or rather doubtful gender, we use often, and with reason; as in ‘cousin’, ‘gossip’, ‘friend’… when they may be of either sex; and so on. This is a clear example of the uncertainty arising from the impinging upon arbitrary grammatical concepts of the everyday idea of meaning. Similarly today, if formal grammar has changed by being dubbed “functional”, it is chiefly to become more pertinent to present usage at the cost of dwindling coherence and assuredness. The position of grammar in the schools is reflected in the confusion of the meanings of some of its titles. Thus “functional” grammar may indicate a) that aiding in correct writing or speaking, and limited in study to items effective in this – a usage of the term primarily American; b) that whose terms are defined according to function – the present most common British usage; c) useful rather in the first sense, but not excluding much general linguistic background, and implying also a more positive approach than the simple correction of error. This is probably the most progressive modern view. Similarly the word “pure” and the word “formal” as applied to grammar have fluctuated in meaning, and still retain diverse meanings. “Pure” grammar may be

a) grammar as a “science”, careless of practical ends though not necessarily excluding them – thus non-functional;

b) the same, but especially intending “universal” grammar, that is, the grammar common to a few European languages; c) grammar applied to helping correct writing and speaking – see functional (a) above.

Again, the word “formal” at first was applied to Latin grammar and to English grammar taught in Latin terms. It may also be used for that part of grammar whose significance is realized chiefly by the external form of words, though some grammarians talk of internal form also. And finally, it is used in the present work to represent that hotch-potch of grammatical terms at present taught mainly by functional (b) methods in British schools.

School grammar books are still insufficiently English, that is, insufficiently concerned with forms specific to our language, and insufficiently formal to be useful in teaching. They rely too much on the dictionary, on the meanings of words being understood before the grammatical function can be given, and too little on the formal signs and structures by which we recognise the relationships of words without necessarily knowing the words’ meanings – indeed, signs by which we often are aided to guess the lexical meanings. But this is to anticipate the discussion, whether the more profitable approach to children’s study of language be through formal signals or through meaning; if the former, the signals must be true to the facts of the language; if the latter, the names of the signals are largely tautological; if the two methods are to be combined, it cannot be in the indiscriminate way of the average school textbook. That the present enquiry might have been uncalled-for had a truly formal and truly English grammar existed, may excuse the preceding paragraph’s insertion at this stage. There are signs, in the work now for the first time being attempted in some English schools, as in the adaptation of C. C. Fries’s ideas to class-teaching, as well as in the pronouncements of such educationalists as I. Michael, or even in the evasions of some English teachers who affect to teach no formal grammar and yet follow a course of it in their school texts – that a new relevance is being given to descriptive English grammar. But these signs are few, tentative, and so inevitably uncertain that it remains true to say that in the English lesson it is still traditional formal grammar with a functional bias that is mainly taught. Since definition may be impossible to reach, it may be as well to say here that by traditional formal grammar is meant, at the ages of twelve to fourteen, such material as: analysis of the sentence into subject and predicate; direct and indirect object; simple complement; recognition of phrases and clauses, though not at this stage their analytical nomenclatures; some extensions of simple parts of speech; naming and recognition of parts of speech; including in the verb the terms transitive and intransitive, some tenses, indicative and imperative mood, active and passive voice, participles and infinitive; the adjective, quality and quantity; in the pronoun, personal, interrogative, relative, and possible demonstrative, with subject and object forms; in the noun, the possessive case and the plural; normal word order; concord. If these terms of equivalent ones are used in teaching English, then for the purposes of this enquiry formal grammar is taught.

Many teachers would concede that they do use much of this terminology. Most would no doubt hope and feel that they succeed in linking the terms to the practical business of composition. This would seem to be a crucial justification for retaining formal grammar in English instruction, and it is to test whether in reality any transfer exists between the terms recognized by pupils and the compositions written that this work sets out. Defences of formal grammar have been proposed other than this of its practical value in teaching correct writing and in making the business of class correction more efficient. These defences are such as: that knowledge of formal grammar helps us to understand literature; that it disciplines the mind, i.e. trains it in concentration, generalization, and logical connection; that it prepares for the study of other languages; and that it gives command of a useful terminology. These have all been debated and I think refuted at one time or another. Mention is made in Chapters II and III of some of the particular objections. But since some of these defences are still current, a short note should perhaps be made on them here. For reasons which will be stated in Chapter III, it is probable that none of the common justifications of formal grammar has been finally and completely refuted. Nevertheless, the objections to formal grammar on some counts are such that no one defense would suffice to maintain the subject in the classroom, were it not for the pressure of examinations and for the lurking expectation that there really must be some practical connection between grammar and writing. Understanding of formal grammar may indeed be able to further understanding of literature – to sense the timelessness of nouns, and the flux of time in verbs, may open a window on this or that passage of prose; but is the key of the grammatical terminology essential to this function? A logical and coherent pattern of formal grammar may, being studied, train students in the qualities it embodies; but not only have experiments shown some doubt that any transfer exists between the learning of grammatical logic and the use of a parallel logic in other disciplines, but formal English grammar itself has been only too often seen to lack the qualities of logic, coherence, and descriptive truth. If English grammar is a good training for the study of other languages, many English teachers must query whether they should spend valuable English time for this purpose, and many would maintain with their foreign-language colleagues that the differences between the grammars of different languages are too great or too confusing to make English a useful tool for this purpose. Finally, whether the terminology is useful depends on the resolution of the doubts entertained above, and in particular – if “useful” is to retain a narrow but practical sense – on an affirmative answer being given to the query of this thesis: whether formal grammar helps to train children in writing correct English.

The main intention of the present work is to evaluate formal grammar in the training of the younger secondary school children. It is commonly believed, if one may judge from conversations with English teachers, that although formal grammar may lack immediate value to children in the primary school,, it should nevertheless be clearly pertinent to the work of secondary children, who will have had time to digest a large proportion of its terms and to apply them frequently to their own written work. That this belief is of uncertain validity can be suggested by a scrutiny of some of the marks awarded to candidates in the London General Certificate examination in 1956. Of just over 700 sets of scores seen by the writer, 285 contained a mark for the question on formal clause-analysis and parsing. In this examination, candidates had in addition to answer three questions based on composition, comprehension, and summarizing. The marks gained by the 285 grammar candidates on the other three questions were correlated with their grammar marks. The figure obtained was low: r = +0.365+_0.222 This suggests that there is very little connection between the clause question and the rest of the questions combined. This is perhaps more noteworthy when it is considered that, in general, instruction in clause-analysis is reserved by schools for cleverer candidates. It may well be that only in the sixth form can the full value be obtained from formal grammar – when there is time and ability for a full discussion of the descriptive validity and historical growth of the language’s grammar. As with the term “formal grammar”, so certain rough limits have had in practice to be set to the denotation of the term “correctness”.

A list of some of the errors commonly found in the writing of children is given in Appendix 1. This list was derived largely from the written work of ten-year-old children, and was in its final form agreed as being representative by a number of English specialists. Nevertheless, since the aim has been to measure the effect of formal grammar on correctness, those errors (namely 1a-b-2a-2b) which most clearly belong to the sphere of convention and usage rather than to logic and grammar have been omitted from the calculation of correctness. There can be great disagreement even among teachers as to what is or is not an error. In the long run, the concept of “correctness” meaning mere “absence of certain errors” is insufficient. It ought to be feasible, for example, to train a child to write a series of almost identical, mechanically correct simple sentences, and in an error-count to outstrip another child who has striven, albeit with inaccuracies, to compose in a varied and maturing style. “Correctness” and grammatical control therefore, in the sense used in the present work, is intended to embrace a certain maturation of style, of which such measures as the variety of sentence structure and an increasing proportion of complex sentences are indicative. These measures were in part suggested by previous research but largely derived from the structure differences between the written work of children of ten and fifteen, and are again restricted in the event to those measures which are objective and common. Thus such important matters as the appositeness of style to content are not considered here. The frequency of the common errors in Appendix I has been shown as only one of eleven measurements of correctness. From the earlier comments in this chapter on the defences of formal grammar, and from the reports in Chapters II and III on previous work on this topic, it will be seen that the value of formal grammar in English teaching has to be accepted already with considerable reserve. That the doubts already expressed have not had a decisive effect on teaching itself may be explained by the inconclusiveness in such a subject of experiments of a relatively short duration and by their ineffectiveness in challenging habits long established. It is difficult to believe that a course one has been following for any length of time, and in public, has been wrong from the start. In consequence, although the approach of grammar teachers to English has been modified since Latin days, and the word “functional” is now in vogue instead of “formal”, the actual time and effort devoted to what is still basically formal grammar has changed very little over the past twenty years. Secondary schools, in the present writer’s experience, still commonly devote one quarter to one fifth of their English time to instruction in formal grammar, and one must add to this the use of formal grammatical terms in the teacher’s corrections.

A brief survey of the first books of fifteen courses of English, all but two of them published or revised since 1950 (see Appendix II) reveals the average number of pages spent on formal grammar to be 35, ranging from 14 to 78. This represents an average of 19.3% of the space available, ranging from 12% to 41%, which corresponds roughly to the time spent in class on the subject, and indeed perhaps explains it. A detailed summary of the grammatical terms used in these books is presented in the Appendix. Here it may be sufficient to note the general and main ones, on the one hand, and the scope of divergence, on the other. The parts of speech were mentioned in some degree by all the books: nouns, pronouns, adjectives, verbs, and adverbs by 15; conjunctions by 13; prepositions by 11; and interjections by 7. These parts were subdivided to a very varying degree: abstract, common, proper and collective nouns received frequent mention, but only three books gave verbal noun and gerund. While five books taught gender, only one classified it as masculine, feminine, common, and neuter. “Case” occurred in five books, in which dative and vocative were included once. Altogether, twelve subdivisions of the concept “noun” were given.

No one type of adjective was named by a majority of books, but twelve subdivisions of the adjective appeared. Of pronouns, most books gave personal and relative, but ten subordinate points or groups were presented. Adverbs of time, place and manner were often found, though not in a majority of the books, and five other types of adverb were explained. Conjunctions were subdivided by two books into five types. Verbs, as one would expect, received fairly exhaustive treatment, voice and tense scoring numerous mentions, mood far less, and weak/strong, auxiliary, and defective had a place. Seven books mentioned the paragraph. The sentence was divided into subject and predicate in fourteen of the books.

A detailed examination of simple sentences in column or graph was practiced in six books, and simple, compound, complex and compound-complex sentence structures were named by one or another. Twelve books had the direct object set out, and five the indirect, but the complement and structures in apposition were also named. Phrases, of which five types were discussed altogether, were named by thirteen books, clauses by four, six types of clause being given, with one book showing all types of adverbial clause.

A few more or less difficult grammatical terms were thinly scattered: antonym (six books), synonym (seven), homonym (one), ampersand (one) antecedent (one). It is evident that very few authors would agree precisely as to the composition of a course on grammar to fit any age group, but nevertheless it is equally clear in these fifteen modern books that a fairly large body of formal grammar is bought and presumably taught by one means or another. Is it all worthwhile? That is the question one hopes to answer with regard to the written correctness of work in the early years at the secondary school.

CHAPTER II - THE CLIMATE OF OPINION 1890-1960 and THEMES OF PREVIOUS RESEARCH

A. THE CLIMATE OF OPINION

At the turn of the century, English was a new subject on the curriculum, and its

treatment was formal in an attempt to establish the value of the subject as a mental discipline in challenge to the well-proven Latin. Historical and analytical grammar was regularly pursued in the secondary school, and parsing and formal analysis were the routine. Education in general was not “child-centred” but prescriptive, and the teaching of grammar shared in this common trait. Rather than the common errors and difficulties of children’s writing, it was the forms and sources of Milton’s vocabulary and the anomalies of Shakespeare’s grammar that were studied. Teachers in England would doubtless have agreed with their American colleagues on the Committee of Ten and the Committee of Fifteen, who maintained that “the study of formal grammar is valuable as training in thought” and that “grammar demonstrates its title to first place by its use as a discipline in subtle analysis, in logical division and classification, in the art of questioning, and in the mental accomplishment of making exact definitions.” This emphasis on grammatical training as a pattern for clear thought weakened rapidly, at least in the minds of teachers of English, and recently had dwindled almost to nothing outside the Grammar schools. This process was encouraged partly by the publication of experimental work on the transfer of training, most of which went to show that little or no transfer existed between competence at the mental skills which grammar was supposed to discipline and competence at grammar. A greater impetus to decay was undoubtedly given by the enormous spread of compulsory education, in which teachers were faced increasingly with the problem of helping native speakers of English to reach an elementary competence in speaking and writing their own language to a level at which grammar might usefully be discussed. Faith in grammar as a mental discipline never entirely vanished; it became absorbed by and linked with the advocacy of the study of formal grammar as a “pure science”, as when the Board of Education in 1921 recommended the study of “not English Grammar, but pure or functional grammar, including the elements of phonetics, analysis, and a little parsing.” The confusion of the terms “pure” and “functional” is not entirely resolved in the report of the Board. “Pure” would seem to represent the grammar common to a certain limited number of European languages; “functional” has no clear sense in the context, although it became important as a label intended to describe the later developments of grammatical teaching here and in America. What teachers were in fact anxiously looking for was not a mental discipline – it is difficult to discipline through language the inarticulate and the illiterate; and it was not a “pure science” – a further complex, and controversial subject to be assimilated by pupils already over-laden in the effort to carry the ordinary English skills. How complex and controversial English grammar was – and is – may be seen from the efforts of the 1911 Committee on Terminology to arrive at an agreed vocabulary of grammar, and from the subsequent reluctance of many grammarians of note to accept all the recommended terms Dissatisfaction with the traditional mode of analytical description has indeed progressed so far that attempts have been made to advocate not only modified terms, as by Jespersen, but entirely new ones as by C. C. Fries. Teachers were seeking rather a solution to the problem mentioned above, and rapidly dominating the work of the schools – the problem of teaching English children to read, speak, and write English. Teachers wanted a tool to assist in this instruction.

It was obvious to many of them, and was felt to be possible by many others, that grammar was such a tool. In so far as the current of opinion has any ascertainable trend in the Twentieth Century teaching of English grammar, it is in this general direction – away from abstract, disciplinary, scientific grammar, towards grammar as a correcting device in the teaching of English, and especially of written English. This generalization cannot bear too much weight; recent work has tended to revive the study of structure for its own sake, though not without the hope in the schools that such study would prove effective in improving children’s written work. Furthermore, the period is noteworthy as much for contradictory opinions about the value and effect of grammatical study as for agreement on these points. Nevertheless, it is generally speaking true that teachers of English, at least until the middle 1950s, did expect grammar to hold a place in the English syllabus, and to justify that place primarily by its functional success in helping to correct and polish the written and spoken English of the pupils.

The ancillaries – grammar as a mental discipline, as an aid in studying foreign languages, as a self-sufficient science – these remained present but in the background. It was the link with composition, the training in the proprieties of language, the justification of usage, the general and logical attack on slovenly incorrectness, that were stressed; in short, the efficiency of grammar as a tool in fashioning linguistic skills, an efficiency which has for so long been assumed as self-evident in the opinion of teachers of English. Grammar as a tool is indeed the grand figurative cliché of this branch of English teaching in the first half of the century. So much may be inferred from the reports of professional and public bodies on the teaching of English; from the contents of large numbers of commonly-used textbooks; from individual publications in articles and books by educationalists of various colours; and today from the conversation and practice of contemporary teachers and examiners. Some illustrations of the general trend and of the inherent contradictions of opinion may be of value here. In 1893, the American ‘Committee of Ten” had recommended that: The teaching of formal grammar should aim principally to enable the pupil 1) to recognize the parts of speech, and 2), to analyse sentences both as to structure and as to syntax. Routine parsing should be avoided and exercises in the correction of false syntax should be sparingly resorted to … The teaching of formal grammar should be as far as possible incidental and should be brought into close connection with the pupil’s work in reading and composition. This seems to give a very fair picture of the aims of grammar teaching in many English schools at the present day. But equal weight is not generally given to other points which the Committee had stressed so long ago – that, for example, “a student may be taught to speak and write good English without receiving any special instruction in formal grammar”, and that “the study of formal grammar has only an indirect bearing on the art of writing and speaking.” These opinions of the Committee, at first sight puzzling to reconcile, are not in fact inconsistent, for the Committee was concerned to propose formal grammar not as a tool for writing but as one for the training of clear and logical thought. In the 1960s, however, such a reconciliation is not so feasible, since the disciplinary value of formal grammar has ceased to figure as a prime motive in study. Although official reports published on the teaching of English in the last 50 to 60 years have seldom underlined the importance of grammar in the syllabus, they have never questioned that grammar should be taught. On the whole, the Inspectors’ reports cling to formal methods of training in language both in writing and in grammar. The 1921 report on “The Teaching of English in England” had, while developing the value of practical pursuits in education and of direct experience in human relationships in the classroom, recommended the study of “pure” grammar, with some phonetics, analysis and parsing included.

The 1931 report on the Primary School stated that juniors “must be made conscious of the functions of words and of the correct structure of the sentence, and must learn the grammatical terms arising there-from.” The stage which official opinion has reached today may be well represented by the words of the 1954 Ministry of Education pamphlet entitled “Language; some suggestions for teachers of English and others”. “It is only reasonable to assume”, states the Ministry, that a knowledge of the structure of sentences is useful at a certain stage in learning to write. To this knowledge most experienced teachers of English would add an acquaintance with the parts of speech and their functions. Experienced teachers of English, however, find that, in grammar schools at least, the accidence and syntax needed for the purpose of composition can usually be mastered by pupils of average ability in the equivalent of one weekly lesson for about three years; in other kinds of school more time would probably be necessary. Professional groups such as the I.A.A.M., the English Association, or in America the National Society for the Study of Education, have normally supported the general case for the inclusion in the syllabus of the study of grammar. Adverse opinion was sometimes heard, up to about 1930. Thus, in the discussion detailed in the English Association pamphlet no. 56 (July 1923), several speakers voiced the unease of teachers who would have preferred less English grammar than was at that time commonly found in the classroom, or at least “to teach the elements of grammar without using any technical terms”. An American committee on the “Articulation of the Elementary Course in English with the Course in English in the High School” (English Journal 1914, vol. III no. 5, pp. 307-8) contained in its report an emphatic rejection of formal grammar and its “useless terminology”, and demanded “grammar in the sense of correct use … not to be terminologised.” But such comment from professional groups, as distinct from rogue individuals, is rare, and most often grammar is recognized as part of the Establishment. Thus, the prevalent view of the 1923 English Association pamphlet mentioned above was expressed by S. O. Andrew, who recommended the study of “universal” grammar. By eleven or twelve the child “should understand what is meant by the several parts of speech, subject, object, complement, transitive and intransitive verbs, phrase and clause and their chief varieties, relative pronoun and conjunction.” Following this universal grammar, the child should study English grammar, including inflexion in English where it is still formally visible. The child was thus faced with two grammars to learn instead of one, and the teacher had no belief in either as a means of inculcating correct usage. For in 1923 the study of grammar was still proposed not as a tool in writing but for its disciplinary and scientific functions: “It is often said, and rightly” explains the Association, “that a study of grammar will not teach correct English.” Instead, the child would gain, it was claimed, a scientific method, a good introduction to foreign languages, and an enhanced ability to study literature. By 1946, official expectations had changed, and correct writing rather than a mental discipline was expected from the study of grammar; but the grammar was very much the same, and the age at which it was expected to be mastered was the same.

An article by Pamela Graden on “The Teaching of Grammar” in the English Association symposium published in 1946 (The Teaching of English in Schools) represents the general view. It reflects fairly well the present aims and criticisms of English grammar teaching in schools, together with the apparent inability to draw the expected conclusions from the adverse comments. The article stresses that the average grammar teaching suffers from a confusion of historical and descriptive grammar. The teaching tends to maintain dead distinctions and invalid rules of thought and expression, as that grammatical categories are unchangeable or that thought distinctions and formal distinctions are co-terminous. Graden indicates that the laws of correct writing base their authority not so much on grammar as on rhetoric, that is, on standards of elegance. She recommends that schools should bring grammar up to date, when it would become effective in helping children to write correctly, and grants as a preliminary to establishing this that as an aid to learning other languages or as a logical discipline grammar may be ignored. Graden lists the errors in children’s writing under three heads:

a) misuse of words, particularly prepositions;

b) technical errors due to ignorance of accidence, such as who/whom; none …

are; ‘than’ as a preposition and ‘like’ as a conjunction; ‘there is twenty dogs in the garden.’

c) faulty syntax, as for example sentences without verbs. Those errors under heading (a) may be corrected “only by constant reading and practice”. There is no place here for formal grammar. Those under (b) are a small percentage only, and can be corrected by explanation of the term “convention”. The syntactic errors are partly psychological – thus, children are said to express differences more easily than similarities, and to find certain relationships hard, as say the idiom expressing causal connection. All these difficulties may be overcome, in Graden’s view, and in language transfer of training is possible from exercise to general expression, by means of instruction in the basic principles of grammar and syntax. Similar assertions, and an attitude less critical of traditional grammar, may be found in the 1952 and 1956 publications of the Association of Assistant Masters and of the Association of Assistant Mistresses. Of course the connection between grammar and linguistic skills on which Graden relies for effective transfer of training has still to be established. The “basic principles” embodied in her recommendations wear a familiar look: they are noun, verb, preposition. The child should master the more formal parts of grammar between the ages of ten and twelve, because the adolescent finds difficulty in concentrating on an analytical and traditional subject. In America as in England it would appear that little essential difference to teaching practice ensued as a result of experiments testing the validity of grammar as a means of improving English skills. The 43rd year-book of the National Society for the Study of Education (Part 2, 1944) has an article on grammar by A. Broening. The author stresses that grammar is a tool to facilitate expression, but all the examples given blunt the sharpness of the analogy. Thus each correction offered is really a specific correction by convention, but masquerading as a grammatical rule. For example on the who/whom difficulty:

Child: Who are we going to have next?

Teacher: You are going to have me next. What word in your sentence takes the place of “me”?

Child: “Who”.

Teacher: You need an object for the verb “to have”. “Whom” is the object.

Child: Very bad.

Teacher: You need the adverb “badly”.

Evidently, one is still in a topsy-turvy world where meaning is used to grasp the form and the form is then supposed to explain the meaning. The whole process is aimed to teach the terminology, which is then averred, though how is never explained or demonstrated, to promote better writing skills. The direction of current practice can be seen in the following quotation: “The child’s writing will often go beyond the point at which he has received grammatical training. Through discussion focused on what the child means … the teacher frequently gets the correct form from the child.” But if only from such discussion, what purpose has the grammar? This we are not told. Individual pronouncements on the value of grammar while less satisfactory to show any general climate may help to underline the contradictory and uncertain nature of our contemporary knowledge and practice in the teaching of English. To multiply examples would be tedious – two or three may suffice. Thus in 1921 S. E. Burton, a witness speaking to the Board of Education, states that “immense harm has been done by the well-meant discouragement of formal grammar in the elementary schools.” In the same year, Dr P. B. Ballard writes, “I have convinced myself that in the elementary school formal grammar:

a) fails to promote a general mental training,

b) does not enable the teacher to eradicate solecisms, c) does not aid in composition,

d) takes up time which could much more profitably be devoted to the study of

literature.”

In 1957, C. S. Bishop in a doctoral thesis on the philosophy of language suggests that “the premature (i.e. under 14) study of analytical grammar has probably done much harm to English teaching”, and recommends the gradual acquiring of a control of complex subordination by means of constructional exercises rather in the manner employed in the teaching of English to foreigners. But he states further that “in the course of his work on composition the teacher is bound to find faults of grammar which must be explained. They must be explained in grammatical terms…”. Or again, H. Blamires in 1951 (English in Education) complains that “Too many teachers are teaching grammar when their pupils should be learning to write or to read … As a direct aid to composition, the abstract grammar lesson teachers something that does not belong…”. But he adds elsewhere in the same book “a certain amount of grammatical knowledge… a minimum theoretical equipment for the writer of correct English (is) recognition of the parts of speech – familiarity with subject, object, verb, direct object, complement – some knowledge of the various phrases and clauses…”. The majority of textbooks, as one might imagine, follow the established body of opinion, and ignore the uncertainties and heterodox comment of some individual critics. An analysis of the grammatical contents of fifteen of the commonly used textbooks is given in Appendix II. An extreme example holding the mirror up to strict educational opinion may be seen in Our Living Language by J. H. G. Gratt and P. Gurray. Published in 1923 and reprinted in 1953 - so stable are some aspects of tradition - this book for children aged 14 to 16 sets out to rehabilitate grammar by effecting a close conjunction between study of the “forms and ordering of words” and “observation of the shades of thought and feeling which those words … signify.”

There is an authoritarian weight in the pronouncements defending grammar: “The study of style must be preceded … by conscious and accurate observation of grammatical constructions. There can be no dispute that the position of grammar in the curriculum is justified.” The book recognizes much more than did many earlier grammars the variety and flexibility of English, and its prescriptiveness is justified when possible by a genuine description of linguistic habits as they are. The attempt nevertheless to express the new flexibility in the old but here extended terminology leads to some confusing expressions, e.g. “a grammatical subject is a word (or group of words) which calls the receiver’s attention to something which the transmitter wishes him to think of” beside “the grammatical subject of ‘run away’ is a tone.” In addition, the complexity of the terminology far exceeds that which many children can be expected to master, in particular as the pupil is sometimes involved in the grammarians’ wrangle over terms (as to attribute, epithet, or qualifier, for example – p. 49). As an instance of the complications, there are 28 pages to deal with the concept of “case”, which for an almost uninflected language seems excessive at the elementary level. Mentioned are nominative, accusative, dative, genitive (or appertinent) – the last subjective, objective, possessive, defining, partitive, and its rivals (compound nouns, substantive + adjective – e.g. bank-clerk), external or adverbial cases (including “satellite association”, a term not envisaged in the 1922 report), the vocative, predicative, the instrumental, the agent case, and caseless nouns. It is not enough to say, as the authors do, that the pupil should “not imagine that in these or indeed any classifications, there are hard and fast lines” – and by this to expect that an over-elaborate and pedantic categorization has become effective and intelligible. The child, his needs, his actual growth and usage, is as lost and remote as ever. It is a far cry from Our Living Language to some of the simplified textbooks published for the use of the less intelligent or less linguistically gifted classes of children, but these books, too, more often than not, preserve a sizeable fraction of their space for the “elements of grammar” while often abandoning the honest enquiry and the struggle to categorize the refractory elements of language. As may be seen in the appendix mentioned, the average fraction given over to work devoted to the correct learning of grammatical terminology is roughly one fifth; and in the majority of cases the vocabulary of this terminology conforms strictly to the usage of 50 years ago. Into these traditional if somewhat ill-defined preserves inroads have been made in recent years, so that the climate of opinion is perhaps less settled than it has been in the past. Randoph Quirk, for example, in “The Teaching of English” (Studies in Communication 3, 1959) and in “English Language and the Structural Approach” sympathises with the teacher’s mistrust of the “old grammar”, and notes the lack in Great Britain of “realistic and functional teaching grammars”. He proposes the study of language “in linguistic terms” as a remedy [i.e. a prescriptiveness based on genuine description of observed forms], “for it is not that the Latin terms of the ‘old grammar’ do not suit English, it is that the non-linguistic frame of reference does not suit language.” He emphasizes the structured system of English, and the part played by the context of situation in aiding interpretation. That grammar is an important dimension of meaning is demonstrated by nonsense words put in structured patterns. Here, as with traditional grammar, the teacher has to be sure that what he is doing – making an intuitive grasp of structure into an intellectual and conscious one – is worthwhile as an end in itself, is within the power of schoolchildren, and is a means to a further valuable end. That, beyond its interest to people as linguists, any further end is served remains to be proved, though this interest is perhaps sufficient end. Clearly, the teacher cannot fail to welcome any attempt to classify with realism and honesty and present structure of the language; and the new grammar, working as it does often by the practical tasting of structures against frames of reference and much less by the bemused juggling with definitions, offer more hope than did the old of a direct and beneficial effect on writing skills. However, assumptions, such as seemed reasonable to people in their time but now are evidently false, cannot be expected to persuade teachers to jettison traditional material, albeit their own practice is founded on earlier and even acknowledge misapprehensions; and one awaits with interest some experiment showing not only that the “new grammar” is superior as a teaching instrument to the old, but is itself necessary. This experiment would answer the question, “does anyone other than the teacher need to know the names of signals and structures?” No-one has ever doubted that English has signals and structures; but it may be that these are best or only understood in class, at least in the younger classes of the secondary school, by means of direct and specific practice and habit formation, with no call to master the larger theoretical background. Thus one may demonstrate the structures of Quirk’s nonsense “sentence” by asking questions in terms of “lexical” meaning: “Plome the pleakful croatations ruggle polanians ungleshably in the pit.”

Question: What did the pleakful croatations do?

Answer: Ruggled.

Question: How did they do it?

Answer: Ungleshably.

School children will usually arrive easily at these answers, when the term “adverb” (or whatever substitute a new grammar may use) is wrapped in mist for them. In similar fashion, the concept of marked and unmarked members (e.g. “old” – unmarked; “young” – marked) is within the reach of non-terminological teaching, and indeed, especially in the matter of word-order and style, is usually taught by direct example in the context of literature or of the children’s own writing. But that the teacher needs to understand the generalizations of his examples if his work is to be systematic cannot be doubted, and it is to the teacher that the correct revitalization of English grammatical classifications is likely to be of fundamental value. They have to be taught first to the teacher, so that grammar at present is likely to be a matter for the training college rather than the school. Be that as it may, it is clear that opinion about the value of grammar in teaching English is, although not rigidly crystallized in its older form, far from the likelihood of being diverted to new and experimental forms in the general usage of the classroom. Traditional formal grammar, however sporadically attacked, survives in the curriculum largely through the nourishing of custom, and through the absence of any cogent answer to the problems of teaching children how to write correct English. Even so liberal an educationalist as I. A. Gordon, writing for the New Zealand Council for Educational Research in 1947 on the problem of grammar, requires, while he advocates direct working from children’s compositions rather than from a textbook on grammar, numerous technical terms to be known and used: the parts of speech; common and genitive, nominative and objective case; subject, predicate, indirect and direct object; transitive, intransitive, passive and various tense names; gerund, infinitive; even clause analysis. His programme is more rational, clear and economical than that of the conventional school grammar book; but it will be apparent from enquiries such as those discussed in Chapter II (b) below that it is at the least optimistic. But no generally accepted replacement for the habitual modes of teaching formal grammar has been found; and equally, no sufficiently convincing proof has been offered to teachers that the grammar of tradition is ineffective in the classroom. What are the comments of research on these last points? CHAPTER II (B)

THEMES OF PREVIOUS RESEARCH

Three main themes emerge from previous work related to the present undertaking. They are:

1. Can grammar be successfully taught to younger and middle school children

in the secondary school?

2. In so far as grammar can be taught, what is the aim of the teaching and how

well does the grammar fulfill its intended functions?

3. What are the really important parts of grammar for schoolchildren, and

what errors do children commonly make?

Theme I. Can grammar be successfully taught to younger and middle school children in the secondary school?

One of the satisfactions in taking a formal grammar lesson with a class of enthusiastic eleven-year-olds derives from the impression that positive and visible steps forward are being made in the acquisition of knowledge –especially if this impression is shared by the teacher and the taught. There are few branches of English teaching in which the facts to be learnt are so demonstrable, so apparently clear in classification, and so ready to be apprehended as the reward of diligent application, as those of formal grammar. Add to this the conviction that the facts so acquired sharpen and order the mind, halve the labour of learning a foreign language, lend subtlety and depth to the understanding of literature, ease the labour and double the effectiveness of marking, altogether serve to purify, to strengthen, and to safeguard the forms of language and the skills of the would-be writer, and it is clear that nothing, absolutely nothing, can expect to supplant grammar in the affections of the teacher or the pages of the English syllabus. Children themselves are more pragmatic, and do not always appreciate information of which they cannot at once perceive the use. In an enquiry of the pupils of ten forms of eleven and twelve-year-olds in five London schools the present writer found that of the five common English lessons (composition, plays, reading, grammar and poetry) grammar invariably was ranked last in popularity. Poetry was its close rival. This of course is no argument against grammar. Children do not like cod-liver oil, but it is good for them. Nevertheless, such an enquiry may hint the possibility that the satisfaction of the grammar lesson can be at times all on one side. And this may cause the experienced teacher to suspect the quality and efficiency of the learning. That grammar and its terms can be learnt by the average child has always been taken for granted, as will be obvious from the recommendations of the various authorities to whom reference has been made in the previous chapter. Some authorities comment that the learning may take rather longer for children who are less bright, but little doubt is ever encountered that the basic elements of the traditional terminology can be mastered by twelve or so. Without exception in the present writer’s experience these elements always include in the textbooks the parts of speech, and nowadays the child is always expected to realise that a word is given its title of noun, verb, and so on according to its function in a group. Yet the assumption that the parts of speech may be recognized safely at the early secondary school stage has been explored in research, and found false.

One of the more dramatic but none the less neglected of the enquiries into grammar in Great Britain is that entitled The Difficulty of Grammar by W. J. Macauley (1944). Macauley’s experiment concerned the courses of grammar in Scottish schools between the ages of seven and a half years and fifteen or sixteen. Children by the age of twelve in Scotland had had an extensive grammar course, studying the subject for 30 minutes each day. Thus at the age of seven and a half they learnt the noun, singular and plural, and the verb; at eight, adjectives; at eight and a half, personal pronouns and the tenses of verbs; at nine, analysis of simple sentences, conjugation of verbs, kinds and cases of nouns; at nine and a half, auxiliary verbs, and adverbs; at ten, prepositions, conjunctions, relative pronouns, and the interchange of phrases and clauses. The results of all this as tested by Macauley showed a complete lack of understanding at secondary school entry. Altogether, Macauley tested a thousand pupils in various grammatical points, such as identifying the parts of speech in sentences, requiring as a pass standard a score of 50%. At the end of the primary course outlined above, one child in 257 reached 50% in noun, adjective, verb, pronoun and adverb. Only the recognition of common nouns, not abstract, and of simple finite verbs following immediately on their subjects, were safe. By the age of fourteen, that is after six years’ grammar, 4 children out of 397 could score 50% in recognizing simple examples of the five parts of speech mentioned above. In the senior secondary school, only the top boys’ and girls’ class scored 50% on all five parts of speech. Thus at the end of a three years’ senior secondary course, recognition of the simplest parts of speech by their function is still too difficult for the median child. “Educators”, concludes Macauley, “are merely following a historical tradition.” The results of Macauley’s work, although surprising, were not incorporated into classroom practice, nor indeed followed up experimentally until 1957, when F. Cawley in assessing the difficulty of English grammar for pupils of secondary school age (Manchester University M. Ed., 1957) repeated and elaborated on Macauley’s test to 1008 pupils of secondary school age from a random sample of ten schools – one grammar, three secondary modern, three Roman Catholic voluntary schools, two Church of England voluntary schools and one “county all age school”. First and fourth year pupils were used. All ten schools agreed that grammar should be taught, the majority suggesting that between seven and eleven years of age was the correct time to begin instruction. Seven schools favoured the use of formal grammar teaching, two a mixture of incidental and formal work, and one an approach to grammar by incidental reference. Usually one period a week was given to the grammatical instruction. The grammar test was finally put into a battery of nine other tests: Raven’s Progressive Matrices, a General Ability Test, Mechanical Arithmetic, Problem Arithmetic, Spelling, English Usage, Capitalisation and Punctuation, Vocabulary, and Sentence Structure; and an essay was written. The essay was marked by four people, each giving marks for general impression, mechanics of English, spelling, and vocabulary/ style/ structure.

The grammar test was checked for reliability by the split-half method giving r = 0.922. No significant differences, where t exceeded 3, were found between the achievements of the different sexes in the grammar test, nor between the types of school with the exception of the grammar school compared with the others. There was a large difference in achievement in the test between the grammar school children and others: thus, the mean score by the first year grammar school children was 30.984 (SD 6.261), whereas by the first year children from the other schools combined it was 16.06 (SD 7.313) and by the fourth year children 17.671 (SD 7.347). The gain in achievement between the first and fourth year in the grammar school was from a mean score of 30.984 to one of 37.364 (SD 7.375) (t = 7.131). Between first and fourth year children in all other schools it was from a mean of 16.06 to one of 17.761 (t = 3.015). It is fairly clear from this that at least for all the children except those in the grammar school a plateau of achievement in formal grammar was reached quite early, and that a mean increase of 1.6 points in 50, although significant, is very small for the fruits of three years’ work in the secondary school on top of several years before reaching the secondary age. The grammar school scores were fairly high, starting already at eleven years of age considerably above the score finally reached by the other children. Thus Cawley, as Macauley previously, suggests that for all except the bright children no level of attainment likely to be valuable can be reached in formal grammar, and even bright children fail to reach anything like complete success in the fourth year in identifying five of the parts of speech. Few teachers, in the present writer’s experience, are willing to agree that their own teaching of grammar is as ineffectual as that of the teachers, except those in the grammar school, in the above enquiries.

It is difficult to see why, otherwise, a message so simple and clear could be so widely and consistently ignored. That the grammar question in the G. C. E. is usually easy for clever children, and offers a safe handful of marks, coupled to the fact that it is possible for children to retain the abstract knowledge required for a short period of time, may explain the perseverance of formal grammar in school, which is well seen in the more or less unanimity of the schools in Cawley’s work in teaching the subject. It is worth noting in passing that it is unlikely that it is the inherent difficulty alone of formal grammar which inhibits effective learning – [it is also the uselessness]. The present writer found an illustration of this in a discussion on December 8th 1959 with 16 teachers of English, all of more than two years’ experience. On December 1st the sentence “Thinking it would be late, the man ran to the house” was analysed in a passing comment to the same group of teachers, and at the second meeting a week later they were asked to analyse into clauses the sentence “Thinking it would be late, the man ran to the house where his friend lived.” This, of course, quite apart from the benefit of the preceding week’s practice, is a very much easier task than that normally set to G. C. E. Ordinary Level candidates. Nevertheless, only 4 of the 16 teachers managed to provide a correct answer; and among those who failed were one or two staunch defenders of the continuance of formal grammar teaching in school.

It is thus at least probable that a disincentive arises from the lack of a clear link between the material of formal grammar and the language situation. To establish such a link between the study of language and the control/ response situation is the main problem of language teaching, and is touched on in the conclusion to the present work. Macauley recommended that grammar should be taught no earlier than Sixteen years of age, and then only in the best classes. He does not show why it should be taught there at all; and he does not show that even if learnt it has no effect on correctness, but only that it is not remembered. Cawley similarly points out that in learning grammar, ability is at a premium, teaching at a discount, and that the ability to do well at a grammar test has far more affinity with verbal intelligence than with the ability to write good clear English. This means that the secondary modern schools will not be failing in their duty if, instead of concentrating upon formal grammar teaching, they concern themselves more with direct creative English expression. For grammar means analysis of functions, and most secondary modern pupils lack the ability to do this well.

A conclusion with which Robinson (1959), whose work on grammar and composition is referred to later in this chapter, concurs. This may conclude the review of the conclusions of research on the first theme, that formal grammar is for all practical purposes not teachable in the first years of the secondary school except perhaps to the brightest children.

Theme 2. In so far as grammar can be taught, what is the aim of the teaching and how well does the grammar fulfil its intended functions? Cawley, however, goes rather further than Macauley had done, and adds to his conclusion the view that (even for the brighter children) “what still remains doubtful is whether the children really need to be taught grammar.” This, which is really the second major theme of research, is not developed in Cawley’s work, but one or two pointers emerge. Thus, in the essay set to the children the marking was divided as follows: 20 marks were given for general impression; 10 for grammar and mechanics; 10 for spelling and vocabulary; and 20 for style. Not only did scores on the grammar test correlate more highly with those gained in the arithmetic and verbal intelligence tests than with those gained on the essay as a whole, but also than with those gained on the grammar section of the essay mark. “Factor analysis shows that [ability in grammar] has little connection with ability in essay writing.” The second theme of research in this field – to enquire whether grammar in so far as it can be taught fulfils a useful function in strengthening English skills or in other ways – has been treated more fully by other writers than by Cawley, but not yet conclusively. Cawley’s work is not of course conclusive in this respect. It pictures the results of grammar teaching as it is, not as it should be, and does not show whether these children would have done better or worse without their grammar lessons not only in the formal grammar test but in the essay. It is weak also in assessing the actual quality and possibly the quantity of the grammar teaching – an interview with the headmaster may not have given a sufficiently detailed and thorough picture of the classroom situation. It does not sufficiently question whether the mean gain in the grammar test by the grammar school children of 7 points represents any equivalent gain in a real value or skill closely related to the command of English. Finally, no sufficient check of the value of grammar in the schools could be made because no course of grammar was undertaken whereby the grammar taught in one school could be compared with that in another or even with that in another form in the same school; the same children were not involved in the first and the fourth year tests. These are weaknesses still apparent in other work which will be discussed below, but which it is hoped to overcome to some extent in the present work.

Adverse criticism of the effectiveness of formal grammar in developing

expressional skills or in providing a satisfactory form of mental discipline (that is, a discipline worthwhile not only in itself but in extending to other and related subjects a parallel control of the processes of thought) was made early in the century in America. It may be said to have culminated in 1936 in the report by M. R. Shuttuck and W. Barnes, entitled “The situation as regards English” in the Ninth Year Book N. E. A. Department of Supervisors and Directors of Instruction. Shuttuck and Barnes reported that their own and previous experimenters’ findings led them to believe that:

a) the disciplinary value of formal grammar was negligible;

b) no more relation existed between knowledge of formal grammar and the application of that knowledge in a functional language situation than existed between any two totally different and unrelated school subjects;

c) evidence did not support the value of knowledge of English grammar in achievement in a foreign language;

d) knowledge of grammar was of no value in reading skills;

e) the contribution of grammatical knowledge to the formation of sentences in

speech and writing has been exaggerated;

f) grammar was difficult if not impossible to teach to the point of practical

application.

The last point has already been exemplified in the work of Macauley and

Cawley. Of the others, items a), b) and e) are perhaps the most important for the present work. T. H. Briggs republished in 1913 (Teachers College Record, vol. XLV no. 4) his work on “Formal English Grammar as a Discipline”. By 1913 in America it was fairly commonly recognized that formal grammar had little functional value, but it was still greatly advocated for its general disciplinary value. This disciplinary effect was held to be most apparent in helping in the making of rules and definitions, in the selection of data and the testing of arguments. Briggs devised or borrowed 54 tests of the abilities noted in these claims for formal grammar. He had two classes of seventh grade (twelve-year-old) children, with 29 children in each class. These he taught for three 30-minute periods a week for three months, the one class having lessons in formal grammar and the other in general language, vocabulary work, and composition. The roles were then reversed for a further three months. The classes had very similar attainments and conditions of work, but the experiment was short and even with himself doing all the teaching he found that he taught one form better than the other. With a possible exception in ability to see likenesses and differences, no improvement due to formal grammar could be traced in the event. The values tested were held to be:

1) The testing of likeness and differences

2) the testing of differences critically

3) the amplification of a definition thoroughly

4) the making of a rule or definition

5) the testing of reasons

6) the selection of relevant data

7) the ability to demand new data before reaching a conclusion

8) the ability to reason in other fields

9) the ability to reason syllogistically

10) the ability to detect “catches”.

In all categories except no. 1, the non-grammar group showed gains over the grammar group. As the concept of grammar as an abstract discipline fell into disfavour, attention was drawn to the need for a more functional type of grammar, one that might help to transfer grammatical abilities into skill in composition and expression. The amount of technical grammar was reduced greatly in school courses in America, where in 1936 the Curriculum Commission of the National Council of Teachers of English recommended that “all teaching of grammar separate from the manipulation of sentences be discontinued.” In fact, this recommendation was never implemented, and furthermore the trend of sympathy towards it was never felt so strongly in England as in America. Indeed, with the word “functional” grammar regained its status as it had never lost its place, and was propounded as a tool. This in England introduced many further complexities for the child to master, and many inconsistencies which however ill at ease in a pure science could in an applied one be allowed to pass. The downfall of formal grammar as a discipline has thus in practice rather strengthened than weakened the position of grammatical terminology in the classroom.

A fair amount of work had been done as early as 1906 in America on the

relationship between knowledge of formal grammar and the ability to write or to interpret correctly. In that year F. S. Hoyt published “The Place of Grammar in the Elementary Curriculum” (Teachers College Record VII, Nov. 1906). He devised three tests, in grammar, composition, and the interpretation of a poem, and administered these tests to 200 ninth grade pupils. The correlation co-efficients found between grammar and composition achievement were 0.18; between grammar and interpretation 0.21; between interpretation and composition 0.18; all these being positive. Hoyt draws the conclusion that “the same relationship exists between grammar and composition as between two totally different subjects.” Other workers, such as Rapeer, found similar or even lower correlations. In Rapeer’s case, the grammar/compsition co-efficient was 0.23, and grammar/interpretation 0.10. Segal and Barr gave tests comparing achievement in formal grammar and in applied grammar - roughly equivalent to the British “functional” grammar. There were 104 cases, and average scores suggested that “no more relationship exists between these two sorts of grammar than there is on the average between any two of the high-school subjects of any curriculum” – the figure was however 0.48, a higher figure than those previously mentioned, which is not unexpected in view of the similarity of the terms of formal and applied grammar. Many echoes of these early experiments are to be found in the work of Robinson on the relation between knowledge of English grammar and ability in English composition (1959). She administered tests of grammatical knowledge, of general ability and of skill in composition to 145 second year and fourth year pupils in four Manchester grammar schools. The mean intelligence quotient of her sample was 120.2, SD 10.33, and like the bright children in Macauley’s and Cawley’s work these pupils did manage to make progress in learning grammar. The measurement of general ability was taken from scores made by the children in the Manchester General Ability Test (Senior) 1, a 45-minute test of verbal intelligence for children of thirteen years of age. Both twelve and fourteen-year-old children took it. Their grammatical knowledge was assessed by means of 8 ten-minute tests, 7 directed at the parts of speech and one at the analysis of sentences, the reliability of these tests being judged by the split-half method, giving a reliability co-efficient of 0.963. However, since the items in the tests were not arranged in order of difficulty, the validity of the split-half method of assessment is questionable. In Composition, the children wrote three essays of 30 minutes each on descriptive, reflective, or expository themes. These essays were scored by three examiners whose total marks were added to give a figure for General Impression; and one examiner counted the errors in punctuation, grammar and spelling in the first 200 words, and, for a third measure, recounted the grammatical errors only. Co-efficients of correlation were calculated between the scores on the grammar tests and ability in composition as judged by the General Impression marks (giving a figure of 0.572); between scores in the General Ability test and in the grammar test (with r = 0.459); between the grammar test score and the total error scores, that is, the errors in punctuation, spelling, and grammar combined (r = 0.493); and between the grammar test score and the grammatical errors only (giving r = 0.307). A significant difference was established, by means of the analysis of variance, between the performance of the second year and fourth year children as a whole, both in their performance as essay writers as marked for General Impression, and in their knowledge of grammar. But more germane to the present argument is the fact that although there were differences between the schools in Robinson’s work – differences probably due mainly to the teachers’ attitude and skill – these do not affect the evidence that the relationship between knowledge of grammar and ability in composition was not very close. The figure for r of 0.572, higher as it is than those reported by earlier workers (dare one guess that this may be because the grammar of 1950 was more liberal than that of 1920), is not high enough to warrant any great degree of association. It is particularly noteworthy that the lowest correlation was found where one might have expected the highest had the teaching of grammar been

affecting writing skills: in grammatical accuracy, the second error count, the coefficient sank to 0.307. Robinson’s conclusions are expressed in these terms:

There is no evidence … that pupils from schools where there is a higher degree of association between knowledge of grammar and general ability, or pupils from schools where there is a significantly different level of achievement in grammar, obtain as a group, scores in composition which differ significantly from those obtained by pupils from schools with neither of these characteristics. Knowledge of grammar is unlikely to lead to correct usage. Valuable as all this may be, it is open to objections similar to those raised on earlier pages in reviewing Cawley’s work as an argument against the efficiency of grammar. The static nature of the correlation analysis could justify the grammar teacher in laying little emphasis on the conclusions drawn. The correlation itself of 0.572 between grammatical knowledge and the General Impression scores in composition, although not high, is not totally negligible; nor is the low figure for the relationship of grammatical knowledge and grammatical accuracy (0.307) necessarily of vital importance. A defence of grammar in class could still perfectly well argue that it is the dynamic process that matters, the long hard study which later bears fruit. The critical question of a comparison of methods is not attempted; thus there is no way of knowing whether the children would have done better or worse without their grammar, or what influence if any the actual process of study in a long teach-test teach experiment would have had on their control of language.

The absence of any general transfer of reasoning ability to other branches of study reported in Briggs’ and others’ work is presumably not surprising if even within the framework of English skills no specific transfer is demonstrable. Nevertheless, since the feeling that formal grammar should be of general mental value lingers on, it may be worth-while suggesting why there is no general transfer of training. This would seem to be because of a combination of three factors. F. A. Cavaneagh, in the “Report of the Committee on Formal Training” made with C. Burt, R. L. Archer and T. H. Pear for the British Association in 1930, suggested that any subject studied so as to create habits of perseverance, application of previous knowledge, independent attack on problems, is valuable; studied in the opposite way, is not. This, however, is in itself not a practical generalization for education. It is not in practice possible to study any and every subject in this beneficial way. As H. R. Hamley pointed out in “Formal training: a critical survey of experimental work” (B. J. Ed. Psy. 1936), “the conative and affective aspect of learning” – i.e. the interest of the learner – “is too little considered.” Children do not usually like the formal grammar lesson. As Hamley further indicated, the stimulus in a cognitive activity has to allied to a need, which determines whether the stimulus will be accepted or not. The need to learn formal grammar, although often asserted by the teacher, is not felt by the child, who fails to find sufficient links that he can understand between formal grammar and skill in composition. Finally, as between formal grammar and composition there are few common usable elements, so to a greater degree the specificity of grammatical relationships hinders grammar from being of any general value. These relationships may possibly be freed in the higher grammatical study of the type proposed by P. E. Bartlett’s work (see page 83), but this is far beyond the powers of school children, in particular because the conventional exposition of the terms of grammar in school is in itself often illogical and ill-disciplined. Although all this may have been generally suspected in the United States by 1930, grammar was still highly esteemed in the correction of common errors. Yet in H. N. Rivlin’s “Functional Grammar” (Contributions to Education no. 435, 1930, Teachers College, Columbia University) Rivlin mentions that in 530 periods of English teaching he found that grammatical nomenclature was not once used to elucidate an author’s meaning, let alone to criticize style. In 502 lessons (that is, excluding 28 grammar lessons), only seven grammatical items received ten or more mentions; these were:

-agreement of verb with subject

-use of don’t for doesn’t

-wrong tense or sequence

-unity and coherence of sentence

-omission of comma after adverbial clause at beginning of sentence

-wrong preposition

-double negative

Rivlin nevertheless was of the opinion that grammar ought to be taught, and principally as an aid in composition. He proposed to make this possible by the elimination of all superfluous material from the grammar course, and this aspect of his research is discussed below.

Certainly grammar still persisted in the classroom. This is not at all surprising in Britain, since American work has in this subject had a negligible effect on the practice of English teachers. Indeed, it has seemed to the present writer in conversation with teachers of English to be almost entirely unknown. The allocation of space in English textbooks remains based more on the complexity of a grammatical point under discussion than on its importance. Practising teachers do not easily find time and opportunity to consult foreign works, and rely more on their own experience in a particular school and locality to indicate the relative importance and order of the matter in the syllabus. Where teachers are not for one reason or another ignorant of foreign studies, they may feel reluctant to place much weight on conclusions reached

with children who after all are not English children, and who speak only a similar, not the same language, in an environment very different from that of Great Britain. If the Americanness of American research has militated against the acceptance in Great Britain of observations even though based on large samples, another factor has strengthened the tacit rejection of the inferences to be drawn from experimental work casting doubt on the worth-whileness of formal grammar as an educational subject. As a general rule, American experiments have been, although large in sample, of short duration. Formal grammar is a long-established and traditional subject; its virtue is often felt to be that of a homeopathic medicine, of which a little taken every day will in time cure all the ills of language, of mental disorderliness, and of undisciplined

character. It is not to be expected that people cherishing this feeling will or should be ready to accept without skepticism the criticisms of a short period of testing or of correlation analysis. Even the results of such an enquiry as Macauley’s do little to weaken this position. That after five years the children have not learnt formal grammar will certainly show that they cannot be applying it to polish and refine their English structures, but it will not show that in the struggle to apply it and to master it they may not inadvertently have benefited. It was perhaps with this in mind that F. M. Symonds investigated the effect of “Practice versus Grammar in the Learning of Correct English Usage” (Journal of Educ. Psych. XXII, no.2, pp. 81-95, Feb 1931, Teachers College, Columbia University). He comments adversely on the static nature of correlation analysis, and suggests that teachers might prefer to trust an experimental test of the value of grammar. He therefore conducted test-teach-test experiments to determine what influence learning grammar had on usage. He used Grade 6 pupils in four New York City elementary schools. The groups were not strictly equated (Symonds experienced here the same practical difficulties as the present writer – forms could not be disturbed) but after correlating mental age and IQ scores on a first and final test, the author concluded that “the fact that classes differed in initial ability had no appreciable influence on the results.” As the constitution of the groups was similarly predetermined in the present work, the conclusion encourages the belief that a similar lack of exact equation was of minor importance here. A test, consisting of 40 sentences having a grammatical error to be corrected by the pupil’s rewriting each sentence, was given initially and finally to the groups. In the teaching between tests, six experimental procedures were tried: a) simple repetition of correct forms

b) repetition of correct beside incorrect forms

c) learning of rules, definitions, and principles of grammar, with some examples, taught for 15 to 20 minutes a day for two weeks (or more)

d) exercises in recognition and naming of grammatical constructions, e.g.

“Most boats carry life preservers”

“Most” is an ---

and modifies ---

e) practice in choosing correct constructions

f) the whole programme.

Procedures (b) and (f), fairly closely followed by (e), gave the best results. Improvement was registered to a lesser degree in all procedures, and brighter children found more value in merely learning rules than did duller ones.

The author concludes that “direct attacks on usage” are certainly more

profitable than the learning of grammar. Grammar is “the summary or epitome of the usage which has already been learned directly” rather than “a tool for guiding the learning of correct usage.” Such an experiment as this gives some prospect of acceptance to the theme of the main conclusion of the present work, that grammar has little or no value in teaching children to write correctly.

The author’s suggestion that a dynamic experimental method of enquiry be used to test the idea is valuable in carrying conviction to teachers who have always so far tended to ignore doubts expressed as to grammar’s utility. Nevertheless, Symonds’ work is not entirely convincing to the practising teacher. A strong attempt was made to control variables in the material of the teaching/course by limiting its content and by precise instructions to teachers (e.g. rules were taught, but not discussed); and in this respect the present work is less exact. This limitation has been accepted however for what may seem a different reason: the test of the effect of grammar cannot be made after a fortnight of 15 minutes’ a day instruction; the degree of change and the stability of the results are too uncertain. The methods of instruction used by Symonds were too formal and desiccated to be applicable in a true, natural, and sustained teaching situation, in short could not adequately test the teaching of grammar. No test was made of the available skills of the children as shown in normal, continuous writing; and with these deficiencies in mind, the variables still remaining - as for example the incalculable differences of teaching-skill and of class relationships – are to be considered. In the opinion of the present writer, such variables cannot be excluded from this type of experimental procedure; they can only hope to be limited and to be overborne by a larger element of difference between the experimental and control groups. It is essential to give to the experimental procedure a sufficient duration and sufficiently normal atmosphere to offer a genuine measure of the quality being tested. It may be that even the two years of the present experiment 0 and those not easy to request of schools – are none too long: but they are, it is hoped, sufficient to overcome the main disadvantages indicated above.

Teachers and educational authorities in Britain, if they knew of the

conclusions reached by Symonds and others, were not affected by them to the extent of seriously modifying classroom practice; and research in England has been devoted rather to the wish to rehabilitate grammar than to dispense with it. Thus in 1936 V. M. Brown produced “An experimental study of a new method of grammar teaching and its bearing on composition” (M.A., London 1936) which was an attempt to prove that grammar trains expression “because words have a dynamic as well as a functional value”. Three groups of children were used, all in their first year at the secondary school. One group was taught the new grammar, another an older formal grammar, and the third no grammar at all. Tests were set in composition and language at the beginning and end of the year. The author examines claims still made in 1936 for formal grammar – that it disciplines the mind, prepares for the study of foreign languages, gives command of a useful terminology, enables one to write better English, and aids in the interpretation of literature, and so on. She discards four out of five of these as false or doubtful, and suggests that the instruction is ineffective because definitions are too rapidly reached to be assimilated, so that children have to rely upon dogmatic assertion instead of upon critical reasoning. No real link, she asserts, has been established between functional grammar or grammar as a pure science and skill in writing; and her aim is to rectify this fault. Even in “functional” grammar, the direction of work is always towards the terminology, not towards an understanding of what the author calls the “dynamic” value of grammar. It is not easy to follow what is meant by the term

“dynamic” here, unless it means - as would seem to be the case if one judges from the examples of linking in action – that the children should learn how to choose the most apt and suitable word for a given context. This is of course an aim of all study of language and of style. Grammar as a distinct subject, a science on its own, is dismissed as unacceptable to children, and no expectation of transfer of training is looked for from its study. In the results, the new grammar group made better progress than the old, and both did better than the non-grammar group. It is at first sight surprising, in view of the results of the present work, that of Brown’s three groups the one having no grammar at all made no improvement in the course of a year. One would have expected some progress to be made as a consequence of the normal English lesson. That this has not happened might suggest that, as has been found in other fields, any group which is deliberately ignored will feel this as a slight and will react by making less effort than an experimental group engaged even in an unpromising line of study. For this reason, it is doubtful if this “control” group did in fact serve such a function. In the present work, it was carefully explained to the non-grammar group that they were doing something positive in place of grammar, and not simply dropping grammar. Clearly, one cannot expect progress simply from a vacuum. As might be expected, the group learning by the new method made greater improvement than that engaged in the older and more rigid formal grammar. Brown’s method of teaching by play, by the stimulation of curiosity, abandoning definitions, diagramming, and formal exercises is quite in line with that found to be effective by the teachers of the non-grammar forms in the present work. Such a method is likely when applicable to lead to a quicker and livelier grasp of the names of the parts of speech – as indeed of any other topic. The basic grammar material studied by the “new” form was: the parts of speech; sentences; transitive and intransitive verbs; finite and non-finite verbs. The pattern of the lesson was a) the learning of the name of the part of speech, very often by some ingenious play or puzzle device; b) practice in the right stylistic use of this part of speech; c) a grammar

test that the names had been learnt.

There is every probability, one would say, that the essential part of this

procedure is (b) – the practice in context of the search for the right word (e.g. in the sentence “The soldier was putting on his overcoat”, the children were asked if “overcoat” was the right word. This was an exercise on nouns). This is a stylistic and semantic problem, at the heart of progress in the control of English. The grammatical term was taught side by side with this practice, and indeed a great deal of time and skill was spent on teaching it; but that any link existed was far from evident. A long exposition of the complement, for example, was given, but no evidence appeared either that the children could understand it, or that if understood it would enable to express themselves “purely” and “readily”. In treating the sentence concept, the author advises the teacher not to define the sentence but to handle it. Such exercises are given, and only such, as would be proposed by a teacher not using grammatical terms at all. Thus for example the children are asked to write the conversation of children aged from four to six years, and then of others aged ten to twelve, and in doing this to see the growth of connected thought, and the advantages of clearer construction. It is at least arguable that had the whole time been devoted to this type of work, the form would have made still greater progress. “Side by side” is not necessarily together: that it is supposed to effect a link is the general misapprehension of the compilers of nearly all the textbooks listed in Appendix 2 of the present work.

A further aim of V. M. Brown’s method, that grammar should show children that material can be classified and experience generalized, revives the earlier and discarded idea that a transfer exists between formal grammar and general mental discipline. It would seem, however, that more direct methods, and more immediately accessible experience, can be better used to provide this demonstration than the terminology of formal grammar. Some such material is mentioned at the conclusion of the present work, and was used in certain of the non-grammar work done during the present experiment. In 1951 E. G. S. Evans reintroduced the idea that a formal approach to the teaching of English, one including a fair proportion of traditional grammar, was the most effective. His thesis An Experimental Study of Methods of Teaching English in Secondary Modern Schools (Birmingham PhD, 1951) involved a comparison of three methods: the project, formal, and oral methods. One thousand and 457 pupils aged eleven to thirteen, from 36 classes in six Birmingham secondary modern schools, took part in the work. These pupils were given English attainment tests, a general information test, an attitude questionnaire, and a group test of intelligence. The tests used were mainly Schonell’s, and included spelling, capitalization, punctuation and vocabulary, and a test of sentence structure by combining sentences into one whole. Written expression was measured by four compositions of 30 minutes each, marked by a schedule giving 12 marks for content, 7 for sentence structure and paragraphing, and 6 for spelling and grammar. The results of the experiment were interpreted by analysis of co-variance.The children were taught for six months, and included in the formal course was a fairly large proportion of grammar teaching, comprising the traditional recognition of the parts of speech in a “functional” way, and sentence analysis. From preliminary enquiries made of 100 teachers, Evans found that such formal instruction was the normal thing in most schools, and that grammatical drills and exercises, parts of speech, subject and predicate and so on figured largely, especially in the work of the upper streams of the secondary modern schools. Evans notes the uncontrolled factors which must be a feature of any work of this sort – the varying levels of culture, intelligence, and achievement of the pupils, of skill and enthusiasm in the teachers, of general atmosphere in the schools, and of extra-curricular activities or home background. An effort to impose some limits upon the procedure of the experiment was made by providing weekly lesson assignments and by limiting the teaching method in each group strictly to the particular style adopted – formal, project, or oral work. The general conclusion is reached that there is “some evidence that regular systematic exercises and drills in capitalization, punctuation, sentence structure and composition writing resulted in a higher level of achievement than that obtained by methods based on activity and interest.” Thus the average number of words written by the formally taught children in 30 minutes was 209, as against 159 by the oral group and 96 by the project group. The formal group also used a higher proportion of complex sentences than the others – 31 out of 100, as against 17 and 14 respectively. It has been generally agreed that fluency (to some extent) and the ability to use

complex sentences correctly are marks of maturation in the management of English, so that there would seem to be in Evans’s result some signs that the grammatical approach judged by its results deserves its high and habitual place in the English syllabus. However, on closer examination of the experiment, increasing doubts make themselves felt. The first weakness of the experiment would seem to lie in the apportionment of the teaching work among the 36 teachers involved. Of these, 11 were experienced teachers, 12 inexperienced but normally trained, and 13 emergency trained. Of the 11 experienced teachers, 7 chose to use the formal method, and 4 the oral; one taught by the project method. Such a weighting of experiences must one feels have swayed the results; and indeed Evans notes that the least effective method was the project – the one in which no experienced teachers participated. A second critical weakness would seem to lie in the rigid separation of the one method of

teaching from another, although this may have been essential to the design of the experiment. In the real classroom situation, a mixing of methods is always apparent, and indeed inevitable if a response is to be evoked from the varied modes of understanding of the children. It is not perhaps surprising that children practising entirely by oral methods should fail to make satisfactory progress in writing correctly, or that children concentrating on describing a limited project should lack fluency when asked to write a general essay. The teaching by the three methods continued over a period of six months, and the author notes that “other conditions being satisfied … the duration of the experiment is probably more important than increasing the number of schools and classes”. It is possible that a greater superiority for the formally taught groups might have appeared if the six months had been extended to a year or two, although of course one cannot be sure of this – in fact, six months is too little time for the slow process of a scheme of education to ripen. But if such had been the outcome of a longer spell of formal work, it would have been only what one might have foreseen, bearing in mind the unbalanced nature of the teachers’ qualifications.

In the present work, this type of criticism has been met to some degree by ensuring that in the same school a grammar and a non-grammar class were taught by the one teacher over a period of two years. The teachers, though few, were all very experienced and most were highly qualified as English specialists. Up to the present, therefore, one can only say that the efficacy of the formal method which incorporates grammatical terminology and instruction remains an uncertain quantity, and it remains true to say that there is little positive and no entirely satisfactory experimental evidence to support it. It rests, as before, on custom and tradition. Although rarely, instruction in formal grammar has been advocated as an aid in the criticism and appreciation of literature; P. E. Bartlett’s thesis on The Influence of Learning Grammar on the Comprehension of Modern English Prose (M.A., London 1953) is an instance of this. The thesis is an enquiry whether it is true that parts of speech and syntax contribute overtones of meaning to linguistic expression that need to be recognized as elements in its comprehension. Four passages from modern novels were given with comprehension questions to pupils and students who had finished a four or five year course in English grammar. The passages were from Virginia Woolf, James Joyce, Henry James, and Ivy Compton-Burnett. In addition, numerous grammar textbooks published between 1940 and 1950 were examined and commented on.

This thesis of Bartlett’s is so persuasive a plea for the value of formal grammar in facilitating and enriching understanding of the full meaning of complex modern prose that one immediately hopes to find in it a sufficient justification for the years spent in studying the subject at school. There is little doubt that anyone who has the same grasp of grammatical detail and terminology as Bartlett will find in it one means of expressing a part of his apprehension of a literary passage, and especially of one in which a complex or unconventional use of structure underlines the significance of a particular contribution of form to total meaning. The level of the criticism of the pieces from Joyce, James, Woolf and Compton-Burnett, as of the pieces themselves, is far above that approachable by the pre-G.C.E. child, and indeed part of the tests was abandoned by a “general” sixth form in Bartlett’s work. It is thus an academic question whether work of this type is valuable in school below any but a good sixth form; in fact, in the work, Training College students did worse than sixth formers. Its possibility is encouraging for the idea that the proper place for the study of grammar is in the highest reaches of the school, where the rudiments and general knowledge can be mastered quickly and the intricate detail put to some practical effect, critical rather than creative. This idea is not, unfortunately, firmly supported by the thesis, although many valid points are made on the way as to the most fruitful approach to grammatical function. In the work, for example, the assumption is repeatedly seen that knowledge of grammatical terms would help in understanding complex structures – the delayed subject, for instance, very common in Henry James. The fullest answers in the tests were given by those who could note structure. But the effect of a reorganization of sentence structure can be felt and recognized by attentive readers or by listeners in the theatre as part of the meaning without any need to be able to name the structures – can be held, in short, by the intuitive grasp of language which comes by long acquaintance and an alert and sympathetic intelligence. This grasp is immediate; and furthermore, it necessarily precedes the “full answer” in grammatical terms. How, for instance, can the “delayed subject” be found unless already the meaning of the sentence, and hence the abnormal word-order, is recognized? Certainly, ability to generalize this apprehension of a particular passage into grammatical terms may be of value to fixing a literary device in one’s mind, as it would certainly facilitate a parody or a pastiche of such styles as depend in large measure on the distortion of the normal word-order; but basically it is constant familiarization with tone, sense, feeling and intention (to use I. A. Richards’ terms for the components of meaning) rather than the knowledge of the names of forms, that matters most. It is not improbable that, in a mind insufficiently mature to seize intuitively on the meaning of a passage, the full description in grammatical terms may be a substitute for the experience which it is supposed to represent; as in classroom patterns which are more complicated than is called for by the learners’ stage of maturation. Such imitation, and such description, can be a mere parrot-like disguise for actual ignorance, and out of keeping with an organic understanding of language.

In P. E. Bartlett’s work this possibility was not seriously to the fore, since the grammatical knowledge required was often lacking in the people taking part. Thus only 9 out of 47 were able to answer a conventional grammar question on the Henry James passage, and not only did the results of a grammar test show “that comparatively few people had learnt grammar with a view to relating it to work in expression” but even by these few only the simple recognition sections of the grammar test were well done. That it was however possible both to understand and to express understanding of a passage without resorting to grammatical terminology was evident, for example, in the answers to the Compton-Burnett passage. There was no significant correlation between scores in the grammar test and scores in the comprehension questions (r = +0.24, +0.20, +0.21, and -0.06 in the four passages). Nevertheless, although the students’ work was not convincing, it was clear from Bartlett’s own critical discussion of the four passages for comprehension that grammatical terms could be used, even if they may not be essential, to discuss prose style. This is especially so if the general effect of a passage is understood before critical discussion begins – indeed, Bartlett accepts Piaget’s statement that the whole is understood before the parts are analysed. If grammar is to be taught, much of Bartlett’s cautionary advice is of great moment. Thus, “Truly descriptive grammar begins where language itself begins – with the relations between man, things and customs in actual life. The detailed study of language must be centred less on isolated terms and more on what the forms represent …”. And again, “The implications of word patterns cannot be judged apart from the meanings of the words that compose them.” “Very few writers have come to terms with the context-of-situation-theory of language as outlined by Malinowski.”

In the present work, it is suggested that many of the aims of P. E. Bartlett may be realized in a method of teaching based on situation, and abandoning for pupils the grammatical terminology, to replace it by the use of their native skill in practising speech patterns relevant to the control of or response to social situations. Valuable to show the present condition of the English textbook, and linking in this to the notes in Appendix 2 of the present work, is Bartlett’s list with occasional comment of a number of textbooks published between 1940 and 1950. This shows that many books use vague and misleading expressions in an attempt to make grammar easy, and that many fail to distinguish between words and things. They express a definite hope that grammar will improve composition, shape thought, and improve normal word order. Only nine of the books mentioned grammar as an aid to literary interpretation, and of these nine, three exemplified the principle by asking for nouns in a poem to be parsed. Most of the textbooks examined as a by-product of the present work conform to the same pattern of authoritarian chaos, failing to distinguish the relative importance of different aspects of formal grammar, and usually mentioning though quite failing to demonstrate the link between formal grammar and composition or general control of language. Needless to say, these strictures apply only to the treatment of grammar – many of the books contain admirable material of various types. One may mention here a few of the better and the worse books: thus, one contained 180 pages on formal grammar, to 90 other pages; another gave 60 pages of grammar, 10 of punctuation, 6 of spelling, and 30 on style, vocabulary, syntax and composition together. On the other hand, some – as for example Moon and McKay’s An English Highway (Longman) - distinguished between things and words, or made a real effort to select and grade grammatical knowledge, as for example Patterns in English by W. H. Mittins (Allen & Unwin). Although the case for the retention of traditional formal English grammar is at best unproven, it is necessary to say that there are teachers who are aware of this and who would welcome the introduction of a totally revised and revitalized form of grammar in the classroom. One may for this reason conclude this section of the argument with a note on The Structure of English by C. C. Fries (Longman, 1957).

The work of Fries, although not directly testing the value of traditional formal grammar, is of moment here as a development from the assumption that the traditional terminology has little justification since it presents too obscure and confused a picture of the true structure of English. Fries’s approach is a behaviourist one, seeking evidence of structure as external as possible, without taking intention or meaning into account. One should, I feel, keep in mind when considering this approach the words of Jespersen in The Philosophy of Grammar: “It should be the grammarian’s task always to keep the two things (form and meaning) in his mind, for sound and signification, form and function, are inseparable in the life of language.” Other grammarians before Fries have disagreed, however – Sweet, for example, (NEG I, 204) says that it is “not only possible, but desirable, to treat form and meaning separately – at least, to some extent.” The present writer’s doubts on this matter, which is indeed one of the most important considerations for the shaping of a language course for children, are further expressed in Appendix 4.

The Structure of English is based on 50 hours’ recorded telephone

conversation. Sentences are classified in it by their effect: the utterance getting an oral response, a response of action, or simply one of continued attention. In the first category come greetings and calls, and above all questions. Commands fall under the second heading, and under the third come statements with such signs of attention as grunts, “yes” and so on. Fries notes also such non-communicative utterances as exclamations of surprise pain, even laughter – these are not grammatically important. He finds in the end his “minimum free utterance” – “a linguistic unit independent by structure of any other linguistic unit”. He points out the tautological nature of the teacher’s instructing the child in difficult terms for something he already knows. Most of our grammatical categories, he explains, depend on meaning – but we do not ask how the structures we use communicate meaning to us. He thus emphasizes the distinction between lexical and structural meaning, and stresses that grammar should be a study of the signals of structural meaning. Thus word order, or variations in the forms of words, are signals; certain kinds of words, or parts of speech, make significant patterns, and other words, function words, are means of arranging them. Fries categorises four parts of speech, very large classes containing 93% of the total words used. Function words although comprising only 7% of the total, are used often, occupying about one third of the speaking time. Structural meaning is conveyed by the shapes of words, the patterns of sentences, and function words combined, and Fries seeks to eliminate lexical meaning by for example using nonsense syllables. The four parts of speech are identified by inserting words into frames of reference:

-The good concert was good.

-The clerk remembered the tax suddenly.

In these frames, “concert” is a Class 1 word; “was” and “remembered” are Class 2 words; “good” belongs to Class 3; and “suddenly” (though not anything normally called an adverb of degree) belongs to Class 4. Only 154 function words are found in all the material, and these have to be individually known. They are classified by letters:

a) the, a, no, few, much, etc

b) auxiliary words (may, etc)

c) not

d) adverbs of degree

e) co-ordinating conjunctions

f) prepositions

g) do (in eg, “Do the boys do homework?”)

h) introductory “there”

i) question words (when, why, etc)

j) conjunctions

k) well, oh, now, why (but only at the beginning of response utterances, when they are sequence signals)

l) yes, no

m) may, look, listen (call words)

n) please

o) let’s…

Teaching on the basis of Fries’s work is now being attempted in some English schools, and the results should be of considerable interest to all those concerned to bring about a revival of grammar. It would seem probable that if a consistent and logical description of English could be formulated, which was at the same time not too difficult for schoolchildren to assimilate, there might be some hope of an effective transfer of training both of a general and of a particular nature. However, many obstacles remain to be overcome before this can be established. Teachers themselves have as yet very little understanding of or even acquaintance with Fries’s formulations. It is uncertain whether the attempt to divorce lexical from structural meaning will attract children whose main concern with language is perhaps not linguistic but rather the immediate desire and need to manipulate the extra-linguistic world. The wish to clarify and focus attention on pure structural signals may in other ways be inappropriate to school practice. Understanding is largely effected as an apprehension of total meaning before analysis. When ambiguity occurs, it is resolved and explained by reference to meaning and context. Thus, in a sentence such as: “The second man looked harder than the first” a Fries explanation of the ambiguity might be that “hard” can be a Class 3 or a Class 4

word. This is however very much the same as to say that “harder” could be a comparative adjective or a comparative adverb. Neither explanation is so helpful to children as the reference to meaning: “Did the man seem to be harder? Or was he looking with more concentration? Which do you mean?” Even the nonsense sentences sometimes used to demonstrate the force of signals are not really nonsense: signals depend for effectiveness upon a whole web of contexts, upon meanings related so often that the relationship comes to affect the meaning, so that when we see nonsense words we look in fact upon the shadows of other meaningful words which could take meaningful forms. The study of structure in isolation is for such reasons likely, as is the critical reassessment of formal traditional grammar, to be best attempted at the top of the school, rather than in the general work of the lower school. With reference to the present work, it would appear that no grammatical picture, however exact and teachable, will be necessary to teach children to write correctly, since such correctness is established by the habit of imitation, by analogical extension, and errors are not felt as important unless either the break with a convention offends a group in which the children wish to mix, or meaning is obscured.

Theme 3. What are the really important parts of grammar for schoolchildren, and what errors do children commonly make? We come therefore to consider the third theme in which research has been of relevance to the present discussion. What are felt to be the really valuable elements for children in formal grammar; and equally, what are the major errors which children make in writing, and the major difficulties they experience in achieving technical correctness? Surprisingly little attempt has been made to decide what measure of agreement exists as to which items of grammar are most relevant to the major difficulties experienced and errors actually made by pupils. An important move in this direction was however made in 1930 by H. N. Rivlin in the work mentioned above. He rated, but by experts’ estimates, and not by the study of children’s actual achievements or shortcomings, the importance of numerous grammatical points. Ratings of 0 to 3 were assigned to the items of grammar by four experts. A rating of less than 1 “means that the item is so unimportant as not to merit general class teaching”. Thus, if it is taken that a rating of 2 or more is “important”, it is found that 18 out of 157 items having functional value receive this rating from each of the four experts: proper noun

concept of verb: predicate verb, object of verb, irregular verbs, tense conjunction: taught with sentence structure, not with parts of speech simple, compound, complex sentence analysis of sentence – but informal (it is not clear what informal means here) complex subject, complex predicate subject, and predicate position phrase, clause and subordinate clause. Many items are still found in English textbooks which were felt by Rivlin to have no functional value, e.g. noun as substantive, abstract noun, person and case of noun, participial phrases, principal clause. The rated list of grammatical items in Rivlin’s work could be a most valuable aid for anyone wishing to recast the grammar syllabus of a school, or to decide which structures teachers must keep in mind to teach by non-terminological devices. (“There is”, comments Rivlin, “no objective evidence to prove that functional grammar does influence the pupil’s ability to read and write correct English.”) The relatively small number of items rated as “important” is an indication of the economies which could be effected in classroom teaching even by those teachers who would retain formal grammar.

It fits well with pointers from other work, such as that recorded by E. L. Thorndike and others in “An inventory of English constructions with measures of their importance” (Teachers Coll. Record, Geb. 1927, vol. XXVIII no. 6). This work lists 438 constructions and measures their relative importance by scoring the frequency (the number of recurrences of any particular construction to 100,000 statements in ordinary word order) and the range (the variety of material in which the construction is found, the total possible in the work under consideration being 45). The following list shows the comparatively few constructions that need to be mastered to arrive at a general correctness: Frequency Range statement in ordinary word order 100,000 45 adjectival clauses 30,430 39 adverbial clauses of time 19,115 45 subord. clauses introduced by that 13,480 45 commands 11,000 39 noun clauses 10,280 43 As with the number of grammatical items that might be considered functional, so with the major errors that are made by children, and the major difficulties they experience. They are widespread, persistent, but not many.

D. Bagley, in “A critical survey of objective estimates in the teaching of English” (B. J. Ed. Psy. Vol. 7, 1937) suggests that there is sufficient agreement between English and American lists to show that children generally are troubled by four or five main difficulties:

1. omission of capital letters and full stops strings of sentences loosely joined by “and” unrelated participles vague “so” and “only” clauses

2. agreement in number between verb and subject (especially “Everybody….”)

3. capital letters for proper names and adjectives

4. case forms of pronouns, especially who/whom

5. correct tenses and sequence for verbs, and use of past participle for past

tense (“I done it”).

Certain of these errors, as hardly strictly grammatical, were not counted in the present work, but others such as the omission of the full stop were very common. The list as presented here obscures however the great difference in frequency of the different errors; and the brief nature of the list indicates perhaps how vital it is for the teacher in practice to consider closely the local and individual errors of his classes. The author reports general agreement that the formal review of grammar should be placed “at the end of a cumulative learning experience”, when “the principles and usages to be formally organized have been, for the most part, practiced and understood by the pupils of some time.” This clearly envisages a return to the idea of grammar as a science, and is in any case equivalent to its abandonment in the lower part of the school. Numerous investigations into correct usage, and counts of common errors have been completed in America, and some in Britain, and such examinations of the types of error of grammar and punctuation have influenced in the present work the decision to limit errors scored to those few shown in Appendix 1, since these were expected to be of a general nature. These surveys of oral and written English were made in the early 1920s by Lyman, Charters and Miller, Stormzand, Pressey, and others. In Britain, P. B. Ballard reported in 1922 (“Group Tests of Intelligence”, ULP) on the error counts of nearly 14 million words written by children in the upper half of 60 elementary schools. 153,000 errors were discovered – one in 89 words. Errors scoring over 7% of the total were:

1. those due to similarity of sound, e.g. is, his, to, too. 14.5%

2. wrong sequence of tenses 7.5%

3. omission or misuse of apostrophe 11.0%

4. miscellaneous errors (chiefly omission or misuse of capitals) 18.5%

5. omission of full stop 21%

In the American work the major grammatical mistakes were generally found

in verb forms, and especially in punctuation, the use of the full stop and comma. Lyman, for example, lists six studies showing in oral English that 49% to 62% of errors were in verb forms. Charters and Miller tabulated errors under 27 headings, and found 47% of written errors were made in punctuation, the major one being the lack of the full stop. Conclusions of the present work are fully in agreement with this. Betz and Marshall analysed 112 papers, and found 55% of errors to be in punctuation – 21% in the use of the period and comma. Sumner’s study of the written work of 8,618 Louisiana children showed two main groups of errors: first, in punctuation, capitalization, and sentence form; and second, in syntax. The most common errors in the first group were lack of the full stop, lack of an initial capital, and the indiscriminate use of the “and” clause. In the second, the major faults were in verbs, followed by pronouns and by confusion of adjective and adverb. Sumner, in common with other workers such as Johnson, Lyman, Stormzand, and Armstrong, noted that errors were often specific to a locality, and of course Americanisms were frequent as in the use of “most” for “almost”; but such main errors as those indicated above are shared by the pupils in the present work.

Pressey in a statistical study of errors in 980 papers of grades 7 to 12, pointed out that errors which obscure meaning are more vital than those which merely offend propriety. Of the former, errors in sentence structure are of great importance. They are of a few frequently recurring types. Thus, failure to make proper sentence division as in “stringy” sentences and fragments, and omissions of words or phrases, make up nearly half of the errors. Faulty reference of pronouns, and repetition or redundancy, were also common, forming another 20% of the total. Studies were made to show which types of error persist. Johnson, for example, took three compositions – narration, description and exposition – from each of 132 high-school freshmen and from 66 college freshmen, and found their errors constant in type. By combining orders of persistence with orders of prevalence he suggested that emphasis in cure should be on mistakes in capitalization, in the use of the apostrophe s, in punctuation, confusion of adjective and adverb, faulty spelling, in pronouns and verbs, in ambiguities of meaning.

Stormzand, using the concept of the “error-quotient”, found a similar persistence of error from the sixth grade to the upper class in the university, but with a regular general improvement in that the ration of errors to words was 1 to 11.5 in Grade 6, 1 to 12.2 in the high-school freshmen, and 1 to 22.9 in the university – still, one would have thought, a high rate of error. The error-quotient is really an attempt to weight errors in gauging their importance, and involves manipulating the frequency of use of a construction, the number of words needed to produce opportunities for error, and the frequency of error proportionate to the frequency of use. The quotient will usually disturb the simpler ranking by frequency only, so that for example the failure to put a period ranked first by Charters becomes in Stormzand’s estimation a poor fourth, whereas the failure to put the apostrophe to denote possession, fifth in Charters, rises to first place in Stormzand. In the present work, the older and simpler criterion of importance has been retained, and the “error-quotient” merely sampled occasionally. From the point of view of the practising teacher, it is not the rank of an error that matters, but its massiveness and generality. An error that is common is common, though perhaps rarer than another in relation to the opportunities for making it. For all practical purposes, the massive errors found by Stormzand are the same as those found by earlier workers and by the present writer, although they may be ranked in different order. Some conclusions noted by Lyman are relevant here. The number of technical difficulties, he points out, is small in comparison with traditional lists. The immediate needs of pupils must be catered for and remedial work must be largely individual. It is highly questionable that a child’s ability to recognize an error and to correct it is indicative of his own present and future use of the particular form. The last point is well seen in the high scores made even by eleven-year-old children on Tests A and B in the present work (see Appendix 3); the former ones are recognized in the recommendations embodied in Appendix 4.

There is very general agreement in research as to the likely signs of increasing maturation in writing-skill, and this agreement has been valuable in the present work in demonstrating what to pick out as measuring devices to gauge the rate or degree of progress made by children. The sign commonly found by most observers is the increasing use of the complex sentence with increasing age. This is reported for example, by C. W. Kimmins (“Methods of expression used by London children in essay writing at different ages”, Journal of Experimental Pedagogy, vol. 3 no. 5, June 5, 1916, Longman, Green & Co) in whose work 1,570 girls and 1,511 boys aged eight to thirteen years from typical elementary schools were allowed fifteen minutes to write about the war. A sentence analysis of their essays was made which wielded a table showing the changes in the percentage of simple and complex sentences during the five years between eight and thirteen years of age: the percentage of unrelated simple sentences rose from 14.0 to 48.9. Similarly in Stormzand and O’Shea’s work (“How much English Grammar?”) (Warwick and York, 1924) a growth of complex and a decline of simple sentence structures was noted. The authors analysed contemporary writings from classical prose to light journalism, children’s compositions, and letters and articles by adults. Ten thousand sentences were selected to determine frequencies of usage. The authors showed that sentence length increased regularly through the various levels of school life and university, the average lengths at 4th grade being eleven words, and in university upper class 21.5 words, although long sentences often gave evidence of lack of sentence mastery rather than of the reverse. This growth in length is allowed for in the present work in the use of the length of the correct simple sentence as one of the measuring instruments. The following table illustrates the changing proportions of simple and complex sentences used by the growing students:

Simple Complex

University upper class 22.0% 54.9%

High school seniors 37.6 42.0

Sophomores 44.4 37.6

Elementary school Grade 8 48.4 37.2

Grade 6 59.0 36.5

Other writers, such as L. Labrant (“A study of certain language developments

in children”, Genetic Psychology Monographs no. 5, Nov. 1933, Clark University, Worcester, Mass.) and A. F. Watts have noted a similar trend. Watts points out that the mere number of dependent clauses is no guarantee of maturation, partly because increase in number can be obtained without variety, and partly because a mature style often condenses clauses into phrases. In the present work, the variety of sentence patterns as well as the number of dependent clauses is used in the measures, and a particularly important type of phrase is included in one of them.

It is probably true to say that relatively little purely observational and

descriptive work has been attempted on children’s language in England, save perhaps in word counts and phonetics, and critical work has been largely concerned to refurbish the methods of teaching grammar rather than to question the basic assumption of its value. Such critical comment as has been made, for example by Ballard, has been rather the expression of strongly felt opinion than a conclusion based on evidence. There is thus room for an experimental reassessment of the place and function of formal grammar in the English lesson.

CHAPTER III

APPROACH TO THE EXPERIMENT

Changes in the control of a complicated instrument such as language are not likely to establish themselves quickly, and for this reason it was decided to organise the experimental work over as long a period as possible. It was hoped in this way to establish positive and measurable differences, ones which would become well fixed in the children’s habits and skills, and to avoid stressing any chance and local variation in development. Previous work had involved either the testing of different populations at various points in the age range, or a relatively short test-teach-test experiment with one population for at most six months. The value of ordinary written work over a long period, as distinct from specific tests, lies in the natural context and connected style, and in the elucidation of available skills rather than in more recognitive correctness. However, the long period entailed inevitable disadvantages and limitations. It is rare to find schools willing and able to devote two years to an experiment which must interfere the whole time, though more at some times than at others, with the normal and traditional syllabus of the English Department and with the administrative arrangements in staffing and timetabling of the school.

The virtual impossibility of precise measurement in work of any length and complexity has no doubt defended many undesirable pedagogic habits, allowed mere opinion to arrogate to itself a decisive power, and supplanted by the doubtful criterion of examination success any genuine enquiry into measurement. The examination is designed to fit the teaching, and vice versa: it is not surprising that they do fit, though surprising that the fit is in formal grammar so bad as was suggested by the G.C.E. marks given in Chapter I. But clearly, neither teaching nor examination can test the wider efficiency of the other. Thus, the present work’s major virtue, in the writer’s eyes – that is, the two years long duration of the grammar/non-grammar comparison – is the source also of much of the organizational fallibility. To interfere for two years with the routine of a school is to make a radical demand on staff and on children. For the very best of reasons – namely, the concern lest any experimental work should prejudice the chances of children in a school busy with its established method and syllabus in English – four schools felt compelled to withhold their active co-operation for every one that volunteered to take part in the work. In the end, despite all efforts, only five schools were found who felt justified in taking the risk of participation. One would naturally have liked a larger sample, to have included not only more schools but more of each type of school. Furthermore, within each school it would have been welcome to have had four forms running in pairs instead of two: but there again, no practical school can be expected to turn over two-thirds of a year-group to an experiment. It was not possible to have that complete control over the situation which is frequently available in short-term and more rigidly experimental procedures. A number of variables had to be accepted without adequate control, in the hope that the difference between the work done by the experimental groups would be sufficiently large and clear to counterbalance in the results uncertainty due to uncontrolled variables or to the lack of random or representative sampling. The schools were all London schools: two from West London (Isleworth and Dormers Wells), one from nearer the centre (Christopher Wren), and two from rather East (Shoredith and Owen’s). Two were grammar schools, two technical/comprehensive, and one secondary modern. The schools necessarily decided the groups of children who could be used, and in this there was administratively no possibility of selecting two ideally equated groups, either in intelligence, background, or attainment. Fortunately in each school at the earlier age-level there was no streaming, and the pairs of forms were in fact roughly equated on the basis of their entry or summer examination results. (See Appendix 8). As was shown by the results of the grammar test given at the start of the experimental work (p. 170), and of the analysis of the first essays (p. 278), the paired forms started roughly equal in English attainment. There was in the second year some reorganization of form groups, with the consequence that some children were lost from the experiment, but in general Head Teachers were extremely co-operative and whenever they could avoided reshuffling groups. Had the groups been seriously streamed at first, one would have wished to run overlapping courses in each school, so that if in one year a bright form took grammar and a less bright one had no grammar, the position in a succeeding year could have been reversed. Thus:

Year 1. Year 2. Year 3.

Bright form 1 grammar ---

Dull form 1 non-grammar non-grammar ---

Bright form 2 --- non-grammar ---

Dull form 2 --- grammar grammar

Such a process would have involved the school for three years instead of two, and have added staff complications in view of the overlap in year 2. It would not have sufficed to reverse the roles of the groups half-way through the course in an attempt to equate them. For firstly, the value of the long term would have largely disappeared; and secondly, the latter part of the work, particularly of the form first studying grammar, would have been almost valueless, since there is no estimating when or to what degree the previously learnt grammar would bear fruit – perhaps after lying fallow for a while.

Another variable to be considered was the teacher and his approach. It was luckily possible to ask one teacher to take each pair of forms in all the schools except the secondary modern, where the grammar form was instructed by a teacher favourably disposed to formal grammar. This general arrangement, incidentally, would not have been feasible had it been necessary to extend the course for a third year. To some extent, the unity of the teacher’s approach was thus established, although no certainty exists that each teacher was absolutely impartial in his attitude to the teaching or abandonment of formal grammar, or indeed equally adept at the one approach as at the other. There is no regulating the quality of teachers or of teaching within fine limits. But, as far as could be judged, all the teachers were willing to be convinced by evidence, desirous of finding out the truth about the matter in hand, competent and practiced in the present grammar of the language.

The general method of grammar teaching among the five teachers was

surprisingly uniform. As had been the writer’s experience in talking with teachers from other schools who finally were unable to take part in the work, he found in the participating schools no-one who aimed at a rigid, abstract, or deliberately Latinised system of grammar. The words “liberal” and “functional” were the descriptive titles generally offered to the grammar taught. Grammatical categories were closely tied to function and meaning, in what it seems fair to call the general fashion of the day so far as this may be gathered from textbooks and teachers’ comments. And although the union of form and meaning may have led to some of the difficulties experienced by pupils in school, it does lend a certain uniformity to the process of the experimental teaching. As with the approach, so with the speed of the work. This was not finally governable, either as to the pace maintained within the course, or as to the final goal reached. Experiments have been reported (as in that of E. G. S. Evans, p. 81) in which a strict allocation was made of material to be taught to a number of forms of varying standards and ability and attainment. Such a procedure is unrealistic in terms of actual teaching, since if the material is adequate to keep the faster groups busy it will be beyond the capacity of the slower. Thus one can do little more than follow the example of the ordinary Senior English Teacher, who gives to all the members of his department a syllabus showing the main features to be discussed and the level of attainment to which he hopes they may bring some of their children – but who knows that the level will have in practice to be adjusted to the quality of the pupils and of the staff. It is possible to mention everything in teaching to a rigid and firmly detailed syllabus; it is not possible to teach everything. In the present work, each group save one studied the same text and course as far and as fast as it could. Thus the Modern school did not manage to cover all the ground in Book 2 of the Active English Course, though it did get far enough to show a significant change in attainment in Test C by the grammar form. Whether the level of grammatical knowledge thus achieved was high enough to affect their writing cannot, however, be decided. The main themes of relevant research – can grammar be taught; does what is learnt of it affect writing skill – emerge in different degrees with different types of children. One can only say of the Modern school that to these children a subject seems unprofitable which after two years’ work leads to no ascertainable gain in writing-skill. But the question whether sufficient grammar ever is learnt in schools is discussed elsewhere in the present work (pages ).

Mechanical matters were regulated more easily. It might be expected, despite some evidence to the contrary in the work of E. E. Wiswall (“A study of sentence structure in Eighth Grade Composition”), that the style of writing used by children would have a demonstrable effect on the type of sentence structure, and thus that an essay calling for a linked argument and connected reasoning might predispose the writer to use let us say more adverbial clauses of reason than would a simple descriptive passage. With this possibility in mind, the titles of the essays, although permitting some variation in order to draw out the interests of the writers, were limited to a general theme of a narrative or descriptive nature, and each child wrote on the same subject in final as in first essay. The time given to the writing was the same in each case. The content of the grammar course was defined by reference to Test C, the formal grammar test, and to the List of Common Errors and the relevant sections of the Course of English. The content of the non-grammar course closely excluded any reference to formal grammar and specified attention to practice in writing and imitating conventional forms and structures. Even in this, the caveat has to be entered that in the grammar and comprehensive schools some foreign languages were studied, in learning which formal grammar was used. This may account for the slight increase in scores in Test C made by the non-grammar groups, although this is more likely to be due to recollection. The increase is so slight that it does not suggest that any great transfer existed between the foreign tongue and English.

CHAPTER IV

A PILOT EXPERIMENT

In order to get some experience of the best way to organize the material, a course of three months was given by one teacher to a pair of first-year forms at a Technical school. Here, as later in the main work, each form’s English syllabus was identical save for the grammar with which the one form replaced a lesson in writing. There were six 40-minute lessons each week.

The shared work included practice in summarizing, in comprehension, in reading and composing verse, in composition, in silent reading, reading aloud and discussion. The grammar work attempted by the one form was that contained in Book 1, term 1 of the Active English Course, the course book used in the main experiment by four out of the five schools. It contains chapters on sentence-types, subject and predicate, the noun, verbs, the direct object, the adjective, the adverb, and the idea of function. The form omitting grammar replaced it by the writing of an adventure story, which concerned the exploits of a boy who was separated from his parents in the war, passed through a number of hair-raising situations largely to do with counterespionage, and was eventually reunited with his parents.

A lesson was spent discussing possible alternative themes, such as a desert-island story, an animal story, a flying tale. These were voted on and the most popular was taken for general use. One or two pupils wished to make up their own story but abandoned the effort after a week or so when it became apparent that all the others were getting a lot more help by working together than they would have done as lone wolves. A synopsis was devised by class discussion and much rubbing out on the blackboard, and each week some time was allotted to working out details of the next chapter. The boys’ attention was directed to the need for inherent plausibility rather than factual accuracy; to various devices for maintaining suspense; to the use of the effective detail in suggesting character and in creating atmosphere. No class and group work was undertaken in the direct imitation of structures. Such work might of course have helped, and was indeed much used in the longer courses, but the story project and the little time available forced its exclusion here. Individuals, however, were given help in for example rephrasing a sentence to produce a better emphasis or rhythm in a passage with which they might be having difficulty; and such improvements were often brought to the notice of the class.

It was not uncommon for children to continue their story at home, though they were always asked to keep within the chapter the class was working on. In this way the majority of the class finished the story together by the end of the term. The stories were then loaned to the members of another class. Each member of this class was asked to read three versions of the story, and to give marks on a separate piece of paper for the excitement or suspense, the characterization, and the vividness of the scenes, and was asked to mention any particularly good piece of description or convincing incident. Thus each script received a mark out of 30. These verdicts were then discussed and the “specially good” descriptions read out and compared. An indirect approach, as this was, to the problem of encouraging writing-skills takes considerable time, but not more than was given to the formal grammar of the parallel form. It showed itself able to elicit not only enthusiasm but also a certain self-criticism and purposive modification of habits of writing.

The two forms, of 22 and 24 boys respectively, were of similar attainment in English, as judged by their school examination results. Both forms tried the formal grammar test C (Appendix 3) at the beginning and

end of the term’s work. As was to be expected, the grammar form made much better progress in this than did the non-grammar form, as the following scores indicate: Table 1. Test C. Difference between means: 7.435 S.E. 1.349 t 5.45

The final ways of measuring change in correctness had not yet been arrived at, and so a fairly simple method of scoring was used. Both forms wrote two compositions on descriptive themes, “A day in the country”, “A day at the seaside”, one at the beginning and one at the end of the course. They were allowed as much time as they wanted in class to do this – all had finished in less than two periods of 40 minutes each. The work was then marked off at the nearest stop to 500 words. Two counts were made: first, of the number of words per error as listed in Appendix 1; and second, of the structures used. These structures were envisaged merely as complex or non-complex sentences, and as sentences correct or incorrect. The error count produced the following scores, favouring the non-grammar form: Table 2. Error count. Difference between means 0.519 S. E. 3.154 t 0.16 This difference, though favouring the one form, is of course quite non-significant. Both forms improved their scores; both, knowing they were experimenting, may have paid more attention than usual to the correctness of their writing. The structure count, here re-scored according to the final method, whereby wrong examples are deducted from correct ones, scored as follows: Table 3. Structure count. Difference between means. S.E.T

a – number of complex sentences 0.784 0.460 1.70

b – number of sentences correct 0.560 0.886 0.63

(In both cases, the scores favour the non-grammar form)

By these scores, one was encouraged to feel that the difference between the work attempted by the two forms was sufficiently clear to produce measurable changes, and that this system used over a larger span of time should be fairly trustworthy. The scores show a small but general change in favour of the non-grammar form, but the short period of work and the small number of children involved greatly reduce any significance such a change might be assumed to have. Indeed, as appears later, the changes that take place in correctness are by no means uni-directional, and take a long time to become apparently stable. But at this stage it was still not clear how best one could measure any changes.

The error count on its own would certainly be inadequate. As has beenreported by Rivlin and others, teachers will not always agree about what is an error; and, more important, some way of showing greater maturity of control, as well as mere correctness, was desirable. For a child correctly using a long sequence of simple sentences would hardly be thought a better writer than one making errors in a more complex system of structures. The structure count seemed promising, but needed refining and extending. This done, and a far greater variety of structures used in measurement, the error count could be adopted as one item in a series containing positive as well as negative elements. It was in this direction that the attempt to score the essays eventually led, but only after a number of fallible or cumbersome methods of scoring had been explored and abandoned. These methods may be seen in Chapter V, where figures of the comparison between essays written at the age of ten and of fifteen are given.

CHAPTER V

MATERIAL OF THE MAIN EXPERIMENT

1) Pattern, population and schools

2) The search for a measuring instrument (p. 140)

3) The reliability of the measuring instruments (p. 168)

4) Score of the formal grammar test (p. 170)

5) Scoring of the main essays (p. 172)

6) Results of the error count (p. 179)

1a) PATTERN

The pattern employed was that of a test-teach-test experiment, relying for the test material upon the ordinary continuous writing of the children being taught. Pairs of forms, each from the same school, wrote essays at the start of the work for a fixed time and on a descriptive theme; and at the end of the two years each pupil wrote another essay on the same theme as the first. In three of the schools, an intermediate essay was written after nine months’ work. All these essays were the material from which the relative success of the teaching course with or without formal grammar was to be judged, although in fact attempts were made to devise other and more formal tests which did not however answer the purpose of the experiment so well and which could not have offered such direct evidence as the essay of the children’s progress since writing skills. The teaching courses in each school were planned so that one form had as nearly as possible the same general English work as the other, but with the one

major difference that while one form studied and applied the terms of formal grammar in its composition work, the other used none of these terms and devoted the time saved to direct practice in writing.

1b) POPULATION

The work was done in the first two years at the secondary school. One hundred and nine children completed the course in the non-grammar forms, and wrote a total of 25,801 words in their first essays, and 37,112 words in their third or final essays. These totals were composed as follows (for means see p. 278g):

Table 4 - Non-grammar forms First essay Third or final essay Grammar) Isleworth 26 boys 4126 words 7596 words schools) Owens 25 boys 9026 words 11,752 words Secondary Dormers Wells 21 girls 4950 7569 Modern Comprehensive/ Technical Christopher Wren 23 boys 4798 5722 Chartesey 14 boys 2901 4483 In the grammar forms there were 119 children, who wrote a total of 30,306 words in their first essays, and 40,624 in their third or final essays. Table 5 - Grammar forms First essay Third or final essay Isleworth 29 boys 6910 words 8738 words Owens 21 boys 6583 7865 Dormers Wells 29 girls 8309 10,906 Christopher Wren 21 boys 4259 6150 Chartesey 19 boys 4245 6965

1c) SCHOOLS

Numerous schools showed interest in the work and its outcome, and five were able to take an active part. These were Isleworth School and Owen’s School, both Grammar schools for boys; Chartesey School (later Shoreditch School) and Christopher Wren School – the former a mixed and the latter a boys’ Comprehensive school, from which the children taking part belonged to the Technical stream; and Dormers Wells School, a Secondary Modern school for girls. In all of these it is fair to say that the English teachers concerned in the actual grammar teaching were willing to be convinced that formal grammar was valuable, and four of them taught the subject for the G. C. E. examination, though all expressed some doubt as to formal grammar’s value in wider contexts. Two of the four were senior English teachers, and all were qualified specialists in English teaching. In the two instances where the senior English teacher was not taking the experimental forms, he took an active interest in the work and offered useful criticism and suggestions. In the secondary modern school, two teachers taught the two forms, the one taking the grammar form being interested in formal grammar, the other convinced that it had no practical value. Each form had five 45-minute periods of English a week, which fall into the usual divisions of reading, drama, poetry, composition (two). Each did the same work in four of these periods, but in the fifth, one form took formal grammar, and the other practiced and discussed written work in non-grammatical terms. In correcting the normal weekly homework of the non-grammar forms, teachers were particularly careful to avoid terms of formal grammar. The work attempted in the non-grammar lesson was basically an extension of the usual composition practice, but most teachers gave cohesion and interest to the work by engaging in a variety of longer projects such as the compilation of a diary, a form newspaper, an adventure story similar in general outline to the one discussed in the pilot experiment, or a book of hobbies. All the children knew, both in the grammar and in the non-grammar forms, that they were participating in an experiment directed to the improvement of their writing. In the course of the experiment, a piece of continuous writing similar to the non-grammar “project” was undertaken with the grammar form also, but it was naturally not possible to spend so much time on this sort of study as with the non-grammar groups, since the grammar forms carefully followed the course in composition in their textbook which aimed to integrate the grammatical material studied with the composition lessons.

Chapter V (1) - Main Experiment Population

Since for the non-grammar groups much more time was available than had been the case in the pilot experiment, the teachers were able to combine the indirect approach of encouraging the development of writing-skills by discussion and illustration of the story or diary with a good proportion of direct practice in sentence structures, points of usage and convention, and the ordering of material in paragraphs, and so on. This material was possibly more readily assimilated by the groups for whom such practice was always related forward to the resolution of immediate difficulties in their writing project than by those for whom it was inevitably often related back to the explanation of grammatical definitions and analysis. The most acceptable form for the continuous writing was that of an adventure story, which called out both an imaginative response from the children and a willingness to work in detail; but the diary form was also popular. The basis of such diary or story work was taken in four of the schools from the type of instruction given in Read to Write (C. Austin, Ginn, 1954). In this, a general theme is agreed on – a treasure island story – and the synopsis divided into chapters which are easily spread over a year’s work. An advantage of diary work was that a diversity of themes could be treated. Thus the diary included sections on paying a visit, hobbies, pets, helping mother, Christmas entertainments, my dog, a day in bed, at the zoo, school adventures, swimming, a cricket match, and so on. An illustration of the professional’s writing on the theme was read to the children, and this gave frequent opportunity to draw attention to some device of sentence construction, such as the use of the short, agitated sentence in a breathless climb, or the emphatic inversion. After that, a possible scheme for each paragraph – giving topic and details – was offered to the form, which was then invited to keep to the scheme if it wished, but to add or change whatever it liked. In this way, the less imaginative, more restricted children were given help and encouragement, without the more fluent writers feeling restricted. There were many children who broke away from the details of the scheme, but all were influenced by the attention drawn of necessity to the proper ordering of ideas, and to the linking of these ideas and incidents from paragraph to paragraph. It took a couple of periods to get the story going, less time for the diary; but after that there were fewer delays.

Teachers did naturally keep in mind the elimination of those particular errors listed in the “common errors” in Appendix 1. These were treated by means of example and imitation, instead of by the abstraction and generalization of the approach through formal grammar – which did not itself, of course, exclude the use of examples. The process of correction by example and imitation seems cumbersome at times, and will like all other exercises be barren unless closely related to work in hand, to mistakes actually made in a particular class and felt by them to mar something in which pride is taken. It is in this that the concentration on the use of the concrete particulars of the children’s own writing scores over the vague generality of grammatical categories. Such mistakes as “Jim and me was going into the cave” (errors 5 and 3g) would be tackled in the grammar forms by direction to agreement of subject and verb and to the proper use of the nominative forms of pronouns when the pronouns are used as the subject. First, of course, subject, verb, agreement, pronoun, nominative or subject case, were terms that had to have been absorbed – at least, this was the ideal supposition. Probably the example which followed (“Jim and me were going …”) actually drove home the point, and one hoped that the rule would spread the influence of the example to other parallels.

In the non-grammar forms, there was time for far more examples. The sentence would be looked at in this way:

“Jim and me was going into the cave”

Teacher: Would you say “We was going into the cave”?

General dissent – one or two brave individuals aver that they would or might, and why not sir? But the vast majority of the class bring home to them the weight of convention. They admit to being wrong, which was more than was asked of them.

Teacher: What would you say then?

Class: We were going into the cave.

Teacher: How many is “we”?

Class: Depends how many there are. More than one, anyway.

Teacher: Well, “Jim and I” means more than one. So they must be followed by “were”. Let’s try some more examples…

And the class, and the teacher, provide 20 or 30 examples in five minutes orally. With a group that needed this sort of practice, it was unwise to try to correct the pronoun error in the same lesson. One has to decide on priorities. With this sort of pronoun error, it is in any case advisable to find sentences in which the singular and the plural verb have the same form. The error can then be exposed to a similar pressure of corrective convention as happened in the verb/subject agreement above. For example:

“You and me both like the same sports.”

Teacher: Would you say “Me like the same sport as you”?

Class: “I like the same sport as you” (and usually they add more or less sotto

voce “of course”)

Teacher: In that case, shouldn’t it be “I like” in our sentence – “You and I both

like the same sports”?

This in the same way is followed by numerous confirmatory examples. All this takes time, but not more than the learning of the appropriate grammatical terminology; and it is, as the result of the experiment may suggest, possibly better learned, and more completely established in the children’s writing habits. In similar fashion, many opportunities arose to bring to the children’s attention the utility of specific practice in sentence building and structure. These also sprang out of the work of particular children when each week the teacher went through the scripts of the class. It was best to note special points on a separate piece of paper rather than to mark the actual script; and the practice related to such special points was then often done in the course of the ordinary composition lesson. Such points were, for example, the avoidance of repetitive structures – leading to exercises in varying modes of saying the same thing, or almost the same thing (“John was in the cave by that time; by that time, John was in the cave; John, by that time, was in the cave”); the clarification of the unrelated participle (“seeing” was a common one – “Seeing the hole was too deep, the box could not be raised to the surface”) – this was usefully attacked by parody; the most effective placing of the adverb (especially of “also”, of which children seem very fond, and which could often be simply omitted) – trials of the meanings of the various possible places for an adverb helped here: He only found three ducats (did not dig them up himself) Only he found three ducats (no-one else found three) He found only three ducats (so few). Grammatical errors such as the failure of agreement between subject and verb mentioned above were usually approachable by the appeal to convention which was in fact really the summoning to use of the actual recognition of convention which the children already had but had failed to implement. Sometimes, as with the very common “There was three pirates on the beach” type of error, a new element of dramatic suitability came in. When asked, for example, which would be more likely to say “there was three…” – Crossbones the pirate or the Captain of the naval frigate – the children would recognize the social label of the failure of agreement and distinguish it from the impersonally correct’s “There were three…” – they did not want the Captain to be a ruffian. Errors of the type “Alan and me ran into the forest”, dealt with as suggested above by being broken into their parts, were although easy to correct, not so easy to eliminate as one or two examples may imply, since the “error” was itself often a convention of the speech of some groups. “Alan and me” is felt presumably as a sort of collective noun which might take a singular verb, whereas “Alan and I”, if used, are conceived as separate persons. The correction or rather conventionalization of language forms is of course very much a social matter. If children do not wish to be like the people who they fancy say “Alan and I ran…” they will continue to say “Alan and me ran…” even when they know through grammatical or social learning that the one form is structurally unjustifiable or socially limiting. Nevertheless, the building up of sentences in a generally accepted pattern was a fruitful exercise, especially when the pattern could advantageously be employed in the children’s own written work.

Where errors stemmed from genuine ignorance, instead of from mere habitual acceptance of a local convention of speech, the consideration of meaning was often effective in a correction. Thus in the sentence – “He read several books like Treasure Island” or The Swiss Family Robinson” the children readily perceived the difference between “He read three books including Treasure Island… where in this instance the writer had intended the former meaning. It is not always practicable to appeal to meaning in this way, and the arbiter has then to be simply authority, whose acceptance however is very much a personal matter between the teacher and the form, whether or not a logical basis for the authority can be provided. The very common construction “The reason for such and such is because …” is thus awkward to correct. It can be broken down into “The reason for such and such is that…” and: “Such and such happened because…” but these constructions despite practice may not become established in the children’s usage with their consenting to the teacher’s authority. The normal use of certain prepositions has to be learnt in the same way. The word “of”, for example, seems to be falling out of use in children’s speech in such phrases as “out of the window”, which may commonly be heard as “out the window”, presumably on the analogy of “The bird flew in the window” and similar sentences. It is possible that children will more readily accept and try to employ the conventional construction in such groups if the teacher is able to give ground to them at other times, in say conceding that it is not uncommon to find the word “so” used in such groups as “They were late, so they had no tea.” The teacher may explain that such a usage is still not officially accepted, but may admit that language changes sometimes against our will or judgement. It is more difficult for the formal grammarian to yield such points, except inadvertently.

Such then was the approach in the forms which did not take formal grammar. Work in the grammar forms was based on the school’s English textbook, which in all schools except Isleworth was Book 1 and 2 of the Active English Course by G. S. Humphreys and J. C. Roberts, published by ULP and first printed in 1939. In this course, the lessons in grammar alternate with ten lessons in composition, and “the grammar lessons set out an orderly progression, through the parts of speech, with stress on the function of words and with constant practical application to Composition.” The remaining school used An English Grammar of Function, Book 1, by Allen and Mason (Arnold, 1939). The scheme of grammatical work covered the following points: difference between a sentence and a phrase subject and predicate, including the “understood” subject of commands the parts of speech, with their functions and phrasal equivalents: noun – common, proper, collective, abstract case - possessive and object verb – transitive and intransitive, active and passive, the infinitive, participles, finite verbs, tenses – past, present, future, continuous, perfect, subject and direct object, cognate object, auxiliary verbs, complement

adjective – of quality, quantity, epithet, predicative pronoun - personal interrogative, demonstrative, possessive, indefinite, relative, case in object forms preposition conjunction – co-ordinating and subordinating sentence analysis – graphic or columnar showing direct and indirect object, subject complement, adverbial and adjectival extensions and phrases, and the beginning of the distinction between main and subordinate clauses.

Corrections in the children’s work were indicated and commented on in grammatical terms. Thus the mistake examined above – “There was three pirates…” would be named a failure of agreement between verb and subject: or in “John and me went …” the nominative or subject case of the pronoun would be required.

2. THE SEARCH FOR A MEASURING INSTRUMENT

The subjective fallibility of the marking of essays and of other continuous writing is notorious. It is normally partly guarded against in experimental work by subdividing the marks allocated so that a definite proportion is given to say grammatical correctness, or to quality and quantity of material, to order and coherence, to style, and so on. Markers are often asked first to separate essays into general categories, say A to E before deciding on a particular mark. Again, several people may be asked to mark each essay; or alternatively, essays may be assessed by being compared in a general way with “typical” examples produced by children of comparable age. However, the number of people who can in practice be asked to mark an essay is very limited, and the utility of the marking may not increase as the number increases since it is not easy to derive a general assessment from the variety of marks awarded. Equally, it is not entirely satisfactory to break down the qualities and contents of an essay into any great number of separable units to each of which a fixed maximum mark may be given in an attempt to reduce the subjectivity of the markers, since some of the main qualities of an essay, such as style or even the ordering of the material, are essentially personal and not mechanical. It is clear therefore that in work such as the present in which it was hoped to use the continuous writing of the children as the critical test and measure of success and progress, a large element of doubt, of mere opinion and prejudice, could enter and invalidate the argument if no better way of ascertaining the degree of achievement in such writing could be devised, than had commonly been used in marking children’s essays. Even such apparently mechanical scores as the counting of errors were full of dispute, as may be seen in Leonard & Moffett’s account of the disparity between one linguist’s and another’s view of what forms of usage were actually errors – they found that 45 out of 102 expressions normally condemned by grammarians were accepted as cultivated English by 75% of the linguists participating in the study (“Current Definitions of Words in English Usage”, 1927). In any case, even an entirely reliable error count, or similar mechanical measuring device, would alone be quite insufficient to ascertain the rate of growth of a child’s maturing style. Thus although it was recognized fairly early in the present work that one had to find some valid and reliable way of gauging the differences in correctness between work written at twelve years of age and at fourteen, - Search for Measuring Instrument the full requirements were not at once envisaged, and too much reliance was placed on the limited idea derived from the pilot experiment. At first, some trials were made with two tests, Test A and Test B (shown in Appendix 3). These were designed on the basis of the list of common errors (Appendix 1) and are thus really extensions of the “error count”. Test A contains in haphazard order ten examples of errors 1 and 3, and five of each of the others. Test B contains the same number of errors and in the same proportion as Test A, but is based on a continuous piece of writing instead of on disconnected sentences. These two tests, together with Test C – the formal grammar test – were tried on forms of 23 to 26 boys at the ages of eleven, twelve, thirteen and

fourteen at one school, and Tests A and B gave very disappointing results. The

average scores for Test A (set out in detail in Appendix 3) were: Table 6: Tests A and B Test A 11 year - 83.2 range 68/92 12 year - 81.5 range 60/96 13 year - 82.0 range 70/88 14 year - 81.6 range 72/92 - Search for Measuring Instrument For Test B the results were similar:

11 year - 79.5 range 66/92

12 year - 81.0 range 72/94

13 year - 82.5 range 66/90

14 year - 81.9 range 48/94

Presumably these tests were insufficiently difficult at the lower age, and to judge by the scores of the fourteen-year group, the quality measured would seem to have been carefulness rather than skill – or perhaps (for these boys all took formal grammar in the English course) skill does not increase as grammatical knowledge increases. In any event, the tests as they stood could not differentiate between the older and the younger child’s writing. They tested recognition rather than ability to use new forms independently; they were too closely limited to the error-count; and they did not use the original and continuous written work of the children. Similar difficulties were of course easier to avoid in measuring attainment in formal grammar. It was necessary to have a means of doing this so that the relative knowledge of the paired groups could be compared at the start of a course and at the end, and so that it could be decided whether or not a group learning grammar had absorbed sufficient of the instruction to derive any practical value from it. Test C, the formal grammar test, provided likely data and was retained for further use. - Search for Measuring Instrument Tested by the split-half method (R = 2r), it gave a co-efficient of +0.923 s. e. 0.014, 1+r and its average scores in the four preliminary forms were:

Table 7: Test C

11 year - 8.1 range 0 to 16

12 year - 13.0 range 4 to 22

13 year - 22.8 range 12 to 34

14 year - 33.1 range 19 to 48

However, a way still had to be felt towards the more complicated matter of a means of measuring the essay. A much clearer guide was required than had been provided by the three months’ pilot experiment described in Part 5, or by Tests A and B. Generally speaking, one would expect the older children to write more fluently and with more complexity than the younger. Thus, for example, in her thesis on the language of children between three and five years of age, E. M. Moore gives the average number of words in the sentence of five year old children as 6.5, but by ten years of age this had become 10.8; similarly, the number of subordinate clauses per hundred sentences was 9.8 at five years, but 49.7 at ten, though not every aspect of the sentences changed at an equal rate. Clauses of time and reason, for example, were the only ones common both at five and at ten. But a definitive measurement was needed. To guide towards this, an analysis was made of compositions written by children aged ten and by others aged fifteen. The former wrote 13,736 words and the 25 fifteen-year-olds wrote 11,835, on a descriptive topic. The younger children came from three primary schools (Arno Vale, Lambley, and St Pauls Bentinck) and the older ones from a comprehensive school (Christopher Wren). The measurements to be finally chosen had to be objective, to show a clear difference between the different ages, and to be sufficiently common to be applicable to the majority of pupils. But initial assessments were inevitably almost completely in the dark, since beyond the general agreement that older children use a greater number of complex sentences than do younger children, and probably write more in a given time, little accurate knowledge was available of the signs of the maturing writer. A first attempt was therefore made to break down essays under headings of some detail, which represent as it were a purely photographic snapshot of the essays’ contents. Since it was unknown what guiding lines would emerge from the snapshot, everything had to be included.

(A). A first attempt was made under the following heads:

1. Paper no. Error no. Sentence Clauses used Special points No. of words

(eg word order, or parenthesis)

2. Points concerning the SUBJECT Clause or Phrase Article Genitive Adjective Substantive Pronoun

3. Subject modifiers of some length Phrase Clause Other

4. Verb structure

Adverb Finite Non-finite Participle Infin. Correct tense Other extensions Tense Mood

5. Direct object Clause Adverb Genitive Adjective Pronoun Other modifiers or phrase and adject. Subst.

6. Complement Adverbial. Adverb Article Adjective Subst. Pron. Modifying adject. or genitive Infin + obj. modifiers

7. Sentence type – loose, balanced, periodic.

The reader has to picture all the above headings spread out across one large sheet, as in Appendix 6, with a column for each subheading shown. This style of analysis was too cumbersome, and in particular failed to direct attention to points likely to indicate maturation, being as it were too flatly descriptive, like a drawing without perspective.

(B). A more selective as well as more manageable method was next adopted, as follows:

1. Paper no. Sentence no. Error no. Clauses used Phrases used

2. Sentence types Simple Double Multiple Complex Complex-complex

3. Instances of the use of ; : -

4. List of modifiers between sentences within sentences

5. Other structures (e.g. use of direct speech; modified or double subjects)

6. Certain measurements of proportions:

No. of words over no. of sentences; ideas/ paragraphs; sentences/ paragraphs;

Words/ errors.

These headings, similarly, spread across a large sheet. They are however more

manageable than those under A, and were adopted for the analysis of the essays in the comparison of ten year and fifteen year essays, by the guidance of which the final measuring instruments were selected. The second mode of scoring, B, produced 33 items of information about possibly important differences between the writings of children aged ten and fifteen. These were the following:

Table 8. Scored differences between writings of children aged ten and fifteen

Age 10+ Age 15+

1. Number of scripts 63 25

2. Total words 13,717 11,341

3. Average length of script 218 words 453 words

4. Total sentences 885 560

5. Average length as written 15.7 words 20.2 words

6. Average length corrected 11.9 19.5

7. Total correct sentences 535 487

(61% of total) (87% of total)

8. Sentence types

a) Simple a) total 292 (33.1%) 154 (27.5%)

b) incorrect 61 18

b) Double a) total 155 (17.5%) 96 (17.2%)

b) incorrect 54 17

c) Multiple a) total 47 (5.4%) 19 (3.4%)

b) incorrect 32 8

d) Complex a) total 300 (33.9%) 204 (36.4%)

b) incorrect 137 41

e) Complex-complex (i.e. having more than one subordinate clause)

a) total 91 (10.3%) 87 (15.5.%)

b) incorrect 64 9

f) As for (e) but correct sentences 27 (3.1%) 78 (13.9%)

9. Total sentence patterns in correct sentences 45 75

10. Average number of different patterns per script 4 (highest 9) 10.3 (highest 17)

11. Major patterns used:

(M = main clause; N = noun; J = adjectival; V = adverbial)

M 244 (46%) 140 (29%)

M M 95 79

N M 25 1

M N 23 16

M J 20 40

M V 21 26

V M 20 19

M M M 13 13

V M M 9 6

M V M 7 3

M M J 2 11

M M V 3 11

M M N 4 8

M J J 3 8

M J V 1 7

M V J 1 7

M N V 4 7

12. Sequences of sentence patterns

(doublet = two sentences of same pattern in succession; triplet = three)

a) as written: 1. doublets 103 (12.5%) 37 (6.9%)

2. triplets 23 (2.8%) 13 (2.4%)

b) with full-stops properly inserted

1. doublets 126 44

2. triplets 21 16

13. False sequence of ideas 50 (6.1@) 11 (2.6%) 822 535

14. Genuine modifier between sentences 3.5% 8.2%

1. time 18 22

2. additive 2 4

3. expository 3 12

4. place 6 6

15. Structure of simple sentences:

a) total number 292 154

b) total correct 231 (43.2%) 136 (27.8%)

c) unmodified by a phrase of three or more words 118 36

d) modified by one such phrase 103 60

e) modified by two such phrases 19 28

f) modified by three such phrases 4 16

g) average length of correct simple sentences 8.7 words 13.6 words

16. Other structural points in the sentences:

a) frequency of subject modified by phrase, clause, or at least two adjectives: 43/2081 clauses 121/1265 clauses (2.1%) (9.5%)

b) subject modified by one adjective, excluding articles or possessives: 42 37

c) clause or phrase as subject 1 1

d) correct omission of subject 52 30

e) inversions, mostly “there verb subject” or “adverb verb subject” 45 27

f) subordinate clause preceding or interrupting the main clause 104 71

g) introductory phrase 69 66

h) other word order devices, mostly of emphasis 7 8

i) number of scripts containing marked clumsy repetition, i.e. where three or more words are repeated without deliberate effect in two consecutive sentences 25 2

k) use of subjunctive 0 2

l) use of interrogative 8 6

17. Total number of paragraphs 249 129

18. Scripts of only one paragraph 11 3

19. Total topics 245 123

20. Scripts in which number of topics equalled the number of paragraphs 10/63 13/25

21. Total number of clauses 2081 1265

22. Average number of main clause to each subordinate clause 4.6 2.2 (including as main those subordinate noun objects in direct speech – these figures otherwise become 2.5 1.8)

23. Average number of subordinate clauses to each sentence

a) as written 0.67 0.81

b) correct sentences 0.41 0.84

24. Clauses used:

a) main (including noun object in direct speech) 1709 (82%) 866 (68.4%)

b) adjectival 89 (4.1%) 148 (11.7%)

c) adverbial time 151 92 place 4 8 manner 14 24 reason 53 44 result 8 (5.3%) 1 (8.8%) purpose 2 0 degree/comp. 2 8 condition 25 17 concession 2 9

d) noun object (indirect speech) 9 29 object (direct speech) 218 52 subject 0 1 complement 7 5 apposition 6 13

25. Number of occasions on which a clause other than the main clause starts the sentence (excluding noun objects in direct speech)

a) total 104 71

b) adverbial time 82 34 place 0 2 manner 2 1 reason 2 6 condition 11 4

c) noun object 0 1

d) adjectival 7 20

26. Average number of catchphrases to each sentence 0.62 1.23

27. Phrases used

a) adverbial tine 94 95 place 186 180 manner 69 109 reason 3 6 purpose 50 19 degree/comp 2 2 condition 0 3 concession 0 0

b) noun object 6 3 subject 1 3 complement 0 0 apposition 9 3

c) adjectival 124 266

28. Links between clauses (2081 clauses) (1265 clauses) and 360 167 after 152 also 0 0 as 40 55 because/for 16 27 before 8 4 but 57 37 how 4 4 if/whether 34 18 or 2 4 so 46 7 that 83 106 then 27 0 though/although 1 7 until 4 8 what 18 9 when 86 37 where 7 19 which 40 56 while 6 6 who 26 24 why 1 1 (others) 5 9 29. Instances of less common punctuation correctly used

a) colon 4 0

b) semi-colon 1 6

c) parentheses 2 3

30. A measure of stretch in vocabulary, and of repetitiveness, made by using acount of adjectives and finite verbs:

a) number of unrepeated adjectives and adjectival phrases 1000 1266

b) total number of adjectives 1142 1418

c) proportion of unrepeated adjectives to total 0.88 0.89

d) proportion of unrepeated adjectives to total words 0.073 0.112

e) number of unrepeated finite verbs 1151 778

f) total number of finite verbs 2081 1265

g) e over f 0.553 0.615

31. Number of words per common error (for list of errors see p. 222) 41.3 182.3

32. Frequency of the common errors (only those reaching double figures are included)

1a 19 2

1b 194 28

1c 236 11

1d 41 16

2b 36 7

3d 51 10

3f 39 8

4 36 6

6 20 9

33. Instances of inadequate separation between clauses or sentences 232 34

34. Instances of excessive separation (caused for example by insertion of a comma between subject and verb, or by repeated “and” with repeated subject) 27 2

The scores thus extracted were tentatively grouped as showing growth in

fluency – item 3

coherence – items 13, 14, 18, 20, 18

complexity – items 6, 8a, 15e, 15f, 15g, 16a, 22, 23, 25, 26

correctness – items 7, 31, 33, 34

variety – items 8, 9, 10, 12, 24a, 30.

There was considerable overlapping between the groups, which moreover involved the retention within the general categories of a number of fallible measures giving too vague and blurred a representation of the differences between the two sets of children.

For these reasons, the grouped comparisons were abandoned, and attention was given to selecting as criteria for differentiation between the two sets of essays those points which occurred sufficiently often to give a clear measurement, and which did not overlap too much.

These were:

a) average length of the correct simple sentence (item 15g)

b) instances of the omission of the full stop (items 32, 1b; 33)

c) number of words per common error (item 31)

d) number of different sentence patterns (item 10)

e) number of non-simple sentences (item 8, b to e)

f) number of subordinate clauses (item 23b)

g) total words (item 2)

h) number of complex sentences (item 8d, 8e)

i) number of simple sentences containing two or more modifying

phrases (items 15 e, f, g)

j) number of correct sentences (item 7)

k) number of adjectival clauses and phrases (items 24b, 27c).

It is important to know how these items were scored, since without this knowledge it is well-nigh impossible to make valid comparisons, and writers have not always made clear what for example is meant by a “complex sentence” or by an “error” in measuring scripts. Without such information it is difficult to interpret the errors of one in 89 words found by one author, against the one in 25 words noted by another in the work of older pupils. Equally, the frequent abuse and mismanagement of complex sentences by growing children is not to be counted a mark of increased control; although the use of complex sentences may indicate the struggle for growth in the child, it is the increased use of correct complex sentences which argues his growing mastery of writing skills.

A note on the mode of scoring items (a) to (k) above may therefore avoid later misunderstandings. The aim, which was not always immediately realized in practice, was to find the simplest effective way of scoring, and embodied the following considerations:

a) Average length of the correct simple sentence. “Correct” here means

containing none of the errors in the list of common errors in Appendix 1.

b) Instances of the omission of the full stop. This was scored by subtracting the number of omissions from the number of sentences in which the stop was correctly inserted. In this way it was expected to provide due allowance for the length of a script. Two omissions in one sentence score 1 and a half, 3 score 2, 4 score 3, etc.

c) Number of words per common error. (See Appendix 1 for the list of

common errors.) From this, errors 1a (omission of the question mark), 1d (omission of apostrophe in ’s for possession), 2d (failure to use capital at beginning of sentence), and 2b (failure to use capital in proper noun or adjective) were excluded as being too remote from true grammatical errors.

d) Number of different sentence patterns used by each writer. The use of, for example, different types of adverbial clauses was accounted sufficient to provide a new pattern, as was inversion of the order of similar clauses. Only correct sentences were counted. But if a sentence contained only one common error, this was allowed as correct for (d) or (h).

e) Number of non-simple sentences. From this, the number of simple sentences was subtracted. Correct sentences were added to incorrect in order to allow for the greater length of the older children’s scripts, in which rather than the mere number of simple sentences the proportion of simple to non-simple sentences was smaller.

f) Number of subordinate clauses used. This was originally scored as a proportion showing the number of subordinate clauses to the number of main clauses. It was found, however, that the limit on the time allowed for writing the essay took care of this factor, and that a clearer and simpler score was made by taking the actual number of instances of the item.

g) Total number of words written in a limited time.

h) Number of correct complex sentences. Sentences containing an error were

subtracted from those which were correct, so that frequently a minus score was found at ten or eleven years of age.

i) Number of correct simple sentences containing two or more qualifying phrases, each phrase to consist of three or more words.

j) Total number of correct sentences minus the number of incorrect.

k) Number of adjectival clauses and phrases, the phrase to consist of three or more words. The suggestion here is that adjectival elaboration is likely to be more conscious and deliberate than most adverbial forms of modification.

Some consideration was given at this stage to the list of so-called “common errors” used in calculating item (c). This list contained the following points:

1. Certain errors in punctuation, notably

a) omission of question mark

b) omission of full stop

c) omission of comma between items in a list, words in apposition, main and non-defining clauses

d) omission of apostrophe in ’s for possession.

2. Failure to use capital letters

a) at beginning of sentence

b) in proper nouns and adjectives.

3. Misuse of various parts of speech:

a) adjective or preposition as adverb

b) wrong comparatives and superlatives

c) faulty positioning of adverbs

d) failure to give pronoun a clear antecedent

e) wrong or ambiguous use of prepositions

f) mismanagement of conjunctions

g) misuse of object forms of pronouns

h) failure to use relative pronoun or adverb.

4. Failure to give a finite verb to each clause, error in or omission of any important verbal word.

5. Lack of agreement between verb and subject.

6. Faulty sequence of tenses.

7. Unrelated or false participle.

The composition of the list was agreed on by ten experienced teachers of English, and the errors themselves were all found in the essays of the ten-year-old children. Not all the errors listed can be said to be grammatical, or to be within the likely compass of grammatical training to correct. For this reason, those errors most clearly due to ignorance of a non-grammatical convention or to simple carelessness – as for example the failure to use a capital letter at the beginning of a sentence (errors 1a, 1d, 2a, 2b) were finally discarded in the calculation of item (c). A trial was given to the suggestion of Stormzand that in such matters the opportunities for error should be included in the calculation and an “error quotient” thus arrived at. The error Quotient (of which five specimens are shown in Appendix 5) is in process not only cumbersome, but also rather a misleading refinement of the manipulated scores, and it was therefore decided to use the simple count of errors actually made, for the time and subject matter of each essay was the same for each writer. Thus measurement of opportunity for the common errors is largely a repeated measurement of length, and length is already involved in several of the categories of measurement (e.g. b, e, g, f). Further, in considering each separate error, the error quotient might be important, but in taking the whole group of errors as one, the opportunity factor becomes less important, since if one error is not made another is likely to be in its place. When Stormzand and others indicated the importance of the error quotient, they were in fact measuring the importance of one error in relation to another, which is not the case here. The average number of words per common error is, even in its present simple and direct form, one of the more reliable measurements (see the

reliability test on p. 171).

Items (a) to (k) gave the following results on being applied to the essays of the ten and fifteen-year-old children. In each instance, the difference between the mean score at ten and at fifteen was taken, and from this and its standard error the critical ratio (t) was obtained. Table 9 – Scores on items (a) to (k) in the ten/fifteen comparison Difference between means Standard error t

a) 4.86 0.832 5.84

b) 11.41 1.670 6.79

c) 141.0 22.90 6.16

d) 6.31 1.12 5.66

e) 7.00 1.21 5.78

f) 11.44 1.71 6.67

g) 235.4 23.00 10.24

h) 9.513 0.916 10.38

i) 1.063 0.251 4.24

j) 13.68 1.373 9.96

k) 12.97 1.341 9.67

It will be seen that the items provided clear differences each reaching a good standard of significance. Making the assumption, which perhaps may be allowed, that at fifteen years most children will write more correctly than they did at ten, one may expect to be able to use the differences shown above as valid criteria to measure correctness in children’s writing. The adoption of the measures (a) to (k) led to a considerable simplification of the process of marking the essays, which could now be treated under a few simple headings in a more economical way, selecting only those points necessary for the calculation of the scores. These headings were:

Paper no. Sentence no. Error no. Clauses No. of words S-- S+2 Adj. phrases

In these headings, S—stands for “simple sentence without any qualifying phrase”, and S+2 indicates “simple sentence but with two or more qualifying phrases”. At first, a column showing whether or not direct speech had been used was included, but this point was omitted fairly early as of insufficiently general appearance. A specimen of a first and third essay analysed under this final set of headings follows. The clause abbreviations are:

M = main clause

N = noun clause

N obj = noun clause as object

N appos = noun clause in apposition

V = adverbial claus

V reas = adverbial clause of reason

V conc = adverbial clause of concession

Vt = adverbial clause of time

First and Final Essays of a Child from the Non-grammar Form in School D, with analysis and scores on measures (a) to (k)

Name: Lewis.

Age: 12 years 0 months.

School: Christopher Wren.

Form: 1d, non-grammar group.

FIRST ESSAY

The Day Mother was Ill

When I came home from school one day I found my mother was laying on the

couch with a blanket over her asleep. (1) So I had to get my tea and my brothers when we had finished tea I done the washing up and brought her up a cup of tea. (2) When she had finished that I had to go and get some errands and phone for the doctor, the doctor came around about an hour later and took her temperature which was 102, then the doctor said that she must go to bed at once. (3) In the morning I found when I went downstairs that mum wasn’t up so I got the breakfast lit the fire and brought her up a cup of tea. (4) After breakfast I done the washing up got my coat on and went to school. (5) That evening when I come home and found out that she had gone to hospital for a few days therefore I had to do all the housework. (6) After a few days mummy

came home and said that she has to go to convelesion for two weeks and she had to leave tomorrow so that meant more housework but we were sad she had to leave. (7) Next morning dad went with her to Clacton where she was staying and then came back and also had to stay away from work. (8) Every Sunday he used to go and visit her and bring her something from us all. (9) At last she came home about 6 o’clock on a Monday and brought us lots of presents, but she couldn’t do any work for a couple of days so I had to do it till Thursday then she took over and I was happy and all, but now I know what housework is like even though I do a lot now. (10)

FINAL ESSAY

The Day My Mother was Ill

It was a sad day for me when I came home from school one evening, and found my mother had been rushed off to hospital, through a nervous breakdown. (1)

As I went into the house I found a note on the mantelpiece, reading, Michael, will you please look after the children while I; away, for mammy is very ill. (2) When Christy comes home from school give him the other note, which has a list of messages that I want. (3) I don’t know when I will be home, but look after things till I get back. (4) When Christy came home I told him what happened and he was very upset, so was I, but we had too put up with it, so I sent him round for the messages. (5) When he came back I started to get the tea. (6) After a while it was ready so I went and got my two little sisters from next door (for the lady there was minding them) and told them tea was ready. (7) Soon the tea was finished, my brother (Christy) and I cleared the table and washed the dishes. (8) We then went up to the sitting room and I put the television on, and we all sat there and watched it. (9) Near the end of the film my uncle came in (he had just come from Ireland and was staying with us for a while) and I told him what had happened he too didn’t seem very pleased about it. (10) We then went downstairs and I helped him get the dinner for him and my dad, when he comes home from the hospital, and gives us a report of what is wrong with her. (11) He came home late that night and said she isn’t very bad, but she has to stay in there for a few days. (12)

Analysis of the essays

First essay:

Sentence Error Clauses used No. of words S-- S+2 Adj phrase no. no.

1. 7 Vt M N Obj 23

2. 1b, 1c, 4 M Vt M M 28

3. 1b, 1b, 1c Vt M M M J M N obj 45

4. 3f, 1c M Vt N obj. M M M 30

5. 1c, 4 M M M 15

6. 1c, 3c Vt Vt N obj. M 27

7. 6 1b, 1c M M N obj. N oboj. M M V reas. 38

8. 1c M N appos. M M 24

9. M 15 1

10. 1b, 1b, 1c M M M M M M M N obj. V conc. 62

Second essay:

1. M Vt Vt N obj 29

2. Vt M M Vt V reas. 31

3. Vt M J J 20

4. M N obj. M Vt 16

5. 3f, 3f Vt M N obj M M M M 34

6. Vt M 10

7. 1b M M M V reas. M N obj. 31

8. 1b M M M 17

9. M M M M 22

10. 1b M M M M N obj. M 40

11. 6, 3d M M Vt Vt N obj. 34

12. 6 M M M M 24

Scores: a b c d e f g h i j k

1st. 15 0 16.2 1 9 17 307 -8 0 -6 0

2nd -- +6 38.5 6 12 20 308 +1 -- 0 4

It will be seen that in this particular instance the following changes have been

measured, and show a large degree of progress as the child grew older:

b) Omission of full stops - 6 points gain

c) Words per common error - 22.3 points gain

d) Variety of correct sentence patterns - 5 points gain

h) number of correct complex sentences - 9 points gain

j) Number of correct sentences - 6 points gain

In other measures, there was a small but probably insignificant gain by the time of the later essay:

e) Number of non-simple sentences - 3 points gain

f) Number of subordinate clauses used - 3 points gain

g) Number of words written - 1 point gain

k) Adjectival clauses & phrases - 4 points gain

Measurement (i) does not occur in the later essay, and simple sentences -

measurement (a) - also fail to appear here.

3. THE RELIABILITY OF THE MEASURING INSTRUMENTS

How reliable were the differences thus established? To find this out, two descriptive compositions were set at an interval of one week to a form of 27 thirteen year- old boys. These boys, although from School A, were not actually participating in the main experiment. They were allowed 40 minutes to write on each of the following subjects: a) A house on fire, and b) A foggy night.

A product moment correlation of their scores on each of the measuring categories was made and the significance of each correlation estimated, using the formula

_____

T = r vN – 2

v1 – r2

(a formula from Lindequist for small samples). The following table shows the

correlation co-efficient followed by the standard error, and the reliability quotient with the level of significance taken from Fisher and Yates’ tables.

Table 10: Reliability of the Measures (a) to (k) r s.e. t level of signif.

a) simple sentence length +0.239 0.181 1.23 worse than 0.1

b) omission of full stop +0.574 0.129 3.49 better than 0.01

c) words per common error +0.844 0.055 7.87 better than 0.001

d) number of sentence patterns +0.791 0.072 6.46 better than 0.001

e) non-simple sentences +0.393 0.163 2.14 better than 0.05

f) number of subordinate clauses +0.705 0.097 4.97 better than 0.001

g) total words +0.232 0.178 1.21 worse than 0.1

h) complex sentences +0.685 0.102 4.70 better than 0.001

i) simple sentences with two or more phrases +0.447 0.154 2.49 better than 0.02

j) total correct sentences +0.839 0.057 7.71 better than 0.001

k) adjectival clauses and phrases +0.446 0.154 2.49 better than 0.02 It is evident that measures (c), (d), (f), (h) and (j) are highly significant and reliable,

and that (b), (e), (i) and (k), but especially (b), reach a good level of reliability. Items

(a) and (g) are not by themselves reliable, but may serve to act as pennants in the general wind.

4. THE FORMAL GRAMMAR TEST

Before proceeding to a review of the scores revealed in the essays, it may be pertinent to note the results of the formal grammar test (Appendix 3) to each form at the beginning and end of the teaching course. As was to be expected, the forms studying formal grammar gained significantly higher scores at the second test than did the non-grammar forms. These scores out of 100 are shown in the following table, which gives the difference between the mean gains together with the standard error and the resultant reliability quotient for each school.

Table 11: Scores on Test C, the Formal Grammar Test First Second Gain Difference s. e. t average

School A (Grammar)

- grammar form 33.03 49.71 16.68 16.11 1.955 8.24

-non-grammar form 34.12 34.69 0.57

School B (Grammar)

-grammar form 36.33 66.62 30.29 26.97 2.393 11.27

-non-grammar form 24.88 38.20 3.32

School C (Sec. Mod.)

-grammar form 20.10 34.34 14.24 12.86 1.825 7.04

-non-grammar form 22.09 23.47 1.38

School D (Comp. Tech.)

-grammar form 19.48 39.96 20.48 19.74 1.359 14.53

-non-grammar form 21.09 21.83 0.74

School E (Comp. Tech.)

-grammar form 15.84 32.84 17.00 15.50 1.669 9.29

-non-grammar form 22.71 24.21 1.50

It is evident from these scores that good progress was made by the grammar forms in learning not only the terms made by the grammar forms in learning not only the terms of formal grammar, but how to apply these, for Test C contained questions requiring both recognition and use of the common grammatical terminology at the appropriate level. Few of the scores were really high – only four children reached 80 – and it may be suggested that had the grammar been thoroughly mastered giving average scores of say 80-90, then a beneficial effect on the correctness of written work might have been demonstrable. However, to judge from the work of Macauley and others, as well as from the present experiment, it does not seem practicable to give this thorough mastery of the subject of formal grammar to most children. The present forms would seem to have done as well as can normally be expected (unless perhaps excessive time be given to formal grammar study) and better than the children in the Macauley observations referred to above, in so far as one can compare the two.

The grammar school pupils (Schools A and B) reached a higher level of

attainment than did those of the secondary modern and technical (branch of the

comprehensive) schools.

All forms, even those not taking formal grammar lessons in English, made some improvement in their final scores, due presumably to recognition of the first test rather than to grammar learnt in foreign languages, for the non-grammar forms in Schools A and B did not outpace those in the other schools, where either no or very little foreign language study was undertaken.

5. SCORING OF THE MAIN ESSAYS

a) Scores after nine months

The measures (a) to (k) (see p. 155) were applied to the essays of the

experimental forms of the first three schools at three stages: a first essay at the

beginning of the course; a second after about nine months; and a third or final one at the end of the course, that is, after two academic years. No great value was of course placed on the scores after nine months, and indeed they were taken only to check that things were proceeding according to plan. The remaining two schools, whose course was not able to start until the end of the year, produced essays only at the start and conclusion of their two years’ course. Whatever theme, narrative or descriptive, was used for the first essay was used again for the last, but a change was allowed at the intermediate stage, although the topics were once more such as invited narrative or descriptive techniques.

The following table shows the difference between the means of the scores made on items (a) to (k) by the non-grammar and the grammar forms, followed by the standard error and the critical ratio t.

The formula used for the calculation of the ratio t was:

t = M1 – M2

(Sd12 + Sd22) ( 1 + 1 ) ( = s.e.)

(n1 + n2 – 2) ( n1 n2 )

where d1 = X1 – M1 (i.e. deviation from mean) and d2 = X2 – M2

Table 12: Scores on Items a-k after nine months for schools A, B and C

Where g is placed after a figure, it indicates that this is in favour of the grammar forms; unmarked scores favour the non-grammar forms. Diff. between means s.e. t

a) Average length of simple sentences

School A 4.090 1.507 2.71

School B 3.047 1.355 2.25

School C 5.603 1.649 3.39

b) Omission of the full stop

School A 0.460 3.354 0.14 (g)

School B 2.917 2.922 0.99

School C 2.102 1.124 1.87 (g)

c) Words per common error

School A 28.67 17.36 1.68 (g)

School B 56.17 38.52 1.45

School C 14.24. 11.51 1.24 (g)

d) Different sentence patterns

School A 1.364 0.590 2.31 (g)

School B 1.195 1.183 1.01

School C 3.091 0.847 3.65 (g)

e) Non-simple sentences minus simple sentences

School A 1.038 1.148 0.73

School B 2.247 1.589 1.41 (g)

School C 4.886 5.183 0.94 (g)

f) Subordinate clauses

School A 1.677 1.823 0.92

School B 2.167 2.118 1.02 (g)

School C 2.898 2.345 1.24 (g)

g) Total words

School A 92.47 21.33 4.34

School B 48.17 33.41 1.44 (g)

School C 52.69 39.90 1.32 (g)

h) Complex sentences

School A 1.754 1.091 1.61 (g)

School B 6.111 2.258 2.71

School C 1.401 1.433 0.98 (g)

i) Simple sentences with two or more phrases

School A 0.514 0.595 0.86

School B 1.917 0.670 2.86

School C 0.303 0.562 0.54

j) Number of correct sentences

School A 4.646 2.441 1.90 (g)

School B 17.528 3.786 4.63

School C 3.197 3.910 0.82 (g)

k) Adjectival clauses and phrases

School A 1.025 1.404 0.73

School B 1.584 3.061 0.52

School C 5.148 1.792 2.88

As may be seen, the scores after this relatively short period reached a significant level, where t = 3 or over, on only four occasions. Three of these four went to the non-grammar form’s advantage, and one to that of the grammar form. Another four scores reaching a fair level of significance, with t = 2.70 or over, were all in favour of the non-grammar form.

These were (a), School A; (h) School B; (k) School C. In no category, however, did the non-grammar form score significantly higher than the grammar in all schools, and of the total of 33 scores, 16 went to the grammar side, and 17 to the non-grammar.

The period of nine months or one academic year is too short to expect any stable changes to have been made, and to allow for variation of pace within a course. Thus item 9d) shows a significant gain to a grammar form, but one which when in the perspective of the completed course changes to a significant score in favour of the non-grammar form.

b) Final scores on items a-k after two years

Much more important therefore are the scores obtained from the third essays of the first three schools and the final essays of the last two schools: the essays, that is, written in each case at the very end of the experimental work.

The following table shows, in similar fashion to that based on the essays

written after nine months, the scores made by the pupils of all five schools at the end of the course. The level of reliability of the various measures is represented on the left by the letter V (= very reliable), F (= fairly reliable0 and U (= unreliable); and details of reliability are shown on page 169.

Table 13: Final Scores on Items a-k after two academic years for Schools A, B, C, D & E. Where g is placed after a figure, it indicates that this is in favour of the grammar forms; unmarked scores favour the non-grammar forms. An asterisk is set after each score reaching a significant level. Reliability Item and school Diff. between s. e. t means U a) Average length of simple sentences

A 1.717 1.296 1.32

B 0.975 1.212 0.79

C 3.190 1.944 1.64 (g)

D 1.779 2.065 0.86

E 0.223 2.262 0.10

F b) Omission of full stop

A 2.171 1.873 1.16

B 3.345 3.180 1.05

C 1.766 1.712 1.03 (g)

D 3.168 2.316 1.37 (g)

E 0.046 3.065 0.02 (g)

V c) Words per common error

A 14.26 13.63 1.05 (g)

B 85.21 26.68 3.19 *

C 44.41 13.32 3.33 *

D 38.41 11.92 3.22 *

E 41.46 21.84 1.89

V d) Different sentence patterns

A 0.322 0.685 0.47 (g)

B 3.331 1.042 3.19 *

C 2.778 0.767 3.62 *

D 2.340 0.900 2.59

E 2.824 1.149 2.46

F to U

e) Non-simple sentences minus simple sentences

A 2.120 1.140 1.86

B 3.640 2.056 1.77

C 0.710 1.688 0.42

D 1.374 1.848 0.74 (g)

E 1.105 1.468 0.69 (g)

V f) Subordinate clauses

A 1.174 1.637 0.72

B 1.116 2.233 0.49

C 1.951 2.274 0.86

D 0.036 1.919 0.02 (g)

E 0.255 2.376 0.11 (g)

U g) Total words

A 70.43 21.54 3.27 *

B 47.99 36.63 1.31

C 35.16 31.39 1.12

D 37.02 19.18 1.93 (g)

E 30.16 37.23 0.81 (g)

V h) Complex sentences

A 0.648 1.107 0.59 (g)

B 5.042 1.466 3.44 *

C 4.495 1.185 3.08 *

D 4.363 1.193 3.66 *

E 6.327 1.779 3.56 *

F to U

i) Simple sentences with two or more phrases

A 0.076 1.899 0.04 (g)

B 0.436 2.493 0.17

C 0.889 0.393 2.26 (g)

D 0.417 0.473 0.89

E 0.918 0.793 1.16

V j) Number of correct sentences

A 2.607 2.168 1.20 (g)

B 10.23 4.225 2.42

C 1.807 2.581 0.70

D 8.276 2.451 3.38

E 8.308 3.434 2.42

F to U

k) Adjectival clauses and phrases

A 3.454 1.322 2.61

B 2.791 2.408 1.16

C 2.003 1.820 1.10 (g)

D 4.217 1.861 2.27 (g)

E 2.850 2.767 1.03 (g)

The five very reliable measures (c, d, f, h, and j) gave 25 scores in the final essays. Of those, 10 reached a significant level where t = more than 3 in favour of the non-grammar forms, as against none favouring the grammar forms. Of these 25 scores, only 6 were at all in favour of the grammar forms. The fairly reliable measures (b, e, i, and k) produced no significant scores. Of the 20, 10 favoured the non-grammar forms. In the two unreliable measures (a and g), 7 of the 10 scores were in favour of non-grammar forms – one significantly so.

6. RESULTS OF THE ERROR COUNT

As a considerable amount of work has been done in the past on counting and

categorizing errors made in children’s writing, the subdivisions of item (c) – the number of words per common error – in the measures may be of interest here.

Item (c) included these points:

1b – omission of full stop

1c – faulty use or omission of comma in lists, apposition, non-defining clauses

3a – adjective or preposition used as adverb

3b – wrong comparatives or superlatives

3c – faulty position of adverb

3d – no clear antecedent for pronoun

3e – misuse of prepositions

3f – misuse of conjunctions

3g – misuse of object form of pronoun

3h – failure to use relative pronoun or relative adverb

4 – no finite verb, or omission of any important verbal word

5 – lack of agreement between verb and subject

6 – faulty sequence of tenses

7 – unrelated or false participles.

In each form, the number of essays containing one of these errors was noted, together with the total number of instances of the error. This information is expressed in the tables below for each form in each school in the first essays, the third or final essays, for the grammar forms combined and the non-grammar forms combined in the first and in the third essays, and for all the essays combined. In each case, the count is shown as the number of essays containing the error, followed by the total number of each error.

The schools wrote the following number of essays at the first and at the third

or final attempt:

School Grammar Form Non-grammar Form

A 29 26

B 21 25

C 29 21

D 21 23

E 19 14

Total: 119 109

a) Table 14: Error Count of First Essays

Column 1 gives number of essays containing examples of the error.

Column 2 gives total instances of the error.

In each school, the grammar form’s score precedes that of the non-grammar.

A B C D E_______

1.2. 1.2. 1.2. 1.2. 1.2. 1.2. 1.2. 1.2. 1.2. 1.2.

Error

no.

1b 26/84 16/35 16/37 21/96 28/190 21/110 20/86 19/83 17/61 10/36

1c 23/143 12/41 18/48 22/82 24/68 20/53 15/35 21/74 16/52 9/25

3a 1/1 0/0 0/0 0/0 0/0 1/1 0/0 0/0 0/0 1/1

3b 0/0 0/0 0/0 0/0 0/0 0/0 1/1 1/1 0/0 0/0

3c 4/4 4/4 0/0 0/0 0/0 0/0 0/0 3/3 2/3 1/1

3d 13/19 6/7 4/5 10/13 9/10 8/10 5/5 10/14 7/9 8/10

3e 8/8 5/5 5/6 7/10 3/3 2/4 4/4 0/0 1/ 2 1/1

3f 7/7 4/5 8/15 14/24 12/28 7/12 11/18 11/19 15/21 8/11

3g 5/5 1/1 0/0 3/ 4 0/0 0/0 0/0 3/ 4 2/2 1/1

3h 0/0 0/0 0/0 0/0 0/0 0/0 0/0 0/0 0/0 0/0

4 15/20 4/4 3/3 11/15 8/12 7/9 9/15 9/10 2/3 6/13

5 5/6 2/2 2/2 4/6 2/2 2/2 2/2 3/3 7/8 3/3

6 0/0 0/0 0/0 2/2 0/0 0/0 0/0 0/0 0/0 0/0

7 4/4 3/ 4 4/5 7/12 12/17 8/11 7/7 5/5 3/5 4/8

b) Table 15: Error Count of Third or Final Essays

Column 1 gives number of essays containing examples of the error.

Column 2 gives total instances of the error.

In each school, the grammar form’s score precedes that of the non-grammar.

A B C D E_______

1.2. 1.2. 1.2. 1.2. 1.2. 1.2. 1.2. 1.2. 1.2. 1.2.

Error

no.

1b 29/74 27/35 12/25 17/59 26/149 16/100 18/71 17/55 18/86 9/28

1c 23/92 21/51 19/58 15/28 26/104 13/44 19/104 19/26 17/68 6/13

3a 1/1 0/0 0/0 2/2 3/3 3/3 1/1 0/0 2/3 0/0

3b 0/0 0/0 0/0 0/0 0/0 0/0 1/1 1/1 0/0 0/0

3c 3/3 0/0 3/3 3/3 3/4 0/0 3/3 1/1 1/2 0/0

3d 13/24 6/8 9/9 9/15 16/28 11/15 15/29 8/10 14/35 9/9

3e 5/6 0/0 10/11 7/10 10/14 2/3 15/21 4/4 8/22 0/0

3f 12/17 7/9 7/12 4/4 24/44 15/28 8/14 10/14 12/24 3/3

3g 3/3 1/1 1/1 0/0 2/2 2/3 6/7 3/3 3/3 0/0

3h 0/0 0/0 0/0 0/0 0/0 0/0 0/0 0/0 0/0 0/0

4 8/13 4/4 6/6 3/3 9/14 7/8 16/29 6/6 11/26 1/1

5 9/11 6/6 4/4 6/12 4/4 0/0 3/3 4/7 8/26 8/8

6 0/0 0/0 0/0 0/0 1/1 0/0 5/9 0/0 0/0 1/0

17 1/1 2/ 3 5/5 6/8 15/22 4/7 10/21 5/5 2/2 1/1

c) The following table gives the total errors of the combined schools for the first and final essays respectively, and for grammar and non-grammar forms separately. That there are sometimes more errors in the final essays is of course accounted for by the greater length of these essays. It is noticeable that the same errors continue to be of the widest range and most frequent occurrence in the final as in the first essays, and that some errors occur very seldom indeed.

Table 16: Total Error Count for Combined Schools, but showing first and final essays respectively and grammar/ non-grammar

As before, the errors are shown as fractions, with

1. the number of essays containing the error

2. the number of errors.

There were 119 grammar essays, and 109 non-grammar.

First Essays Final Essays

Grammar Non-grammar Grammar Non-grammar

1. 2. 1. 2. 1. 2. 1. 2.

Error no.

1b 107 458 87 360 91 405 76 277

1c 96 344 84 275 103 426 74 162

3a 1 1 2 2 7 8 5 5

3b 1 1 1 1 2 3 1 1

3c 6 7 8 8 13 15 4 4

3d 38 48 42 54 54 125 44 57

3e 21 23 15 20 48 67 19 25

3f 53 89 44 71 63 111 39 58

3g 7 7 8 10 15 16 6 9

3h 0 0 0 0 0 0 0 0

4 37 53 37 51 50 88 21 22

5 18 20 14 16 28 38 22 33

6 0 0 2 2 16 25 12 17

7 30 38 27 40 36 52 18 24

d) The final table of this section shows the combined total count of errors for the 456 essays written. The great proportion of error borne by numbers 1b and 1c is clear; important also are 3d, 3e, 3f, 4, and to a lesser degree 5 and 7. Other errors are relatively individual, and 3h did not occur at all though present in the essays written at ten years of age.

Table 17: Combined Total Error Count on 456 Essays Column (a) shows mean instances in essays containing a particular error. Column (b) shows mean instances in all essays.

Type of error No. of essays Total errors a b

1b – omission of full stop 361 1500 4.16 3.29

1c – faults in use of comma 357 1207 3.38 2.65

3a – adj. or prep. as adverb 15 16 1.07 0.04

3b – wrong comparative or sup. 5 6 1.20 0.01

3c – faulty position of adverb 31 34 1.10 0.08

3d – no clear antecedent for

pronoun 178 284 1.60 0.62

3e – misuse of prepositions 103 135 1.31 0.29

3f – misuse of conjunctions 199 329 165 0.71

3g – misuse of object form of

pronoun 36 42 1.17 0.09

3h – failure to use relative 0 0 0 0

4 – no finite verb, or omission of an important verbal word 145 214 1.48 0.47

5 – lack of agreement between

verb and subject 82 107 1.30 0.23

6 – faulty sequence of tenses 30 44 1.47 0.10

7 – unrelated or false participles 111 154 1.39 0.34

No attempt is made here or in the conclusion strictly to compare one error with another; nevertheless, the massive importance of errors 1b and 1c stands out for practical classroom purposes, and some comments on this and the more refined Error Quotient are included in the Appendix on the Error Quotient on p. 269. Although there are differences between schools, the generally similar distribution of error is apparent.

e) The error count in the comparison of methods

The scores tabulated in parts (a) to (d) of this section suffice to give a picture of the general stability and relative massiveness of the errors in the schools. It is possible also to give a rough indication of the relative effectiveness of the grammar nd non-grammar methods in eliminating particular errors. The value of such an indication is discussed in the conclusion.

The errors worth considering individually, that is those of most common

occurrence, were:

1b – omission of full stop

1c – faulty use or omission of comma in lists, apposition, non-defining clauses

3d – no clear antecedent for pronoun

3e – misuse of prepositions

3f – misuse of conjunctions

4 – no finite verb, or omission of any important verbal word

5 – lack of agreement between verb and subject

7 – unrelated or false participles.

The comparisons were based on the numbers of each error in the first and final essays. The scores in the table which follows were computed by simple addition of errors made, with no allowance for opportunities for error or for length of script.

Table 18: Significance of the Differences of Mean Changes between First and Final Essays by all Non-grammar as compared with Grammar Pupils in errors 1b, 1c, 3d, 3e, 4. 5. 7 (i.e. the most common errors)

Error Diff. between means s. e. t

1b – omission of full stop 0.449 1.276 0.35

1c –omission of comma 1.725 0.470 3.67

3d – no antecedent for pronoun 0.619 0.179 3.44

3e – misuse of prepositions 0.324 0.104 3.11

3f – misuse of conjunctions 0.304 0.101 3.01

4 – no finite verb 0.560 0.160 3.49

5 – lack of agreement 0.005 0.167 0.03 (g)

7 – unrelated participles 0.265 0.140 1.89

Where the number of instances of an error was sufficiently large, a similar

comparison was made within each school, as the following tables show:

Tables 19-23: Significance of the Differences of Mean Changes between First and Final Essays by Non-grammar as Compared with Grammar Pupils in Errors 1b, 1c, 3c, 3d, 3f, 4, 5, 7 (i.e. the most common errors) within each school

Table 19: Error 1b – Omission of Full Stop

Diff. between means s. e. t

School A 0.340 0.622 0.55 (g)

School B 0.910 1.241 0.73

School C 0.938 1.298 0.72 (g)

School D 0.503 0.90 0.56

School E 1.887 0.666 2.83

Table 20: Error 1c – Omission or misuse of commas

School A 2.145 0.771 2.78 (g)

School B 2.636 0.588 4.48

School C 1.670 0.659 2.54

School D 5.380 0.733 7.62

School E 1.699 0.674 2.52

Table 21: Error 3d - No antecedent for pronoun

School A 0.143 0.268 0.49

School B 0.110 0.293 0.38

School C 0.383 0.292 1.31

School D 1.217 0.283 4.29

School E 1.439 0.444 3.24

Table 22: Error 3f – Misuse of conjunctions

School A 0.189 0.221 0.86

School B 0.657 0.364 1.80

School C 0.210 0.503 0.42 (g)

School D 0.027 0.266 0.10

School E 0.839 0.316 2.66

Table 23: Error 4 – No finite verb

School A 0.241 0.231 1.04 (g)

School B 0.623 0.268 2.33

School C 0.117 0.287 0.41

School D 0.841 0.290 2.90

School E 1.910 0.294 6.49

Table 24: Results of comparisons of errors 1b, 1c, 3d, 3f, 4 within schools

Significant gains by non-grammar forms: 5

Significant gains by grammar forms: 0

Gains by grammar form Gains by non-grammar form

School A 3 2

School B 0 5

School C 2 3

School D 5 5

School E 0 5

The significant gains were made in:

1c – omission of the comma (2)

3d – lack of antecedent for pronoun (2)

4 – lack of finite verb in sentence (1)

f. Opportunity for error

In any detailed comparison of the importance of one error with that of another,

it is not easy to find a satisfactory way of making due allowance for the combination or frequency of an error, number of opportunities for making that error, and number of words written to produce those opportunities. Some writers have based the estimate of an error’s importance on the relation of frequence of error to opportunity; others, more commonly, have related importance to frequency and the number of words written. The advantages and disadvantages of these methods are mentioned below in

Appendix 5. In the main measures of the present work the total errors were counted as one Item of measurement, rather than compared individually; and for the immediate purpose it is probably sufficient to show simply that the general opportunity for error was not less for the non-grammar pupils than for the grammar ones. The following figures may be useful in supporting this view. The non-grammar pupils wrote on the average a greater number of correct complex sentences than the grammar. 190. The difference between the mean increase from first to final essay was 3.694, which had it been feasible to amalgamate the scores of all schools would have yielded a s.e. of 0.616 and a t ratio of 6.00. They also produced an average increase of 17.28 words more than their rivals (see p. 126); of 12.30 clauses as compared with 10.50, or if all

non-grammar pupils are placed by the top third of the grammar pupils, of 12.30

clauses as against only 8.24. None of these differences is significant, but all the factors represented in them directly contribute to the opportunities for error in the very common errors.

CHAPTER VI

CONCLUSION AND DISCUSSION

The need for an investigation of the present type has been proposed in Chapter

I. The advantages of the long-term educational experiment which may be allowed to compensate for the inherent difficulty of controlling variables have been discussed in Chapter III. It is thus now possible to examine any conclusions which may be drawn from the work described in Chapters IV and V.

1. Measurable changes in children’s writing

A first outcome of the work has been the enumeration of a series of objective and reliably measurable changes that take place in the mechanics of children’s writing as the children grow older. This enumeration has been made possible by a preliminary comparison of essays written by children aged ten years and fifteen years, yielding a list of eleven measures in which the older group showed significant gains over the younger. The measures were tested for reliability, and the evidence provided by the preliminary comparisons was substantiated by the results of the later ones between the first and third or final essays of all the children in the main experiment. In addition to the eleven main measures, a number of subsidiary items gave further indications of the changes, albeit these items – which are largely referred to in the following three paragraphs – are not individually significant. They do, however, support the trends marked by the eleven main measures. Major changes were generally apparent in mechanical correctness, in the

coherence, complexity or ordering of thought, in the search for a more mature and varied style, and in fluency.

a) The children showed an increasing ability and willingness to conform to the

conventional practices in usage and punctuation, both of a gross and elementary nature, as in the correct use of certain prepositions and conjunctions. Thus at ten years of age, mistakes in the placing or omission of the full stop occurred four times as often as at fifteen; of commas, ten times as often. “Then” was given the force of a conjunction in about one in 30 sentences at ten years of age – not at all at fifteen. An idea of the improvement in general correctness may be gained from the 61% of correct sentences written at ten as compared with the 87% at fifteen; and there were only 41.3 words to each common error at ten, but 182.3 at fifteen.

b) Older children showed that they could attain to a closer expression of thought relationships. At fifteen, for example, there was more evidence of genuine modifiers between sentences – such links as “At this time,…”, “In spite of that,…” occurred between 8.2% of the sentence as against 3.5% in the younger children’s work.

Conclusions - (1) - Measurable changes

At ten, less than one sixth of the writers could paragraph their essay so that the

number of main topics equaled the number of paragraphs; but at fifteen more than half could manage this. In 6% of the sentences of the ten year olds, a pronoun would be found which had no clear antecedent; but in those of the older group this fault occurred in only 2% of the sentences. Again, a faulty sequence of tenses appeared twice as often in the younger children’s scripts as in those of the older ones.

c) This increasing ability to manipulate thought relationships demonstrated

itself in greater extension and complexity also. Correct simple sentences became, on the average, longer – from 8.7 words they rose to 13.6; and the average length of all correct sentences developed from 11.9 words to 19.5. Instances of the growth in correctly handled complexity of thought were numerous. Thus the average number of subordinate clauses per sentence rose from 0.41 at ten to 0.84 at fifteen; at ten, the proportion of correct sentences having two or more subordinate clauses to each main clause was 3.1, but at fifteen it was 13.9 (p. 148). Similarly, the proportion of simple sentences decreased with age from 43.2% at ten to 26.67%, while the number of modifying phrases in each sentence rose from 0.62 to 1.23. Two percent of simple sentences at ten years of age were modified by three or more phrases, but 12% at fifteen.

Again, at ten, only 2.1% of the subjects of sentences were modified by phrases, clause, or a group of two or more adjectives; at fifteen, this proportion was 9.5%.

d) Coupled with this increase in intellectual control was a growth in the

variety of patterns and structures used. The average number of different patterns of sentences correctly used rose from 4 at ten to 10.3 at fifteen, 9 being the highest number of different structures used in any essay by the younger group as compared with seventeen by the older. Furthermore, the younger children used 12.5% of “doublets” – i.e. groups of two sentences of the same pattern in succession – whereas the older children had only 6.9% of these. Similarly, the older children showed an enhanced ability to escape from the inevitability of time sequences in a narrative and to introduce a greater number of artistically deliberate colourings into their scripts – the 4.1% of clauses which at ten were adjectival had become 11.7% at fifteen; the number of adjectival phrases per sentence rose from 0.13 to 0.5.

e) Fluency also – the ability to extend ideas generally and to associate them

freely – was well marked in growth by the age of fifteen, but was as a measure

characterized by extreme instability and variability. But the average number of words written at ten was 218; at fifteen 453.

Naturally, not all the numerical items referred to in paragraphs (a) to (e) above are individually reliable. But the increase in the number of words per common error, in the variety of sentence patterns used, in the number of subordinate clauses and complex sentences correctly written, in the power of generally subordinating structures, and in the production of a greater proportion of correct sentences – these were all changes shown by the most reliable measuring instruments: c, d, f, h, j of the eleven used. Those measures which were only fairly reliable (b, e, i and k) showed more instability within forms, but a similar general tendency – thus full stops were omitted less frequently as the children grew older, simple sentences figured less in the total, and among them those with an inner phrasal complexity increased. A number of pointers mentioned in the comparisons above were rejected as measuring devices either because too great an element of subjectivity came into their assessment, or because although clear they were not sufficiently spread to be valuable as measures. But all tend to build up the picture of greater ability with increased age – to recognize and implement conventional usage; to control relationships of ideas more closely; to control relationships of ideas of greater complexity; to sense and to employ more varied sentence structures; to engage in deliberatestylistic colouring beyond the almost inevitable demands of a narrative; to extend and to elaborate general ideas.

2. How much formal grammar is learnt and applied?

a) How much formal grammar is learnt?

Previous work, and in particular that of Macauley and Cawley (reported on pp.

53-56) has given the strong impression that no real likelihood exists of successfully teaching formal English grammar to any but bright children. In Macauley’s work, only the top boys’ class and the top girls’ class were able to score 50% in identifying five parts of speech; in Cawley’s, although a significant mean gain (t = 3.015) was made over four years by the children of all schools, the gain by the children from other than grammar schools amounted to a mere 3%, and the highest level of their grammatical achievement after the four years was still only just above a half that of the grammar school children at the beginning of the first year. The grammar school children improved their score by about 15%, but still, as with Macauley’s Scottish groups, were far from attaining complete success in recognizing the five parts of speech by their functions.

Scores made by the forms studying grammar in the present experiment do on the whole confirm the earlier findings. Only in one form (that in Grammar School B) was there a final percentage of over 50 in the Grammar Test C after two years, and the final scores of the technical and secondary modern children were hardly higher than the initial scores of the grammar school forms.

Here as before, the less bright children reach a degree of precision in their grammatical achievements which may well be too uncertain to have any likely value as a constructive tool in shaping writing skills. This is especially so when the lack of effective tie between a relatively high grammatical score and improvement in the measured items of the essay is noted, as may be seen in the insignificant correlation of the top third of scores made in the final grammar test with changes in score between first and final essays. Here the highest positive correlation was +0.229 ± 0.138. Forms studying grammar in the present experiment appeared to work at least as successfully at their grammar as those reported in the earlier experiments. Their gains as compared with the relative stagnation in grammatical matters of the non-grammar forms were quite large – the lowest, indeed, that of the Secondary Modern School, was 12.86% (s.e. 1.835), a mean gain decidedly significant (t = 7.04). This higher rate of progress than might have been expected is perhaps due to a competitive element in the experiment – could the grammar forms outshine the non-grammar?

The formal grammar test used is not, of course, strictly comparable with the type used by Cawley and by Macauley, being much wider in scope and covering as it did most of the grammatical topics in which the children had been instructed over the two years. The apparently simple demand in Macauley’s tests for recognition of the separate parts of speech, with which most learners of English start.

Conclusions - (2) - How much grammar learnt and applied may be in fact far from simple compared with such less analytical items as the recognition of types of sentence, or of clauses as contrasted with phrases.

The low ceiling reached in grammatical attainment by the secondary modern and comprehensive/technical schools, and the far from high one of the grammar schools, may be seen as a major factor throwing doubt on the advisability of studying formal grammar in the early part of the secondary school. However, it is not in itself sufficient to exert vital pressure against grammar teaching at this stage, partly because the level of achievement itself, though low, may be sufficient to have practical value; and partly because not the direct achievement but the process of study and struggle to attain it may be the efficient feature in training the children to write adequately. It is presumably considerations such as these which have prevented the results shown by previous workers from modifying at all seriously the general practice of the English classroom.

b) How much grammar is applied?

We now come to what may be the main interest of the present work: the enquiry how effectively the grammar studied, whether it be finally grasped or not, influences the children’s competence in writing.

It will be recalled that eleven measures were used in judging the essays written at the beginning and end of the course, and that their validity as measures was established by a comparison of essays written by children aged ten and fifteen. The measures were:

a) average length of the correct simple sentence (not reliable)

b) instances of the omission of the full stop (fairly reliable)

c) number of words per common error (very reliable)

d) variety of correct sentence patterns (very reliable)

e) number of correct non-simple sentences minus correct simple sentences (fairly reliable)

f) total number of subordinate clauses (very reliable)

g) total words (not reliable)

h) number of correct complex sentences minus number incorrect (very reliable)

i) number of correct simple sentences with two modifying phrases (fairly reliable)

j) total correct sentences minus incorrect (very reliable)

k) number of adjectival clauses and phrases (fairly reliable).

There were thus five very reliable measures (having a level of 0.001 in Fisher and Yates’ tables); four fairly reliable (0.05 to 0.01); and two in themselves not reliable (0.1 in the tables). With each school considered separately, the five schools participating in the experiment produced 55 scores, 25 of which were very reliable, 20 fairly so, and 10 unreliable if considered on their own.

The results of these measures showed that in 10 out of the 25 very reliable scores, significant gains (i.e. where t = more than 3) were made by the non-grammar forms, as also in one of the 10 unreliable measures.

No significant gains were made by forms studying grammar. Again, of the 23 changes in score between first and final essays where the critical ration although less than 3 is higher than 1.5, 10 favoured the non-grammar form, only 4 the grammar. This is shown in tabular form below:

Table 25: Number of Instances of Scores on First and Final Essays where the Critical Ration exceeds 1.5

Significant gain Non-significant gain

(t = 3+) (t = 1.5+) by non-grammar by non-grammar by grammar form

Grammar School A 1 3 0

School B 3 5 0

Second. Modern C 3 3 2

Comp/Tech. School D 3 4 2

School E 1 4 0

Where an intermediate assessment was made by scoring an essay written after

nine months of the two year course (that is, in three schools, giving 33 scores) the change of score between the second and the third or final essay favoured the non-grammar form on twenty-two occasions.

The ten scores showing significant changes favouring the non-grammar forms in the five very reliable measures were made in all types of school taking part – three in the grammar schools, three in the secondary modern, and four in the comprehensive/ technical schools. The failure to profit from instruction in grammar is thus not confined to any one educational environment or category of children. It had been hoped to treat the amalgamated scores of all the schools by means of the analysis of variance. Some examples of this process are shown in Appendix 7, where items (c), (d), and (h) are tabulated, showing in two cases an F ratio at the 5% level significantly in favour of the non-grammar forms. However, it is probable that none of the other amalgamated measures would have yielded a significant ratio either for or against the grammar method as the interaction of methods by schools would have been too large, and systematic differences between the schools, or other uncontrolled variables, would invalidate the method of comparison. Furthermore, the analysis of variance would have been unsuitable for the comparison occasionally employed of the top third of the grammar pupils with all the non-grammar pupils, as the necessary equalization of the classes would have reduced their size excessively in at least three of the five instances where schools had in them only a small proportion of the top grammar pupils. Some testing by the within-classes variance of the variance of the methods interaction was undertaken, and confirmed this view (see Appendix 7b, p. 277).

Equally, the amalgamation of scores of all schools for each of the eleven main measures in order to produce an estimate of the significance of differences between grammar and non-grammar methods in the shape of a t ratio would, in view of the lack of randomization and of the probably important systematic differences between the schools in this sort of calculation, have been invalid – though it was undertaken.

Nevertheless, the main point remains, that the grammar lesson in these five schools was unreliable as a means of securing a greater mastery and control in children’s writing than could be secured with the entire neglect of grammar in English lessons, and its replacement by some form of direct practice in writing.

The gains reaching an important degree of significance, where t = 3+, were made by the non-grammar forms in five aspects of linguistic skill:

1) in the number of words per common error. Three forms, from grammar,

technical, and modern schools, gained here.

2) in the variety of sentence patterns use. There were two gains here, in a grammar and a modern school; but if a level of significance of 2+ is considered, the two non-grammar forms from the technical groups could be

included.

3) in the total words written. One grammar school scored here, or both non-grammar forms from the grammar schools if a level of significance of 2+ is counted.

4) in the number of correct complex sentences used. Four gains were made by

the non-grammar forms, from a grammar and a modern school, and the two technical schools.

5) in the number of correct sentences written. Here, one technical school

scored, and, if the 2+ level of significance is included, a grammar and a

technical school in addition.

These gains cover a wide field. Mechanical, conventional correctness (as in the number of words per common error), maturity of style (as in the variety of sentence patterns used), the control of complex relationships (as in the number of correct complex sentences), as well as general overall correctness (seen in the total number of correct sentences), were all improved significantly in groups practising direct writing skills as compared with the groups studying formal grammar.

Further evidence for the inadequacy of grammatical instruction to produce advantageous changes may be found in the scores made by all pupils in the counts of individual errors of common occurrence. These scores are shown in Tables 14 to 17. The most common errors were:

1b – omission of full stop

1c – faulty use or omission of comma in lists, apposition, non-defining

clauses

3d – no clear antecedent for pronoun

3e and 3f – misuse of prepositions or of conjunctions

4 – no finite verb, or omission of any important verbal word

5 – lack of agreement between verb and subject

7 – unrelated or false participles.

In all these, when the combined non-grammar pupils were compared with the combined grammar pupils, there was greater improvement by the former; and in four of them the improvement reached a significant level (p. 186):

1c t = 3.67

3d t = 3.44

3e t = 3.11

4 t = 3.46

Such amalgamated scores obscure the differences between schools, and will support no weight. However, when comparisons of performance in each error are made within each school separately, a similar picture emerges to that derived from the comparisons on the main measures. The five commonest errors yielded twenty-five comparisons. Of these, twenty showed an advantage to the non-grammar forms, of which five were significant, with a t ratio exceeding 3.0. No significant gains were made by the grammar forms.

Although the improvements indicated here are simply a greater decrease or lesser increase in the numbers of any particular error in the final essay as compared with the first essay, they may be held to represent a real gain. For the non-grammar pupils might have been expected to make more mistakes than did the grammar pupils.

As has been noted (p. 190) they wrote more clauses, even if not significantly more than the grammar pupils; thus they are likely to have had more opportunity for errors 1c, 3f, 4 and 5.

They wrote at greater length in total words, and used a greater number of sentences than the top third of the grammar pupils; thus they are likely to have had at least an equal opportunity for error in the omission of the full stop, lack of an antecedent for pronouns, and misuse of prepositions or participles.

The table of errors on p. 186 (Table 18), although showing gains by the nongrammar pupils, is thus weighted in favour of the grammar pupils. It is interesting therefore that in each instance save one in the scoring of the total comparisons the advantage, although statistically not reliable, is to the non-grammar pupils, and it seems sufficiently likely that the grammar lesson did not enable its students to overcome errors of the sorts described more efficiently than did its replacement in the work of the non-grammar students by direct practice in writing. The results enumerated above seem overwhelmingly to favour the nongrammar forms, but such a conclusion has to be treated with reserve. The reservation is not based on the impossibility of controlling variables adequately in this type of experiment. This has already been discussed, and the view expressed that the weaknesses in design are not decisive here. There was not, for example, any critical need to equate exactly the groups in each school, since it was possible to devise a means of measuring the progress of each group and comparing the results.

Nevertheless, in general attainment and in English in particular the pairs of forms were roughly of the same standard, as can be seen from the results of their first grammar test and their scores in the first essay analysis (p. 288). Within wide limits, it was not necessary to define the content and order of the grammar and non-grammar syllabus, since it was likely that there are many methods of encouraging non-formal progress in English writing skills, and since formal grammar itself has a vague and fluctuating meaning in present usage. Groups and teachers, furthermore, vary and must be allowed to do so in their pace of work. Limits were nonetheless set, both to the order and final amount of grammar to be studied, and a certain minimum content was stipulated and achieved. As far as possible, the teacher taking a grammar and a non-grammar group gave his best skill and attention to both, being conscious of the danger that he might not do so. All the teachers were experienced, and where two teachers took the forms, the grammar form studied under one who was keen to see grammar succeed. In particular, the difference between the work done by forms was large and simple, and could be expected after the considerable length of time to show through the obscuring influence of the ill-controlled variables an end in favour of or against formal grammar as taught in two liberal and progressive grammar schools, in one equally adventurous secondary modern school, and in the technical branches of two comprehensive schools.

It is thus for other reasons that one must express doubts about accepting

unreservedly the implications of the significant gains made in eleven instances by the non-grammar forms. For one thing, it is apparent that in no measure, either after nine months or after two years, are there significant changes in all five schools to the advantage of either grammar or non-grammar; nor does a gain made say by a non-grammar form after nine months always increase or even remain after two years. Similarly, an advantage gained by grammar forms in the second essays may have vanished or been reversed by the final ones. This last happened on only one occasion when a good level of significance was reached in both second and final essays’ scores. This was in measure (d) in the secondary modern school (School C), where a significant advantage to the grammar form in the second essay changed over to a similar advantage to the non-grammar form at the end. The work of the modern school was generally less reliable than that of the grammar or technical/ comprehensive schools, but this particularly large swing was more likely to have been due to the attention of the grammar form having been specially drawn to the topic measured under (d) – variety of sentence structure – just before the second essays were written, and to this matter having been discussed in isolation only later with the non-grammar form. Formal grammar was, in addition, a new subject to the modern school, where it is noticeable that of the eleven measures in the second or intermediate essays seven went in favour of the grammar form, as compared with five and three respectively in the grammar schools. The form was responding to the stimulus of some special attention. In the third or final essays, the figure dropped to four out of the eleven, as against five and nought in the grammar schools. The thrill of the experiment had perhaps worn off, and the difficulty of the grammar increased. Thus all that may probably be said with safety is that in five varied schools a form was taught formal grammar for two years, and fairly successfully. Yet in no school and in no measurement did the essay writing of these forms show any significant superiority over that of forms whose grammar lesson had been replaced by one giving direct practice in writing English. To say this is perhaps to say enough. That significant gains were made by the non-grammar forms is less to the point here, but that such gains commonly existed need cause very little surprise when one considers that an extra writing period in place of grammar must in fact probably double the time given each week to the actual written work in class. It seems safe to infer that the study of English grammatical terminology had a negligible or even a relatively harmful effect upon the correctness of children’s writing in the early part of the five secondary schools. Such a conclusion is reinforced by a further point, mentioned in Chapter I of the present work. This was that no high degree of correlation was found to exist between the marks gained by 285 G. C. E. candidates for their answers to the grammar question in the examination and their marks for the other parts of the paper – essay writing, précis, and comprehension. The correlation co-efficient will be recalled as 0.365 ± 0.022.

c) The value of teaching time

At this point one has to ask the crucial question – have teachers been wasting a

quarter to a fifth of their English teaching time, and are they still doing so? If the value of formal grammar as an instrument in helping pupils to write correctly is abandoned, is the rest worthwhile? Teachers have either to rebut the evidence, or to show that it has been misinterpreted, or to accept its verdict. Or of course they can ignore it and plead examinations. They can escape into the comforting belief that they teach grammar so much more effectively than did the people in the present experiment – I know that these would themselves be the first to say that this may be only too true, but in this I should certainly discount their opinion. Lack of pragmatic value is, in any case, no matter for grammarians to despair. A pure science has a fascination of its own. Jespersen remarked, probably tongue in cheek, that “the study of grammar is certainly quite useless, but it is extremely fascinating”. The lovers of grammar – and I think all English teachers must be this in some degree – are left free to chase definitions and functions as Apollo pursued Daphne, and to the same unfading if lamentably wooden end. No illusions. We are surrounded by a universe of facts, and we choose to remember that “the” and “a” always accompany nouns. This, as between consenting adults, is no harm – but have teachers the right to teach these things to children?

If there is in these reflections a polemical note which ill befits a thesis, it may be excused by the reflection which the present writer finds never remote from the contemplation of a lesson in formal grammar. This is, that the total teaching time available for the grammar lesson is no more than five days in five years; yet this little is one fifth of all the time given to English. One period a week, of 40 minutes, for 40 weeks – 135 hours in five years. Deduct 15 hours – a minimum – for loss due to revision, examinations, inattention, and occasional holidays, and five days are left. This may explain the ill-success of grammar teaching as reported by Macauley’s work (p. 53) in bringing pupils’ knowledge of formal grammar to the point of practical application. Added to the common distortion forced on syntax by the process of clause-analysis, it may show why such a poor correlation exists between success in a grammar question and success in comprehension and expression. It could, of course, provide an argument for more, not less, formal grammar, save that children seem soon to reach a ceiling in the heights to which they can rise in the subject.

W. H. Mittins (The Teaching of English, pp. 106-32) demonstrates by quotation the weaknesses of structure to be found in the writing of people ranging from schoolboy to professor. He envisages the establishment of the new grammar emphasizing “function and pattern”, dispensing with much of the traditional apparatus of rules and definitions, and with “the business of the sub-classification of parts of speech… and other complexities that derive from a preoccupation with the separable word rather than the sentence.”

Through such a grammar, descriptive rather than prescriptive, “a repertoire of the structural resources of the language… reinforced by sensitive and meaningful contact with the rich verbal patterns of expert literature”, the student may develop a sense of “appropriateness in language”. Thus a grammar systematically and regularly taught on the basis of current linguistic theory with its greater accuracy, consistency and realism, might prove fruitful.

One can only agree that such might be the case. Such a defence of grammar is

of course equivalent to the abandonment of the “old grammar”, despite the retention of some of its terms. Negligibly few teachers are as yet competent to teach, or indeed do understand, exactly what is the new grammar. The old tradition, however enfeebled, is that which like a much pampered invalid demands so much of the average schoolchild’s time at present. Perhaps simply during an interregnum while one awaits the new rule of the “realistic and practical teaching grammars”, teachers themselves should concentrate on becoming better informed about language and structure, and impart the subsequent skill by a sort of imperceptible osmosis, and not by a resented injection, to their pupils. In due course, it may be evidence will be available that a class taught by the new grammar can outfly a non-grammar class. Until then, with the failure of half a century of practice before our eyes, we cannot safely assume that this will happen. The present writer would make only the conservative reservation that of the normal allocation of time to the teaching of English in schools it would seem unwise to spend 20% on a subject which in the time is certainly unteachable and probably, as the work under consideration suggests, unprofitable to correct writing.

3. Some possible replacements for the formal grammar lesson

It would not be right to conclude, from the poor showing of the grammar forms in the present work, that the grammar period’s time should be given over, as in

this case, to direct written practice. No attempt was made to assess the most generally fruitful means of using the time made available to the non-grammar form, and it is extremely probably, in the opinion of the present writer, that a more profitable educational use could be made of the period by a wholly revised scheme of instruction in language. It is obviously not enough simply to omit grammar and to hope for the best. In the present work, the aim was to stress written correctness, and to that end the grammar period was filled in by practice specifically related to written skill. Stories, essays, magazines, newspapers, expositions of hobbies, and other matter of like nature were written up and discussed in class, with on the whole a satisfactory result as far as the limited aim went. But more is needed.

Writing is no mere technical exercise, but the expression of a large background of study in general linguistic and experiential matters. In the same way, one objection to time’s being spent on formal grammar is not simply that formal grammar does not transfer to other linguistic skills, but that its whole outlook is at present too narrow to form a valuable experience in itself, just as it is too remote to aid the children in controlling and responding to linguistic situations. To find a form of teaching which makes and makes clear the bridge between language teaching and the situation in which the child finds himself is the central problem of language teaching. It has never been satisfactorily solved. A secondary but related problem is to ensure that language is taught in a truly educative form, that is, in a form that does not simply prescribe but demands from the pupil active co-operation and choice, a sharing of responsibility for the organic growth of language itself.

The purpose of language is often said to be “to communicate” – but this is too large an overcoat to fit anyone properly. The common Civil Service ambiguities, and what Malinowski calls “phatic communion”, do not communicate in any ordinary sense of the word, any more than does the large element of repetitive and collective monologue in Piaget’s children under seven years of age.

It is safer to consider the purpose of children’s language as to control or to respond to a situation, if “situation” may be allowed to connote a reader, a listener, or even the material of an exposition. Children need to order their world, and will co-operate in learning to this end. Thus the study of language in school is primarily the practice of control through words of social situations – an experience of society, a study of manners. Teachers sometimes forget, in the hunt for errors, the relative expertise of their pupils in English. These have, for example, a practised knowledge of much of the basic syntax of language. “By five years of age, normal children have a remarkable grasp of all the main English sentence structures, and have mastered the principles of arrangement and word order.” (E. M. Moore). They will sometimes regularize a few past tenses which are irregular in common usage – though not nearly so often as the inventors of grammatical examples would have us believe. Such faults as the failure of agreement between verb and subject are not infrequent, but they occur in two main situations: first, where the child is trying to use a sentence too long and complicated for his level of attainment and the uncertain clarity of his thought processes, and second, where such everyday devices as the introductory “there” are used. Children have a deplorable habit of ignoring the objective use of “whom” – but in such things adults are not guiltless themselves. Religions thrive on guilt, and prescriptive grammar is the guilt of linguists.

As educated speakers, our own errors, if such they be, are not so very dissimilar from those of the children, but we call them usage. “To whom are you talking, Smith?” frowns the Latin master; but “Who are you talking to, Smith?” remonstrates his English colleague.

Children’s errors are mainly crudities and inelegancies, habitual slips of the tongue. Children usually know the basic structures, but need to be taught how to employ them, to vary them, and so to control the situation they are faced with.

The classroom should be a sort of laboratory for the manufacture of linguistic situations. These exist in every drama, poetry, composition and reading lesson, and indeed the practice of English in a situation has a strong effect on the learning of structure, rather than vice versa. The vital thing in English is in fact to speed maturation, to deepen the thought and to broaden the experience of children. New thoughts and experiences will need to express themselves in new words and forms. It is of little value trying to teach pupils sentence structures which are too complex for the thought relationships the pupils are able to master. At the early secondary stage, language structure studied without reference to the control and response needs of the children is a rational equivalent of the glossy paperbacks – it is language without commitment, without life.

What material then will satisfy these needs? Essentially, I would repeat, language in action. The pupils have to examine the processes of language: how do words arise, change, decay; what gives them stability; what part have dialectal forms to play, and what is standard English; what of slang and idiom; what is symbolism, abstraction, and what are their uses and dangers; how do we avoid boring a reader – how insist, or suggest, or invite, rebuff, demand, puzzle, or inform; does spelling matter; what parts of language are free for experiment and what parts must be learnt; what history is still living in the full meanings of words, and what are the links between our history and our speech – these are the sorts of question the syllabus could contain. Names children must learn; the names of names are not at this stage pertinent. The analogy between a language and a tool is often made; more relevant is that between language and muscles. As these are trained by exercising, so is language learnt. It is learnt by the manipulation of whole meanings – not lexical meaning simply, but structural meaning; not the parts of speech, but speeches. When we are run over by a car, we do not first ask if it was a Ford or a Rolls – we cry for help. The application is the learning.

In Appendix 4 are included some suggestions for the treatment of the study of language on these lines, for the making as it were of a “grammar of situation”.

4. The contents of the textbooks

To the degree to which the content of the English syllabus and the approach to

it might need to be altered if the main conclusions of the present work be accepted, it would also be only logical to revise the contents of textbooks and courses of English. But even supposing the conventional grammatical terms are still taught, there is much evident room for improvement in their presentation in the books. The major difficulty these present to children lies in the equal stress given to vastly differing topics – the need the compilers seem to feel to cover, say, all the points relating to the noun and thus to obscure the major importance of such themes as word order and relationship. Nobody knows for certain which grammatical terms, if any, are valuable in teaching, and which have no tie with practical work or one so remote as to be valueless. Terms such as “singular” and “plural”, “statement”, “question”, “command”, are thus on the same footing as “noun attribute”, “gerund”, clause of concession”, and so on. This is not the place to rewrite all the textbooks, but if 90% of the present grammatical terminology were to be abandoned it is conceivable that the grammar section of the books could be reduced to let us say some graded lists of the “formulas”, idioms, terms of greeting etc, to one or two basic rules of word order, and perhaps such terms as “singular and plural”.

That a revised language course would call for a new type of school grammar

book, or “language book”, is evident. In such a book there might well be a certain proportion of material with which the English teacher himself might not always, as things are at present, be thoroughly acquainted. When the importance and the volume of basic language learning is contemplated in the setting of the five days available, it would seem astonishing that so much effort and so many hours are devoted to traditional formal grammar.

Further discussion of this topic has been relegated to an Appendix (pp. 246- 68), but that it has to be considered is an important consequence of the present work.

5. The nature and importance of common errors

One of the major difficulties in compiling a textbook in grammar and in assessing its value is that little organized knowledge is accessible in Britain of the relative importance of the errors children make in language or of the difficulties they experience. Little work has been done in this field, compared with the massive studies of American writers. However, the hints that may be drawn from the present work go to confirm the general conclusions of these writers, that the number of different errors is in reality fairly small, and that local errors need specific attention. The common errors used in the present work were selected from the writing of the ten-year-old and of the fifteen-year-old children who provided the preliminary essays from which the measuring instruments were adapted, and were then confirmed as being “common” by a group of teachers of English. With one exception, these errors did indeed recur throughout the work of the five schools, but not by any means in equal proportions. The following list shows in what proportion of essays the errors were found in the combined schools (but see pp. 181-4 for more detailed scores):

Table 26: Total Error Scores in 456 Essays No. of essays No. of occasions containing the error on which the error is made

________________ ______________________

1b 361 1500

1c 357 1207

3a 15 16

3b 5 6

3c 31 34

3d 178 284

3e 103 135

3f 199 329

3g 36 42

3h 0 0

4 145 214

5 82 107

6 30 44

7 111 154

The main errors here are in the use of the full stop and comma (1b and 1c), failure to give pronoun a clear antecedent (3d), mismanagement of conjunctions (3f), omission of an important verbal word (4), wrong use of prepositions (3e), and to some extent the lack of agreement between verb and subject (50, and the use of unrelated verbal particles (7). Other errors are made relatively seldom and by relatively few people. These figures tally quite well with those reported by previous workers. Clearly, not all errors are equally important, and not all are equally spread even within the schools in one town. It may well be that the local variability is greater than any communality of error. In particular, there appeared a noticeably greater weakness of the Secondary Modern School in the basic errors 1b, 1c, and 3f (full stops, commas, and conjunctions). It would be difficult to devise a textbook which would be usable on the country-wide scale that is necessary for commercial publishing. It is likely that certain main errors such as 1b, 1c, and 3f above would be found in a large variety of places and schools, but the need in the present state of linguistic knowledge for teachers to compile their own case-book for each class is apparent.

APPENDIX 1

A list of common errors in children’s writing

1. Certain errors in puctuation, notably:

a) omission of question mark

b) omission of full stop

c) omission of comma between items in a list, words in apposition main and non-defining clauses.

d) omission of apostrophe in ’s for possession.

2. Failure to use capital letters

a) at beginning of sentence

b) in proper nouns and adjectives.

3. Misuse of various parts of speech:

a) adjective or preposition as adverb

b) wrong comparatives and superlatives

c) faulty positioning of adverbs

d) failure to give pronouns a clear antecedent

e) wrong or ambiguous use of prepositions

f) mismanagement of conjunctions

g) misuse of object forms of pronouns

h) failure to use relative pronoun or adverb.

4. Failure to give a finite verb to each clause, error in or omission of any

important verbal word.

5. Lack of agreement between verb and object.

6. Faulty sequence of tenses.

7. Unrelated or false participle, e.g. “laying” for “lying”.

In the error count referred to in the pilot experiment, and in the final reckoning

of the third measuring instrument, errors 1a, 1d, 2a, and 2b, were omitted from the calculations, as being too remote from grammar as such, and too simply mere matters of convention which training in formal grammar should not be expected to help in eradicating.

APPENDIX 2

Grammatical content of some courses of English published or reissued since 1944 In the table of contents which follows, it has not always been easy to say

exactly how much of the book can be included in the category “formal grammar”, since exercises could often serve a general as well as a formal purpose. However, an attempt has been made to include only those exercises, in estimating the number of pages given to formal instruction, which are directed primarily to the recognition of terms rather than to the general practice of linguistic skill. Naturally, much would depend on the teacher’s use of the exercises.

Date No. Title a) Publisher Total Grammar % of

b) Author pages pages grammar 1958 1. English for Schools a) Bell 206 26 12.6

b) Venables 1956 2. English with a Purpose a) Nelson 148 26 17.6

b) Paterson 1956 3. Understanding English a) Arnold 180 30 16.7

Book I b) Cunningham 1958 4. A First Eng. Course a) Longman 240 77 32.1

b) Burton 1957 5. Eng. for Lower Forms a) Harrap 145 30 20.6

b) Lamb 1954 6. Active English Course a) ULP 189 78 41.3 Book I b) Humphries 1944 7. A First Modern English a) Dent 228 30 13.2 Course b) Jones 1950 8. An English Highway a) Longmans 178 30 16.9

b) Moon 1952 9. English on the Anvil a) Murray 199 54 27.1

b) Britton 1950 10. Everyday English for a) OUP 150 18 12.0 Seniors b) Sussams 1948 11. English Today a) Ginn 208 40 19.2

b) Ridout 1957 12. Queen’s English for a) Wheaton 122 14 11.6 Seniors (a very simple book for secondary modern schools) 1952 13. English Explained a) Schofield & Sims 151 23 15.2

b) Barnes 1959 14. Progressive English a) McDougalls 118 25 21.1 for Seniors b) Ogilvie 1958 15. Background English a) Arnold 120 27 22.5

b) Roche Average 35 19.3%

Where more than one author has co-operated on a book only the first name is given above. There follows a summary of the grammatical terms used in these books.

Table 27: Grammatical Terms Used in the Fifteen Named Courses of English

(It will be remembered that these books are intended chiefly for the first or first and second years of the secondary school.) Parts of speech Number of book

a) Noun All

abstract 1, 3, 4, 5, 6, 7, 11, 13, 14, 15

common 1, 2, 3, 4, 5, 6, 7, 11, 13, 14, 15

proper 1, 2, 3, 4, 5, 6, 7, 11, 12, 13, 14, 15

collective 1, 2, 3, 4, 5, 6, 7, 11, 12, 13, 14, 15

concrete 4, 14

-- as adjective 1

verbal noun or gerund 1, 8, 14

gender of noun 3 (masculine, feminine, common, neuter)

11, 13, 14 (masculine and feminine)

15 (masculine, feminine, neuter)

number 2, 3, 6, 13, 14

compound 15

-- of multitude 14

-- of material 14

b) Definite and Indefinite

Article 4, 7, 11, 13, 15

c) Adjective All

quantity/ quality 4, 6, 13, 14

epithet or descriptive 4, 7

predicative 4, 6

of number 14

possessive 4, 8, 15

proper 14

interrogative 4

distributive 15

demonstrative 4, 7, 14, 15

as noun 1

comparison 2, 7, 9, 14

superlative 2, 7, 14

d) Conjunction All except 10 and 12

subordinating 14

co-ordination 14, 15

compound 14

introductory 14

double 15

e) Pronoun All

personal 2, 4, 5, 6, 7, 13, 14, 15

emphasizing 14

possessive 8, 13, 14, 15

reflexive 14

interrogative 1, 4, 6, 14

distributive 14

demonstrative 4, 6, 14, 15

relative 2, 3, 5, 7, 8, 11, 13, 14

of quantity 4

antecedent to 14, 8

f) Preposition 1, 3, 4, 5, 6, 7, 8, 9, 14, 15

g) Exclamation or

Interjection 4, 6, 7, 8, 9, 11, 14

h) Adverb All

time 2, 4, 5, 6, 7, 11, 13

place 2, 4, 5, 6, 7, 11, 13

manner 2, 4, 5, 6, 7, 11, 13

degree 4, 7

of negation 4

of number 4

interrogative 4

i) Verb All

active 1, 7, 8, 14, 15

passive 1, 7, 8, 14, 15

transitive 1, 5, 6, 7, 11, 14, 15, 8

intransitive 1, 5, 6, 7, 8, 11, 14, 15

impersonal 8

finite/ non-finite 1, 5, 7, 8, 9, 14

weak/ strong 15

regular/ irregular 14

defective 15

auxiliary 1, 6, 8, 11, 14

mood –

indicative 1, 8, 14

imperative 8, 14

subjunctive 8, 14

tense –

present 2, 3, 5, 6, 7, 8, 10, 11, 12, 13, 14, 15

present continuous 8, 11

past All except 4, 9

past continuous 1, 6, 8, 11

past perfect 1, 8

future 2, 6, 7, 8, 13, 14, 15

future perfect 8

future continuous 6, 8, 11

participle 1, 2, 3, 7, 8, 10, 14

infinitive 5, 8, 14

person/ number 1, 6, 7, 8, 11, 14

agreement 1, 6, 7, 8, 11

Phrases

participial 1, 7

prepositional 1

adjectival 1, 4, 5, 6, 8, 9

adverbial 1, 4, 5, 6, 7, 8, 9

noun 7, 8

Clauses

noun 7, 14

adjectival 7, 9, 14

adverbial 7, 9, 14

main/ principal 14

subordinate 14

Types of sentence

simple 2, 4, 7, 14

compound 7, 14

complex 7, 14

compound-complex 14

Analysis of simple sentences

into columns 4, 5, 7

graphic 6, 8, 11

direct object 1, 3, 4, 5, 6, 7, 8, 10, 11, 14, 15

indirect object 4, 6, 8, 14, 15

Parts of speech

subject and predicate All except 12

complement 6, 7, 8, 15

apposition 1, 8, 15

Other points

Case

possessive 1, 8, 11, 14, 15

nominative or

subject case 1, 6, 8, 14, 15

dative 8

accusative or

object case 1, 6, 8, 14, 15

vocative 15

Antonym 4, 7, 11, 13, 14, 15

Synonym 1, 4, 7, 11, 13, 14, 15

Homonym 7

Ampersand 4

The change in textbooks towards attempting a closer relationship between grammar and composition is noted by Rivlin (p. 38) as is the increasing percentage of space given to exercises rather than to rules. The exercises in British textbooks, however are often primarily devoted to the recognition of terminology rather than to the acquiring of linguistic skills. “During the past four years”, writes C. D. Poster (“The case against grammar”, The Use of English, vol. XI, no. 2, 1959), “I have reviewed a number of language study textbooks. Without exception, the authors of those for the primary (and secondary) school appear convinced that the ability to recognize and name – often in some detail – the seven parts of speech is essential”.

By comparing the list of the items given in this appendix with those of the four experts in Rivlin’s study (see p. 58) the large proportion of material probably functionally irrelevant in the British books may be seen.

APPENDIX 3

Preliminary test scores

Table 28: Scores on Preliminary Tests A, B, and C, made by children aged 11+, 12+,

13+ and 14+

TEST A TEST B TEST C

11+ 12+ 13+ 14+ 11+ 12+ 13+ 14+ 11+ 12+ 13+ 14+

82 96 76 82 74 94 98 86 10 14 17 24

84 84 74 84 68 72 88 86 8 15 29 40

92 86 84 82 92 90 76 86 3 13 25 33

82 82 80 92 84 84 80 86 13 14 13 38

88 82 86 78 82 88 94 78 9 12 18 27

84 90 74 84 88 80 74 84 1 19 21 30

78 82 88 86 76 80 70 82 10 22 29 38

90 82 70 78 92 88 82 80 13 4 20 32

78 74 82 74 86 76 74 88 2 19 21 33

88 80 86 82 84 82 82 76 14 8 28 33

86 84 82 90 70 88 84 88 10 13 31 45

78 74 78 82 66 76 72 66 11 15 12 25

68 84 84 80 66 82 84 78 6 16 23 34

82 82 86 72 84 92 82 48 7 10 23 19

80 82 80 82 78 76 82 88 1 8 18 31

84 84 84 80 84 82 92 84 0 12 21 19

76 80 86 74 84 72 66 88 18 9 21 29

92 86 88 82 66 72 86 86 15 23 20 44

84 68 88 90 92 82 70 86 6 8 34 30

82 72 82 82 88 72 78 82 11 5 19 32

88 88 86 86 68 72 86 94 7 8 30 48

84 80 82 76 70 86 84 80 3 8 22 23

84 74 76 80 88 80 94 78 9 7 26 34

84 86 84 88 94 84 26 12

78 86 76 82 100 90 12 38

88 72 88 15 19

74 68

Averages

83.2 81.5 82.0 81.6 79.5 81.0 82.5 81.9 8.1 13.0 22.8 33.1

Spread

68/92 68/96 70/88 72/92 66/92 72/94 66/100 48/94 0/18 4/22 12/34 19/48 Tests A and B were designed to estimate ability to recognize common errors,

but seem in effect to have tested carefulness rather than skill, since the fourteen-year old children did no better than the eleven-year-old. These two tests were therefore scrapped, in favour of measurements of the ordinary and connected written work of the children, in which it was expected to test available skill and not simply care and recognition.

Test C, on the other hand, was a grammar test, and showed general improvement in knowledge over the years. This was therefore retained, and administered to the children in the experiment at the beginning and end of the course. Copies of these three tests follow this page.

TEST A

In each of the following items, a choice of words or punctuation is offered.

One choice is correct, the other incorrect. In each item, cross out the incorrect form. Work carefully but as fast as you can, and if there is any item you cannot

manage, move on to the next. Read these two examples before you begin the rest:

Example (a) monday.

We go to school on Monday.

Example (b)

He only paid for one book, but he was given two.

He paid for only one book, Now do the rest yourself.

worse

1.   This one is the of the two.

worst

_____________________________________________________________

what

2. The man he saw was a negro.

that

_____________________________________________________________

he

3. The rain continued. wished he had not come.

He

______________________________________________________________

yet?

4. Is it time

yet.

______________________________________________________________

are

5. That sort of book hard to understand.

is

______________________________________________________________

like

6. They tried to win they always had done.

as

______________________________________________________________

Rain was

7. They trudged along. falling heavily all the time.

Rain

______________________________________________________________

8. As they have been successful,

the prize will be awarded.

Considering their success,

______________________________________________________________

can

9. The boy asked whether he open a window.

might

______________________________________________________________

10. Will

you come if you could?

Would

______________________________________________________________

11. Hoping they would see him again,

when it was all over.

They hoped they would see him again

______________________________________________________________

12. The Red Indians and the cowboys were fighting. The Red Indians were in the

they

right, but said the Indians were to blame.

the cowboys

______________________________________________________________

However,

13. the boy was eating as if he would never stop, hoping to finish before

While

the others were ready.

______________________________________________________________

and

14. He came along in a rage an overcoat.

and was wearing

______________________________________________________________

rabbit’s

15. A tail is short.

rabbits

______________________________________________________________

were

16. If air poisonous, no-one would want to breathe it.

was

______________________________________________________________

football

17. Boys often play in the winter.

Football

______________________________________________________________

didn’t do

18. All the quack-doctor’s medicines no good.

did

cost.

19. How much did that

cost?

______________________________________________________________

food, ammunition, beads, cameras.

20. The explorers carried

food ammunition beads cameras.

______________________________________________________________

John’s house.

21. They went to

Johns house.

______________________________________________________________

after

22. We were invited to the party, and we went to bed.

afterwards

______________________________________________________________

was

23. The boy running. He wished he were home already.

is

______________________________________________________________

is

24. A general and a captain soldiers.

are

______________________________________________________________

any

25. They didn’t have at the shop.

none

______________________________________________________________

26. He was running

along the road.

Running

___________________________________________________________

because

27. The reason for the error was nobody had seen John.

that

______________________________________________________________

can

28. He thought he manage to be in time.

could

______________________________________________________________

stopped the

29. The horses train went on.

stopped. The

______________________________________________________________

its

30. The dog chased tail.

it’s

______________________________________________________________

Frenchman.

31. He knew that his enemy was a

frenchman.

______________________________________________________________

paris

32. He wanted to go to for his holidays.

Paris

______________________________________________________________

to

33. Come my house and have a game of cards.

up

______________________________________________________________

quick

34. They walked to the shore.

quickly

______________________________________________________________

35. This idea, although not a new one,

had not occurred to them.

This idea although not a new one

______________________________________________________________

twelve. The

36. The clock struck dial had been polished.

twelve the

______________________________________________________________

will

37. “If I gave you a million pounds, you give me a toffee?”

would

______________________________________________________________

will have corrected

38. When you finish this text, you fifty items.

are correcting

______________________________________________________________

39. As they wandered

along the road, the tractor was to be seen ploughing the field.

Wandering

______________________________________________________________

is

40. There three rabbits in the field.

are

______________________________________________________________

41. He and I

are going to play football.

Him and me

______________________________________________________________

42. Bill Smith my best friend

lives in London, as I do.

Bill Smith, my best friend,

______________________________________________________________

only try once.

43. You are not allowed two tries. You must

try only once.

______________________________________________________________

was

44. “I wish I a king”, said the poor man.

were

______________________________________________________________

“Hallo there!” cried a voice.

45. They waited.

“hallo there!” cried a voice.

______________________________________________________________

46. He clenched his fists, for the

pain seemed hard to bear.

Clenching his fists, the

______________________________________________________________

was

47. If I you, I would try again.

Were

______________________________________________________________

48. “May

I be excused, please?” asked the child.

“Can

______________________________________________________________

was

49. The men planning to go on strike.

Were

______________________________________________________________

was

50. None of the soldiers given a medal.

were

______________________________________________________________

TEST B

In the story which follows, a choice of words or of punctuation is often given.

Read the story. Then read it a second time, and cross out the incorrect form each time a choice is given. Death and the Robbers was Three angry robbers ready to fight anybody, even Death. They went were challenged older out and an old man to tell them where Death was to be found. The challenge oldest

_______________________________________________________________

Grimstone, me,” robber, whose name was kicked the old man. “Why do you harm grimstone, me?”

______________________________________________________________

was he asked. “I do not know where Death lives. But if I to tell you where you might were

_______________________________________________________________

would alone? find him, you leave me Can I depart safely then?” will alone.

_______________________________________________________________

“Yes, for sure”, the robbers promised.

_______________________________________________________________

The old man told them that Death lived by a cave beyond the town wall. Being

_______________________________________________________________

eager to find their enemy, they soon left the wall behind them. Grimstone wanted to the wall was soon behind them.

_______________________________________________________________

thorough.

do the job He ordered the younger robbers to act as scouts – Marner the

thoroughly.

_______________________________________________________________

Scotsman french, He

To the left, and Salaud, who was to the right. did not

scotsman French, Salaud

_______________________________________________________________

Grimstone’s after quarrelled

understand instructions, and they about it. The

Grimstones afterwards quarrel

_______________________________________________________________

being, because finding

reason the robbers were not very clever. However, a cave

was that they found

_______________________________________________________________

soon, but Death was nowhere to be seen. Instead, the robbers found a bag full of gold

_______________________________________________________________

doubloons moidores sovereigns.

coins –

doubloons, moidores, sovereigns.

_______________________________________________________________

They as

acted you would expect: those coins drove all thought of

The robbers like

_______________________________________________________________

Death out of their minds. They yelled and danced for joy and forgot Death.

_______________________________________________________________

He was waiting

where they least expected him.

Waiting

______________________________________________________________

do?” Grimstone the chief was

“What shall we asked after a while. There three

do,” Grimstone, the chief, were

_______________________________________________________________

him and them,

hours to sunset, and they planned, to have a feast. Then, after dark, they

he and they,

______________________________________________________________

secretly sent

would creep into town and share the loot. Marner they off to buy wine

secret send

_______________________________________________________________

and food.

______________________________________________________________

“What you thinking?”

asked Grimstone suddenly. Salaud was wishing that

“What are you thinking?”

_______________________________________________________________

were

there only two of them to share the gold – and that Marner was dead. “And

was

_______________________________________________________________

not?” “who on?”

why snarled Grimstone. is he to waste good money

not.” “Who by?”

_______________________________________________________________

Salauds,

“Yes – we’ll knife him when he gets back! Then it will be my gold,

Salaud’s

_______________________________________________________________

else’s.”

and yours, and no-one

elses.”

_______________________________________________________________

He was thinking

Marner, meanwhile, was on his way to town, out a plan of

thinking

_______________________________________________________________

was being

his own. If three to become one, one would have quite a lot to spend – such

were were

_______________________________________________________________

his thoughts. So he bought poison, and put it into two of his three bottles of

_______________________________________________________________

wine one should be,

without poison he tied to his belt in case there by any chance, a

wine. One would be,

_______________________________________________________________

were dividing

mistake. Sure enough the other two quick to set on him and the

was to divide

_______________________________________________________________

As they seized

spoils. the two poisoned bottles, the gold seemed to glitter in their

Seizing

_______________________________________________________________

If were

greedy eyes. they hadn’t drunk some of the wine, they all right.

if would have been

_______________________________________________________________

The not only greedy

There was no chance of that, you can be sure! two robbers, for

the greedy not only

_______________________________________________________________

was

the gold, just as greedy for the wine. Down it went and down they went with it!

were

_______________________________________________________________

Down among the dead men! They had found Death after all, and just where the old

man had promised.

were

But even if there no mistake in this story as I have told it, you

was

_______________________________________________________________

enjoy

more the version in Chaucer’s Canterbury Tales.

would enjoy

_______________________________________________________________TEST C

There are four sides to this test.

Answer as many questions as you can.

You may answer them in any order.

Write your answers in pen or pencil in the space provided at the right-hand

side of the paper.

Put your name, age, and the name of your school at the top.

Q. 1. Name the part of speech of each of the underlined words in the following

sentences:

a) The ship sailed. a)

b) It went on a long voyage. b)

c) The pirate crew landed on an island. c)

d) They searched for treasure. d)

e) They wandered over the island. e)

f) They sat down exhausted. f)

g) This may be right or wrong. g)

Q. 2. Write out the subject of each of the following:

a) The boat is aground. a)

b) Who ate the cake? c)

c) There was no chance. c)

d) A soldier and a sailor were there. d)

Q. 3. Write out the direct object in each of the following:

a) Give him the book. a)

b) You should read this. b)

c) The two friends spoke the truth. c)

d) I know he is a friend. d)

Q. 4. State which of the following sentences is a command, a question, an

exclamation, or a statement.

a) Stop at the gate. a)

b) There is no time to do that. b)

c) Confound it! c)

d) Why should anyone do that? d)

e) Come with me. e)

Q. 5. Name three types of noun, and give one example of each:

a) a)

b) b)

c) c)

Q. 6. Name three types of verb, and give one example of each:

a) a)

b) b)

c) c)

Q. 7. Name two types of pronoun, and give one example of each:

a) a)

b) b)

Q. 8. Name two types of adjective, and give one example of each:

a) a)

b) b)

Q. 9. State whether each of the following underlined groups of words is a phrase or a

clause:

a) He knows when to stop. a)

b) They will soon find out what he thinks. b)

c) I did not realise that he lived there. c)

d) Do try to keep from laughing. d)

e) Say when you mean. e)

Q. 10. State whether the verbs are used transitively or intransitively in each of the

following sentences:

a) He kicked the ball. a)

b) There was no hope of a goal. b)

c) The gardener was growing old. c)

d) The roses looked splendid in bloom. d)

e) The weather seemed to improve. e)

Q. 11. Name the tense and the mood of the verb underlined in each of the following sentences:

Tense Mood

a) He is at home. a) a)

b) He will be there. b) b)

c) Give the boy his cap. c) c)

d) If only it were true. d) d)

e) Nobody was at home. e) e)

Q. 12. Explain, in grammatical terms, what is wrong with each of the following

statements or word-groups:

a) Walking down the road, the aeroplane was seen to crash. a)

b) There was not more than two sentries on duty. b)

c) The man hoping to find out his error. c)

d) They ran quick to the shop. d)

e) He gave the book to my friend and I. e)

Q. 13. State whether the underlined words in the following sentences are direct object, indirect object, or complement:

a) The thief turned informer. a)

b) He turned the book over. b)

c) The gardener grew some radishes. c)

d) The carpenter grew infirm. d)

e) Give the man a chance. e)

f) Take the man with you. f)

g) If you allow him leave he will go. g)

h) He said “him” not “her”. h)

i) The civilian seemed the man he knew. i)

j) It is I. j)

Q. 14. State whether the underlined infinitive in each of the following sentences is being used as a noun, adjective, or adverb:

a) He was hoping to come. a)

b) He had no coat to wear. b)

c) They were certain to win. c)

d) Ask the Librarian to lend you a book. d)

e) That one was difficult to understand. e)

Q. 15. Name the part of speech of each of the underlined words in the following sentences:

a) The ball is round. a)

b) It rolled round the corner. b)

c) It span round and round. c)

d) The cow jumped over the moon. d)

e) A wicket fell in the third over. e)

f) “Over” is a common word. f)

g) This is better than that. g)

h) That book is well illustrated. h)

i) Fancy that! i)

j) Pick your fancy. j)

k) Here is a well without a bucket. k)

l) He spoke well. l)

m) The water may well up through the soil. m)

n) Whatever you do I will do. n)

o) That is no good whatever. o)

p) The goat will butt the boy. p)

q) “But” is an unpleasant word in a treaty. q)

r) The truth will do but nothing else will. r)

s) Paul was a Roman soldier. s)

t) A Roman never faltered. t)

APPENDIX 4

Some suggestions for the treatment of the study of language intended to replace the traditional instruction in formal grammar in school The teacher’s best approach to language in the English lesson is by the control/ response situation. In what follows, I hope to outline the material which I feel should be covered in a language course before the O-level G. C. E. examination. It is not enough to study English in isolation from the topic of language in general. I do not mean that children should be set to chase the will o’ the wisp of a universal grammar by comparing the grammatical structures of English, French, Latin, and so on, but that they should be asked to think about the nature of the instrument which is language:

The study of language is only too often regarded as being a matter of examining such things as pronunciation, spelling, vocabulary, grammar, and sentence structure… but, as we all know from everyday experience, learning language is not simply a matter of learning words: it is a matter of correctly relating our words to the things and happenings for which they stand, and begins properly with a study of what language is about. I have quoted here from S. J. Hayakawa’s book Language in Thought and Action (Allen & Unwin, 1952) to which I am indebted for many of the ideas that follow.

A. Material for the study of language in general

A language is a system of agreed signals, the most complicated and adaptable

of such systems that human beings use. Although other systems may have advantages on particular occasions, the limitations of gesture, flags, lights, smoke, notched sticks, knotted ropes, whistles, and drums, are very severe. All, compared with speech, are inflexible, limited in the range of meanings they can carry, and in the relationships they can represent. Language has to be a system; a few code sounds invented by a schoolboy are not language. It has to be intended to convey meaning to someone else, to be a signal: totally private symbols are hardly a language. And it has to be agreed, that is, to have at least two participants; for without this prior agreement, communication is not possible; the signals will not work. Since everyone recognizes the value of international understanding, children might be asked to think why Esperanto and similar structures do not flourish – perhaps belonging everywhere the artificial language belongs nowhere. A really successful language must therefore have not only a good system of agreed signals, but must have also the deeply involved personal concern of its speakers. Real language (as distinct from mere exclamations) is the means and expression of human co-operativeness. Through language, by becoming less individual we become more ourselves; we give up some of our differences to gather what is valuable to each of us from the similarities between us. It is not difficult to demonstrate to children the limitations of their own, or of one’s own knowledge, experience, and skill, and then to show how books and speech can enlarge for them the world to which they can respond. Our nerve endings finish where language falls silent. Children like to speculate about the origin of speech, and their guesses may be very similar to the common theories – what Partridge calls the bow-wow, the poohpooh, the ding-dong, and the ho-ho-ho theories of origin. Whatever theory they arrive at, it is not hard to show that agreement about the denotation of the sounds, or that imitation which is natural agreement, is essential to the establishing of effective language. Influenced maybe by the order of chapters in courses of English, they may hazard a guess that the first words were names of things. However, in watching infants one may notice that a relatively complex situation is among the primary names

on occasion. “Bye-bye”, the first recognizable word that my elder daughter spoke, meant roughly “I see that you are going and I am filled with desolation, anger, fear.” Children can learn much from trying to write down what other children say. Speculation and observation of this sort should lead to an examination of some of the assumptions that we make about language, and of the process such as symbolization and abstraction that enable language to work smoothly or not at all. It may seem that such a discussion is above the heads of children. But provided clear examples are chosen, and provided pupils are asked to think out like instances for themselves, no difficulty arises that is not well repaid by results, if only by the realization that “human fitness to survive means the ability to talk and write and listen and read in ways that increase the chances for each of us and fellow members of our species to survive together.” Two assumptions that need deliberate investigation are: that there is an identity between words and the things they denote; and that words and ideas can be set in order without the intervention of words or symbols. The former assumption is dangerous because it leads to a confusion of fantasy and reality; the latter because it relieves anyone making it of the necessity to set his words in order.

Human beings can by agreement make anything stand for anything. But children will often feel that there is an inherent connection between the object and the symbol. Piaget gives an example in his book The Child’s Conception of the World: “Could the sun have been called moon and the moon sun? – No – Why not? – Because the sun shines brighter than the moon. – But if everyone had called the sun moon and the moon sun, would we have known it was wrong? – Yes, because the sun is always bigger. – Yes, but the sun isn’t changed, only its name. Couldn’t it have been called … etc. – No, because the moon rises in the evening and the sun in the day.

As they grow, children become enslaved to the power of the symbol, in

advertisement, propaganda, education. As there is a break between language and reality, so also there is a positive relationship, and we can tell truth only when our language conforms to that relationship. We don’t see things exactly as they are, but abstract a few general resemblances between one and another. It is this which makes general thought and calculation practicable. Hayakawa uses an “abstraction ladder” to illustrate this process:

wealth

asset

farm asset

livestock

cow

“Bessie”

the cow we see

More and more of the perceived and individual characteristics are left out as the degree of abstraction increases.

Abstractions are useful as we recognize them for what they are – things remote from the extensional or real object. If we forget this, false judgements, prejudices, rumour, may spread from our omission. We may use emotionally charged words, such as Russian, chorus-girl, teddy-boy, Jew, nigger, in a hostile generalization, forgetting that each of these abstractions is only one of a thousand that may be applied to a particular individual. Children in this way will often think in terms of good/bad, hero/villain, and so lapse from reality. Even expert grammarians may fall into the same error – as when Eric Partridge, following Jespersen, talks of English as “manly and vigorous, masculine, virile”, on the basis of such remarks as that “The Englishman does not like more words than are necessary … or to comomit himself to being too enthusiastic”. So much can and should be said before bringing the children to English as a particular example of language..

B. Study of English language

1. The spread of the language

Some knowledge of the geographical spread of English and of its historical development check parochialism and intolerance in children. Evidently, it is unlikely that the 50 million population of Great Britain can expect to be the only arbiters of language for a tongue spoken by about 350 million people. This wide spread sharpens the need for standardization and simplification of forms: some regularisations are still mere vulgarisms, but others are continually becoming established. “The bird flew out the window” may now be commonly heard, on the analogy of “in”; and the blurring of distinctions as between “shall” and “will” is part of the cost of exporting and universalizing a language. Just saying “no” is vain: Partridge writes that the form “don’t” is “no longer permissible”, for example.

2. History.

To tell children that many modern languages arose only recently means little, and the hypothetical forms of the Indo-European languages fall on deaf ears. But a brief resumé of the family and personal history of English can be important in children’s study of language. It reveals to them the perpetually changing process of language and their own indebtedness to history. Simple evidence, for example, can be assimilated, of links between the Romance languages and between the West Germanic, with particular reference to English relatives in these groups. But the essential is not the mastery of much detail, but the realisation of the general trend, and of the ostrich-like digestion of English. The minimum detail should be regarded as the maximum, always provided that sufficient example are given to bring historical process to convincing life. The landmarks must be clear – the coming of Christianity, the Scandinavian invasions, the Norman conquest, the influence of American and colonial forms. Children should grasp the enormous enrichment of vocabulary, the simpler patterning of accidence, the move towards a standardized form of pronunciation and spelling, and the increasing subtlety of the general structure of sentences, especially the larger proportion of subordinated groups in modern English as compared with say that of the Anglo-Saxon Chronicle.

The perception of change and growth realizes for children the organic tie between language and the changing circumstances of life around them. It persuades them that they are the present instruments of change, and must share the responsibility for it. They should therefore study its causes, and the means also of stability.

3. Agreement and disagreement.

Children need to know something of language’s contribution to these topics. It is worth asking them to consider words not as parts of speech but as areas of meaning or intention: for example: [central “round” links on the one hand to applause/ spherical/ robin; and on the other, to golf/ table/ the corner] Children are curious to find, say, that the New English Dictionary gives over 200 main senses of the word “set”. In most such instances, the right division will be distinguishable by context. But when the context does not help, confusion results. This happens commonly with many abstract words, words covering a large area of meaning, as well as with words broken into numerous barely related meanings whose ambiguities may at times be resolved by grammatical form. If a word of small area is taken, a simple linguistic molecule, disagreement is unlikely:

of applause of golf

spherical

round

table

robin the corner

hydrochloric acid

U.S.A. U.S.S.R.

But a word of larger area is another matter:

Love

Here there is considerable overlap, but plenty of room for misunderstanding. Or take:

Marriage

G.B. Hollywood

France

There are points of contact, but that is all. Freedom is another word in which the area outside agreement may be greater than that within:

Freedom U.S.S.R.

- although of course each concept may have a local validity.

Disagreement arises too from our calling the same thing by different words, reflecting thus our various attitudes:

swot praiseworthy boy rat catcher pest eradication officer

G.B.

Turkey

France

Freedom U.S.S.R.

Freedom U.S.A.

Freedom G.B.

guide, philosopher and friend

teacher

schoolmaster

schoolteacher

schoolma’am

pedagogue

pedant

instructor

a regular Squeers

Then of course we are quarrelling about our attitudes, rather than about our inability to understand It is still true, however hard for linguistic reformers to swallow, that human beings fight about things before they fight about language.

From this, it may seem to children that words of small area, such as

“hydrochloric acid”, are the “best” because they cannot lead to disagreement. But of course situations not unambiguous can hardly be described in unambiguous language. There are varied styles of speech and writing. Vocabulary depends on wishes, technical needs, and social aspirations.

4. Vocabularies

By the vocabularies we choose, we select our social groups – both those in

which we would be included, and those we would exclude. Children growing up are in continually changing social situations, especially in their relationship to adults. Technical terms however complicated are not really so difficult for children to learn as the apparently simpler tonal language of everyday speech, and this is because they are of small area.

Everyday terms, on the other hand, need more experience of life for their adequate manipulation. It is not hard to help children to grasp that they do use words that relate only to their own feelings. They do so every time they make a judgement without having a clear idea of its grounds: “This is a smashing book”; “Maths is a lot of rot”. They benefit from a few examples of the temperature chart of certain words – how from a datum line these have risen or sunk in the estimate of history: words such as “silly”, “cunning”, “intellectual” – or King James on St Paul’s Cathedral: “amusing, awful and artificial”, that is, pleasing, awe-inspiring, and skilfully done. Less important than the lexical meaning of words are the spelling and pronunciation. Within ordinary limits, pronunciation does not affect intelligibility, though localisms can serve, whether in Oxford or Bow, to define social groups to protect or exclude. Tape recordings of a passage of dialogue between people speaking a wide variety of dialects are valuable for bringing home to pupils the great tolerance of the language for variation in sound, and one would guess that educated speech is willing to accept a wider range of pronunciation than used to be the case when a certain crusading zeal attached to the words “standard English”. The pronunciation that calls for any special study is of course that which is natural to the particular group one is teaching, and that largely in proportion to the limitations it imposed on the group. “Muvver, fing, sumfink, wa?er”, are crudities even in Cockney. Just when a dialect form loses the dignity and colour of dialect and becomes simply a class-conscious restriction is hard to say. Probably it is only when local sound is combined with other limitations – ignorance, thin experience or vocabulary, fear of competition, a kindly or a rude complacency. The change from local to general language is delicate for children, who in attempting it are open to the scorn of their fellows and the ridicule of the educated speaker, and certainly a patronizing attitude to the local dialect on the part of the teacher bars success. The dialect should be described rather than judged. When children feel safe, then ambitious, they will imitate to acquire and not to mock; they will judge for themselves. Spelling is refreshingly simple to treat after this. The most chauvinistic local patriot will be the first to comment on any unconventionality of spelling as “bad” spelling. That which is taken for granted by professional people, and insisted on by commerce and bureaucracy, must be a crucial subject for children growing up into these communities. Some resent it as tyranny, others welcome it as a further protective device. Children learn to spell by practising specific examples and extending these analogically, rather than by rules.

Here is one of the three chief rules of English spelling, as set down by G. H. Vallins in a popular work: When a monosyllable ends in a single vowel plus single consonant, the consonant is doubled before a suffix beginning with a vowel; similarly when the final syllable of a disyllabic or polysyllabic word ends in a single vowel plus single consonant, the consonant is doubled only if the final syllable is stressed.

Naturally, there are exceptions to this rule. Some understanding of the link between spelling and etymology reconciles children to the rigidity of the conventions and the inconsistencies which elude rules, and they learn to see the word as an eye-symbol, a window on history.

5. Structures

In structures, as in spelling, there are unalterable arrangements which the

children just have to learn. These “formulas” are effective in language because the situations in which they are used are continually recurring. They are situation controllers in the purest sense, and hardly units of communication at all. It is not usual to reply “Very poorly” to the apparent question “How are you?” The formulas of greeting, parting, of superficial acquaintance, are of great value to children, though mentioned in textbooks less often than the written differences between say “Yours sincerely” and “Yours truly”.

Between complete formulas and what seems complete freedom there are many

degrees. The complete formula can have nothing added or interpolated, though

intonation or stress may sometimes alter (“How d’you do?”, “How d’you do?”). And in any society there are subjects as distinct from linguistic formulas – the weather, politics. Silence is to most human beings an intolerable threat. Nonsense is better – “Is it really you?” someone will say.

There are no completely free structures. They would be meaningless. The patterns of intonation, the stresses, word order (especially in written English), the few inflexions – have to be observed. All that is really free is the ability to choose which rules to obey, and which lexical units to put into relationship. To do this, children have to learn the fixed units which are used in relatively free structures. These are the more common regular and irregular endings which signal agreement, comparison, and so on; word order; and common prepositions and conjunctions and the words they accompany. Those need separate practice which are often used faultily, but the great majority of these units are fully absorbed only by using them in context. Oral work is more valuable than written in this, since twenty examples can be rehearsed in speech

for one in writing, and this can be decisive in setting up the correct habit. One can usefully start by showing children how much they do know. They are really quite advanced by twelve in their acquaintance with the signals and structure of English. Words such as “subject and predicate” are not of course tabu, but it isn’t helpful to say that the subject is what names the thing being discussed: “It is a tiger.”

It is as well, since many school grammar questions are simply disguised

comprehension questions, to ask one’s grammar questions in semantic terms: “They are throwing the ball.” Who are? – They are. “The soldier who was going on leave left his pass behind.” – Which soldier left his pass behind? – The one who was going on leave. “Give him the book.” – What was given? – The book – Who to? – Him. The children are perfectly at home in all these constructions, though they may never have heard of an adjectival clause or an indirect object. What is wanted is a sharpening of their powers, not a breakdown. It is possible by means of example and analogy to teach more complicated patterns, especially if these can be shown to be useful in improving the conciseness or balance of a particular piece of writing a child has done. An example of a pattern is built up on the board, and the class imitates by substituting new clauses for those given. Children will readily perceive, for example, how the situation gets out of control – how, say, the reader becomes bored – it the sentence patterns in a story are not varied. But even so, recognition of new sentence patterns precedes the power to use them, and there is no point in anticipating the need for complicated extensions of complex sentences.

6. A grammar of situation

Learning language through situation, making as it were a grammar of

situation, is simply to adopt one well-known method of teaching as a basis, and to organize language material in its terms. It is the method whereby language is treated as an element in a total context. When teachers complain that the specimens of sentence structure used for analysis in G. C. E. papers are contextless, they usually mean that they are without linguistic context. But children do not see language linguistically nor learn it for linguistic reasons: they learn it to manage life. The non-linguistic context, the situation, is thus the motivating factor in their study. Often, it is hard to provide a genuine situation in the rigid and bare framework of the classroom, and one has to search for the tie between a form of language and the outside world. In a cynical mood, one might say that Stephen Potter’s One-Upmanship is the best English grammar book in print. But children are friendlier that Mr Potter’s gamesman. They wish to respond as well as to dominate.

No work is wasted with a class when any language need is satisfied by example, but systematic teachers are wary of chasing hares, and will want to classify a syllabus of situations. Only the most general classification is practicable. There are two main groups of situations in which the pupil will repeatedly find his problem. First, that situation in which the speaker is always aware of his audience and of social effect; and second, that in which he is as a writer alone with his material, and in which his main consideration is for the nature of things – scientific reports, how to work a model train, and so on. In the latter, the difficulties are intellectual, and the aims are clarity and the proper subordination of ideas. In the former, the aims are psychological, and the aim is to control the audience’s response: to persuade, request, calm, anger, cajole, encourage, to greet and say farewell.

Situations of the second group are, for example, all the “How to do something” titles. One exercises the language by examining the content. The question “Why so and so?” brings out an adverbial clause; “Couldn’t you do so and so?” stimulates a conditional or concessive clause. Once a topic has been concluded, one may study some of the fragments of structure that need reshaping before they will fit into the clear picture – unmix any mixed structures, for example, or show what happens when limiting words are divorced from the ideas they limit, or are moved about in a group. If an example occurs in a talk, that must be the seed example; if no mistakes are made, there is no need to develop the theme. Such a sentence as “He showed his friend a motorcycle in the shed that was for sale” may occur, and once a real example has been spotted, it is sensible to build others on it, and the children will see the connection. It is no use starting with “Wanted, a Pomeranian for an old lady with a good pedigree.” If structures are portmanteau’d in the actual work, one can pull them apart. “The reason we caulk the seams is because the yacht won’t sail if it leaks” – this becomes either, “The reason we caulk the seams is that…” or “We caulk the seams because…”. If tenses are jumbled, try out some more, and let the children speculate about possible meanings. Which covers the longest spell of time – “It is 5 o’ clock”; “It is Tuesday today”; “Twice two is four”; “He is playing patience”? What does “I forget…” mean in “I forget what he said?” What does “I catch the 8.15 on Saturday” mean as distinct from “I am catching it at 8.15”? Which is a true statement: “He read a book as he came into the room” or “He was reading a book…”? Which relates to farther back in time: “Did you read it?” or “Have you read it?” Different classes will work at different levels, and a rigid syllabus is out of place. The golden rule is, use the actual work in hand; if certain mistakes are not made, it is pointless to correct them – but only too often teachers draw the children’s attention to errors they would not otherwise commit.

The purely informative statement is quite common to children’s speech and usually without error in structure. But the series of consecutive and carefully proportioned statements conveying a complex piece of information is comparatively rare, as it is in adult language. It is the more dramatic language, which is immediately face to face with another person’s reaction, that is constantly to be heard. There is a type of public examination which shows very well how to treat this need. In it, the candidate may be asked to imagine that he is, for example, in charge of an office, and that he has to reprimand the staff for their increasing tendency to lateness and to get them to be on time in the future. What does he say? Similarly, situations in which the children might find themselves can be postulated and dramatized in class, and the effect of various structures assessed. Two children from different schools meet in a train. What do they say to each other? (This is not easy for the teacher, if he is rash enough to suggest the dialogue. Teaching of this sort demands some literary imagination and a close contact with the particular social needs of the children, rather than with the textbook.) In the train, there is silence for a while. Each child reads his comic. What is the effect of continued silence? Discomfort. The one child says, “Want a sweet?” Why does he do this? (What might an adult say in similar circumstances? – “Any idea when we are due in?”) “Thanks” – willingness to be friends – or ‘Thanks, what are they?” – a more prickly character. And later on, “D’you like lessons?” – exploring the possibility of decent civilized communion. “Not much” – i.e. “I am quite normal and on your side”, or, “They’re not too bad” – slightly apologetic, but he does like them and although willing to be friendly does not want to start off on the wrong foot; or, “I’ll say I do – they’re smashing!” – this could lead to the friendship of a lifetime, or to one of those little comic-strip galloons inscribed “Thinks – blimey!”, and picks up comic. Other situations are such, as a prefect dressing down a small boy, and aiming to make him feel small or alternatively to feel that he can do better; or a conversation on the dance floor at the sixth-form Christmas party; or that between members of opposing football teams before the match – what, for example, is the effect of “Have you won any matches this year?” as compared with “Have you lost any matches this year?” For the teacher, it is perhaps easier to deal with the situations in which the child is facing an adult. A form has been threatened with the loss of a privilege because of some misdemeanour. A child is selected to approach the Head to ask for the form’s reinstatement in favour. What does the child say? A rather touchy aunt has promised to take the child to the theatre. The child would rather have the money. How does he say so? The child has been taken out to tea with a friend of his father’s. He knows that the intention is to show what a good boy he is. The stranger asks, “And what is your favourite lesson, eh?” What is the comparative effect of:

Maths, it’s fine.

Oh, English, I suppose.

Chemistry – it’s smashing to turn on the gas when old Stinks isn’t looking

I hate the lot. Such exercises as these bring out the differences between demand and request, command and invitation, decent reserve and cold rebuff. They can lead to the detailed consideration of particular structures – to the loading of sentences, for example:

Do you have grammar at your school? (a simple question)

Don’t you have grammar at your school? (surprise)

Do you still have grammar at your school?

Perhaps most important of all, they make the children realise that other people exist, must be considered. In making a syllabus on this basis, the material cannot of course be divided horizontally – that is, Year 1 – spelling; Year 2 – history of language. The division has to be made vertically, for some of the material is applicable to all ages in the secondary school, and the division lies primarily in the detail of the examples and in the degree of difficulty of the applications, situations, and exercises.

The reparative gesture which these suggestions embody does not pretend to

include everything which probably should be studied in the field of language at

school, but may it is hoped show that the active co-operation of children may be assured if their needs to control and respond to a situation are satisfied. They become responsible for language as they perceive the link between language and their life.

In conclusion, a summary of the argument of this appendix is offered, together with a few of the types of question which may be usefully asked of children studying language in the way proposed.

SUMMARY

Study of language in school

Aims:

to link study of language with total situation.

to enable pupil to be responsible for linguistic choices.

A. Study of language in general

1. what language is – a system of agreed signals

2. what it does for us – co-operation, extends boundaries of self.

3. its origin

4. assumptions about it that need examining –

a) that the word is the thing

b) that thought is independent of symbols

5. how language works –

a) symbolism

b) abstraction – uses and dangers.

B. Study of English language

1. Geography and its effects – simplification, tolerance.

2. History –

a) show by example general trends – simplified accidence

standardized spelling and pronunciation

growth of proportion of subordinating structures

enrichment of vocabulary

b) link with main events -

Christianity Renaissance

Northmen American and colonial

Normans

c) change – organic tie between Language and Life –

causes – outside world; inner attitude;

mechanics

3. Agreement and disagreement –

the word

a) as an area of meaning and intention

b) as a mirror of attitude

4. Social or class function of vocabulary, spelling, pronunciation.

Protecting or excluding function of local forms – describe, not judge

Pronunciation

Spelling – value as history

5. Structures

a) formulas

b) fixed units in “free” structures – specific example and oral work to establish habit

c) knowledge of simple structure – demonstrate semantically

d) more complex structures – practise according to need

6. Situations

a) the writer/speaker and his material – exercise language by examining content, e.g. mixed structures; order of modifiers; tenses.

b) the writer/speaker and his audience – exercise by inventing

situations

1. child/ child

2. child/ adult

C. Summary of content and method

Content: 1. nature and function of language

2. history and geography of English language

3. forms of change and stability

4. areas of agreement and disagreement

5. pronunciation, spelling, vocabulary and structure developed in concrete situations

Method: 1. avoid tautology – hence little place for use of systematic grammatical terminology

2. work to level of maturation, not generalizing or anticipating

unnecessarily

3. use motivating experience of control/ response situation. By specific examples related to actual work, satisfy the children’s present needs, rather than the teacher’s.

APPENDIX 5

The Error Quotient

No completely satisfactory formula for estimating the relative importance of

errors has been found, even when agreement as to what constitutes an error has been reached. Earlier error counts, as for example those by Charters and Miller, relied chiefly on ranking the errors in a large sample of writers by the frequency of occurrence in a given number of words or sentences, and by the proportion of a given number of writers who made any particular error. Stormzand and O’Shea pointed out in How Much English Grammar that the relative importance might be better derived from a combination of the frequency of occurrence of an error and the number of opportunities for its appearance. Thus one error might occur 645 times in 1000 opportunities, another only 145 times. The former would rank higher than the latter in importance, judging by its error quotient. However, by the earlier mode of calculation another rank might be obtained. Thus if the second error were found say 50 times in 500 sentences, it might be considered more important than the first error found to occur say only 10 times in the same sentence – for want of opportunity. These exaggerated and hypothetical examples perhaps serve to highlight the dilemma. The error quotient has been strongly recommended by some American investigators for the purpose of rating errors, even in the simplest form where number of errors give the quotient. number of opportunities Lyman, however, although supporting the use of the quotient, considers that it may be misleading unless allowance is made for the number of words needed to produce the opportunities for error. Thus a formula such as:

number of errors plus opportunities number of opportunities number of words

has been proposed for the error quotient. But such as formula involves an excessive degree of manipulation of the simple scores, which could well distort these and obscure the difference between one essay and another. For such reasons, the importance accorded to the error quotient by Stormzand, Lyman and others is hard to understand and can be deceptive. It serves chiefly to underline the unsatisfactory nature of previous attempts to compare one error with another. Under the formula, the importance of opportunity is excessively magnified. For example, the quotient of an error appearing once in an essay, provided only oneopportunity for it occurred, would be _E = 1 = 1 0 1 If another error were to occur 100 times, its quotient, given 200 possibilities, would be 100 = 0.5 200 Using the longer formula and allowing for 500 as the number of words written, the relative quotients would become 1.002 or 0.9. Such a ranking is quite artificial in the classroom, where the massiveness of an error – that is, its generality and frequent occurrence – is the crucial factor in deciding which difficulty to treat by group instruction. The error quotient is furthermore inapplicable when a group of errors is being considered as a whole. To some errors, as for example the omission of the full stop or of the comma in lists and non-defining clauses, children’s habits and the demands of English itself give great opportunity, even though this be not always seized; that other and rarer errors are individually more certain to occur is irrelevant here. It is possible to allow for opportunity as well as length by scoring an error as is done in the present work in measurement (b) (omission of the full stop) in Chapter V above, where examples of faulty usage are subtracted from correct examples in the child’s script. The resultant score is clear and easy to use. But this method too has to be used with discretion, since it can give overmuch weight to length when an error is comparatively rare. In all probability, the practising teacher has to assess the needs of each particular class quite separately, treating the rare errors individually, and the common ones by group instruction. If this is so, the expression “rank” or “importance” applied to errors is question-begging.

In the present work, the problem of the error quotient does not look very

large, since a whole group of errors was being employed as a compound item of measurement, and since the performance of one set of pupils was being compared with that of another, rather than one error with another. However the contrast between the massive and the rare errors was quite clear.

Error quotients were occasionally taken as a matter of curiosity, and the following table show five instances worked from essays by ten year old children, giving for each error the quotient based on the simpler formula not allowing for length of script.

Table 29: Error Quotients Based on Separate Errors for Each of Five Ten Year Old Children Figures show opportunities/ errors/ EQ in each case.

Error Essay 1 Essay 2 Essay 3 Essay 4 Essay 5

1b 24/0/-- 16/4/0.25 15/0/-- 25/3/0.12 54/15/0.28

1c 35/26/0.74 13/11/0/85 20/18/0.9 18/16/0.89 33/30/0.91

3d 6/1/0.17 6/1/0.17 0/0/-- 4/3/0.75 3/2/0.67

3e 48/0/-- 17/0/-- 14/0/-- 13/0/-- 20/3/0.15

3f 16/0/-- 6/1/0.17 9/0/-- 8/0/-- 12/1/0/08

3h 7/0/-- 4/1/0.25 2/0/-- 4/1/0.25 2/0/--

4 39/0/-- 34/0/-- 36/1/0.03 47/0/-- 59/2/0.03

7 2/0/-- 1/0/-- 0/0/-- 1/1/1 0/0/--

Total words 402 223 210 259 329

Total errors 27 18 19 24 53

Words per Error 14.9 12.4 11.1 10.8 6.2

APPENDIX 6

Examples of headings used at different stages in analysing essays

A. A first attempt was made to break down some essays under the following heads: Error number against:

Points concerning the Subject: sentence number; clauses used; special points; number of words; clause or phrase; article; genitive; adjective; substantive; pronoun. Subject modifiers of some length: phrase; clause; other. Verb Structure: adverb; finite; non-finite; mood; participle; infin.; correct tense; other extensions.

Direct and indirect object: clause or phrase; adverb and adjective; genitive; adjective; pronoun; other modifiers; subst.

Complement: adverbial; adverb modifying object; article or genitive; adjective; subst.; pronoun; infin + object; modifiers.

Sentence type: loose; periodic; balanced.

This style of analysis was too cumbersome, and in particular failed to direct attention to points likely to indicate maturation, being too evenly descriptive.

B. A more selective as well as more manageable method was next adopted, as

follows:

Sentence number against:

Error number:

Clauses used; phrases used (Sentence types): simple; double; multiple; complex; complex complex

Instances of the use of ; : –

List of modifiers between sentences/ within sentences Other structures (e.g.use of direct speech; modified or double subjects) Certain measurements: no. of words over no. of sentences; ideas/ paragraphs; sentences/ paragraphs; words/ errors.

C. Then after certain trials and modifications these last were suitably defined, the following and third system of heads was used. This, as may be seen, was more economical in every way, selecting as it did only those points necessary for the calculation of the scores. Sentence no. Error no. Clauses used No. of words S – S + 2 Adjectival phrases

In these headings, S – equals “simple sentence without any qualifying phrase”; and S + 2 equals “simple sentence but with two or more qualifying phrases”.

At first, a column indicating whether or not direct speech had occurred was included, but this point was dropped out early from the measuring instruments as of insufficiently general appearance. Headings for essay analysis. APPENDIX 7

Tables 30-34

A. Tables showing some results of analysis of variance of the eleven main amalgamated measures

Table 30: Item (c) - Number of words per common error df Sum of squares Variance

Methods 1 75896.1 75986.1

Schools 4 29349.1 7337.3

Methods by Schools 4 47101.7 11775.4

F = 75896.1 = 6.45 11775.4

(This figure for F would need to be 7.71 to be significant at the 5% level)

Table 31: Item (d) - Number of different sentence patterns used df Sum of squares Variance

Methods 1 210.0 210.0

Schools 4 67.9 16.98

Methods by Schools 4 99.5 24.88

F = 210.0 = 8.44

24.88

(significant at the 5% level)

Table 32: Item (h) - Number of correct complex sentences used

Methods 1 699.6 699.6

Schools 4 256.7 64.18

Methods by Schools 4 344.9 82.23

F = 699.6 = 8.11

86.23

(significant at the 5% level)

In each of these three cases, the difference between the means was to the advantage of the non-grammar forms. In all other cases no significant difference emerged, and an excessively large interaction of methods by schools threw doubt upon the utility of the method for comparing the amalgamated scores.

B. Two examples of F ratio for within-classes variance in Item (j) and Error (3d)

Table 33: Item (j) – Number of correct simple sentences exceeding the number of incorrect df Sum of squares Variance

Total 227 26149.2

Classes 9 4247.1

____ __________

Within Classes 218 21902.1 100.5

Methods 1 1442.5 1442.5

F = 1442.5 = 14.4 100.5

(For df.1 and 218 (3.89 or 6.76)

Table 34: Error (3d) - wrong or ambiguous use of prepositions df Sum of squares Variance

Total 113 169.7

Classes 5 60.0

____ ________

Within Classes 108 109.7 1.1

Methods 1 36.9 36.9

F = 36.9 = 33.5 1.1

(For df.1 and 108 (3.94 or 6.90)

APPENDIX 8

Mean I.Q.’s of pairs of forms

Table 35

a) Mean I. Q.’s of pairs of forms at the beginning of the work

Type of school Grammar form Non-grammar form

I.Q. S.D. I.Q. S.D.

A. Grammar 122.0 7.59 120.7 6.70

B. Grammar 116.7 7.12 114.2 7.61

C. Sec. Mod. 96.52 7.91 96.81 7.47

D. Technical 104.5 8.11 105.3 7.72

E. Technical 101.2 7.23 100.1 6.98

b) Mean scores of pairs of forms on the eleven main measures in the first essays

School A B C D E

a) grammar 9.40 11.77 9.21 9.61 11.67

non-grammar 9.90 10.00 10.67 10.50 11.00

b) grammar 10.20 15.00 9.52 5.52 6.01

non-grammar 9.00 16.00 8.48 4.79 7.00

c) grammar 30.02 45.31 37.53 34.76 39.37

non-grammar 36.23 40.03 32.82 30.83 32.24

d) grammar 4.07 5.73 4.52 3.52 3.37

non-grammar 4.46 5.96 4.29 2.54 2.77

e) grammar 7.93 12.19 11.24 8.12 8.32

non-grammar 6.69 14.92 9.81 6.79 8.12

f) grammar 8.17 11.19 10.14 7.01 7.90

non-grammar 6.65 13.12 8.95 7.50 6.94

g) grammar 238.3 313.5 286.5 202.8 223.4

non-grammar 158.7 361.0 235.7 207.7 207.2

h) grammar +1.69 +1.66 - 0.89 - 0.52 - 0.26

non-grammar +1.12 +1.28 - 0.95 - 1.26 - 0.93

i) grammar 1.10 1.29 0.52 0.91 1.42

non-grammar 0.85 1.40 0.24 0.88 0.82

j) grammar +2.07 +4.48 - 6.00 +1.24 +0.27

non-grammar +2.69 +3.04 - 4.57 +0.42 +0.18

k) grammar 7.52 12.95 5.45 5.48 8.90

non-grammar 5.85 9.72 4.00 5.92 8.47

APPENDIX 8 B

Mean I.Q.’s of pairs of forms

Table 35

a) Mean I. Q.’s of pairs of forms at the beginning of the work

Type of school Grammar form Non-grammar form

I.Q. S.D. I.Q. S.D.

A. Grammar 122.0 7.59 120.7 6.70

B. Grammar 116.7 7.12 114.2 7.61

C. Sec. Mod. 96.52 7.91 96.81 7.47

D. Technical 104.5 8.11 105.3 7.72

E. Technical 101.2 7.23 100.1 6.98

Appendix 8 - mean I.Q.'s b) Mean scores of pairs of forms on the eleven main measures in the first essays

School A B C D E

a) grammar 9.40 11.77 9.21 9.61 11.67

non-grammar 9.90 10.00 10.67 10.50 11.00

b) grammar 10.20 15.00 9.52 5.52 6.01

non-grammar 9.00 16.00 8.48 4.79 7.00

c) grammar 30.02 45.31 37.53 34.76 39.37

non-grammar 36.23 40.03 32.82 30.83 32.24

d) grammar 4.07 5.73 4.52 3.52 3.37

non-grammar 4.46 5.96 4.29 2.54 2.77

e) grammar 7.93 12.19 11.24 8.12 8.32

non-grammar 6.69 14.92 9.81 6.79 8.12

f) grammar 8.17 11.19 10.14 7.01 7.90

non-grammar 6.65 13.12 8.95 7.50 6.94

g) grammar 238.3 313.5 286.5 202.8 223.4

non-grammar 158.7 361.0 235.7 207.7 207.2

h) grammar +1.69 +1.66 - 0.89 - 0.52 - 0.26

non-grammar +1.12 +1.28 - 0.95 - 1.26 - 0.93

i) grammar 1.10 1.29 0.52 0.91 1.42

non-grammar 0.85 1.40 0.24 0.88 0.82

j) grammar +2.07 +4.48 - 6.00 +1.24 +0.27

non-grammar +2.69 +3.04 - 4.57 +0.42 +0.18

k) grammar 7.52 12.95 5.45 5.48 8.90

non-grammar 5.85 9.72 4.00 5.92 8.47

Appendix 8 - mean I.Q.'s BIBLIOGRAPHY

Abercrombie, D. The social basis of language. English Language Teaching 3, no. 1, 1947.

Allen, E. E. and Mason, A. T. An English Grammar of Function. Book I. London: Arnold, 1939. 351 pp.

Andrew, S. O., Hawer, A., and others. The problem of grammar. English Association pamphlet no. 46, 1923. 23 pp.

Archer, C. P. Research in English composition. Review of Educational Research 19, no 2, 1949, pp. 135-151. (Detailed survey with useful bibliography. Supports the trend from formal to direct.)

Asker, W. Does knowledge of formal grammar function? School and Society 17 (LSE), January 1923, pp. 109-111.

Association of Assistant Mistresses in Secondary Schools. Memorandum on the Teaching of English. London: U.L.P., 1956. 71 pp.

Austin, C. Read to Write. Books 1 and 2. London: Ginn, 1954. 246 pp.

Bagley, D. A critical survey of objective estimates in the teaching of English. British Journal of Educational Psychology, 7, 1937, pp. 57-71.

Ballard, P. B. Group Tests of Intelligence. 2nd edn. London: U.L.P., 1949. 252 pp.

Bartlett, P. E. The Influence of Learning Grammar in the Comprehension of Modern English Prose. M.A. thesis. London, 1953.

Betz, A. and Marshall, E. Grammar based on errors. English Journal 5, September 1916, pp. 491-500. (Still “grammar for its own sake”, “to think clearly” – emphasis on error and drill.)

Biaggini, E. G. The teaching of English. English in Australia. Educational Research Series no. 29. Melbourne University Press and Oxford U.P., 1933, pp. 108-28. (à la I. A. Richards)

Bishop, C. S. The Philosophy of Language and its Implications for the Teaching of English. Ph.D. thesis. Nottingham, 1957.

Blackwell, A. M. A List of Researches in Education no. 1, 1918-48 (and subsequent lists and supplements.) National Foundation for Educational Research in England and Wales. London: Newnes, 1950. 172 pp.

Blamires, H. English in Education. London: Bles., 1951. 159 pp.

Board of Education. Memorandum submitted by the Standing Committee on Grammatical Reform, 1919. Published as Appendix IV, The Teaching of English in England. H.M.S.O., 1938.

Board of Education.

a) The Teaching of English in England. Report of Departmental Committee. H.M.S.O., 1921, pp. 278-294.

(pro func. grammar – description required)

b) (the same). H.M.S.O., 1938, pp. 50-51, 76-78, 106-113, and Appendix IV.

(Chaos – a little of everything – grammar hardly mentioned.)

c) Report of the Consultative Committee on the Primary School. H.M.S.O.,

1931, p. 290.

Briggs, T. H. Formal English grammar as a discipline. Teachers’ College Record 45 no. 4, Sept. 1913, pp. 251-343.

Broening, A. Grammar. Forty-Third Year Book, Part 2, National Society for the Study of Education, 1944, pp. 175-181.

Brook, G. L. A History of the English Language. London: Andre Deutsch, 1958. 224 pp. (Has a useful descriptive bibliography.)

Brown, V. N. An Experimental Study of a New Method of Grammar Teaching and its Bearing on English Composition. M.A. thesis. London, 1936.

Cast, B. N. D. Efficiency of different methods of marking English compositions. British Journal of Educational Psychology 9, 1939, pp. 257-69; and 10, 1940, pp. 49-60.

Cawley, F. The Difficulty of English Grammar for Pupils of Secondary School Age. M. Ed. Thesis. Manchester, 1957. (See also British Journal of Educational Psychology 28 no.11, June 1958, pp. 174-76.)

Charters, W. W. Language and Grammar. Third Year Book, Dept. of Superintendence of N.E.A. (Washington), 1925, pp. 278-96. (How to construct a grammar curriculum based on error; a study of previous surveys – only 4 written. Errors mostly reported by teachers. Relies entirely on definitions to teach, eg “clauses can be made plain quite easily”!)

Charters, W. W. and Miller, E. Minimal essentials in elementary language and grammar. Sixteenth Year Book, Part I, National Society for the Study of Education. (Bloomington, Illinois: Public School Pub. Co.), 1917, pp. 85-110.

(Collected 4 weeks’ spontaneous scripts in Kansas and tabulated errors: verbs

and tenses esp.; recommend focus on local errors.)

Crawford, C. C. and Royer, M. M. Oral drill versus grammar study. Elementary School Journal 36, 1935, pp. 116-119. (7th grade – 60 pupils; methods rotated dfor 8 lessons; oral better, but no test of permanency; focus on specific local errors.)

Diack, H. A re-examination of grammar. The Use of English 7 no. 4, 1956, pp. 251-55.

Edmiston, R. W. and Gingerich, C. N. Relation of factors of English usage to composition. Journal of Educational Research 36, 1942, pp. 269-71. (r = 0.56 ± 0.04 at grade 7)

Evans, A. A. Grammar, language and style. The Use of English 5 no. 1, 1953, pp. 3-9; no. 2, 1953, pp. 76-81. (Plea for later start on grammar; good résumé of present confusion.)

Evans, E. G. S. An Experimental Study of Methods of Teaching English in Secondary Modern Schools. Ph.D. thesis. Birmingham, 1951.

Finlayson, D. S. The reliability of the marking of essays. British Journal of Educational Psychology 21 no. 2, 1951, pp. 126-34. (The more markers the better, provided consistent.)

Fleming, C. N. Research and the Basic Curriculum. London: U. L. P., 1946. 120pp.

Fries, C. C. The Structure of English. New York: Harcourt, Bruce, 1952. 304 pp.

Frogner, E. Grammar approach versus thought approach in teaching sentence structure. English Journal 28 no. 7, 1939, pp. 518-26.

Gardiner, A. H. Theory of Speech and Language. Oxford: Clarendon Press, 1932. 332 pp.

Gordon, I. A. The Teaching of English. Educational Research Series no. 27. Wellington: New Zealand Council for Educational Research, 1947. (See esp. pp. 108-118.)

Gradon, P. The teaching of grammar. English Association Symposium on the Teaching of English in Schools. Oxford: O.U.P., 1946. 166pp.

Gratton, J. H. G. and Gurray, P. Our Living Language. London: Nelson, 1925. 323pp. (Reprinted 1953).

Greene, H. A. and others. English language. Review of Educational Research 7, 1937, pp. 474-81. (a review of 67 monographs; stress on audience and reality situation in composition.)

Greene, H. A. Contributions of research to special methods: English usage. Thirty-seventh Year Book, Part 2, National Society for the Study of Education (Bloomington, Illinois: Public School Pub. Co.), 1938, pp. 115-21. (survey to 1937; pro local detail and socio-functional method)

Greene, H. A. English language, grammar and composition. Encyclopaedia of Educational Research, 1950. New York: Macmillan, pp. 383-96.

Gunn, M. A. and Barlow, E. R. English composition. Review of Educational Research 22, no. 2, 1952, pp. 96- 101. (really just a list)

Hamley, H. R. Formal training: a critical survey of experimental work. British Journal of Educational Psychology 6, 1936, pp. 233-49.

Harris, C. W. and Liba, M. R. (eds.). Encyclopaedia of Educational Research. 3rd edn. New York: Macmillan, 1960, pp. 383-96.

Harris, R. J. The teaching of formal English grammar. The Use of English 6 no. 4, 1955, pp. 224-29.

Hatfield, W. (chairman) An Experience Curriculum in English. Report of the Curriculum Commission of the National Council of Teachers of English. New York: Appleton Century, 1935. 323 pp. (stresses authority of usage; the positive rather than errors; habit formation through repetition; but still expects terminology to be “picked up” by hearing it referred to)

Hornby, A. S. A Guide to Patterns and Usage in English. Oxford: O.U.P.,1954. 216 pp.

Hoyt, F. S. The place of grammar in the elementary curriculum. Teachers College Record 7, Columbia Univ., Nov. 1906, pp. 467-500. (Attacks the 4 defences – a) as discipline – but no transfer; b) other languages – but English “grammarless”. 300 pupils, 2 markers; tests in composition, interpretation, grammar.)

Hughes, V. H. A study of the relationship among selected language abilities. Journal of Educational Research 47, 1953-54, pp. 97-105. (332 grade 5; IQ 90-110; word-meaning vital; usage; sentence sense -)

Humphreys, G. S. and Roberts, J. C. An Active English Course. Book 1 (191 pp) and Book 2 (230 pp). London: U. L. P., 1939. Incorporated Association of Assistant Masters. Memorandum on the teaching of English. 2nd edn. Cambridge: C. U. P., 1957. 196 pp.

Kimmins, C. W. Methods of expression used by London children in essay writing at different ages. Journal of Experimental Psychology 3 no. 5, 1916, pp. 289-95.

King, H. V. Oral grammar drill. English Language Teaching 14 no. 1, 1959, pp. 13-18.

Leonard, S. A. and Moffatt, H. M. Current definitions of levels in English usage. English Journal 16 no. 5, 1917, pp. 345-59. (Jury of linguists disagrees with grammarians on rating certain usages as errors.)

Leonard, S. A. The doctrine of correctness in English usage 1700-1800. Studies in Language and Literature 25 (Univ of Wisconsin), 1929, p. 361.

Lewis, M. M. The everyday use of the written word. In Language in School. London: U. L. P., 1942, pp. 37-64.

Lindquist, E. F. Statistical Analysis in Educational Research. Cambridge, Mass: Riverside Press, 1040. 266pp.

Lyman, R. L. Summary of investigations relating to grammar, language and composition. Supplementary Educational Monographs 36 (Univ. of Chicago), 1929. 302pp. (A most important source of information for American work up to 1929.)

Lyman, R. L. Language, grammar and composition. Review of Educational Research 2 no.1, 1932, pp. 34-42.

Macauley, W. J. An Investigation into the Difficulty of English Grammar with Particular Reference to Pupils in Primary and Junior Secondary Schools. Ed. B. thesis. Glasgow 1946. (See also British Journal of Educational Psychology 17, 1947, pp. 153-62.)

McKerrow, R. B. English grammar and grammars. Essays and Studies of the English Association 8, Clarendon Press, 1922, pp. 148-57.

Miller, H. R. “What if they don’t know grammar?” English Journal 40, 1952, pp. 525-26.

Ministry of Education. Language: some suggestions for teachers of English and others. Ministry of Education Pamphlet no. 26, M. H. S. O., 1954. 169 pp.

Moore, E. M. A Study of the Language of Some London Children. M. A. thesis. London, 1953.

Nice, M. M. Length of sentences as a criterion of a child’s progress in speech. Journal of Educational Psychology 16, 1925, pp. 370-79.

Palmer, H. E. The New Method Grammar. London: Longmans, 1938. 216 pp.

Piaget, J. (transl. Warden, M.) The Language and Thought of the Child. New York: Harcourt, Bruce, 1926. 246 pp.

Pooley, R. D. Teaching English usage. English Monograph no. 16, National Council of Teachers of English. D. Appleton Century, 1946. 225 pp (esp pp. 223-40). (Contains useful objections to many prescriptive rules; grammar not valuable below 14, i.e. until child is able to generalize.)

Poster, C. D. The case against grammar. The Use of English 2 no 2, 1959, pp. 92-95. (Correct by analogy)

Pressey, S. L. A statistical study of children’s errors in sentence structure. English Journal 14, Sept. 1925, pp. 529-35. (980 papers grades 7-12. Errors few, frequently recurring, esp failure to make proper sentence division; older children use more complex; 1) complex sentences; 2) omissions, 3) pronouns.)

Quirk, R. From descriptive to prescriptive. English Language Teaching 12 no.1, 1957, pp. 9-13.

Quirk, R. and Smith, A. H. (eds.). The Teaching of English. Studies in Communication no. 3. London: Secker & Warburg, 1959. 192 pp. (word order, intonation, morphology, v. word types and definitions)

Rapeer, L. W. The problem of formal grammar in elementary education. Journal of Educational Psychology 4, March 1913, pp. 125-37. (A verification of F. S. Hoyt with 11+ beginners at High School; some low correlations; recommends leave grammar till grade 8; real situations; use “functional”.)

Rivlin, H. V. Functional Grammar. Contributions to Education no. 435. Teachers College, Columbia, 1930. 93pp.

Robinson, N. Formal English Grammar and English Composition. M. Ed. Thesis. Manchester, 1959.

Schonell, F. J. The development of educational research in Great Britain. British Journal of Educational Psychology 18, 1948, pp. 1-5 and 53-66.

Seegers, J. C. Grammar and usage – some current thoughts. Schools Review 58, 1950, pp. 468-74.

Segel, D., and Barr, N. R. The relation of achievement in formal grammar to achievement in applied grammar. Journal of Educational Research 14, Dec. 1926, pp. 401-7. (Tests in formal applied to 1000 sophomores – like Test B in the present work – choose correct form. No transfer with any other subjects.)

Shattuck, M. E. and Barnes, W. The situation as regards English. Ninth Year Book, N. E. A. Dept. of Supervisors and Directors of Instruction, 1936. 193pp.

Smith, D. V. English language, reading, and literature. Review of Educational Research 8, 1938, pp. 25-33. (Composition, public speaking, vocabulary, grammar, spelling and handwriting.)

Stormzand, M. J. and O’Shea, M. V. How Much English Grammar? Baltimore: Warwick & York, 1924. 224pp. (Errors studied in 288 compositions from grade 6 to univ: 58,196 words.

Summers, A. M. The Relation between Knowledge of Formal Grammar and Ability in Written Composition. M. A. thesis. University of Illinois, 1930.

Symonds, F. M. Practice versus grammar in the learning of correct English usage. Journal of Educational Psychology 22, Feb. 1931, pp. 81-95.

Thorndike, E. L. and others. An inventory of English constructions with measures of their importance. Teachers College Record 28 no. 6, Feb. 1927, pp. 580-610.

Thorndike, E. L. Mental discipline in high school studies. Journal of Educational Psychology 15, 1924, pp. 1-22 and 83-98. (Gain due to original IQ)

Tucker, S. Grammar – still a problem. The Use of English 10 no. 2, 1958, pp. 85-89. (A plea for simplification and non-Latinization.)

Vernon, P. E. The Measurement of Abilities. London: U. L. P., 1940. 308 pp.

Warner, P. C. and Guiler, W. S. Individual versus group instruction in grammatical usage. Journal of Educational Psychology 24, 1933, pp. 140-51.

(3 groups – individual; group; no English at all!)

Watts, A. F. The Language and Mental Development of Children. London: Harrap, 1934. 354 pp.

Westgarth, G. N. Grammar and its concepts. Researches and Studies 5 (Univ. of Leeds Institute of Education), Jan. 1952, pp. 7-23.

Westaway, F. W. The Teaching of Grammar. London: Blackie, 1933. 276 pp.

Wiswall, E. E. A study of sentence structure in eighth grade composition. Elementary School Journal 26 no 6 (Chicago), Feb. 1926, pp. 441-48. (Does topic of composition influence sentence structure? – hardly. 200 children; 4 types of composition-) Editor’s note: some handwritten author’s comments on the works have been typed in.