L.B
JOURNAL OF EDUCATIONAL RESEARCH MONOGRAPHS
༤༧ སཱམ – '', ཡ ༠༣ལོ ༥༧
B. R. BUCKINGHAM, Editor
15 73
Number 8
N
TEACHING BEGINNI
READ
1925
TEACHING BEGINNERS TO READ IN
ENGLAND:
Its Methods, Results, and
Psychological Bases
By
W. H. WINCH, M. A. (Cantab)
Chairman of the Committee of the Education Guild of Great Britain
and Ireland on Psychological Research in Schools, Author of
Problems in Education, German Schools, When Should a Child Begin
School, Inductive vs. Deductive Methods of Teaching, Children's Percep
tions, Fatigue in Evening Schools, etc.
PUBLIC SCHOOL PUBLISHING COMPANY
BLOOMINGTON, ILLINOIS

1837
ARTES
LIBRARY
SCIENTIA
VERITAS
OF THE
UNIVERSITY OF MICHIGAN
DOUS UNUM
TUEBOR
SI QUÆRIS PENINSULAM·AMⱭNAM
CIRCUMSPICE
PLAJASI
EDUCATION
LB
1573
.W76
JOURNAL OF EDUCATIONAL RESEARCH MONOGRAPHS
B. R. BUCKINGHAM, Editor
Number 8
1925
TEACHING BEGINNERS TO READ IN
ENGLAND:
Its Methods, Results, and
Psychological Bases
By
W. H. WINCH, M. A. (Cantab)
Chairman of the Committee of the Education Guild of Great Britain
and Ireland on Psychological Research in Schools. Author of
Problems in Education, German Schools, When Should a Child Begin
School, Inductive vs. Deductive Methods of Teaching, Children's Percep-
tions, Fatigue in Evening Schools, etc.
PUBLIC SCHOOL PUBLISHING COMPANY
BLOOMINGTON, ILLINOIS
Copyright 1925 by
Public School Publishing Company
Bloomington, Illinois
PRINTED IN U. 8. A.
02
25-27 H HII,
Part. Schol
> -->
12621
4 crp.
EDITOR'S INTRODUCTION
This is not just another book on reading. In treatment and spirit
it is quite unusual. It exalts no method, yet it is devoted to method.
It is neither hortatory nor didactic, yet it teaches a great lesson. It makes
no commitments save commitment to truth.
Its distinguished author is an Englishman and it refers to conditions
in English schools. Yet the American teacher will find nothing in it
which will fail to apply in the American school room except a few super-
ficial matters of terminology. Quite evidently the problems in teaching
reading to beginners is the same on both sides of the Atlantic.
As an evidence of the universality of the problem-the fact that it
looks the same through English and American eyes—the reader will note
almost in the first paragraph the belief, deliberately arrived at and sol-
emnly expressed, that "young children in London do not read or spell.
as well as they did some fifteen or twenty years ago." Have we not
heard the same in New York, Chicago, Denver, and San Francisco?
Have we not heard it at every cross-road? Business men affirm it and
teachers themselves echo it. Again, the author notes a tendency in
England "toward diminishing the requirements of the infant school
(kindergarten and early primary) and introducing more play, more
rest, and less work." Do not the American schools even so?
The author distinguishes certain methods of teaching reading. Pair
by pair he sets them one against another to find out which is best. These
are the phonic, syllabic, alphabetic, and look-and-say methods. Have not
these methods had vogue in America? Do not the phonic and look-and-
say methods variously mixed constitute the accepted approach to the
teaching of reading? To be sure, the particular phonic method of this
book is the Dale--which sounds strange to our American ears; and the
syllabic method is the Sonnenschein-which seems even more strange.
But in their substance these methods are old friends. Listen! The
author expresses the view that the phonic system he has chosen has, in
spite of its popularity, "unnecessary elaborations and that its reading
matter is not always suitable for young children." Does not this criti-
cism fit some phonic methods with which American teachers are familiar?
I think, therefore, the teachers of the reading of the English lan-
guage, no matter where they live and work, will find that the essential
conditions out of which this book arises are the same as the conditions
with which they themselves are dealing.
1
But what I like about this book is its wholesome attitude-an atti-
tude which to me would be interesting even though the conditions to
which the attitude referred were quite unfamiliar. The author be-
lieves in fact takes it as a premise, else why the inquiry?—that meth-
ods of teaching reading differ in value. He is prepared to say that a
good teacher with a poor method may, and probably does, get better
results than a poor teacher with a good method. This I think would be
generally admitted. Moreover he thinks that any method consistently
adhered to is better than a hodge-podge of methods; and I think his
American friends will agree with him. But under reasonably equal con-
ditions, it is nevertheless his judgment that methods of teaching reading
are not of equal merit. And then he says-and I call your attention
to the dignity, simplicity, and clear-eyed truth of the passage :—“But,
if I say I do not think all methods are equally good, I should be able,
if I am called upon, to give grounds for my opinion. Those grounds
may be rational, i. e., based on accepted principles of child-psychology,
or they may be based on scientific evidence. The former method of
argument has had a long run for many, many years; but we seem as
far from convincing ourselves as to the best methods as ever. Some
of the principles appealed to are true; some are false; but nearly all of
them are tainted by what I have previously called the diminished-adult
view of the child-mind. Our own psychological processes are put into
the child, diminished in strength, but similar in form. We are getting
old and worn, many of us. We do not like the mechanical acquisition
of new things; it is hard for us; so we say children do not like it. As a
matter of fact, they do. Repetition bores us; so we say it bores the
young child. As a matter of fact, he loves it. We pride ourselves on
our intelligence rather than on our memory. I am not sure that we
should come out really well if our intelligences were tested. We have
probably, most of us, passed our upward limit years ago, but are not
aware of it; though we do know that our rote memories are faulty.
Without abundant associations we remember very little. Straightway we
assert that the young child is like us. We teach by methods so intelli-
gent that he does not understand us; we associate his knowledge by so
many bonds, or try to, that he loses sight of what he really has to re-
member; and we say his memory cannot improve, knowing that our
own does not."
Accordingly, the author approaches his subject from the strictly ex-
perimental point of view. To us he seems to exhibit the spirit of today
in education, a spirit which has no local habitat. Its country is the
world of science; its creed, the love of truth; its method, the method
of research.
The particular method which Mr. Winch adopts is one which has
long been associated with his name. He had already been using it for
several years when it first came to our attention in 1910 in the newly-
founded Journal of Educational Psychology. It is the method of equal,
parallel, and highly correlated groups. According to this method the
effort of the investigator is directed toward a high degree of refine-
ment in controlling variation among the learners with whom he is deal-
ing. Large numbers of cases are not so much his concern. It is only
when a refined technique of this sort is observed that valid experimen-
tation may dispense with large numbers. Under properly controlled con-
ditions, however, it can. As a matter of fact, it is not too much to say
that in theory significant results, so far as the number of cases is con-
cerned, might be obtained from just two pupils, provided they were
equated in all relevant respects at the moment of beginning the experi-
mental procedure. These results, however, would be significant only for
pupils having the equated characteristics of these two pupils. They
would be significant but not general.
In order, therefore, that conclusions may have a useful application
we must have enough pairs of children to represent the different levels
of aptitude likely to be found in the classroom. If we were limited, let
us say, to three perfectly equated pairs we should want one pair to con-
sist of two average children—that is, average in the function with which
we were concerned-while the other two pairs were composed of pupils
at the 75- and 25-percentile. If we could have still other equated pairs
we should want them well spaced all the way from the best to the
poorest levels of ability. When we have this condition we have Mr.
Winch's "equal, parallel, and highly correlated groups." One of the
groups contains all the first members of each pair and the other group
contains all the second members. The groups are equal in the usual
sense of having the same average; and they are parallel as well as
highly correlated because each pupil in one group has his equal in the
other.
This device is a powerful instrument for research purposes. Other
precautions, however, must be taken than those relating to the selec-
tion of pupils. One may have the most satisfactory groups in the world,
yet if subsequent conditions are not controlled one's results may be
worthless. If, for example, in a learning experiment-and the experi-
ments in this book are essentially such-the same or equal teaching
3
ability or other opportunities to learn are not provided, the experiment
breaks down. Of course, it goes without saying that as careful a worker
as Mr. Winch does not fail to provide for these additional controls.
But his principal contribution to the methodology of research is
his system of grouping pupils. And every teacher can use it with his
own class. He must first have a valid test upon which to divide his
pupils into groups—a test, in other words, which measures the thing he
wants to measure. This is the main difficulty but it is not generally in-
superable. Indeed, for the purpose in hand the test a teacher may de-
vise will often serve better than one that has been made and published
by somebody else. On the basis of this test the teacher may then divide
his class into two "equal, parallel, and highly correlated groups" with
respect to the function he has in mind. He is then ready for the real
experiment. He may wish to find out whether it is better to teach
physics by a lecture-demonstration method or by a method involving
experiment by the pupils. A primary teacher may wish to find out
whether it is better to have children use the long division form rather
than the short division form in dividing by a one-place number. Or
again, as in this book, the teacher may wish to try out the value of two
rival methods of teaching reading. Having tested the pupils and ar-
ranged the groups on the basis of readiness, interest, and aptitude for
learning physics or division by one-place number, or reading from print,
the experiment may proceed.
What we wish to point out is that the teacher who reads this book
will not only obtain some rather accurate notions about methods of
teaching reading, but will also be learning a technique of experimen-
tation which he can apply in his own work.
B. R. BUCKINGHAM
4
TABLE OF CONTENTS
CHAPTER
I. Introduction. Early Reading in England...
II. Observations in 1905-1906..
PAGE
7
11
17
21
29
59
III. The Work of Other English Experimenters...
IV. Shapes and Sounds and Their Association..
V. Phonic versus Look-and-Say Methods...
VI. Phonic versus Alphabetic Methods...
VII. Phonic versus Phonoscript Methods.
VIII. The Influence on Spelling of Phonoscript versus an Ordinary
Phonic Method of Teaching Reading in Its Early Stages...... 107
IX. On the Loss Due to Transference from Phonoscript to Ordinary
82
Print after One Year of Teaching the Phonoscript Method.... 120
X. A Further Experiment on Loss When Children Taught by the
Phonoscript Method Are Required to Read from Ordinary Print 129
XI. On the Loss after Eighteen Months of Teaching When Trans-
ference Is Made from Phonoscript to Ordinary Print.
139
XII. An Experiment to Check the Results Given in the Last Chapter. 149
XIII. Final Tests in Reading..
158
XIV. Final Tests in Spelling....
167
XV. Summary and Some Important Conclusions.
174
•
APPENDIX
I. Some Suggestions Made by Teachers and by the Writer in
Connection with Phonoscript..
177
II. A Suggested Introduction to the Phonoscript System of Read-
ing in Its Very Early Stages.
183
5
LIST OF TABLES
TABLE
L'AGE
I.
The Results of the Examination Following the First Lesson....
24
II.
Summary of the Results of the Examinations Following the
Second, Third, and Fourth Lessons..
25
III. The Total Marks Obtained by Each Child for the First, Sec-
ond, Third, and Fourth Examinations.
27
IV. The Results of the Examination in Shapes and Sounds, the
Children Being Divided into Two Equal, Parallel, and Highly
Correlated Groups on the Basis of the Marks Then Obtained
(Maximum Possible Marks, 48).....
33
V. The Marks Obtained in the Final Tests with Times Taken by
Groups A and B Compared.
53
VI. The Marks Obtained by the Two Groups on Ballard's One-Min-
ute Reading Test..
58
VII. The Results of the Examination on Shapes and Sounds and the
Division into Two Equal Groups...
62
VIII. The Speed of Groups A and B Compared....
79
80
IX. The Summarized Individual Results...
103
X. The Work of Group A and Group B Compared Section by Section 80
XI. Individual Results in the Preliminary and Final Tests...
XII. Individual Marks in Spelling of Group A and Group B Com-
pared after a Year's Learning to Read; Group A by an Ordi-
nary Phonic Method, Group B, by the Phonoscript Method.... 112
XIII. The Individual Results of Group B, the Phonoscript Group,
Tested after Twelve Months' Teaching in Reading.
XIV. Individual Results after Six Months' Instruction in Phonoscript
(Maximum Mark, 25).....
•
123
132
XV. Individual Results for the Phonoscript Group, when Tested after
Eighteen Months' Teaching (First Reading from Phonoscript
Print, Second Reading from Real Print)..
144
XVI. Individual Results for the Phonoscript Group, when Tested after
Eighteen Months' Teaching (First Reading from Real Print,
Second Reading from Phonoscript Print).
151
XVII. Individual Marks on the Final Tests in Reading.
164
XVIII. Final Results in Spelling..
6
172
CHAPTER I
INTRODUCTION-EARLY READING IN ENGLAND
School attendance in England is compulsory at five years of
age. At this age a child enters an infants' school or class; and
somewhere between six and seven years of age is expected, in
normal cases, to enter Standard I, which may be placed in an
infants' department, or in a boys' or girls' department, or in a
junior or senior mixed department. The standards range from
I to VII, and compulsory attendance ceases at about 14 years
of age.
In London the stages of the infants' schools are called grades.
Grade III, the highest grade, immediately precedes Standard I;
Grade II is next below; and in schools in which children enter
at three years of age (a permissible, but not a compulsory at-
tendance) the lowest class is called Grade I. The American
Grades I to VIII correspond fairly closely with Grade III to
Standard VII. When a grade or standard is large, it is usually
taught in two or more classes, divided on the basis of scholastic
attainment. The upper half or section is usually called the "a"
division. Thus, Standard Ia means the upper section of Stan-
dard I.
In England children begin to learn to read in the infants'
schools, at which, as has been said, attendance is compulsory from
five years of age, and optional from the age of three, if there
is adequate school accommodation in the neighborhood. In Lon-
don, if there is no Standard I class (the beginning of the senior
grading) in the infants' department, children leave the infants'
school roughly at an average age of six years nine months and
are promoted to the senior or boys' and girls' departments. In
schools where there is a Standard I class in the infants' depart-
ments the promotion age is required to average about seven years
nine months, if the children are then fit to begin the work of
Standard II. There are complications of these regulations due
7
8
TEACHING BEGINNERS TO READ
to conditions of accommodation; but the above are the basic
rules.
The problem, then, of beginning to learn to read is primarily
an infants' school problem. What is the position today? There
is, of course, the usual dissatisfaction with the reading and
spelling of young children-of older ones too, for that matter-
I discount that as a general feature; it has always existed and
probably always will. But it is the considered opinion of a good
many able and experienced mistresses that young children in
London do not read and spell as well as they did some fifteen
or twenty years ago-an opinion which, to some extent, I share.
How is this to be accounted for? Is it because our infants'
schools have, as some of our senior schools certainly have, a
curriculum which is overburdened? I can hardly subscribe to
that view, since all recent tendencies in English education have
been toward diminishing the requirements of the infant school
and introducing more play, more rest, and less work. Though I
ought to say, that there has been, in the last few years, a revival
of interest in reading and a desire for its early progress, which
indicates some reversal of the general current. Moreover, the
Froebelian mode of thought has lost ground, and it is no longer
thought an essential of sound method to put off learning to read
and write as long as possible. The Montessori influence has
operated in the same direction, with its claim that children can
be, and have been, taught to read successfully at an early age.
So much for theoretical tendencies. In practice, our infants'
schools do teach reading fairly early, probably as early as is de-
sirable-a problem on which, for the moment, I suspend judg-
ment. Many varieties of method are used in these schools. There
are phonic methods, in which the sound-values of the letters are
taught first; though, nowadays, the names of the letters are
not neglected. Of these the Dale System has earned, and rightly
earned, many enthusiastic advocates; though I think, personally,
that it has unnecessary elaborations and that its reading matter
is not always suitable for young children. There are also syllabic
methods, of which Sonnenschein's is the best known, where the
INTRODUCTION
9
syllable is taken as the unit of language, and analysis stops short
at the syllable. Again, there are look-and-say methods in which
there is no analysis of the word at all, and in which the word is
sounded as a whole. Some head teachers say they vary their
methods; and in other schools, I regret to say, there is one method
in one class and one in another.
Fortunately, the young child is not as bothered by this
kaleidoscopic progress as an adult would be; but, nevertheless,
it wastes his time and delays his advance. And as an adminis-
trator, I have often had to say, being asked for advice by head
teachers, "Get your staff together, discuss the matter fully, and
then decide on your method. Having adopted it, maintain it
unchanged for a period of years and see that all your teaching
in every class and by every teacher conforms to it." I do not
say this because I think all definite methods are equally good;
but because I think that any method, definitely adhered to, is
better, far better, than no method at all.
But, if I say I do not think all methods are equally good,
I should be able, if I am called upon, to give grounds for my
opinion. Those grounds may be rational, i. e., based on accepted
principles of child psychology, or they may be based on scientific
evidence. The former method of argument has had a long run
for many, many years; but we seem as far from convincing our-
selves as to the best methods as ever. Some of the principles
appealed to are true; some are false; but nearly all of them are
tainted by what I have previously called the diminished-adult
view of the child-mind. Our own psychological processes are
put into the child, diminished in strength, but similar in form.
We are getting old and worn, many of us. We do not like the
mechanical acquisition of new things; it is hard for us; so we
say children do not like it. As a matter of fact, they do. Repeti-
tion bores us; so we say it bores the young child. As a matter
of fact, he loves it. We pride ourselves on our intelligence rather
than on our memory. I am not sure that we should come out
really well if our intelligence were tested: we have probably,
most of us, passed our upward limit years ago, but are not aware
10
TEACHING BEGINNERS TO READ
of it; though we do know that our rote memories are faulty.
Without abundant associations we remember very little. Straight-
way we assert that the young child is like us. We teach by
methods so intelligent that he does not understand us, we asso-
ciate his knowledge by so many bonds, or try to, that he loses
sight of what he really has to remember; and we say his memory
cannot improve, knowing that our own does not.
No doubt all this is better than no psychologizing at all.
But it is not child-study; and its conclusions are in complete
variance with the facts. We may argue, in fact we often do,
that if the child is not naturally like us, he ought to be made so.
This is an argument which may have ethical or philosophical
grounds: it certainly has no basis psychologically. So we will
leave it aside. We will suppose only that children, most of them,
can learn to read. Our problem shall be, When and How?
CHAPTER II
OBSERVATIONS IN 1905-1906
I have used the term "observations" deliberately. I tried to
discover which of the prevailing methods of teaching reading to
young children gave the best results in the infants' schools
themselves, before the children were promoted to the senior de-
partments. But I made no specific arrangements, nor laid down
any conditions as to the mental equality of the children tested.
In other words, I made observations, but did not experiment.
I thought it might be possible, in so large a school system as
that of London, to find the conditions of valid comparison,
arranged as it were, accidentally; and, in many ways, I was not
disappointed.
From my own inspectorial knowledge, aided by that of one
or two of my colleagues, I selected three infants' schools, each
of which had an extremely high reputation for the teaching of
reading and whose successful results were well known. One
taught exclusively by the "Dale," an elaborate phonic system;
the second by the "Sonnenschein," a definite and well-worked
out syllabic system; the third by the old Alphabetic, not phonic,
spelling system. This last system was fast disappearing in
infants' departments; but we have to remember that it is still
almost universal in senior schools. The head mistress had often
been told she was out-of-date, a terrible accusation in pedagogy,
where the fashions change every few years. But she was a
strong-willed woman; and since the reading of her school was
admittedly of the highest order, she was not interfered with.
As to the social class of the children attending them, the
schools were much, but not quite, alike. They were all placed
in good neighborhoods, one in the south, one in the southwest,
and the other in the southeast of London. The southwestern
school had a slight advantage, the southeastern school came next,
and the southern last; but there were no very serious differences.
11
12
TEACHING BEGINNERS TO READ
The southwestern and southeastern schools had good systems of
home reading, the southern school had not.
All the schools were well built, with halls for assembly and
other modern appliances. The classrooms were all what we
now call large, accommodating 60 children (boys and girls) and
all the schools were very full; so full that children could not
be promoted from the infants' departments at the proper age.
The teachers of the classes were all thoroughly efficient and the
head teachers were women of experience, capacity, and strength
of mind. Each of them "swore by" the method she had adopted
and gave many reasons therefor. Moreover, the time-tables were,
for reading purposes, practically, though not quite, the same: each
school gave approximately the same time in hours and minutes
per week.
It looked as if I had fairly valid conditions of comparison
arranged for me, and I was not without hopes that I should
obtain some useful conclusions. So in 1905 and in 1906 (during
the intervening period I was in America), with the full co-
operation of the head teachers, I tested every child above a certain
age in the three schools from an unseen reading book, copies of
which I took with me. The book selected was Alice in Wonder-
land, admittedly very hard for young children. But, quite obvi-
ously, reading can only be really tested from an “unseen” reading
book; and also quite obviously, if we wished to obtain a wide
range of marks—and no others are of much or, at least, of easy
use for statistical purposes-the book chosen must be difficult.
I tested each child individually in a room apart, the head
teacher and the teacher of the class being present. My maximum
mark was ten, given for accuracy and fluency. The head teacher
and the teacher of the class called my attention to any case in
which the mark given did not correspond with what was thought
to be the real ability of the child; and in such cases I gave
further tests. In fact, the head teacher and I marked inde-
pendently and compared our results after each case. No bad
pedagogical times, such as Monday morning, were chosen; and
I felt that we had done all we could to ensure a fair comparison.
OBSERVATIONS IN 1905-06
13
The results were carefully tabulated and I lectured on them to
teachers in the course of a series on Pedagogical Methods in
Schools, delivered by me for the London County Council from
1909 to 1913, in which latter year I resumed my inspectorial
duties.
I think I ought to say at once that the differences between
the average marks, age-group for age-group, in the three schools
respectively, were not very high, and were probably statistically
invalid. Such as they were, they came out thus: the Dale system
first, the Alphabetic system second, and the Sonnenschein system
third. But all the reading was of the highest order; and I frankly
told the teachers who came to my lectures, that I should not
recommend any of them to use effort and thought and expend
municipal funds (for new books might have been necessary), to
change their methods on the evidence of the facts supplied by me.
A more detailed statement of the results follow, together with
some incidental observations, which, in this case, may probably
be of more value than the summarized statistics.
Every child of seven years and upwards was examined in each
of the three schools, the Dale school, the Sonnenschein school,
and the Alphabetic school. Although the schools were by no
means the same size, all the classes had 60 children on the roll or
closely approximated to 60, and there were no "weak" teachers.
The classrooms in all the schools were very full and there were
children over eight who could not be promoted to the boys' and
girls' departments; because they, too, were very full. The time
given per week to reading (including word-building and spelling)
was carefully noted for every class in the three schools tested.
The Phonic school gave three hours weekly in all classes from
Standard Ia to Grade II inclusive. The Sonnenschein school
gave three hours thirty minutes per week in all classes above
Grade II, but much less in that grade. The Alphabetic school
gave three hours thirty minutes in Grades II and III, but rose
to four hours thirty minutes in the Standard I class. It cannot
be said that there is any serious difference; though the somewhat
smaller time throughout in the Phonic school and the extra hour
in the standard class of the Alphabetic school are worth notice.
14
TEACHING BEGINNERS TO READ
It was, of course, useless to summarize the results for children
of different ages in different schools. Indeed, hardly any of the
children over eight should have been in the infants' schools at all,
and would not have been, had not the senior departments been
so crowded. Nor would it have been of any value to compare
standard by standard, and grade by grade; for, to take one in-
stance only, the Grade III of the Alphabetic school contained a
very large number of children who, in the Phonic school, would
have been classed as Standard I. But, by taking every child in
each school at 7 years 0 months to 7 years 11 months inclusive,
from whatever standard or grade he happened to be, we had a
fairer basis of comparison.
The length of time each child had attended his present school
was calculated and all children noted who had not attended school
before or had passed some time in another school. Their marks
were, at first, tabulated separately; but, so soon as it was seen
that they would not affect the comparative results, they were put
in with the others. They were, generally speaking, a little lower
than the average for the school; but that was equally true for all
the schools tested. The "floating" population among children
attending elementary schools is rarely the best; but as I have
said, each school was equally affected, and none of them greatly.
In the 7-8 year groups we had 82 children in the Phonic
school, averaging 7 years 5 months, who scored 6.7 marks for
reading out of a possible maximum of 10. In the Alphabetic
school we had 58 children, averaging 7 years 8 months with an
average mark of 6.4; and in the Sonnenschein school we had
117 children of an average age of 7 years 6 months who gained an
average mark of 5.8. All these are very high marks for an
examination based on Alice in Wonderland for children of this
age.
Standard variations (o's) were calculated and the probable.
errors (P. E.'s) of the differences between the means or averages.
The difference between the Phonic and Alphabetic averages
is not statistically reliable; for the difference is 0.3, and
its probable error is 0.38. The difference between the
OBSERVATIONS IN 1905-06
15
Alphabetic and Sonnenschein average is hardly reliable, for the
difference is 0.6 and its probable error is 0.39. The difference
between the Phonic and Sonnenschein school is most likely re-
liable, for the difference is .9 and its probable error is .36. (A
mean difference, to be reliable statistically, should be from two
to three times its probable error.)
The reader will therefore see that by an observational or
inspectorial method, even though we worked with very great
care, we have obtained a result which is far from conclusive;
though, as far as it goes, it is corroborative of my later work. I
proceed now to note a very few points which interested the
head teachers with whom I worked, and which may interest other
teachers. In the Sonnenschein school there had been, contrary
to the recognized system, some phonic building-up of the syllabic
bases in one of the lower classes; this was quite evident during
my examination.
In the Phonic school there were more errors than I should
have expected in small words which the children really knew.
I was not then aware that the child on any system apprehends a
well-known letter and sound, and then "jumps." The Standard I
class in this school had already read from an ordinary reading
book not arranged on the Dale system. I noted that the Grade
III children who were 7 and upwards, only 8 in number, knew
a few sounds only. I thought the rate of the Dale reading was
very slow; every word seemed to be a problem; and the system
does not stop a child from making bad shots, though I found out
later that it minimizes them. And I noted, too, that the sounding
of the component letters did not, by any means, always give the
word, and often not the syllabic constituents of it. It seemed
to me that in throwing away syllables, the system deprived itself
of a valuable aid. Let any educated adult, for example, sound
the word "utility" letter by letter, without syllables, then try to
pronounce the word, and he will readily grasp my meaning.
Of course there were many words in Alice in Wonderland quite
outside the Dale system as far as known; and this was true for
the Sonnenschein system as well, and also for the Alphabetic
system now to be commented on.
16
TEACHING BEGINNERS TO READ
In the school which used this method, the head mistress
urged me also to hear the children from their own reading books.
This I did, selecting parts they had not yet read. Of course
the children did read better from their own books, but not so
very markedly so, considering the difficulty of Alice in Wonder-
land. There was naturally a good deal of alphabetic spelling of
unknown words and this made the reading rather slow. One
point impressed the head teacher and me very much. Many of
the most fluent readers of their own books could not attack new
words as well as some of the poorer readers. They were work-
ing, for the time, below their highest level of endeavour; and,
not usually meeting difficulties, were forgetting how to deal with
them.
Distinctness of utterance was most marked in the Phonic
school and least in the Alphabetic school; but the results from
the latter were so high that I felt fairly sure the method was a
good one. Teachers in senior schools practice an alphabetic
syllabic method almost universally and assert with much
unanimity that this method gives the best results in spelling, which
has, of course, to be learned sometime or other. When I point
out that a phonic system is also a spelling system and, perhaps,
a better one; they retort that children so taught cannot spell, and
they do not refer merely to the irregular words. Anyway, an
alphabetic method is not one which should be discarded because
it is unreasonable, that is, as judged by an adult mind. It may
very well be reasonable enough to a child. However, I cannot
decide with any certainty on the above evidence. These comments
are opinions, not proofs, and I offer them only as suggestions.
CHAPTER III
THE WORK OF OTHER ENGLISH EXPERIMENTERS
Following on these observations of mine, Dumville, who had
attended my lectures on pedagogy and had rightly pointed out
that I had not included a school which really taught by the look-
and-say method, (for in Sonnenschein's system the look-and-say
element is confined to the syllable, and does not embrace the whole
word), began some experimental as distinguished from observa-
tional work. He read his paper at the Psychological Section of
the British Association, published it in the Journal of Experi-
mental Pedagogy, and subsequently, I believe, offered it for dis-
cussion at a meeting of the Committee of the Teachers' Guild on
Psychological Research in Schools.
Dumville took a number of groups, at least two of which, by
means of Binet Tests, he had demonstrated to be approximately
equal in general intelligence. The others were estimated as
fairly equal by the opinion of teachers. After a period, during
which the equal groups were taught, one by a look-and-say
method rigidly observed, and the other by a phonic method, also
rigidly observed, he tested them in accuracy, speed, and com-
prehension. His results were somewhat irregular, and it is very
doubtful whether the differences between his means were suf-
ficiently greater than the propable errors of their differences, ex-
cept in speed. In one school, his phonic group was superior both
in speed and accuracy, but on the whole, his look-and-say groups
were superior in accuracy as well as speed. His figures showing
comprehension were expressed in percentages of the whole pos-
sible marks, and I could not calculate from them. Apparently
there were three schools involved, with equal groups in each,
each tested two or three times and, granting the equality of the
groups at the outset, there was certainly a presumption in favour
of the look-and-say method as against the phonic in speed, ac-
curacy, and comprehension. An effort had been made to get
J
17
18
TEACHING BEGINNERS TO READ
teachers of equal ability to teach the equal groups; and, generally
speaking, this research was an advance on work previously done.
The argumentative part of Dumville's paper gave the usual
grounds for the look-and-say method, namely, the irregularity
of our pronunciation and the fact, or alleged fact, that children
really, in whatever way we teach them, judge of the word, as
adults are said to do, by its general form, and not from its com-
ponent letters or syllables. The argument as to irregularity had
been countered long ago by Sonnenschein in his little book for
teachers on the Teaching of Reading, though he might be re-
garded as an interested party pleading for his own system, which
was based on regularity.
Valentine followed Dumville, also by an experimental method,
and working with supposedly equal groups. He used an ad-
mirable device for minimizing the influence of previous method
in teaching; for he used Greek letters for English words. His
averages (the probable errors were not given) were contrary to
those of Dumville as regards accuracy; for he regarded his re-
sults as disproving the conclusions from Dumville's work. He
was prepared, however, to concede that with very young or
feeble-minded children, the look-and-say method might be better.
He argued, like Sonnenschein, that the irregularity of the lan-
guage was much over-stated and that young children were in-
terested in learning to read as such, without what we should call
interesting reading matter. Valentine's groups were not demon-
strably equal and a criticism was made that his figures were not
statistically valid.
About this time Ballard was making one of those more
general observational surveys for which he is now so well known.
By means of his One-Minute Reading Test, he collected results
from a whole district and found a phonic method (the Dale) the
most successful. Criticisms were made that his one-minute test,
consisting of words, not sentences, did not test ability in reading
and that many of his schools tested had no specific and definite
method at all. No one had asserted that comprehension of read-
ing matter was tested; and, on the other count, it could at least
WORK OF OTHER ENGLISH EXPERIMENTERS
19
be said that a definite method had proved the most successful.
Of course the general intelligence of the children in the schools
tested varied greatly; but his inspectorial knowledge enabled him
to allow for that, and he might fairly claim that for pedagogical
results his comparisons embraced every variety of school child.
I do not myself adopt a method of general survey; it is too rough
to determine nice and difficult issues, but I should not hesitate
to ascribe value to it.
Sometimes, for example, general-survey results are quite un-
mistakably clear, as in those, also collected by Ballard, in which a
comparison is made between the accuracy of the decomposition
and equal-addition methods in subtraction. Schools consistently
superior in addition, multiplication, and division were found in-
ferior in subtraction, a result which could only be ascribed to a
known difference in method. Unless, of course, ability in sub-
traction is naturally negatively correlated with that for addition,
multiplication, and division, a condition which is quite contrary to
competent English research and opinion.
But let me get back to the problem of reading. About this
time, Drever in Child-Study was asking for a better laboratory
analysis of the mental functions involved in reading, and point-
ing out that equal groups could hardly be obtained from chil-
dren who had already learned some reading; they were already
sophisticated. We must start with children who knew no read-
ing and had learned none, either at school or at home. Meanwhile,
I had been elaborating my method of equal, parallel, and highly
correlated groups. I had applied it, among other instances, to an
attempt to solve the disputed question of inductive versus de-
ductive methods of teaching demonstrative geometry. But how
was I to obtain groups of children really equal, not in reading,
but in ability to learn to read? Could I get them by means of
a test in general intelligence, following Binet's methods? I
thought not results on such a division are too jerky and irregu-
lar. I could get them in reading itself, but that would hardly
solve the question of teaching to read in its early stages; and
moreover, we should have to allow for the possible obstruction
20
TEACHING BEGINNERS TO READ
to a new method by the persistence of an earlier one, a condi-
tion I wanted to avoid.
I then remembered that for some years, after preliminary
trial in 1906 in the Child-Study Bureau at Chicago, I had tried
to find out whether a child was fit to begin to learn to read by
using Greek letters and sounds to see: (1) whether the child
could adequately perceive and distinguish unmeaning sounds or
not; (2) whether he could adequately perceive and distinguish
unmeaning shapes or not; and (3) whether he could associate
these arbitrary sounds and shapes correctly or not.
Finally, I discarded Greek shapes, and made some new ones,
then I discarded the Greek sounds and took some which were
English, but which to very young children did not mean any-
thing. Here, I thought, I have the means of testing whether
young children who know no reading are fit to begin to learn
to read. These are the underlying psychological conditions:
To perceive, discriminate, and remember arbitrary shapes and
sounds: To be able to associate them accurately. Children who
can perform these functions can profitably begin to learn to read:
if they cannot, they cannot and ought not to be taught or rather,
to suffer an attempt to be taught to read. Moreover, the success
of their undertaking will be a measure of their natural ability to
learn reading in its early stages and I should be able to form my
groups equal in natural ability for the processes to be taught. To
a description of the work undertaken to see whether my fore-
casting was correct or not, I now proceed.
CHAPTER IV
SHAPES AND SOUNDS AND THEIR ASSOCIATION
This experiment began on Friday, July 10th, 1914, with a
small number of children, 5 boys and 5 girls, of estimated aver-
age ability, from a Grade IIa class in an infants' school situated
in a poor neighborhood in the inner ring of the southern suburbs
of London. Grade IIa is the upper section of a Grade II class,
which contains children whose average age would approximate
to 5 years 9 months at the end of the educational year, then
about 8 or 9 months distant. All the teaching and testing were
done by the head mistress, who had had years of experience in
experimental work and who was, moreover, an able and sympa-
thetic teacher. The object was to discover whether children of
this age and ability could perceive, discriminate, and associate
shapes and sounds which were new to them, and remember them
over a period of time. Above all, would they show results which
were steady and consistent? The following were the original
shapes and sounds used for the purpose:
TABLE of Shapes and Sounds
66
called ra,
mo,
a
sounded as in "father."
66
66
기미
​Φ
7
+]~@\68
66
66
66
66
66
66
dell,
mon,
re,
si,
boo,
ki,
un,
ba,
fu,
lat,
"00"
(6
i
a"
u
a "
(C 66
"so."
e
"tell."
"on."
""
66
66 66
e
"me."
i
66
(6
66
"site."
""
""
"boom."
""
66
66
"kill."
""
<<
u
"under."
66
""
66
"baby."
"fume."
(6
66
"fat."
21
22
TEACHING BEGINNERS TO READ
METHOD OF TEACHING the Shapes and SOUNDS¹
Each child was provided with a piece of chalk, a mill-board
to draw on, and a slightly damp duster.
First step.-The teacher clearly and distinctly said the sound
"ra," and the children repeated it once collectively. The teacher
repeated the sound, and the children again said it collectively.
The teacher pronounced the sound a third time; on this occa-
sion, the children repeated it individually.
Second step.-The teacher drew the shape
with chalk
on the blackboard, and the children copied it three times on their
mill-boards (cleaning off the first before drawing the second, and
the second before drawing the third).
Third step.-The teacher pointed to the figure drawn on the
blackboard and said, "This is the shape that says 'ra.''
Fourth step.-The teacher pointed to the drawn figure, and
asked each child, individually, “What does this shape say?”
Fifth step. The blackboard and mill-boards having been
cleaned, the teacher said, "Draw the shape that says 'ra.'"
Sixth step. The fourth and fifth steps were repeated a sec-
ond and then a third time.
“Mo,” A, and then "dell," A, were dealt with in ex-
actly the same way.
Three shapes, with their associated sounds, had been dealt
with in the first lesson; but we were by no means sure that the
children had attained success, or, if they had, what measure of
success. So they were tested in the following manner during
the same morning, each child individually in a room apart from
all the other children.
METHOD OF TESTING THE RESULTS OF THE LESSON
First step. The teacher drew the shape
on the black-
board and asked the children to give the associated sound. She
then dealt with
and A in the same way.
1 An outline of this chapter and Chapter V was presented to the Teachers' Guild
Committee on Psychological Research in Schools on January 9th, 1915.
SHAPES AND SOUNDS
23
Second step.-The teacher said "Draw the shape that says
'dell,'" and then dealt with "mo" and "ra" in the same way.
METHOD OF MARKING THE ANSWERS
Obviously, in answer to the first questions, a child might give
one of the sounds which had been taught, but not the one as-
sociated with that particular shape. One mark was allowed for
such an answer; he had perceived and discriminated the sound,
though he had not correctly associated it with the corresponding
shape. But he might also give no answer or a sound which
had not been taught, in which case he received no mark. Or
he might, having given a correct sound, repeat the same sounds
for the other shapes; but for the second or third enunciation
of the same sound he received no mark. Or he might give the
sound correctly and correctly associate it, in which case he re-
ceived two marks, one for the accuracy of the sound and one
for its correct association. Likewise, in answer to the second
question, the child might fail to answer; he might draw a shape
which had not been taught; he might draw correctly, i.e., dis-
tinguishably, a shape which had been taught, but he might im-
perfectly associate it; or, finally, he might draw the shape cor-
rectly and accurately associate it. Mutatis mutandis, the marks
were allotted as before. This scheme therefore gave a maximum
of four marks for each shape and sound.
Unfortunately, a maximum of only three marks was tabu-
lated for this test on the ground that for reading purposes we
needed to pass correctly from "shape" to "sound," but not from
"sound" to "shape." No mark, therefore, is given in the fol-
lowing table for the correct association of shape to sound,
although we did this subsequently.
RESULTS OF FIRST EXAMINATION
I may say at once that the results were unexpectedly good.
Table I shows them.
As I have said, the results were unexpectedly good; 69
marks out of a possible total of 90 were obtained. As we ex-
24:
TEACHING BEGINNERS TO READ
pected, the association of sound with shape, or-more exactly
phrased the passage from shape to sound, gave most difficulty,
but 15 marks were scored out of a possible total of 30. To give
a correct sound seemed more difficult than to give a correct shape,
but we cannot be sure of that on the basis of these results alone.
At any rate, the children were fairly well strung out, though
there were rather too many who obtained nine marks.
TABLE I. THE RESULTS OF THE EXAMINATION following the
FIRST LESSON

NAME
AGE ON
JULY 31,
1914
(Yrs. and
mos.)
MARKS MARKS
MARKS
FOR
"ra"
FOR
"mo"
FOR
"dell"
TOTALS
FOR
ra, mo,
and dell
SUMMARY
TOTALS
I II III I II III I II III I II III
Ray S...
· ·
Frank F....
Henry R.
Arthur T...
Daniel D...
Ivy H..
Emily F....
Florence S..
Violet S.
Ada H.
•
Totals.
5 - 3
4
32+3 TILO LO
5
5
5
5
LO LO LO LO LO LO LO LO LO LO
5
•
1
1
1
1
1
1
1
1
0
0
1
1
1
0
0
1
1
0
1
0
0
0
3
23
1
1
3
1
1
1
1 3
0
0 1
1
1
1
1 1
3
0
1
3
0
0
1
1 3
♡100NMÍMII
∞ ∞ ∞ ∞ CD CD CD OMN
3222 MM I∞ no co
3
3
9
3
6
5
COLO
∞
3
3
3
3
3
2
976
6
7
3 9 10
6
9 10 6 25 29 15
69
In the columns marked I are shown the marks for a correct sound, in those
marked II the marks for a correct shape, in those marked III the marks for a
correct passage from shape to sound.
I have dealt in detail with this first lesson, with the immedi-
ate examination on it, and with its results because, if this is
clearly apprehended, there will be no difficulty in grasping the
further stages of the experiment. I suggest to any teacher in-
terested that she take a few children through the steps so far
described; the further account will then have a reality other-
wise lacking.
This first lesson was given on Friday, July 10th, 1914, at
11 A.M., and lasted exactly 15 minutes. A second lesson, also
lasting 15 minutes, was given at the same time on the morning
of Monday, July 13th, and the results were examined as before.
In this lesson "non," "re," and "si" were taught. A third les-
SHAPES AND sounds
25
son of the same length and an examination took place at 11 A.M.
on Tuesday, July 14th, in which "boo," "ki," and "un" were
taught; and a fourth and final lesson of the same length and a
subsequent examination were given on Wednesday, July 15th,
at 11 A.M. At this time "ba," "fu," and "lat" were taught.
The shapes and sounds had now all been taught and exam-
ined on, the examination for each set of six, three shapes and
three sounds, being given after each lesson on the same day. It
will not be necessary to give the results of the second, third, and
TABLE II.
SUMMARY OF THE RESULTS OF THE EXAMINATIONS
FOLLOWING THE SECOND, THIRD, AND FOURTH LESSONS



Second
Lesson
Third
Lesson
Fourth
Lesson
Name
(initials only)
Marks for
"mon"
Marks for
Marks for
Totals Summary
"boo"
"ba"
for
of
"re" "si"
"ki" "un" | "fu" "lat"
Totals
R. S...
F. F.
•
•
H. R..
A. T…….
D. D
I. H..
E. F.
F. S..
V. S.
A. H...
•
Totals.
3 2
3
3 1 3
1 1
3 3
3
3 3 3
3212 MMM232
I II III I II III I II III I II III
3 3 3
oo oo no co co co co co co co
∞∞∞ - CD CD C122
3 3 3 3 3
3 3 3 3
3 3 3 3
∞∞∞∞ ∞ MD MD MMN
∞∞∞∞ ∞ ∞ ∞ Q32
9876O
9873 a
9
27
3
25
3
9
23
1 3 0 2 3
2
5
7
15
3
3 9
3 3 9
9
27
9
27
6
24
2
5
18
3 1
8
9
23
1
6 8
сл
5
19
228
3
3 3
2
3 3
3
3 3
1
3 2
3
2 3 3 2 3 1 2
24 127 124 124 30 21 27 129 122 75 186 167
Column I, as before, gives the marks for correct sounds; column II, those for
correct shapes; and column III, those for the correct passage from shape to sound.
fourth tests in so much detail, though I am sorely tempted to do
so. Perhaps the following summary will suffice, but I strongly
urge any interested teachers to try the second, third, and fourth
sets of shapes and sounds and make out each child's record for
herself.
As in the first lesson, the results are unexpectedly high; 75
marks are obtained in the second lesson; 75, in the third lesson;
and 78, in the fourth. The same relationships are shown be-
tween the different mental functions; to give a correct shape
seems very easy, to enunciate a correct sound seems easy, and
26
TEACHING BEGINNERS TO READ
to pass correctly from a shape to its associated sound seems
rather more difficult. But after all, what have we proved? Only
that on the same day as that on which they were learned, young
children can remember shapes and sounds and associate them
when they are taught and tested, six at a time. What will hap-
pen if they are tested subsequently?
To determine this, we held a complete examination on all the
shapes and sounds dealt with on Thursday, July 16th, at 11 A.M.
The examination was conducted exactly as before, except that
instead of having to give three correct sounds and three correct
shapes and pass from shape to sound, as on each of the previous
occasions, the child now had to give twelve sounds, twelve shapes,
and twelve associations on the same occasion, and these were
asked for in a "chance" order. The method of marking and
tabulation was exactly as before. We did not expect much loss
in an examination following so soon after the completion of the
lessons, but we were hardly prepared for an improvement. The
relations between the functions remained the same, as we totaled
100 marks for sounds, 111 for shapes, and 90 for associations
as against 99, 115, and 82 in the first examination. The total
mark was now 301 as against 295 when the testing was done
immediately after each lesson and over a small range.
This is interesting and useful to know; but what I was
much more concerned about was the answer to the question "Are
the individual children steady?" I can say at once that they
were, and I shall later give a short table which shows this clearly.
But we were not yet out of the wood, for the knowledge of
these shapes and sounds was of little use for my purpose unless
it represented not a mere flash-in-the-pan due to novelty, but
remained constant over a considerable period of time. Conse-
quently a third examination was given on Wednesday morning,
July 22nd, on the seventh day after the last lesson had been given
and six days after the second examination. On this occasion we
tabulated a fourth mark for each associated couple, that is, a
mark for the passage from the sound to the shape (this is really
a spelling-problem), as we had arranged to do at first. This ap-
peared to present no more difficulty than the passage from shape
SHAPES AND SOUNDS
27
to sound, but I shall continue to show the results excluding the
fourth mark, so as to compare fairly with those of the first and
second examinations. The total marks were now 110, 109, and
102, showing a rise on the whole on the former record.
Finally, after the summer holiday was over-a short one, as
war had broken out in August—we gave another examination
on Tuesday, August 18th, on the twenty-eighth day after the
last examination. Again we gave four marks for each associ-
ated couple; but, as before, I shall use three only. The marks
were now 113, 114, and 106, showing a rise in every case from
TABLE III. THE TOTAL MARKS OBTAINED BY EACH CHILD FOR THE
FIRST, SECOND, THIRD, AND FOURTH EXAMINATIONS
(Maxima, 36)
་


Name
(Initials only)
R. S..
F. F..
II. R..
A. T..
D. D..
I. H..
E. F.
F. S.
•
•
V. S
A. H...
Total..
•
Age
(Yrs. and
First
Exam.
Second
Third
Fourth
Gross
Lesson by
Exam.
Exam.
Exam.
Totals
Mos.)
Lesson
July 16
July 22
Aug. 18
5
5
5 -
3243 — — LOBOI
10 LO LO LO
10 10 10 LO LO
36
31
31
28
20
2228
32
34
36
138
35
34
131
35
36
130
26
30
32
108
1
35
33
33
34
135
36
36
34
36
142
27
30
36
36
129
27
29
30
34
120
30
25
21
122
32
31
29
122
23
26
95
295
301
321
333
the pre-holiday record. Of course, as every teacher knows, the
examinations had really acted as a means of fixing and improv-
ing the knowledge already gained from teaching, though no cor-
rections, praise, blame, or other observations or suggestions had
been made by the examiner. I must confess, however, that I
was prepared for a greater rate of obliviscence to offset the
favorable effect of the examinations. But I have hardly yet
shown the steadiness and reliability of the individual results,
which I shall do now.
Steady improvement is the rule of Table III and the total
marks string out the pupils fairly well. The marks give no vio-
28
TEACHING BEGINNERS TO READ
lent jumps, so it would seem that by this method we could ob-
tain really equal groups, well paired, if our numbers were larger.
But the lessons are really a little easy for Grade IIa, for we
have no less than four children with full marks in the final ex-
amination, and the boys are a little better than the girls. They
average a month older, and this fact, at the age of five, may count
in their favor; but the numbers are too small to draw any valid
inferences concerning the relative ability of boys and girls.
CONCLUSIONS FROM THE EXPERIMENT WITH SHAPES
AND SOUNDS
Quite obviously, the tests on sounds and shapes and their as-
sociation have given us reliable results; and it is highly probable
that, on the results of such teaching and testing, a class can be
divided into two equal, parallel, and highly correlated groups—
equal, that is, in their capacity for the fundamental psychological
operations involved in beginning to learn to read. It is also
obvious that the tests are rather too easy for this Grade IIa, and
that these children are more than ready to be taught to read. So
for the next experiment, in which we proposed to compare the
results of two definite systems of learning to read, we chose the
class below this, namely, a Grade IIb class. To an account of
this experiment I now proceed.
CHAPTER V
PHONIC VERSUS LOOK-AND-SAY METHODS
THE PRELIMINARY TEACHING OF ASSOCIATED SHAPES AND
SOUNDS
A whole class, technically known as Grade IIb, was chosen
for the purpose of this experiment in a school of poor neighbor-
hood, but not of the very poorest. These children were about
5 years 4 months old at the beginning of the work.
It was
a separate class under one teacher and worked in a classroom of
its own.
For the most part it consisted of children of rather low
mental ability. This judgment was not a mere opinion or esti-
mate, for all the 5-year-old boys and girls in this school had
been examined by the Binet Tests¹ as a means of proper class-
grading. There were, however, a few children on a somewhat
higher plane of mental ability; but they were recent admissions,
had not previously attended any other school, and were not fit,
pedagogically, for Grade IIa. None of these children had re-
ceived any previous instruction in reading, and neither the names,
sounds, nor shapes of the letters were known to them. Such
books as any of them had at home would hardly help them in
learning to read. The London County Council rule allowing
"worn-out" school books to be given to the children (an ad-
mirable regulation) was not in operation in this school at this
time.
We proposed to select two groups equal in the psychological
functions underlying the early steps of learning to read, and then
to teach one group in one way and the other group in another
way, and finally to examine the results to find out which was
the more successful. The experiment, therefore, falls into three
parts: the preliminary teaching and testing for the formation
1
¹ They were tested on my Binet Form for Testing Backward Children.—
W. H. W.
29
30
TEACHING BEGINNERS TO READ
of the groups; the teaching by the two competing methods; and
the final testing of the results.
The preliminary lessons were four in number. They were
given by the same head teacher who conducted the former les-
sons in shapes and sounds on Tuesday, Wednesday, Thursday,
and Friday of the week ending August 29th, 1914. Each lesson
was 15 minutes in length, and was given at 11 A.M., immedi-
ately after the recreation interval. The lessons were exactly like
those described in an earlier section of this monograph; the
same shapes and sounds were dealt with in exactly the same
way. A set of six shapes and sounds and their associations
was taught in each lesson, exactly as before. There was no re-
capitulation of back work in any lesson, and consequently, if any
child were absent on one or more of the lessons, he had no op-
portunity, even an inadequate one, of "picking up” what he had
lost. Fifty children were present when the first lesson on shapes
and sounds was given, but ten of these were absent on one or
more of the four lessons, so that only 40 could be finally tested
on the results. In the individual questioning in which sounds
were required during the teaching lessons, it was not possible in
the time available to obtain 50 answers, so that the teacher had
to be content to call on a dozen or so who may have appeared
least attentive.
THE EXAMINATION OF THE RESULTS OF THE FOUR LESSONS
IN SHAPES AND SOUNDS
Four days after the last lesson, namely, on Tuesday, Septem-
ber 1st, the children were tested exactly in the way described
in a preceding section of this monograph. As there were now
40 boys and girls to be examined, the individual testing lasted
over two days. To lighten the work and save time, the experi-
menter had previously drawn each of the twelve shapes in white
chalk on a separate piece of brown paper 10 inches square.
These papers were shown in a "chance" order, one at a time, to
the child, all other shapes being hidden from view. No other
variation was made from the procedure laid down in the previ-
PHONIC VERSUS LOOK-AND-SAY METHODS
31
ous section, except that we now tabulated four marks instead of
three for each sound and shape and their association. We had
come to the conclusion that the maximum of two marks for as-
sociation, one for the passage from the shape to the sound, and
one for the passage from the sound to the shape (though this
really is a "spelling" problem) was the fairer assessment. It
will be obvious that a child could gain four marks for each as-
sociated shape and sound; that is, as there were twelve associated
shapes and sounds, he could obtain 48 marks altogether.
But what is a correct sound? Here there was little difficulty;
the child could say it, if he remembered it, in a distinguishable
manner. But to decide on a correct shape was a little more dif-
ficult. If what the child drew was distinguishable as the shape
in question, we accepted it; mere variation in the proportion of
its parts was not counted as error.
Some remarks made by a few children may be of interest,
though the great majority just "went at it" as an extremely in-
teresting exercise. One girl, at an early stage in the testing,
who could not remember how to draw a shape when the sound
was given to her and who saw the experimenter place an o in
the appropriate column of the mark sheet, said "Oh! is it like
that?" The mark sheet after this was hidden from view. A
boy, when shown the shape which has the sound "ba" (a as in
baby) said, "I don't know what you have that called." Two
other boys, when shown the same shape called it "obey." The
boys, it must be remembered, were not in the testing room at
the same time. A girl when shown the shape to which the sound
“lat” was attached (see page 21), named it "box" (this shape
evidently may have a misleading association). Another girl
when shown the shape known as "un" (see page 21) called it
"honey." A boy called upon to draw the same shape said, "I
can't do that, but I can draw onions and pears and apples."
RESULTS OF THE EXAMINATION ON SHAPES AND SOUNDS
Of the 40 boys and girls who were present at all the four
lessons on shapes and sounds, one boy was absent on the exami-
32
TEACHING BEGINNERS TO READ
nation days, and he was, perforce, excluded from the ensuing
tabulation. Thus 39 children were left for division into two equal
groups, and one other child was omitted, so that we might obtain
an equal number in each of the two groups. We started the
actual teaching of reading, therefore, with 19 recorded children
in each group.
A record of each individual's work was made as shown in a
previous section (see page 31). The shapes, as before, were
remembered better than the sounds, although there were instances
to the contrary. Their associations, of course, were found rather
more difficult. Correct sounds averaged 6.9 marks out of 12;
correct shapes 7.4 marks out of 12; correct associations of
sounds with given shapes averaged 5.9 out of 12; and correct
associations of shapes with given sounds averaged 5.6 out of 12.
As I am strongly of the opinion that the conflicting results
of experimenters have been largely, if not entirely, due to work
done with imperfectly arranged groups, I shall show carefully
how the groups with which we are going to work were composed.
It must never be forgotten that natural ability is more important
than any method of teaching, and that the abler group will al-
ways do better than the inferior group, even though the latter
is taught by the better method. The actual grouping is given
in Table IV.
THE ACTUAL TEACHING OF THE Two Groups
The two groups having been formed, the actual teaching of
reading began. The first lesson of the series was given to each
group on Tuesday, September 8th, 1914, and the lessons contin-
ued on the Wednesday, Thursday, and Friday of this and the
following weeks until Wednesday, October 21st. A week's holi-
day intervened, the two days' usual mid-term holiday having
been lengthened on account of the shortened summer holiday,
and the final lessons of the series were given on Wednesday and
Thursday, November 4 and 5. On three Thursdays the lessons
were omitted since the head mistress, who taught both groups,
was otherwise engaged. No lessons were given on Monday
PHONIC VERSUS LOOK-AND-SAY METHODS
33
mornings, when mental adaptation is low. In all, 25 lessons
were given to each group, covering a period of nine weeks,
which, of course, does not include the preliminary work on shapes
and sounds. One group-Group A-was taught by a phonic
method, the other-Group B-by a look-and-say method.
TABLE IV. THE RESULTS OF THE EXAMINATION IN SHAPES AND
SOUNDS, THE CHILDREN BEING DIVIDED INTO TWO EQUAL,
PARALLEL, AND HIGHLY CORRELATED GROUPS ON
THE BASIS OF THE MARKS THEN OBTAINED
(Maximum possible marks, 48)
PHONIC METHOD
GROUP B TO BE TAUGHT BY A
LOOK-AND-SAY METHOD
GROUP A TO BE TAUGHT BY A


Name
Age on
Sept. 1, 1914
(Yrs. and
Total
Name
Age on
Sept. 1, 1914
Total
Marks
(Yrs. and
Marks
Mos.)
Mos.)
Alfred C.
Edward H...
George B.
Doris M.
Edward C.
·
William B..
Ada P...
Louisa S.
•
Annie T..
Grace J.
Ethel B.
Queenie S.
Thomas Da.
Thomas De.
Leonard M.
Helena A..
Fred P..
Winifred B.
•
•
•
6
Ellen B..
LO LO LO LO CO LO LO HILO LO LO LO TH LO LO T CO ++
5
-11
42
George B....
6 0
5
6
41
Dora K.
5
1
37
Henry P...
6 0
0
36
Vera H..
•
11
35
Elizabeth B...
5
8
30
Arthur B...
5
9
30
Ivy A..
•
4
8
29
Albert L..
5
6
O2O7478
44
40
37
36
34
34
30
29
5
-11
28
Violet T.
26
•
·
•
2
26
Albert C.
4
6
26
4
8
23
Ella T.
23
Nellie M
-10
22
Doris J..
5 — 8
20
Edward S.
LO LO LO LO
23
1
23
2
20
1
20
·
19
Amy H..
20
5
19
Alice E.
6
18
•
0
16
Rose B..
6
7
16
4
-11
12
4 — 9
9
Agnes D...
Lily D...
4
0
16
·
4
9
6
•
Average.. 5 - 4
26.2
Average..
5 — 4
26.2
Each group worked under the same conditions as to the time
of day and the freshness of the teacher. The lessons began at
11 A.M. and each lasted 15 minutes. The group which had its
lesson first one morning had it second the next morning and so
on throughout the series. The group not engaged in learning
to read was occupied with organized games in the school-hall,
while the other group was being taught to read.
34
TEACHING BEGINNERS TO READ
During the whole experiment no reading lessons of any kind
other than those for the purpose of this research were given to
any of the children in the two groups. It was necessary to choose
a basis of some definite reading matter, and this was taken from
a book which the children in the experiment had never seen. It
was not placed in their hands during the experiment, for all the
teaching was done from the blackboard. The actual text was
exactly followed by the group taught phonically; for the look-
and-say group, sentences were compiled based on the words of
the text.
Of course the look-and-say group "got through" more read-
ing matter than the other, which, perforce, went very slowly.
In considering the ground covered during these 25 lessons, it
must be remembered that the children belonged to the lower or
"b" division of Grade II; that their average age was but little
over 5 years; that no previous reading had been done; that
they had not previously learned the names or even the sounds
of the letters; and that in these lessons they dealt with words,
phrases, and sentences.
In the lessons words involving the short sounds of "o" and
"a" were taught (though we did not limit our final examination
to these), and all the consonants were used, except k, u, v, w,
x, y, and z. Infants' teachers will realize that if words contain-
ing "k" had been taught at this stage, the children would have
confused that letter with hard "c." No word was taught in
which "q" was used. "V" might perhaps have been used with
advantage, but the "v" sound might easily have been confounded
with that of "f," which was used. No word was used which
involved "w" or "x" or "y" or "z.”
In each case we started with words. By the look-and-say
method they were not analysed, but dealt with as wholes; it is the
essence of the look-and-say method that they should be so treated.
By the phonic method they were analysed into their component
sounds and shapes; it is the essence of this method that they
should be; but it is, of course, a slow process at first. A more
detailed account of the actual process of teaching follows.
PHONIC VERSUS LOOK-AND-SAY METHODS
35
Method employed in teaching the phonic group.-Just previ-
ous to the 15-minute lesson, outline sketches of appropriate
pictures in connection with it were made by the teacher and the
children were asked to say what they saw. My own opinion is
that this is quite unnecessary in cases where the actual objects
and actions are familiar to the children. Cats, for example, they
already know well; lions and hippopotami they do not know
well, but we were not going to teach the words lion and hippo-
potamus. However, the teacher preferred it in all cases and I
acquiesced. Then the 15-minute lesson began. A picture of a
cat had been shown.
Teacher: "Who can say the first sound we can hear when
we say cat?" The teacher gave the sound of the word very dis-
tinctly, with a slight pause between its component parts. Some
correct answers were given.
Teacher, repeating the word as above: "Who can tell us
the last sound we can hear when I say the word cat?" Some
correct answers followed.
Teacher, repeating the word as above: "Who can tell us
the sound we hear in the middle of the word cat?" This was
more difficult, and only one or two bright children answered.
Teacher: "Who would like to stand by me and help to sound
the word cat?" Three of the volunteers were selected, Annie,
Willie, and Nellie.
"Annie, tell us the first sound of the word cat; Willie, tell
us the next sound you can hear when I say cat (Willie had the
hard one); Nellie, tell us the last sound you can hear when I
say cat."
The three children, standing before the class in a row, were
then required to sound their particular letters, and the class col-
lectively pronounced the whole word. Then the whole class
collectively pronounced each component sound as the teacher
touched the three children in succession, thus, c-a-t, and fol-
lowed up the analysed sounds by combining them into the word
cat, which they were required to say clearly and distinctly.
36
TEACHING BEGINNERS TO READ
At this early stage the teacher thought, and I agreed, that
it was well to call for some collective answering. Later on, when
the children had acquired more confidence, the questioning was
directed to particular children, especially those who appeared
less attentive.
The sounds in cat had now been analysed and compounded,
and the component shapes were next dealt with. The teacher
printed the word cat on the blackboard in the characters of the
printed book. May I be pardoned for a sentence or two of
apparent digression? I use the words "printed book" advisedly;
I do not mean in the characters known as "script-writing" now
in use in many schools. These are not "printed" characters, but
symbols which are a cross between printing and writing. What-
ever view we may have as to their value in writing, the early
lessons in teaching to read have no place for them. All the
characters used here should be in "print."
Pointing to the printed word cat on the blackboard, letter by
letter, the teacher said, "This is the letter which says 'c' (pro-
nouncing the hard sound of c); you say it. This is the letter
which says 'a' (using the sound of a as in bat); you say it.
This is the letter which says 't' (making the sound of t and not
calling it ter, which, unhappily, I hear so often); you say it."
Then, again, pointing to each letter separately, she asked the
class collectively to give the sounds and say the word. This
was done three times. Then children who had volunteered to
give an answer were called on, one at a time, to pronounce the
sounds and say the word, on this occasion not being assisted by
the teacher's pointing.
In the first lesson cat, pat, and sat were thus analysed and pro-
nounced a considerable amount of new matter in a first lesson
for children beginning to read at five years of age. Every new
word which occurred during the course of the twenty-five les-
sons was dealt with as above described.
At a very early stage in the progress of the lessons, phrases
and sentences were used. Thus in Lesson 2 the letter a and the
word fat were prefixed to the word cat, making an intelligible
PHONIC VERSUS LOOK-AND-SAY METHODS
37
phrase; and in Lesson 3 a sentence was introduced, namely,
"pat a fat cat." Not only the meaning, but the jingle, pleased the
little ones. After Lesson 3, beginning with words treated previ-
ously, the teacher printed all at once on the blackboard the words
to be used during the lesson, halting only to analyse new words
by the method above described.
In Lesson 4 the sentence "A man ran" was among those
printed; this involved a new sound of a. When a stands by it-
self the children were told that it is sounded like "a" in amidst.
Then, after the sounds of m, n, and r had been analysed out
and synthesized into the words in which they occurred, as de-
scribed in the first lesson, the whole class collectively went
through the sentence, sounding each letter individually, pro-
nouncing the words and subsequently the sentence. Then one
or two of the children, selected by the teacher, dealt with the
sentence in precisely the same way.
If a child were unable to pronounce the word after sound-
ing the letters (a real step to children because, after all, the
sound of c, plus the sound of a, plus the sound of t do not
really make the sound of cat, though they do when the separate
sounds are modified by composition), or if a child pronounced
the word incorrectly, he was encouraged to sound the letters
again. If he then failed, another child was allowed to help.
When a child, after sounding each letter, could not manage the
whole word, it was common for other children to say spon-
taneously "You can hear the word, can't you, teacher?" Fre-
quently, however, when these children were taken at their word,
and called upon, they also were unable to pronounce the word.
There were many spontaneous comments, similar to the
above, made during the progress of the lessons. One boy of
five years one month, who for several successive lessons had
been unable to pronounce the words, though he correctly gave
the sounds of the separate letters, appeared one day to have an
inspiration, and after sounding the letters pronounced the word
correctly. He was conscious he had scored without being told,
38
TEACHING BEGINNERS TO READ
and his achievement evidently gratified him, for he turned to
the class with a smiling face, and said, "That's comic."
Doubtless, to the educated layman, all this procedure seems
very labored and possibly unnecessary; he is accustomed to say
with a superior smile, "I can't remember how I learned to read;
I just said the words, you know, when I saw them." Perhaps
he did, and perhaps he didn't: anyway, these are poor children
with no literary help at home and they are interested and happy
to learn like this. Whether it is a successful way or not, or how
far successful (for all distinct methods have a measure of suc-
cess), we shall see when we test our results. I now pass to the
teaching method employed by the same teacher for the look-and-
say group.
Method employed in teaching the look-and-say group.—Just
as when dealing with the phonic group, before each 15-minute
lesson outline pictures were drawn by the teacher and the chil-
dren were questioned on them. Unlike the procedure with the
phonic group, however, all the reading matter which was to be
dealt with during the lesson was printed beforehand on the black-
board; but of course it was not shown to the children until the
exact moment for beginning the lesson. We did not print be-
fore the children for the reason that if they watched the print-
ing of the words letter by letter, a process of analysis, though
a very imperfect one, would be involved, and it is the essence
of the method not to carry analysis beyond the word as a whole.
I am aware that teachers often say they employ the look-
and-say method, when, with a word difficult to sound in parts,
they spell the word alphabetically and pronounce it afterwards
as a whole. I may be able to deal with the alphabetic method
later (it is a successful one); but it must not be confused with
the look-and-say, which is founded, rightly or wrongly, on
quite another psychological basis. The children have had the
pictures (which, personally, I do not think necessary unless the
reading matter is beyond the daily experience of the children),
and they have the printed words and sentences before them on
the blackboard. The 15-minute lesson now begins and proceeds.
PHONIC VERSUS LOOK-AND-SAY METHODS
39
The printing on the blackboard being turned to the class, the
teacher, pointing to the word a, said, “This is a," pronouncing
it as a in amidst; "you say it." Then pointing to the word
fat, she said, “This is fat; you say it. That," pointing to the
word cat, "is cat; you say it;" and so on, till all the words
of the sentence “A fat cat sat on a mať” had been repeated after
her, one by one, by the class collectively. This process was re-
peated a second time. Then the words were read again by the
class collectively, the teacher pointing to the words, but not pro-
nouncing them. As a further step, individual children were
now called upon to pronounce one word each. If the child
called on did not know the word, other members of the class
were allowed to help; if they could not, the teacher carefully
and distinctly pronounced the word again. Finally, individual
children were called on to read the whole sentence.
In the first lesson taught by the look-and-say method the fol-
lowing sentences were also dealt with precisely as above: A man
sat on a mat. A man had a hat. A cat sat on a hat. In subse-
quent lessons only the new words in each sentence were pro-
nounced by the teacher as above, if the children were found
able to say the other words by themselves. If not, the teacher
told them. Then the procedure indicated above was followed
with each sentence.
Some specimens of the actual reading matter studied by
Groups A and B in corresponding lessons will now be given.
(As I have already said, Group B got along much faster than
Group A and dealt with more reading matter.) Then the ex-
amination test will be given in full with the results; but, before
I give these, I shall invite attention to the impressions formed
by the teacher during the progress of the lesson.
One or two illustrations of the reading matter dealt with
during the actual lessons now follow. That which was used in
the first lesson of each series has already been indicated.
40
TEACHING BEGINNERS TO READ
Specimens of reading matter-
PHONIC GROUP
pat a fat cat,
an, and, ran,
a pan and a can
THIRD LESSON, THURSDAY, SEPTEMBER 10TH
LOOK-AND-SAY GROUP
Tom had a bat and ran.
Tom had a hat.
Tom had a bat and a hat. Tom had a cat and
a bat and a hat. Tom had a fat cat. Tom
had a can and a cat and a hat and a pan.
A man had a fan.
NINTH LESSON, WEDNESDAY, SEPTEMBER 23rd
LOOK-AND-SAY GROUP
PHONIC GROUP
A man had a hat.
A lad had a cap.
A cat can lap.
PHONIC GROUP
A man can rap on a tap. A lad can pat a
cat. A cat can nap on a lap. A lad can nap
on a mat. A cat had a nap on a lap. A man
had a nap on a mat. A cat sat on a mat and
had a nap. A lad can rap a cat. A man can
rap a hat.
TWENTIETH LESSON, WEDNESDAY, OCTOBER 14TH
LOOK-AND-SAY GROUP
A rod and a cod and a pod.
A man had a pod. A lad
has a pod. A cat sat on a
cot.
A dog has a top. A dog has not a hat, a man
has a hat. Sam had a rag on a bag. A hog
can not lap, a cat can lap. A dog has ham,
and a lad has jam. A cat sat on a lap. A
dog ran at a cat. A cat sat on a rat. A man
had a tap on a hat.
The twenty-fourth and twenty-fifth lessons in both series
were wholly recapitulatory. I shall now give the impressions
formed during the progress of the experiment.
IMPRESSIONS OF THE TEACHER AS TO THE RELATIVE PROGRESS
OF GROUPS A AND B
During the early days of the experiment the group taught
by the look-and-say method appeared to gain more knowledge
of reading than the group taught by the phonic method. At the
end of the individual lessons the former were more ready at
pronouncing words than the latter. The children of the phonic
group often forgot the sounds of the individual letters, and even
when they remembered the sounds, they frequently failed to
combine them, even in such a familiar word as cat. This dif-
ficulty of combination persisted throughout the series, though,
as time went on, in a much less marked degree. As the lessons
proceeded, however, and the words which had been analysed in
PHONIC VERSUS LOOK-AND-SAY METHODS
41
previous lessons were re-introduced, the children immediately
tried to discover what they were by sounding the letters, and not
a few of them were able to do so, smiling when they succeeded.
The children of the look-and-say group, on the other hand,
would generally say nothing if they did not immediately remem-
ber the pronunciation of the word, or else they named a word
whose relation to the word actually printed did not seem obvious.
As the words taught increased in number, they floundered over
the words which they had already learned. They appeared to be
overwhelmed by the difficulty of remembering so many, and gave
a troubled look round the class as if asking for assistance. The
phonic children, on the contrary, started off eagerly to sound the
letters; and, though often incorrect in combining the sounds and
pronouncing the words, were more often right than the members
of the other group.
It may be of interest to add that the words Tom, Sam,
and on were favorites with the children of both groups from
their first introduction; and when the head mistress passed
through the room during the hand-work lessons, quite a number
of the children would spontaneously and frequently call her at-
tention to their mill-boards, on which they had printed one or
more of these words. No printing on mill-boards was done as
part of the experiment, nor had any suggestion been made to
that effect,
Although care had been taken to see that the children knew
the meanings of the words and sentences by means of appropriate
sketches, and although at the end of each lesson the blackboard
showed complete sentences for each group, it was rare for a child
to realize that the words had any connected meaning until it
was pointed out to him. Each word was a distinct unit to the
children, but the sentence was not a connected whole. It would
appear that the actual process of learning to read put out of
action any attention they might have been able to give to the sub-
ject matter. No doubt, however, exists as to the fact that the
children thoroughly enjoyed the lessons. They would quietly
touch the teacher's arm in passing and say, "Will you come and
42
TEACHING BEGINNERS TO READ
take us today?" or "Can we do some more reading?" They also
showed pleasure in seeing the blackboard full of printing.
When half way through the course the teacher had some
doubts as to the advisability of teaching reading to children of
this age; but during the progress of the experiment she modified
her views, which she now gives thus: "Reading can be profitably
taught to children who are able to pass the Binet Tests for five-
year-old children, or a form of the tests revised and arranged
for English children. But for those who are found to be below
this mental level the time and labor involved are not justified
by the results." My own opinion is that a much quicker and
more satisfactory way of finding out whether a child or a class
of children are fit to begin to read arises from the conclusions
of this research. For Binet's tests are tests of general intel-
ligence, and not of the fundamental psychological processes in-
volved in learning to read. I suggest that we deal with the class
by giving the children the lessons on shapes and sounds fully
set forth in a previous section. A child who can secure half
the possible marks or more is certainly fit to begin. I am doubtful
about those who score less than 20 out of a maximum of 48,
though certainly some of the children who scored less than 20
marks made very satisfactory progress. On the whole we might
say that a class is fit to begin if and when it can score an average
of half-marks. I should, if possible, exclude children who come
out very low, say at 12 or under, though even this mark with
very young children might not be quite prohibitive.
I may be pardoned for emphasizing the fact that this class is
a lower division of Grade II; not an estimated lower division,
but one selected by the Binet tests.
Lastly, the mistress was much impressed by the necessity for
short lessons when new material or new processes were being
taught. She found that as the 15 minutes allotted to each lesson
drew to a close, some, even of the abler children, showed a
falling-off in memory. When asked at this period of the lesson
to read words which had occurred several times during the lesson,
and which they had previously pronounced correctly, they could
not do it. We shall now pass to the results of the lessons.
PHONIC VERSUS LOOK-AND-SAY METHODS
43
THE FINAL TESTS
At the end of the series of lessons, four distinct tests were
given to each child individually. The tests were compiled on
the following lines:
Test 1 consisted of a new piece of reading matter of which
both groups had already had all the words. Neither of the
groups, however, had read either the words or sentences in the
order given in this test.
Test 2 was a piece of reading matter of which both groups
had had all the words. Now, in addition, the look-and-say group
had been taught the actual sentences in the course of the lessons,
but the phonic group had not had the words or sentences in the
order of this test.
Test 3 was a piece of reading ш r of which again both
groups had had all the words in the course of the lessons. In
this case, however, the phonic group had previously had the
actual sentences in their lessons, but the look-and-say group had
not had either words or sentences in the order of this test.
Test 4 was composed of an entirely new piece of reading
matter, made up partly of words which both groups had been
taught and partly of words which neither group had seen. The
arrangement of words and sentences was new to both groups.
The actual words of the tests and a very short analysis of
each test follow:
Test 1
Sam and Tom had a cat and a dog on a mat. A
man had a cod on a pan. A lad ran and got a top
and a fan.
In Test 1 there are 30 words, of which 17 are different.
Test 2
Sam ran and got hot. Sam has a bag and a rag.
A lad had not a bag. A lad had a cap and a fat dog
on a mat.
In Test 2 there are again 30 words. Of the 51 words actually
dealt with during the lessons this test contains 17, nine of which
44
TEACHING BEGINNERS TO READ
do not occur in Test 3. The group taught by the look-and-say
method had read these actual sentences as part of their lesson on
October 21st.
Test 3
A man and a dog sat on a log. Tom has a cat
and a dog. Sam has a bat and a pad.
ham. A lad had jam.
Tom had
In Test 3 there are also 30 words, 17 of which are different;
nine of these do not appear in Test 2. The group taught by the
phonic method had read these actual sentences in their lesson
of October 21st.
Test 4
A dog had a bone and hid it in a hole. A fat cat
got it. She ran into a house and put it on a mat at
the door.
Test 4 also contains 30 words: it consists of 17 words which
both groups have been taught, the arrangement into sentences
being new to both groups, and 13 words which neither group
had seen before. Of the 17 words "a" is used six times, and the
word "and" is used twice. Of the 13 new words, "it" is used
three times. Leaving out repetitions, we have 11 old and 11 new
words. Seven words contain two sounds and four contain one
sound which neither group have had either singly or in combina-
tion. The new words were bone, hid, it, in, hole, she, into, house,
put, the, and door. The old words were a, dog, had, and, fat,
cat, got, ran, on, mat, and at.
Of the 51 words taught in the course of the lessons the four
tests contain between them 32. The repetitions of the same word
were required for the purpose of making intelligible sentences.
How the tests were administered.-Each of the four tests
was printed in bold characters on separate sheets of cartridge
paper measuring 8 by 6 inches. Forty copies of each test were
made by means of a hectograph, making in all 160 copies, which
permitted the examiner to have a copy of each test for each
child upon which she could write exactly what each child did on
each test. Upon each sheet appeared the number of the test-
PHONIC VERSUS LOOK-AND-SAY METHODS
45
1, 2, 3, or 4—and the appropriate reading matter. Spaces were
left in which to write the name of the child and the time in
minutes and seconds taken to read the exercise.
The testing began on Tuesday, November 10th and continued
on Wednesday, Thursday, and Friday, November 11th, 12th,
and 13th. In order that the length of time between the last teach-
ing lesson and the examination should be the same for both
groups, the two children who were paired with each other in the
two groups worked the tests in alternate succession. In order
to avoid temporary fatigue which might have caused inaccuracy.
Tests 1, 2, and 3 were taken at one sitting and Test 4, which
was the most difficult, at another. Each child was tested in a
room by himself. The test was placed before the child, and the
examiner upon an exactly similar sheet wrote down precisely
what the child did. Against each word pronounced correctly a
"tick" was placed, and when a word was incorrectly read the
sounds made and the actual pronunciation given to the word
were written by the examiner above the printed word. We had
thus four complete records for the work of each child. The
exact time of beginning and ending each test was also recorded.
Method of marking the results.—One mark was awarded
for each word read correctly. If a child in the phonic group
sounded the letters of a word correctly, but failed to pronounce
the whole word, no mark was given. If a word occurred more
than once in a test, a mark was given each time it was correctly
pronounced. It was thus possible for a child to gain 30 marks
for each test, or 120 marks in all. The time of beginning and
ending the exercise was registered by a watch showing seconds.
At first the watch lay on the table; but after the first test, it was
hidden. One or two children who were being examined had
asked "What's that for?" pointing to the watch. No child in
either group attempted to read any word backwards, and every
child pointed to each word as he read it or tried to read it, and
was able to follow on to the next word and line unaided. With
few exceptions, the children in the phonic group sounded the
letters before pronouncing the word. Possibly they could have
46
TEACHING BEGINNERS TO READ
pronounced many words correctly without doing so; but during
the lessons this method had not been encouraged; neither, it may
be added, had it been discouraged.
It was noticeable, that when the children of either group
misnamed a word they seldom gave, in their incorrect answers,
words other than those they had been taught. There were, how-
ever, a few exceptions. The phonic group introduced got, rag,
bon, lot, dot, god, dog, punt, head, sun, long, lamb, hand, and
lat. The look-and-say group introduced gog, rog, lot, lod,
lan, pet, mouse, wag, rad, her, and lat. (Lat, it will be remem-
bered, was one of the sounds given in the preliminary lessons
on shapes and sounds).
When endeavoring to pronounce a word, a child often
sounded the letters more than once. If he did this, rightly or
wrongly, he never altered the sounds he gave at the first attempt.
The members of the phonic group sometimes pronounced a word
incorrectly, although they had given the right sounds to the
letters. On the other hand, though much less frequently, they
sometimes pronounced a word correctly although they had given
some of the individual sounds wrongly. I have said that we now
had a complete record for every child in every test showing
precisely what he did for each word unless he were quite accurate.
I give one or two examples of the records. The letters separated
by hyphens were sounded and the words pronounced as printed
above. The actual words of the test will be found on pages 43
and 44.
TEST 2
Phonic Group.-Louisa S.-Age: 4 years 8 months. Time: 9 minutes
3 seconds.
man
mot
Sam r-a-n and g-o-t hot. Sam has a bag and a rag.
as
dog
mat
man.
A l-a-d had n-o-t a bag. A lad had a cap and a f-a-t dog on a m-a-t.
Look-and-say Group.—Albert L.-Age: 6 years 1 month. Time: 3 minutes
3 seconds.
Sam at am on not.
Sam Sam a ran at a mat.
A rat Sam on a bat.
A at pat a rat as a rat not on a pat.
PHONIC VERSUS LOOK-AND-SAY METHODS
47
TEST 3
Phonic Group.-Queenie S.-Age: 5 years 2 months.
32 seconds..
gog
Time: 5 minutes
pat
A man and a d-o-g sat on a log. Tom has a c-a-t and a dog.
fat
bat
and had
hat
Sam has a b-a-t and a p-a-d. Tom h-a-d h-a-m. A lad h-a-d jam.
Look-and-say Group.-Nellie M.-Age: 5 years 1 month. Time: 2 minutes
14 seconds.
A has and a bag as on a got. Tom has a cat and a bag
Sam has a has and a got. Tom had has. A got had jam.
TEST 4
Phonic Group.-Fred P.-Age: 6 years 0 months. Time: 9 minutes 1
second.
gog
bon
lad lot lad
lod
lot
A d-o-g had a b-o-n-e and h-i-d i-t i-n a h-o-l-e. A fat cat got i-t.
hos
pan lot
er
rod
tob rat ot
S-h-e r-a-n i-n-t-o a h-o-u-s-e and p-u-t i-t on a mat at t-h-e d-o-o-r.
Look-and-say Group.-Rose B.-Age: 6 years 7 months, Time: 2 minutes
12 seconds.
A dog on a lad and and lad lad a and. A fat and dog lad.
Sam jam had a lad and and lad on a lad on lad jam.
The head mistress had already made up her mind as to the
frequency and kind of errors which were made, for she had
tested each child herself with four separate tests. We now had
128 recorded results before us and I asked her to go through
the records and note those of her impressions which were justified
by the results. It appeared that, except on some occasions with
Tom and Sam, which had been very frequently repeated in the
reading matter used for the look-and-say group, the phonic group
were more accurate on every word of the tests. It appeared also
that the children, when in error, did not take the word as a whole,
but selected individual letters for recognition and then made a
"shot" at the word. This would, of course, to some extent, be
the procedure expected of the phonic group; but it is not the
method of error which would be expected from the look-and-say
group.
48
TEACHING BEGINNERS TO READ
A list of erroneous pronunciations follows:
Actual Words
(all of which had
been taught)
Sam
and
had
cat
dog
on
mat
man
had
cod
pan
lad
ran
am
Errors Made by
Phonic Group
can, as, man
has, hat, bat, ran
fat, man, can, mat
gog, rog, log, hop
not, can
map, cat, man, fat, pat
Sam, can, ran, and, mat
had, can
hog, cot, rod, nod, hod
pat, can, man, an
as, ad, can, ran, had, bat
can, man, rat
god, dog, rod, not
pot, mop
got
top
fan
fat, man, can
Errors Made by
Look-and-Say Group
am, man
lad, at, as, sat
hop, and, cat, has, sat, man
cod, hat, mat, and, on, man, can
bag, rag, lad
can, not, Tom, an, man, log,
hog, at, rog, and
cod, hat, as, pat, hot, cat, lad,
man, not
lad, can, mat, er, fat, as, cat,
at, an, and, has, Tom
had, hot, sat, Sam, and, mat,
can, at, fat, on, hop, sad
rod, can, got, cat, and, cad, sat,
bat, nod
mat, rod, can, hat, man, bat,
and, an, her, has, sop, lad, hot
hat, man, bad, lat, sat, fan, and,
as, rat, not, got, pat
got, rod, fat, pat, on, man, rat,
rod, mouse, jam, and
dog, rod, rag, can, bag, rat, fat
pot, lad, fan, hat, tap, pat, hot,
cot, mop, hop, and, sat
fat, dog, bat, tap, mat
Actual Words
(none of which had
been taught)
bone
hid
it
in
hole
she
into
house
put
the
door
bat, bon, log, dog, cot, not, mop, not, man, sat, an, can, on,
man
lad, hand, fan, had, ham,
ran
lap, lad, at, lat, cat, has,
an, fat
lamb, lad, an, long, nap,
can, man, has
lad, hoe, hat, has, had, hog,
lod
Sam, can, sop, has, sob
lat, top, lad, on, Tom, not,
got, and
has, hos, ham, sun, head, as,
dog, on, hat
pat, pot, an, cat, not,
ran, pan
top,
nap, hat, cat, Tom, has,
had, can, hot
rod, hog, ran
had, and, got, lad, cat
had, lad, hop, hot, on, her, lat,
cot
not, lad, on, at, cat, as, mat,
jam, got, had, man, and
lad, can, nod, on, an, hot, lat
hat, lad, sat, and, on, ham, can,
fat
Sam, rod, lad, sat, hop
lad, and, on, hat, at, mat, ham,
lad, Sam, man, had, not
has, and, sat, Sam, man, not,
cat, lad, can
pot, cap, Sam, hat, mat, tap, at,
pat, on, sop, fat, and, got
cat, ham, sat, and, get, mat,
man, lat, fat, on, can, got
rod, cat, on, and, an, cot, rat,
Sam, jam, dog, not
PHONIC VERSUS LOOK-AND-SAY METHODS
49
Speaking generally, the errors made by the phonic group are
traceable to the perception and memory of the letters whose
sounds have been taught, while the errors made by the look-and-
say group too often look like random shots—any word which
they have learned during the lessons seems to do. But they too,
in a number of cases, seem to be guided by some recognized letter
and to give words in which it occurs. A very few mistakes
appear to have been introduced by "making sense," and a very
few by the general appearance of the word, e. g., hop for dog
in the phonic group, and hot for had in the look-and-say group,
though the latter error might be explained differently.
Consider the pronunciations given to the untaught words
house and put. The letters h and s appear to play a prominent
part in the erroneous words given for house; we have has, ham,
sun, head, hat, sat, Sam, and as; while the o seems influential in
hos and possibly in not. For put, also an untaught word, the p
and t appear to suggest pat, pot, not, top, pan, cat, cap, hat, mat,
at, sop, fat, and got. In the last test the children of neither group
could pronounce bone, another untaught word. The long sound
of o and the sound of e had not been taught. With but five ex-
ceptions, the children of the phonic group sounded the o short
and identified e as c, thus making b-o-n-c, which, as teachers
would say, was a first-rate error. Two of the five exceptions, not
knowing e, left it out, and sounded the word bon. No child in
the look-and-say group achieved anything near the correct word.
And for the word door, while there were, as might have been
expected, no correct pronunciations, the phonic group gave only
three substitutes, rod, rog, and ran. On the other hand, the
look-and-say group gave eleven, viz., rod, cat, on, and, cot, an,
rat, Sam, jam, dog, and not. When trying to sound the letters
of the word hid, also an untaught word, and containing the un-
taught letter i, the phonic group, quite rightly, on the basis of
their knowledge, took i for an I and sounded it is such in every
case. Five other untaught words contained i; and every child
in the phonic group sounded it as I in every case. In so far as
the look-and-say group seemed to be guided by the look of the
50
TEACHING BEGINNERS TO READ
word at all, they too appeared to be influenced by the occurrence
of individual letters, although of course they could not sound
them.
From the foregoing analysis, the phonic group appear to
advantage; they were much more successful than the look-and-say
group. They are really more in advance than our final tables
will show, for we have allotted to each word correctly pronounced
one mark every time it is used, and it is largely on the necessary
repetitions of connecting words that the look-and-say group has
achieved what success it has. Moreover, the phonic group can
attack new words with some measure of success, which the other
group can scarcely do at all. As I hold that no system of teach-
ing is good which does not, where possible, yield high "transfer"
values, if they may be so called, I regard this consideration as
of great importance.
Another consideration seems to me to be worth much. How
far does the method induce spontaneity and interest? I am well
aware that middle-aged psychologizing attaches interest to its
own interests, not to those of the child. Because it is interested.
in "meanings" rather than "words" it ascribes the same interest
to the young child; because it often likes vague and inexactly
generalized knowledge better than a definite and exact system, it
ascribes the same interest to the child; in both cases wrongly,
I think; but let us see what the children say for themselves.
Which method interested them the more and which gave rise to
greater spontaneity? The head mistress herself was quite con-
vinced and the following remarks of the children in the phonic
group support her view.
One boy, before he started to read Test 3, pointing to the
words, said, "There's jam and there's Tom. My bulb what
teacher gave me is growing up." "They're marks, ain't they?"
he said, pointing to the experimenter's paper.
Another boy, having completed Test 2, said, "I'd like to do
five more." Moreover, on the following day, when he had com-
pleted Test 4, he made exactly the same remark.
A girl, upon completing Test 2, said "I like doing this. I'm
getting clever;" and when Test 3 was finished, she said, "I've
PHONIC VERSUS LOOK-AND-SAY METHODS
51
soon done that." Again, at the end of Test 4, in which untaught
words formed part of the exercise, she said, "I like saying these;
I know every one of them." She did not, for she only obtained
17 marks out of a maximum of 30; but she went very willingly
to work.
A boy after finishing Test 2 said, "I like doing this," and
at the end of Test 4 said, "Any more?" Another boy, after
reading the word Sam in Test 2, pointed to a second word Sam
and said, "Here he is again." The same boy began sounding
the letters of the word Tom and said, “What did I say was that
word?" Also, having finished Test 3, he said, "Can I do some
more words?" and having completed Test 4, he said, "What do
you put there?" in reference to the marking on the examiner's
paper.
A girl before starting Test 1, said, "I know that and that and
that and that, "pointing to the words Sam, Tom, mat, and top;
but she did not know top when she came to it in the course of
reading. Another girl, having finished Test 2, said, "I like doing
this," and in Test 4, when she encountered the untaught word
door, said, "I don't know this word, there are two o's in it,"
(giving the short sound of o twice).
A boy who gained 20 out of 30 for his first test said, "I'm
clever;" and before he started Test 2, remarked, "I know all the
words." But he did not, for he too only obtained 17 marks out
of 30 for his exercise. He went on to say, comparing Test 2
with Test 1, "That's what I just learnt, ain't it?" Again, before
starting Test 3, he said, "There's two Toms." Then, pointing
to the paper on which the examiner was entering the actual work
of the children, he said, "That's the cleanest one, ain't it?"
After finishing Test 4, he drew his finger along the line of print,
and said, "I've learnt that and that and that and that, ain't I?"
A boy, having finished Test 4, pointed to the word door
(an untaught word) and said, "Two o's (giving the short sound)
—that's funny." It was the only word with the two o's in the
tests. A girl after finishing Test 1 pointed to seven o's in dif-
ferent words, and said, giving the short sounds, "That's o and
52
TEACHING BEGINNERS TO READ
that's o," etc. Another girl, when reading Test 4, pointed to two
untaught words and said, "I don't know them," and then to
another saying, "That's one I didn't know" (but this last word
was not one previously used). The child said this before she
sounded the letters of the words.
A girl while reading Test 4 came to the word “it” (an un-
taught word), and giving i the sound of 1 said "That's boy."
The examiner thought she was trying to call to mind lad and fail-
ing to do this thought of boy through lad, which was itself sub-
conscious; but, of course, boy may have been immediately thought
of, though it does not make sense. If the examiner's opinion is
correct, it is a rather queer instance of estimating the word from
a single letter.
The only instance of spontaneous assertion or request made
by the look-and-say group in connection with the final tests was
that of a boy who, pointing to a small space unoccupied by
printing on one of the test papers, said, "Can't we have some
more in here?" It was the examiner's opinion that, with the
exception of the length of time taken, the phonic group had
scored all along the line. Let me now set out the final tables of
results and make the necessary calculations.
FINAL TABLES of Results
We started our teaching with 38 pupils; but during the rather
protracted period our lessons occupied, four of them had left
the school, one from the phonic group and three from the look-
and-say group. George B., the top boy in Group B, had left,
so we crossed out Alfred C-., the top boy in Group A. Ada
P., in group A had left the school. Ivy A.-., who paired
with her in Group B, had also left. Edward S. in Group B
had left, so we crossed out Thomas De-., who paired with him
in Group A. We now had 32 paired children available for final
tabulation. This tabulation is presented in Table V.
There were a few absences during the lessons in both groups;
the only serious case is that of V. H., who was away for 14
days in the country. Ag. B. and En. B—. were young recent
PHONIC VERSUS LOOK-AND-SAY METHODS
53

THE MARKS OBTAINED IN THE FINAL TESTS WITH TIMES TAKEN BY GROUPS A AND B COMPARED
GROUP A. THE PHONIC GROUP
Group B. THE LOOK-AND-SAY GROUP
INITIALS
AGE ON
MARKS
SEPT. 1, PRELIM-
1914
(Yrs. and
MARKS FINAL TESTS
AVERAGE
TIME
AGE ON
MARKS
MARKS FINAL TESTS
AVERAGE
PER TEST
INITIALS
SEPT. 1, PRELIM-
1914
TIME
PER TEST
INARY
Mos.)
TEST First Second Third Fourth Total (Min, and
Sec.)
(Yrs. and
Mos.)
INARY
TEST
First Second Third Fourth Total (Min, and
Sec.)
E. H.
5
G. B.
5
D. M.
E. C.
5
W. B
5
10 10 10 30 10
6
41
29
30
30
17
106
3 0
D. K.
5
1
37
28
25
28
16
97
5 -16
H. P.
0
36
26
24
24
17
91
6 -14
V. H.
4
6-11
35
27
25
28
17
97
5.-27
El. B
8
30
20
20
23
13
76
7
9
Ar. B
4
LO CO + LE
—
2
40
21
23
25
16
85
2-32
6
0
37
21
25
25
14
85
3-15
7
36
18
14
18
7
57
3 -49
5
4
34
21
19
7
34
18
16
2223
11
74
3 -32
20
12
66
3 -37
I.. S..
4
A. T
5 -11
G. J
E. B
5
Q. S.
5
∞ 1 ∞ ∞ 2
THŁOŁO LO LO
8
29
28
2
26
5 8
23
2****
20
24
24
24
24
20
25
23
25
** 22 *
24
24
22
24
**2**
14
82
14
86
14
80
24
15
86
23
16
88888
8 -24
A. L
5 -11
V. T.
6
9 56
0
A. C.
8
E. T.
5
0
N. M
5
654 LO LO
1
29
13
11
12
6
26
20
20
21
13
73
43
6
1
74
2-20
6
26
15
12
17
9
53
4 -22
6
23
12
1
23
17
27
14
13
7
46
4 -35
19
21
9
66
3-17
T. D
4 -10
22
20
L. M.
5
1
19
20
H. A.
4
5
19
11
F. P.
6 0
16
W. B
4 -11
12
22=22
16
21
10
67
17
19
11
67
10
11
6
8223
5
4 5 ∞
-23
D. J
5
2
20
14
14
17
9
54
2 -51
6 -36
A. H.
5
1
20
17
14
18
10
59
2-23
8
-12
A. E
6
1
18
17
16
17
10
60
3
3
24.
22
15
83
6 -47
R. B.
6
7
16
13
17
18
11
59
2-27
12
10
14
7
43
7 -47
Ag. B
4
0
16
21
19
24
11
75
2-20
·
En. B.
4
9
9
22
19
24
14
79
4 -35
L. D.
4
9
6
CO
15
14
13
7
49
8
9
Averages..
5
3
25.3 22.2 20.91 22.6 13.5
79.1
6 -17
Averages.. 5
3
25.2
17.3
16.41 18.9 10.1
62.8 3-31


TABLE V.
54
TEACHING BEGINNERS TO READ
admissions to the school, and both of them showed an im-
provability unexpectedly high in all their school work. I shall
now set out the usual calculations, though they might fairly be
dispensed with, owing to the regularity of the figures. The
time taken by Group B is much less than that taken by Group A,
an average of 3 minutes 31 seconds per test as against an
average of 6 minutes 17 seconds per test. The difference be-
tween the means is 166 seconds and the probable error of that
difference is 21, about one-eighth of the difference between the
averages. As the children were not paired on time, but on
accuracy, the formula used for the P. E. was
2
.6745 Vo²+0,³
1
n
On the other hand it might be interesting to note that there is,
in fact, a slight and very doubtful negative correlation by the
Pearson "r" formula of -.149.
In proficiency of reading, though not in speed, Group A has
been victorious all along the line in every test, as the figures given
in the table show. The averages are 79.1 as against 62.8; the
difference is 16.3; the correlation between the final results by
the “r” formula is .241; and the probable error of that difference,
worked out by the formula
P.E. 0.67 Vo? to?
20 σ σ
1 2
n
is 3.19. This is only one-fifth of the difference between the
means. It may be of interest to note that for the two series of
marks in the preliminary tests on shapes and sounds (excluding
the children who had left and their paired associates), the cor-
relation is .83 with a difference between the means of .1, whose
probable error is .87. That is, there is no significant difference.
THE TWO GROUPS EXAMINED BY MEANS OF BALLARD'S
ONE-MINUTE READING TEST
Dr. Ballard's test, for those who do not know it, may be de-
scribed as consisting of 158 disconnected words, starting with
words of two letters and proceeding to others of a more difficult
PHONIC VERSUS LOOK-AND-SAY METHODS
55
character. The specified time allowed for the test is one min-
ute. In that time the child is supposed to read as many words
as he possibly can. I need hardly say, perhaps, that Dr. Bal-
lard's test was constructed without any reference to my research;
he used it for the general survey to which I have referred in an
earlier chapter of this monograph. His test was arranged with-
out any reference to any particular system of teaching reading,
though to me it seems, at least in the earlier stages (with which
alone I am experimentally acquainted), to give an excellent
chance to children taught by the look-and-say method, if what is
claimed for that method be true.
The reading matter for our phonic lessons had consisted of
phrases and sentences and for our look-and-say lessons entirely
of sentences, and now we were going to ask the children to see
what they could do with a number of disconnected words. Our
final tests in reading were completed on Friday, November 13th,
1914; this further test of Ballard's was given on Thursday,
November 24th. Each group consisted of 16 children of the
average age of 5 years 6 months.
How Ballard's test was administered. The examiner
printed all the 158 words on a piece of stiff paper in bold type.
The words were placed in rows in the order of ordinary reading
matter. Enough hectographed copies were made for the exam-
iner to have one copy for her own use with each child, upon
which to record, word by word, how the child worked the test.
Each child was tested in a room alone, and was given one minute
to read as many words as he could. The watch used for timing
was kept out of sight, in order that there should be nothing to
distract the child's attention from the work in hand. Indeed,
the children did not know that a watch was being used, or that
they had only a limited time for their exercise. A copy of the
test sheet was placed before each child and he was told to read
the words. At the end of one minute, carefully timed, he was
told to stop and the test sheet was taken from him. If the child
pointed to the words or letters with his finger he was not dis-
56
TEACHING BEGINNERS TO READ
couraged from doing so; on the other hand, the action was not
suggested to him.
The method of marking the results.-In recording the work,
the examiner wrote the child's name on her copy of the test and
placed a tick above each word pronounced correctly. These
alone counted for marks, one for each correct pronunciation.
When the letters were sounded, rightly or wrongly, they were
written down; and the actual pronunciation given to the word
was also written above the actual word of the test. Mutatis
mutandis, this was done with both groups.
The examiner's comments on the results.-Comparatively
few words were read by the children in either group in the min-
ute allowed for the exercise. The maximum number attempted
by any child in the phonic group was 11, and in the look-and-
say group 13. With the phonic group this was to be accounted
for by untaught sounds and words, and by the delay caused in
sounding the letters. With the look-and-say group difficulty also
lay in untaught words. They, not sounding the letters, read
more rapidly the words they did know, but they hesitated so
much over the untaught words, looking up at the examiner for
help, that they had to be told to go on reading. Of the 11 words
attempted by the phonic group, eight contained one sound which
they had not been taught. The words were is, me, by, so, us, it,
or, and be; the sounds were i, ē, y, ō, ŭ, and ö. The i-sound
occurs in is and it; the e-sound in the words me and be; the
y-sound in by; the o-sound in so; the u-sound in us; and the
remaining o-sound in or. In the 11 words which these children
attempted, there were only three which did not contain untaught
sounds: these words were at, on, and an. All the consonants,
with the exception of y, were known to them, but only the short
sounds of the vowels a and o.
The look-and-say group had been taught to pronounce words,
but not to sound letters; and it was very interesting to see what
they did with the untaught words. Of the maximum number
attempted, only four were words which had been taught, and nine
were untaught. The taught words were at, on, an, and as. The
PHONIC VERSUS LOOK-AND-SAY METHODS
57
untaught words were is, me, by, so, us, it, or, be, and to. Con-
fining our attention to the words which had been taught, we
note that the phonic group was much more accurate than
Group B. For example, only one child in Group A was wrong
when pronouncing the word at, and that one said fat (15 out
of 16 were correct). On the other hand, in Group B 13 out of
16 were incorrect. Three were right and the rest gave vari-
ously, and, fat, cat, sat, mat, rat, not, and an.
With the untaught words no children of either group scored
any marks, though the phonic children got nearer the correct
pronunciation than did those of Group B, as the following ex-
ample will show. In dealing with the untaught word so, every
child in Group A, with the exception of one girl who did not
attempt the word, sounded the letters s-ŏ, giving o the sound
which it possessed in the words they had been taught; seven
pronounced the word so (o as in Tom); three pronounced it
sop; two pronounced it son; one said sond; one called it on,
and one called it Sam, who was a great favorite with both groups.
Fifteen children of Group B also attempted the word so. Four
gave words containing the sound of o, but no child gave a word
which contained both the s- and o-sounds.
Every child in Group A gave the l-sound for the letter i in
the word it, for the l-sound had been taught and not the i-sound;
and they either took the i for a badly printed 1 or thought I was
1
the nearest of the letters they knew. In the word by, every
child sounded the y as j, for the j-sound had been taught, but
not the y-sound. In the word me, every child, identifying the
e with c, gave the hard sound of c for e. Again, in the word
us, all these children, with the exception of one, gave the n-sound
for the letter u. The letter s appeared in eight of the words
given for us by Group A. Two called the word not, involving
the n sound; and six would not pronounce the word at all.
The look-and-say children were more confident; 14 out of 16
made a shot at the word. Ten of the errors made involved the
letter s, three involved the letter n, and one contained no let-
ter found in us, even allowing n for u, which it resembles.
They appear to see a letter they know and dash at the word.
58
TEACHING BEGINNERS TO READ
It now remains to give a summary of the results. I arrange
the groups in sections according to the marks obtained in the
preliminary test on shapes and sounds given months before.
TABLE VI. THE MARKS OBTAINED BY THE TWO GROUPS ON
BALLARD'S ONE-MINUTE READING TEST

AVERAGE MARKS
NUMBER OF
Group A
Group B
MARKS IN
PRELIMINARY TEST
CHILDREN
IN EACH
Prelim-
Prelim-
Ballard's
Ballard's
GROUP
inary
inary
Test
Test
Test
Test
Over 25..
25 to 35..
15 to 25.
Below 15..
Average..
∞ LO CON
3
38.0
3.0
37.7
2.0
29.6
2.6
29.8
1.2
20.3
2.2
20.0
1.3
2
9.5
1.9
11.0
0.5
25.3
2.4
25.2
1.3
The correlation coefficient between the two final series is .41;
and the difference between the means is 1.1, with a P.E. of .13.
The phonic group has scored a complete victory, involving
in this case time as well as accuracy. But why should the look-
and-say group have made a better time record in my own tests,
as they most unmistakably did? I think the explanation lies in
the fact that there was not so much real trying when they did
not know a word: they made a “shot" and passed on quickly to
a word they knew. This would avail them nothing in Ballard's
test; for, as they went on, the words of the test became more
and more unfamiliar. And of course they did not sound the
letters of the words, which takes time in itself. Rapidity with-
out accuracy does not seem to me a desirable combination for
young children in such a study as learning to read.
CHAPTER VI
PHONIC VERSUS ALPHABETIC METHODS
The alphabetic method, from its apparent unreasonableness
to the adult mind, has suffered an eclipse in infants' departments
during the last fifteen or twenty years. Teachers of senior
schools maintain it almost universally; though, because it is old-
fashioned, they are anxious not to parade it as a basis of teach-
ing, however much they may employ it. They assert, and
rightly, that an alphabetic method teaches a child to spell. They
forget that a phonic method is also a spelling method; or, if they
admit that, they assert that it produces poor spelling. Whatever
method is to be employed, it must satisfy two practical condi-
tions, even if it is theoretically perfect to the adult mind: it must
be a method usable by teachers under ordinary school condi-
tions, and it must be suitable to the mental development, or
lack of development, of the children with whom it is to be em-
ployed. An alphabetic-syllabic-spelling method does satisfy these
practical conditions, though it has many theoretical imperfec-
tions, as, indeed, which "system" has not?
In my judgment, however, there is little ground for the
wholesale overthrow of the alphabetic method, except of course,
change of fashion-which I do not deny is a most potent force,
perhaps as potent in education as in dress. I am not asserting
or implying that this method is the best in the early stages of
teaching reading. I think I can prove that it is not; but it makes
a very good show in results, and is continuous with senior-school
methods. Whether it is better in the early stages is a problem
to which I now address myself.
GENERAL PLAN OF THE EXPERIMENT
The school selected for this experiment was the same as that
in which was conducted the experiment described in the last
chapter of this monograph on phonic versus look-and-say meth-
59
60
TEACHING BEGINNERS TO READ
ods; and the same teacher-the head mistress-gave the lessons
and conducted the tests. It will be remembered that these chil-
dren came from rather poor homes, and that at this time-April,
1915-many of their fathers and brothers were absent from
home with His Majesty's armed forces.
The class selected was a Grade II class-a technical expres-
sion used in London implying usually an average age of about
5 years 9 months at the end of the educational year. In this
school the age was lower. This was the first month of the edu-
cational year in London's elementary schools; many Grade II
children would enter later on in the year, and the whole grade
would then be divided into an upper and a lower section. There
is no doubt that most of the children in this experimental class
would be placed in Grade IIa-the upper section-at the end
of the half-year, and they were on the whole young and bright
children, averaging about 4 years 6 months at the beginning of
the experiment. None of them knew anything about reading,
or even the names or sounds of letters.
The first, and perhaps the most important, stage of the ex-
periment was the division of the children into two equal, parallel,
and highly correlated groups-equal, that is, in the psychological
functions required in the very early stages of learning to read.
Then, when the division had been satisfactorily made, one group
was taught by a phonic and the other by an alphabetic method.
Finally, all the children in both groups were tested individually
on the reading matter dealt with in the course of the lessons,
and on new reading matter not so dealt with. Throughout the
course of the experiment, all teaching lessons lasted exactly
15 minutes. A more detailed account of each stage of the ex-
periment follows.
THE LESSONS AND TESTS ON SHAPES AND SOUNDS AND
THEIR ASSOCIATION
Four lessons on shapes and sounds and their association were
given to the whole class of 59 children, boys and girls, on
Wednesday, April 28th, on Thursday and Friday, the 29th and
PHONIC VERSUS ALPHABETIC METHODS
61
30th, and on Tuesday, May 4th. They were given precisely at
11 A.M., immediately after the recreation interval. The same
shapes and sounds were taught as previously described, in the
same order and in the same way; and now, as formerly, there
was no recapitulation in the later lessons of the shapes and
sounds dealt with in the earlier ones.
THE TESTING OF THE RESULTS OF THE LESSONS ON SHAPES
AND SOUNDS
The method of testing was identical with that described in
the first experiment. The testing began on Friday, May 7th,
continued on Monday, May 10th, and was completed on Tues-
day, May 11th. Of the 59 boys and girls with whom we started,
nine had been absent on one or more occasions, and one was
absent from the examination. These 10 children had, perforce,
to be excluded; and as two equal groups were required, another
child was omitted who would not pair. There were, therefore,
24 children in each of the groups. The average age of each
group was 4 years 7 months; each group obtained 807 marks
in the examination on shapes and sounds; and the individual
children were adequately paired. Table VII, showing the re-
sults, follows shortly.
Comments on Table VII.-These marks are high, very high;
and were they not exceptional, I should feel very hesitant about
my own general criterion. This is a good class of children; and
though they are very young, it is fairly certain, as I have said
before, that most of them will be placed in Grade IIa, the upper
division of Grade II, in the ensuing October. But there are cer-
tainly some among them, who, by my own criterion (see page
42), should not yet begin to learn to read. The two groups are
well balanced as to age and ability; the correlation between
the two series of total marks for Groups A and B is practically
perfect; so let us include all the children, dull ones as well as
bright, and see what they can do.
62
TEACHING BEGINNERS TO READ
TABLE VII. THE RESULTS OF THE EXAMINATION ON SHAPES AND
SOUNDS AND THE DIVISION INTO TWO EQUAL GROUPS
GROUP A TO BE TAUGHT READING BY A
PHONIC METHOD
GROUP B TO BE TAUGHT READING
BY AN ALPHABETIC METHOD


Names
Age on
April 30, 1915
(Yrs. and
Totals
Names
Age on
April 30, 1915
Totals
Mos.)
(Yrs. and
Mos.)
Edith S...
Albert B.
Fred B...
•
Helen R..
Millicent C..
Fred B..
Charles K.
William F.
William D..
•
LOT TO H
5
4
4
5
4
11
ONO+++
0
48
Beatrice H....
2
47
Charles R.
·
6
46
David B..
4
46
Margaret R..
4
45
John W.
4445O H
9
48
0
46
6
46
4
3 LO
46
46
■
•
Robert B...
+4
275+∞
44
Sidney S
4
45
44
Henrietta B
4
6
44
•
43
Albert D.
42
•
4
40
Albert P
4
40
·
40
Henrietta P………
9
40
Thomas P. … ….
John S..
4
40
4
- 8
38
•
Ernest K.
5
0
36
George F
4
-10
38
•
Dorothy W.
4
-11
35
Caroline B.
Ernest L.
4
9
31
Dorothy B....
Emma G..
4
- 5
34
4
11
*
•
5 - 5
31
Rose C..
5 - 1
•
•
Gladys F...
4
4
30
Mary I...
4
-11
30
Henry A..
Ivy W.
4
•
Percy R...
Frederic I
Albert A.
444
0
27
Elsie P..
9
27
Lily A.
4
25
Ellen C.
547
•
Marjorie R..
4
George C..
4
Dorothy B...
Thomas H..
4
4
1342
18
Sidney D..
16
Sidney H.
4
14
Dorothy H.
4
4970∞
157
32 22222
33
30
30
29
28
26
26
17
17
13
4
Sarah H..
4 - 9
5
Average..
4
7
33.6a
Average.
4 7
33.6ª
•
a I do not account for these very high marks for these young children. The cause
was neither their exceptional mental ability nor the teaching and testing by the head
teacher.
THE LESSONS IN READING GIVEN TO GROUP A AND GROUP B
Chronology. The two groups having been formed, the actual
lessons in reading began, Group A being taught by a phonic
method and Group B by an alphabetic method. The lessons com-
menced on the Thursday following the Whitsun recess, May 27th,
1915. They were given throughout the course at the same time
of day, in the morning immediately after the recreation inter-
val, each lesson lasting exactly 15 minutes. The children were
PHONIC VERSUS ALPHABETIC METHODS
63
taught in their own classroom by the head mistress, while the
group not engaged in reading was having a number lesson in
another room with their own teacher. The groups were taught
reading alternately thus: on the first morning of the lessons,
Group A began at 11 A.M. and Group B at 11:20 A.M.; on
the second morning of the lessons Group B began at 11 A.M. and
Group A at 11:20 A.M.; and so on throughout the course. Both
the groups of children and their teacher would thus, on the
whole, have equally favorable times for teaching and learning.
The lessons continued on Friday, May 28th, Tuesday, June
1st, and Wednesday, Thursday and Friday, June 2nd, 3rd, and
4th. The lessons were again given on June 8th, 9th, 10th, 11th,
15th, 16th, 17th, 18th, and 22nd. On Wednesday, June 23rd,
the school was closed for "sports," and on the next day the
children seemed too tired to learn much. The next lesson was
therefore given on Friday, June 25th, and proceeded on the 29th
and 30th of June and on July 1st, 2nd, 6th, 7th, 8th, 9th, 13th,
14th, 15th, 16th, and 20th, making in all thirty lessons up to
this date. Then came the mid-summer holidays, and the lessons
were not resumed until Wednesday, August 25th, after an in-
terval of exactly five weeks. Additional lessons were given on
August 26th, 27th, and 31st, and on September 1st, 2nd, 3rd,
7th, 8th, 9th, and 10th, and the final lesson, making forty-two
in all for each group, on Tuesday, September 14th.
The method of teaching the phonic group.—The reading
matter, and the methods adopted for teaching it, were exactly
the same as in the preceding experiment in phonic reading, and
do not need further description here.
The method of teaching the alphabetic group.-Immediately
prior to the actual lesson in reading, appropriate pictures in con-
nection with it were shown, and individual children were asked
to say what they saw. (I have previously stated that this is
unnecessary when the actual objects, actions, and positions are
known to the children as well as the words describing them; but
it was done throughout at the instance of the teacher.) The
64
TEACHING BEGINNERS TO READ
time thus occupied was not included in the 15 minutes allotted
to the actual reading lesson. The lesson proceeded thus:
1st step. The teacher printed at on the blackboard, letter by letter, saying:
"This is 'a' (giving the alphabetic name) all say it." (once). "This is 't'
(giving the alphabetic name) all say it." (once).
2nd step. Pointing to each letter, the teacher said "a-t spells at." The
class then spelled the word collectively, and did this three times.
3rd step. Children were called on individually by name to spell and
pronounce the word. The teacher in this step did not point to the letters.
If the child selected did not succeed, another child was called on to do it,
and the one first selected tried again.
The words at, bat, fat, pat, and sat were thus dealt with
during the first lesson. One more lesson will be given by way
of example the fourth.
Preliminary step. Pictures were shown which were described by the children.
The reading matter dealt with was "a man ran, a rat and a cat, a fat cat
ran,” and the pictures were appropriate.
1st step. The teacher printed the letter a on the blackboard by itself
and said, "Letter a when it's all alone by itself spells a," (giving the sound
of a in amidst) All say: 'a (giving alphabetic name) spells a.'"' (sounding
a as in amidst). This was done collectively three times and then individually.
2nd step. The word man was next dealt with. It was a new word, but
the letters were now known. The teacher printed the word on the blackboard,
letter by letter, and asked individual children to give the names of the letters
(not their sounds). She then said, "m-a-n spells man. You spell it."
The
class spelled and pronounced it three times, and then individual children were
called on to do it, one at a time.
3rd step. The word ran had been taught in a previous lesson. The
teacher printed it, letter by letter, on the blackboard, and without spelling or
pronouncing it herself, or calling for collective work, asked individual children
to spell and pronounce the word.
4th step. The whole sentence "a man ran was by this time printed
on the blackboard, and the teacher called on the class collectively to spell,
pronounce, and read the sentence once, afterwards calling on them individually.
If a child could not do it, another did, and the one who had failed tried again.
When a new letter occurred, the teacher, of course, told its
name, but otherwise the procedure was exactly as given above.
At the end of the 15-minute lesson the blackboard showed all the
reading matter dealt with during the course of the lesson. No
child was asked to pronounce a word without spelling it, but he
was at liberty to do so. These children had had no previous in-
struction in naming letters. As our fathers and mothers would
have said they did not know their alphabet. They were learning
the names and the spellings in their actual uses in words.
PHONIC VERSUS ALPHABETIC METHODS
65
Some observations arising from notes made by the teacher.-
Up to and including the fifth lesson, the children in both groups
seemed slightly less attentive than those who worked in the
phonic versus look-and-say experiment previously described.
This was probably due to their lower age, as they were, on the
average, nine months younger than those who took part in the
former experiment. It was noticed that the phonic group suc-
ceeded with the pronunciation of words already taught better
than the alphabetic group did after spelling them, though the
accuracy of memory for names and sounds appeared to be equal.
On Wednesday, June 16th, the alphabetic group read a few
more words than the phonic group. The phonic group read,
"A lad has a cap." "A cat can lap." "A lad sat at a tap." The
alphabetic group read in addition, "A cat had a nap." "A rat
ran at a cat." It would be rash to infer that therefore they were
making more real progress in reading. Up to this time no child
was able to pronounce any word without first naming or sound-
ing the letters. But it must be remembered that they had not
been called upon to do so, nor, indeed, were they throughout the
course of the experiment. If a child is called on to do this, he
may "jump" from one or two known letters. Thus, it is quite
common for a child called upon to read has to say had, and also
to pronounce the word man as can, etc. We did not, therefore,
press them to "shoot" at the word; we should only have mul-
tiplied errors and delayed real progress.
On Friday, June 18th, the word ram occurred for the first
time. We ought probably to have refrained from using the
word at all at this stage, but there it was. A picture of a ram
had been shown to the children before the lessons started, but
no child in either group knew what it was. Some called it a
sheep, but the great majority called it a cow.
ham was dealt with for the first time, no child recognized the
picture given as that of a ham. It was called a piece of wood,
and a piece of stone, and quite half the children said it was a
fish. This was not the fault of the picture. It was due to a
When the word
66
TEACHING BEGINNERS TO READ
lack of real knowledge which pictures cannot make up for, al-
though in certain conditions they are fair substitutes.
At about this period in the course the word jam was dealt
with, and j was a new letter. Both groups retained the names
and sounds of letters and the pronunciation of the word quite
easily, and subsequently when a new word involving j was shown
there was a strong tendency for the children to jump to the
word jam without naming or sounding the letters. Four days
later, as j had been a new letter in the previous lesson, it was
repeated. No sooner did the letter appear than the children of
both groups-who were taught separately, be it remembered—
thrust up their hands and, contrary to rule, called out altogether
"jam." This was the only instance during the whole course
where the eagerness shown was quite so excessive.
Up to Friday, June 25th, neither group appeared to perceive
that it was reading sentences. And it was noted, when a new
word was printed on the blackboard, or a word which had not
been used for some time, that the alphabetic group floundered
about more than the phonic group. Some of the latter pro-
nounced a new word correctly the first time it appeared. But
no child in the alphabetic group could pronounce the word bat,
for example, after spelling it. It was variously called cat, mat,
and rat.
By Tuesday, June 29th, both groups were giving full atten-
tion, and were steady at their work. On Wednesday, June 30th,
the word Sam, which had been previously used, occurred again.
After several children in the phonic group had sounded the let-
ters of the word, suddenly many of them called out the pro-
nunciation. In the alphabetic group, however, no child could
pronounce the word after naming the letters and spelling the
word; and finally the teacher had to do so. This situation was
not uncommon with other words; and it appeared that while the
alphabetic group seemed to learn only by constant repetition, the
phonic children were often (not by any means always) able to
pronounce a word after having sounded the letters, even though
it were a new word.
PHONIC VERSUS ALPHABETIC METHODS
67
On Thursday, July 1st, a new vowel, the letter o, was intro-
duced in the syllables ot, not, hot, got, cot, and pot. No child
in either group was able to deal with the letter, though it is one
of the easiest in the alphabet, and might, one would have thought,
have been "picked up." No child in the alphabetic group could
pronounce these words. The teacher had to do so, though sev-
eral children tried very hard, after, of course, spelling the words.
They appeared to be guided by the letter t, and gave words con-
taining that letter. For the word got, for example, the pro-
nunciations bat, mat, and rat were given. The best reader in
the group said bat. But in the phonic group one or more chil-
dren could generally pronounce the syllable after sounding the
letters, though they failed with ot and not. The word pot ap-
pealed to the children of both groups. When the word was pro-
nounced there was a spontaneous burst of laughter in both cases,
though neither group had seen or heard the other group deal
with the word.
Generally speaking, the children remembered the names of
the letters and their sounds, but many times failed to pronounce
the word, even though it had been dealt with in previous lessons.
It would therefore appear to be scarcely profitable to spend much
time with separate names or sounds, which are learned rather
easily. The combination into words or syllables presents greater
difficulty.
On Thursday, July 8th, the word ran reappeared in the
reading matter for the first time since June 24th. Eight chil-
dren in the phonic group pronounced the word after sounding
the letters, but no child in the alphabetic group could pronounce
it after spelling it.
On Tuesday, July 13th, the word on was printed on the black-
board for the first time. All the children of the alphabetic group
could name the letters, but none of them could pronounce the
word. Many children of the phonic group could do so after
sounding the letters. The teacher was interested to obtain an
individual result, so she had asked each child to whisper the re-
sult of his or her attempt.
68
TEACHING BEGINNERS TO READ
The word cod was also a new word. No child in the alpha-
betic group could pronounce the word, although the naming of
the letters was correctly given. The word was pronounced
pad, lap, pat, and even man. The children of the phonic group
sounded the letters easily, but only one child could give a really
correct pronunciation. Evidently they could not yet be relied
upon to combine well-known separate sounds into one word,
even though these sounds had, on a previous occasion, been
analysed out from the words themselves.
On Wednesday, August 25th, after an interval of five weeks,
the lessons were resumed. It might have been expected that
there would be a considerable setback in the work. But this
was not the case. Both groups showed practically perfect reten-
tion of the names and sounds, although the pronunciations had
suffered somewhat. Generally speaking, both groups remem-
bered with certainty only those words which had occurred many
times during the course of the experiment. Of the less common
words, the phonic group was fairly accurate in pronunciation,
after sounding the letters several times; but the alphabetic group
was unable to pronounce these words after naming the letters,
or, as it is commonly described, "spelling" the words. May I
say in parenthesis that the phonic group also spelled the
words though their spelling was, of course, "phonic" and not
"alphabetic?"
Such words as Sam, Tom, has, ran, and and were correctly
spelled and pronounced by each group; but words like pad, bag,
get, not, and rod were finally pronounced only by members of
the phonic group. Having sounded the letters they succeeded;
but the naming of the letters did not enable the alphabetic
group to secure the correct pronunciation. So the teacher went
back to Lesson 20, which had been given previously on Friday,
July 2nd. This was the lesson in which the vowel o first ap-
peared, and in which some rather unusual words had been given
in order to include certain consonants. It was found, however,
that the recapitulatory work was very quickly done by both
PHONIC VERSUS ALPHABETIC METHODS
69
groups, and they were able to cover twice as much reading mat-
ter during the given time as when it was first presented.
The final teaching lesson, the forty-second, was given to both
groups on Tuesday, September 9th. These were the sentences
dealt with: Sam has a bat and a pad. A man had a fat dog
and a cat and a cap. A lad ran and got hot. A dog sat on a
log. Tom got a cod on a rod.
This does not seem much to attempt after forty-two lessons
of fifteen minutes each, spread over many months, and it must
not be supposed that every child in both groups could even read
all these sentences fluently. But this I will say: this school,
not now in my own inspectorial district, is one of the best I
know, an encomium which is most fully accepted by the in-
spector in charge of it; the head mistress who taught the chil-
dren is one of the best infants' teachers I have ever met; and
if more cannot be accomplished here under perfectly fair condi-
tions, it cannot by these methods be accomplished anywhere
with children of this class. Perhaps there are quicker and bet-
ter methods; we may deal with that question later. Meanwhile
it is interesting to make certain comparisons within the school
itself.
In the previous experiment, in which equal groups of chil-
dren worked by a phonic and a look-and-say method respectively,
25 lessons only were given. During the present experiment
42 lessons were given, yet the children, taught in both cases by
a phonic method and the same phonic method, only covered the
same ground as in the previous experiment. It is true that
there was a midsummer break in the second case, but this was
not found to be a serious hindrance. The children taught
phonically this time were a brighter group than those so taught
in the previous experiment, and the results of their preliminary
work with shapes and sounds and their association appeared
better. Why then did they make slower progress? The ages of
the earlier group averaged 5 years 4 months at the beginning of
their experimental work, while the present group averaged only
4 years 7 months old. I think, having regard to their age, all
70
TEACHING BEGINNERS TO READ
experienced infants' teachers will feel that they have done well.
I include this apparent digression because one educational friend
of mine, who does not know much about infants, said to me in
a disappointed tone, "Is that all they've done?"
Our present interest, however, lies in a comparison of the
two groups taught by different methods. In the previous ex-
periment the look-and-say group managed to get through more
reading matter than the phonic group. It was found, however,
in the final tests that they had not really learned as much. In
this case the alphabetic group and the phonic group kept well
together throughout, though there were one or two lessons in
which the alphabetic group read more. Whether that was an
advantage or not we shall see when we proceed to our final ex-
aminations which we shall now consider.
THE FINAL TESTS and the METHOD OF MARKING THE Results
The next step in the experiment was to ascertain by actual
tests the knowledge which had been acquired by the two groups,
phonic and alphabetic, throughout the whole course of the lessons.
The results will give us an answer to the original question pro-
posed, namely; which is the better of the two methods for the
early stages of teaching to read?
Accordingly, five final tests were prepared and given. The
testing began on Tuesday, September 21st, and ended on Friday,
October 1st. It will thus be seen that the examination occupied
nine school days. The first four of the tests used in this experi-
ment were exactly the same as those used in the experiment
previously described. Each of them contained 30 words.
Test 4 was much more difficult than the three previous tests.
Seventeen of the thirty words had been read by both groups
during the course of the lessons, though the words were then
differently arranged. The remaining 13 words had been read
by neither group; they were new words. Of the 17 familiar
words, a is repeated six times, and the word and is used twice:
Of the 13 new words it is used three times. Leaving out re-
petitions, there are 11. old words and 11 new words. Of the
PHONIC VERSUS ALPHABETIC METHODS
71
11 new words, 7 contain two "sounds" or "names," and 4 contain
one "sound" or "name" which neither group had been taught
during the lessons. The new letters are e, i, ŭ, and ō. The
old words were: a, dog, had, and, fat, cat, got, ran, on, mat, and
at. The new words were: bone, hid, it, in, hole, she, into, house,
put, the, and door.¹
1
Test 5
dan and nan had a nag and a cab. dan can jog
on. bob had a tag and a sad dog. dot had a lot of
sop on a hob.
There were likewise 30 words in Test 5. Leaving out the
repetitions, there were 12 words which neither group had been
taught, though they had had all the sounds or names of the
letters; and there were 7 words which both groups had been
taught. This test, though containing words which neither group
had been taught, was easier than Test 4 because it contained no
new sounds or names of letters. The following considerations
will help towards a complete understanding of the tests.
1. Each test contains 30 words.
2. Tests 1, 2, and 3 each contain 17 different words.
3. Tests 2 and 3 each contain 9 words which are not common to the two
tests; that is, there are 9 words in Test 2 which do not occur in Test 3, and 9
in Test 3 which do not occur in Test 2.
4. Test 4 contains 11 different new words and 11 different old words,
and four new sounds or names.
5. Test 5 contains 12 different new words and 7 different old words.
6. The words taught during the lessons numbered 51; the above five tests
contain 33 of them.
7. The repetitions in the tests were necessary for the purpose of making
intelligible sentences, as there were only two vowel sounds in the 51 words
that had been taught.
The method of administering the tests and marking the re-
sults was identical with those employed in the experiment on
the phonic versus the look-and-say method, and has already been
fully described.
As previously stated, it took nine school days to complete the
examination. It must be borne in mind that each child of both
groups had to be tested separately, and that it was not possible,
in some cases, to give a child more than one test in one day.
Otherwise a fatigue factor would have influenced our results, a
condition which I was anxious to avoid. There were 44 children
1 Tests 1 to 4 are on pages 43 and 44; test 5 is given on this page.
72
TEACHING BEGINNERS TO READ
tested, 22 in each group. These children worked through the
whole of the experiment including, of course, the five tests.
Each of the 44 children had to be tested five times, and at the
completion of the tests had attempted to deal with 150 words.
The reader will now understand that the work of testing these
children took considerable time, especially since no child was
hurried (hurry and impatience are fatal to correct results in
pedagogical examination) and since, with few exceptions, the
children sounded or named each letter before pronouncing the
words. Even when words were repeated in the tests in order
to make sentences, the child generally dealt with the word letter
by letter before pronouncing it.
The exact time, in minutes and seconds, which each child took
to work each exercise was carefully recorded. A record was also
made of how each child dealt with each word, whether he pro-
nounced the words without naming or sounding them letter by
letter; whether he first sounded or named them before pronounc-
ing the word; or whether he named or sounded the letters without
pronouncing the word (though the last procedure was not ob-
served in Test 5). The exact letters, right or wrong, which the
child sounded or named were written down; and the word given
(if any), rightly or wrongly pronounced, was recorded. I need
perhaps hardly mention that the paired children were tested
alternately, one from each group, to obtain equal intervals be-
tween the lessons and the tests. If a child merely sounded or
named the letters of a word without pronouncing it correctly
no mark was given; but each correct pronunciation carried one
mark.
There was one rather curious case. A girl, Sarah H—, who
was 4 years 9 months old when the experiment began, and who
was placed at the bottom of Group B, appeared, from the high
marks she gained in reading Tests 1 and 2, to have been wrongly
placed by the results of the preliminary tests in shapes and
sounds. It was found, however, that as the tests proceeded she
fell somewhat rapidly away, showing that she had not been
greatly misjudged with regard to her initial ability. It appeared
PHONIC VERSUS ALPHABETIC METHODS
73
to be a case (these are very few) of very rapid learning and
relatively poor retention, although in the aggregate her final
marks were much too high for her position at the bottom of the
list in Group B.
OBSERVATIONS FROM NOTES Made by THE EXAMINER
Naming the names of the letters by Group B did not appear
to assist so much with the pronunciation as did sounding the
letters by Group A. But both groups did well, and were equally
pleased to do the tests, beginning eagerly and with smiling faces,
even though some of their attempts were poor; and no child
showed any anxiety. It was unusual for children of either group
to pronounce a word without first giving the names or sounds
of the letters. Possibly some children could have read many
words correctly without dealing with one "letter-name" or
"sound" at a time. This procedure, however, had not been en-
couraged during the teaching lessons, because the teacher believed
that it led to what is commonly called guessing or making "shots"
from what she called the general appearance of the word. This,
doubtless, is what the adult does often enough; but children
rarely. They appear rather to select some well-known letter or
letters and "shoot" at the whole word from these. In this con-
nection may be studied the errors made, especially in the new or
untaught words. To try this issue, the errors made by both
groups will be given for Tests 4 and 5. These are the tests
which contain both taught and untaught words, the former, in-
deed, contains untaught letters besides.
Test 4. (a) ERRORS MADE IN WORDS THAT HAD BEEN TAUGHT.
ACTUAL
WORD
WRONG PRONUNCIATIONS
GIVEN BY
PHONIC GROUP
log, hog
dog
had
has, his, lad
and
ran, man, lad
fat
cat
got
ran
mat
at
at, fan, got
pat, at, mat, can, hat
dog, hot, log
pan, rat, can, man, an, and
cat, man, hat, and, fat
cat, fat, mat, pat, rat, an
WRONG PRONUNCIATIONS
GIVEN BY
ALPHABETIC GROUP
bag, log, got
has, sat, hot, and, lad
can, has, log, cat, man
mat, lot, cat
fat, can, mat, has, hot, rat
log, dog, cot, cat, not, bag
rod, man, rat, can, an, fan
rod, rat, cat, man, log, sat
mat, cat, got, on, bat, rat
74
TEACHING BEGINNERS TO READ
The alphabetic group have the wider range of error, especially
in the substitution of words with a wrong vowel. In both groups
there are a few cases of error in which the children may have
been wrongly guided by the appearance of the word as a whole;
but most of the errors appear to be due to "jumping" from some
well-known sounds or letters. At this stage of reading, errors
which make "sense" and are due to a misconception of the
sentence as a whole may be disregarded as practically non-
existent.
Test 4. (b) ERRORS MADE IN WORDS THAT HAD NOT BEEN TAUGHT IN-
ACTUAL
WORD
bone
hid
it
in
hole
she
into
house
put
the
door
CLUDING THOSE WITH UNTAUGHT LETTERS AND SOUNDS
WRONG PRONUNCIATIONS
GIVEN BY
PHONIC GROUP
bon, man, on, gon, lad, hat,
cat, cod, frog, dog, can
had, hod, ran, hat, lat, bad
at, lad, can, bad, not
lad, an, lod, on, can
hol, Tom, hod, can, lad, had,
hog, hot, hand, on, log, cat
has, Sam, hot
Tom, on, gon, rod, not,
frog, cot
has, Sam, hod, hat, on, hot,
can, hog, cat
pet, pot, pat, and, at, hat,
pan, sop
hat, cap, cat, can, hot, has,
fan, at, and
rod, ran, can, on, dog, rot,
not
WRONG PRONUNCIATIONS
GIVEN BY
ALPHABETIC GROUP
bat, got, cat, log, has, cot,
cod, pat
lap, hat, had, has, and
lad, cat, had, lat, at
had, lod, lad, cat, on, dog
had, cat, has, log, sat, lad,
got
had, Sam, has
lad, cat, on, cot, log, has,
cap, pat
has, Sam, cat, hot, Tom
pat, sop, cat, bat, ran, has,
log, pan, and, hat
rod, rat, cat, man, log, rat
rod, ran, log, rat
The phonic group has the wider range of error, due, in the
opinion of the examiner, to their greater capacity for attempting
any pronunciation at all. Throughout Test 4, in both taught and
untaught material, the only words that were given by Group A
other than those which they had learned during the series of
teaching lessons were the words frog (given twice by the same
girl), bon, lat (one of the sounds taught with the associated
shapes in the preliminary lessons), lod, hand, and not. The only
new word introduced by Group B was lod.
It must be borne in mind that none of the following sounds
or names had been taught to either group of children: long o, i,
e, oo, sh, th, and w. No child could reasonably be expected to
PHONIC VERSUS ALPHABETIC METHODS
75
pronounce accurately words containing these letters or combina-
tions of letters; but it is interesting and instructive to see what
they actually did with them. No child omitted ō when it
occurred; and every child gave it the name of o or sounded it o
as in dog in the two groups respectively.
The letter i appeared in four of the untaught words. No
child in either group omitted the letter, but every child in both
groups named or sounded it as 1, and there was no hesitation
or uncertainty in thus naming or sounding it. The children
evidently were unobservant of the difference between i and 1.
The letter e occurred in five of the untaught words. Every
child of both groups named or sounded it as c (hard).
The oo in the word door was sounded as ŏ (in dog) by five
children in the phonic group, the second o being entirely omitted.
The remaining 17 children sounded the letters as ŏ, ŏ, the
two o's being sounded separately. By 6 children in the alphabetic
group the double oo was named o, the second o, being omitted;
and by 16 children it was named o, and o, each letter being named
separately.
The sh in the word she was named and sounded by every child
in both groups; and in every instance the two letters were named
or sounded separately. The phonic group sounded the s as in
the word Sam, and then aspirated the h. The alphabetic group
named the letters sh. Every child in both groups sounded or
named the letters s-h-c; but the pronunciations of the word
when given were various. All of them, however, contained the
sounds either of s or of h, or of both of them.
The th in the word the was named or sounded by every
child in both groups as t and h, and with one exception in each
group, the word was sounded or named th-c. The exception
in each case only sounded or named t and h, leaving out the
final e.
These two children knew that they did not know the
letter and did not confuse it with c, which they did know.
Of the phonic group, 15 children sounded the u in the word
put as n, one of them sounded it as ă, and the remaining 6
omitted it altogether. In the alphabetic group, the u was named
76
TEACHING BEGINNERS TO READ
n by 12 children and was called r by one child. The remaining
6 omitted the letter altogether.
Test 5. ERRORS MADE BY BOTH GROUPS WITH TAUGHT AND UNTAUGHT
WORDS WHICH CONTAINED NO NEW LETTERS
ACTUAL WORDS
ALREADY
TAUGHT
and
as
had
can
on
dog
sop
WRONG PRONUNCIATIONS
GIVEN BY
PHONIC GROUP
has, hot, dad
man, cat, and, ran
log, hog
top, has, log, as, pot, pad
WRONG PRONUNCIATIONS
GIVEN BY
ALPHABETIC GROUP
as, can, lad
lad, has, hot, sat, and
bon, cat, had, lad, man,
Sam, mat, cot
log, cod
got, cat, not, mat, bag, can
top, dog, Tom, Sam, got,
and, has, lad, pat, sat, bat,
hot
Quite obviously the range of error is much wider in the
alphabetic group; and the usual method of proceeding from some
well-known sound or name is abundantly exemplified. It is
difficult to see, however, how a child in the alphabetic group
arrived at bat after naming so - p.
TEST 5.
ACTUAL WORDS
NOT YET
TAUGHT
WRONG PRONUNCIATIONS
GIVEN BY
PHONIC GROUP
Dan
Nan
nag
cab
jog
bob
and, Sam, ran, ban
an, can, man, dam, Sam,
ran, and, fan
rag, hag, dog, bag, pag,
gag
cap, can, bag, cad, cat, pan,
lad
jam, dog, log, hog, John
bog, bad, hob, pad, log, lad,
cot, cod, dop, hop, fog, bat,
bon, got, and, dod
tap, rag, hag, bag, dog,
log, lag, lad
das, has, lad, sat, frog, had,
Sam
tag
sad
dot
got, not, hot, cot, hog
lot
not, log, got, lad, hot
of
hob
fan, for, fat, jam, nod, on
hop, hot, hod, log, bag
WRONG PRONUNCIATIONS
GIVEN BY
ALPHABETIC GROUP
cad, had, and, as, sop, can,
man, Tom, log, lap, top,
not, has, lad, sat
ran, man, cot, can, mat, and,
has, lad, Sam
log, got, pod, dog, hot, bag,
had
and, bat, lad, cat, can, bad,
dog, cod, top, bag, had
jam, got, and
lad, bat, cat, got, and, log,
hat, can, cod, has, had
not, got, cot, lad, bag, dog,
log, nod
has, got, cat, lad, and, Sam,
rod
bat, cat, got, and, lad, Sam,
cod, at, can, hat, man
lap, lad, cat, got, cod, sat,
mat, hat, had
sad, fat, as, jam, fan, bat
had, bat, dod, cat, log, and,
has, pot, got, lad, hot, hat
PHONIC VERSUS ALPHABETIC METHODS
77
The phonic group scored some marks-thirty-three-even
with the untaught words of this test. With each of the words,
both taught and untaught, some children sounded the component
letters quite correctly and yet gave no pronunciation of the word.
With the untaught words, no child in this group scored a mark
for nan or for hob; only one child pronounced the word nag
correctly; one pronounced cab correctly, one the word bob, and
one the word tag. The word jog was called jam by 7 children
out of 22; of was called fat by 3 children and fan by 7 of them.
When a word was pronounced wrongly, the word given had
almost invariably occurred during the lessons. The following
are the only exceptions: dad, dam, john, for, and frog. The
word frog was used twice by the same child, once for the word
sad, and once for bob.
In the alphabetic group no child scored a mark for any of
the untaught words, whereas the phonic children scored 33. No
child scored a mark for dan, nan, nag, cab, jog, bob, tag, sad,
dot, lot, of, or hob. For each word, taught and untaught, some
children named the letters correctly, but gave no pronunciation
of the word. No child, rightly or wrongly, pronounced a word
which had not occurred during the teaching lessons. The word
jog was called jam by 18 children out of 22: it is evident that
“j stands for jam" with them, as the old spelling books have it.
The word of was pronounced fat by 12 children, and by one
child fan.
It is noteworthy, I think, that no child in either group pro-
nounced words backwards, with only one exception, in which
dog was called god. Both these methods of teaching follow the
actual order of the letters. In this lies the explanation, for
reversals are common, both in pronunciation and spelling, when
children are taught by methods which select some letter or letters
within the word for treatment first.
It is interesting to note that the relative positions of the two
groups remained unchanged for each of the five tests in reading,
though the tests were of different character.
78
TEACHING BEGINNERS TO READ
The average marks were as follows:
TEST
PHONIC GROUP
3.
20.8
1.
19.6
2.
19.5
5.
15.2
4.
12.2
ALPHABETIC GROUP
18.0
18.0
16.9
11.5
11.0
The regular teacher of the class feared that she would have
uphill work with reading when the two groups were put together
again for reading and handed over to her. She subsequently
stated spontaneously that the alphabetic group (her method was
phonic) was at a disadvantage for only a short time. The head
teacher's term examination at the end of the year in the ensuing
April showed no difference in the work of the two groups,
but it was, of course, not of a nature to detect small differences,
being the usual routine examination made yearly for promotion
purposes. The teacher of the class also stated that the interest
which had been aroused continued to stimulate both groups,
although one boy (Jack S-) in the alphabetic group did not like
the transition, and showed temper, calling out, "I don't want to
do it like that;" but Jack soon fell into line.
TIME TAKEN TO READ THE TESTS BY THE PHONIC AND
ALPHABETIC GROUPS RESPECTIVELY
The differences throughout are in favor of the phonic group;
they are not great, but they are clear and unambiguous. The
average time per test per child for the phonic group was 5 minutes
20 seconds, and for the alphabetic group 5 minutes 44 seconds.
Taking no account of any correlation that might exist between
the two series, which had been arranged, it will be remembered,
to correlate on a basis of accuracy and not of speed, the prob-
able error of the difference between the means (24 seconds)
was 9.82 between 2 and 3 of that difference. But, as a matter
of fact, there was considerable correlation between the two series
Formula used: P.E.=.67 Vo₁² +σ₂
2
n
2
PHONIC VERSUS ALPHABETIC METHODS
79
of times, for the abler readers in both groups were, on the whole,
the quicker ones. Worked out from the individual cases by
Pearson's "r" formula, it amounted to .529. The probable error
of the difference between the means is thereby reduced. Worked
on the formula
2
P. E. = .67 Vºª +0,3
1 2
2 γ σ σ
n
it falls to 6.9, which is not much more than 1/4 of the difference
between the means. I may say that each reading test, taken
separately, shows the same comparative results-the phonic group
is always quicker as well as more accurate. The results in time
for the five tests, averaged, are shown compendiously in Table
VIII.
TABLE VIII.
MARKS IN PRELIMIN-
ARY TESTS ON
SHAPES AND SOUNDS
Over 45..
•
35 to 45.
25 to 35.
Below 25.
•
THE SPEED OF GROUPS A AND B COMPARED

GROUP A, Taught
PHONICALLY
GROUP B, Taught
ALPHABETICALLY
dren
Average Time per
Child in 5 Reading
Tests (seconds)
No. of
Chil-
Average Time per
Child in 5 Reading
No. of
Chil-
dren
Tests (seconds)
4864
305
304
LO CO
5
334
6
343
330
7
353
357
4
375
The individual results in respect to accuracy are shown in
Table IX.
At the conclusion of the experiment the average age of Group
A was 4 years 11 months and that of Group B was also 4 years
11 months. The average mark for each group in the preliminary
tests on shapes and sounds was 33.4. In the final reading tests
the average mark of the phonic group was 87.4, and that of the
alphabetic group was 75.6. The difference between the means is
therefore 11.8. The probable error of this difference is 3.75;
the difference is therefore a significant one, being more than three
times its probable error.
80
TEACHING BEGINNERS TO READ
GROUP A,
TABLE IX. THE SUMMARIZED INDIVIDUAL RESULTS
TAUGHT PHONICALLY
TAUGHT ALPHABETICALLY
GROUP B,

Marks for
Total
Marks for
Total
Name
Preliminary Marks for
Name
Preliminary Marks for
(Initials only)
Shapes and
5 Tests in
(Initials only)
Shapes and 5 Tests in
Sounds
Reading
Sounds
Reading
E. S..
48
A. B...
F. B...
M. C..
F. B..
C. K...
W. F..
W. D...
R. B.
•
•
H. P...
E. K...
D. W..
C. B..
31
E. L..
G. F..
M. I..
P. R.
F. I...
M. R.
G. C...
D. B……….
J. H...
•
•
•
*9** *3999 83528 ONNO ++
131
B. H.
48
114
47
98
C. R..
46
102
•
•
46
94
D. B..
46
65
45
111
J. W..
46
95
44
113
S. S....
45
104
44
95
H. B...
44
43
94
A. D..
42
•
40
90
A. P.
40
..
40
82
T. P..
40
40
97
J. S...
38
36
124
G. F..
38
100
106
D. B...
34
103
E. G...
33
31
120
R. C.
30
•
30
73
H. A.
30
30
27
27
18
16
@ywoo
83
I. W...
29
51
E. P..
28
37
L. A.
26
76
S. D
17
37
S. H..
17
•
14
68
D. H..
13
•
4
40
S. H...
5
***N* 88REG CIRJA AL
88
62
68
72
84
70
41
40
63
82
79
64
57
50
76
Average.
33.4
87.4
Average.
33.4
75.6
•
The results are shown more compendiously in Table X.
TABLE X.
THE WORK OF GROUP A AND GROUP B COMPARED
SECTION BY SECTION

GROUP A,
TAUGHT PHONICALLY
MARKS IN
PRELIMINARY
TESTS ON SHAPES
AND SOUNDS
No. of
Chil-
dren
Mark for Mark for
Shapes and Final Tests
Sounds in Reading
GROUP B,
TAUGHT ALPHABETICALLY
Average Average No. of
Chil-
dren
Average Average
Mark for Mark for
Shapes and Final Tests
Sounds in Reading
Over 45...
46.5
108.5
5
46.2
96.0
35 to 45.
8
40.2
100.1
6
40.3
79.0
►
25 to 35..
6
29.3
77.8
7
30.0
66.1
Below 25..
4
13.0
55.2
4 &
13.0
61.7
a I have already commented on the work of Sarah H-, in this section, whose
quite exceptional advance from 5 marks in the preliminary tests to 76 in the final
reading tests has thrown this group out of balance.
PHONIC VERSUS ALPHABETIC METHODS
81
GENERAL CONCLUSIONS FROM THE EXPERIMENT ON PHONIC
VERSUS ALPHABETIC READING
1. It appears clearly that in the early stages of teaching
reading to young children a phonic method is superior to an
alphabetic method.
2. The difference in result, though quite clearly statistically
significant, is not very large.
3. If, as teachers almost unanimously assert, an alphabetic
method is greatly superior to a phonic method in acquiring correct
spelling, this advantage might outweigh a somewhat earlier pro-
ficiency in reading, but this is a matter for experiment, not for
mere opinion.
4. In careful experimental work ranging over six months,
with excellent teaching, both groups of these "poor" children
have arrived at a certain stage of proficiency in reading. Is
this high enough to justify us in teaching reading at this age?
I incline to think not, and that for many of these children
at least it would have been better to postpone it. I have given
tests in associated shapes and sounds which can be applied in
any individual case; but, generally speaking, I am of the opinion
that it is well to begin the definite teaching of reading in the
first half-year of the Grade II class in infants' schools with the
abler or upper half only, postponing this definite teaching until
the second half of the educational year with the weaker half or
Section b of this grade. Grade II, it will be remembered, accord-
ing to London County Council rules, ends its educational year
at an approximate average age of 5 years 9 months.
CHAPTER VII
PHONIC VERSUS PHONOSCRIPT METHODS
INTRODUCTION
I have used the terms "phonic" and "phonoscript" as if they
were antithetical, because an ordinary phonic method is to be
used in this experiment and its results estimated against those
of the phonoscript method invented by Mr. Hayes. But the
phonoscript is also a phonic system; it is, indeed, the most recent
of the phonic systems with which I am acquainted. And I wish
to say in advance that, though I do not pledge myself to every
detail of the system nor to the order of the words used in the
presentation of its earlier lessons, I am strongly of the opinion
that, whatever method of teaching reading to young children a
head mistress may ultimately adopt for her school, she cannot
fail to derive very great benefit from a study of Mr. Hayes'
system. This is not the place for any attempt to describe it
in extenso. As far as may be necessary, its characteristics will
become apparent in the experiment. Suffice it to say now that,
without destroying its character as intelligible script to those who
have not learned the system (unlike the International Phonetic
Script), marks are placed without or within ordinary literal
characters which give to every letter a definite and invariable
sound. The last proposition is not a reversible one; the same
sound is represented, not by one definite shape, but by various
shapes. This involves a spelling problem rather than one of
reading, and will not be dealt with here. In phonoscript no use
is made of so-called "diphthongs," and much use is made of
silencing marks on certain letters-all of which will become more
apparent as we proceed.
GENERAL PLAN OF THE EXPERIMENT
The general plan was the same as in previous experiments.
A whole class, knowing nothing about reading, was divided into
82
PHONIC VERSUS PHONOSCRIPT METHODS
83
two equal, parallel, and highly correlated groups on the results
of preliminary lessons in shapes, sounds, and their association.
Then one group was taught by an ordinary phonic method and
the other group by the phonoscript method. After months of
careful work—the chronology of the experiment will be given
later the two groups were tested (a) in their own script or
print, and (b) in the script or print which they had not studied.
Comparisons were instituted between the results as a whole and
also when differentiation was made between (a) taught and (b)
untaught words. The impressions of the teacher conducting the
experiment are given, and the final tables of results, involving
time as well as accuracy, are subjected to the usual statistical
checks.
We shall thus be able to ascertain whether the phonic group
can read its own print better than the phonoscript can read its
own script or vice versa; we can find out which group suffers
the greater loss when required to read from a print or script
which it has not been taught; and we can find out which group,
from our knowledge of the procedure adopted to pronounce
words which it has learned, can attack more successfully the
words which it has not been taught.
THE CHILDREN WHO DID THE WORK
The school selected was a municipal infants' school in a rather
poor, slightly suburban neighborhood of South London. The
work was done with the whole of a Grade II class. There had
been, on account of falling numbers common to nearly all infants'
schools after the European war, some reduction of staff, so that
the usual division of the Grade II class into an "a" and "b"
division could not be made. This Grade II class numbered 50
boys and girls of an average age of about 5½ years during the
experiment. The head mistress, who did the whole of the teach-
ing and examining throughout, was well practised and proficient
in experimental work, besides being a very experienced and able
teacher.
84
TEACHING BEGINNERS TO READ
THE PRELIMINARY LESSONS IN SHAPES, SOUNDS, AND THEIR
ASSOCIATION
These were the same, and given in the same way, as in
previous experiments. Neither the lessons nor the examinations
which followed will be again described; they will be referred to
in the chronology of the experiment and a summary of the marks
obtained will be given in the final tabulation of results. One
spontaneous observation was made during the lessons on shapes
and sounds. When the associated sound "dell" was pronounced
by the teacher, one boy-Stanley W—, aged 5 years 4 months-
called out, "That's where the Germans go to." It was probably
a pardonable error in a consonantal sound; but this was four years
after the war, which evidently had not been forgotten. Though
the children showed eagerness, both in drawing the shapes and
enunciating the sounds, no other spontaneous observations were
made.
The first lesson was given on Tuesday, October 24th, 1922,
at 10 A.M. It lasted exactly 20 minutes. Similar lessons were
given on Wednesday, Thursday, and Friday of the same week.
Three shapes with their associated sounds were taught, as usual,
on each occasion. There was no review of back work, so that
absentees on any occasion were, of necessity, omitted from the
concluding examination.
The mid-term holiday (2 school days) followed immediately
after the fourth and last of the preliminary lessons on shapes
and sounds, and we thought it desirable to allow time for for-
getting and time for re-adaptation to mental work after the
holiday before we tested the results. So it was not until the
eleventh day after the last preliminary lesson had been given that
the examinations began, namely, on November 7th. They were
completed within three days, and only good pedagogical times
were chosen for them. Each child was tested separately, of
course; and in order to avoid any possible distractions from new
surroundings he was tested in the room in which he had been
taught. No child was hurried; the average time required for
the test was 8.2 minutes. This is not a long time, when it is
PHONIC VERSUS PHONOSCRIPT METHODS
85
remembered that each child was required to draw 12 shapes, the
names being given, and to give the associated sounds for 12
shapes, both sounds and shapes being presented in chance order.
One very subnormal boy, 7 years of age, who had been
recommended by the school doctor to be examined for a school
for the mentally defective, had learned and remembered only
two names of shapes, "ra" and "boo," and he gave these alter-
nately, as each of the twelve shapes were presented to him.
His drawings of the shapes, however, were above the average;
and he recalled five out of the twelve, but only two were associated
correctly with their names or sounds. Two children said “a
pipe" when the shape "fu" was shown, and three others said
"a worm" for the shape associated with the sound "mon." I
had evidently not quite succeeded in avoiding all possible mis-
leading associations.
The boys and girls did not know that they were ever going
to be examined on the shapes and sounds which they had been
taught. They enjoyed the tests and appeared quite contented,
even when they scored very low marks, of which they were not
informed. These children did not bother about the marks they
got; they did not even see that what they were doing was being
registered. Children somewhat older are very keen about their
marks; but these averaged only 5 years 1 month in age at this
time and were too young to be concerned. By both boys and
girls, the shape and sound of "dell" were given most correctly
and the shape and sound of "ba" least correctly. The boys
obtained an average mark of 22.9 and the girls of 22.5. More
marks were scored by both boys and girls for the shapes than for
the sounds; they are probably easier, and, in the testing, the
shapes were presented before the sounds, and so may have been
resuscitated in memory. With a possible maximum of 48 marks,
it seemed very likely, as indeed was the case, that children were
being included in the groups who were really not competent to
begin to learn to read, though one or two with low preliminary
marks actually did make good progress.
86
TEACHING BEGINNERS TO READ
THE ACTUAL LESSONS IN READING
As I have said before, two equal groups were formed, based
on the results of these preliminary lessons, and also with regard
to the ages of the children. Group A was taught by a phonic
method and Group B by the phonoscript method. Group B was
taught the reading matter in Mr. Hayes' Introductory Phono-
script Primer, Lessons 1-20. Group A, of necessity, had lessons
specially compiled for them. They were required to cover the
same ground, but of course their method of approach was dif-
ferent. They had to follow an order of presentation suitable to
a phonic method which is without distinguishing marks for the
different sounds of the same letters, though they were to reach
a stage enabling them to deal with the same words as the phono-
script group by the time the teaching lessons ended and the ex-
amination of the results began.
The lessons to both groups were given by the head mistress,
to one group at 10 A.M. and to the other at 11 A.M. alternately,
each lesson following a recreation interval. The lessons were
given on Tuesday, Wednesday, Thursday, and Friday of each
week, beginning on Thursday, November 16th, 1922, and ending
on Thursday, May 17th, 1923. During this period there were
three considerable breaks: a Christmas holiday of about two
weeks, an Easter holiday of about ten days, and another break
of a week or more before Easter in which the mistress was con-
ducting her term examination throughout the school. Every les-
son was recorded as to substance, day, time, and date; and in all
70 lessons, each of exactly 20 minutes length, were given to
each group.
In order that no child should see any reading matter not
connected with its own particular method, no letters of any kind
were written on the wall-boards except in connection with this
teaching, and no writing of any kind was done by either group
during the progress of the experiment. While what is now
known as "individual work" was in progress, no "reading" or
"letter" apparatus of any kind was given to these children. All
the reading matter used for both groups was written on black-
PHONIC VERSUS PHONOSCRIPT METHODS
87
boards, and no school books of any kind were accessible to these
boys and girls.
Method adopted in teaching the phonic group.-The method
used in teaching reading to this group was identical with that
described in a previous experiment. The matter, however, was
different and more difficult. It was specially compiled to include
all the letters, all the sounds, and all the words, and no others,
which were included in Mr. Hayes' Introductory Phonoscript
Primer, Lessons 1-20 inclusive.
The reader may, perhaps, be reminded here that these children.
had to be taught not only sentences and words, but also the
sounds and shapes of the letters. First, words and sentences
were taught which contained only what are usually called the
"short" vowel sounds and consonants; next the "long" vowel
sounds; then double o (oo) as in good; then the au-sound as
in taught; the ow-sound as in how; the ea-sound as in teach,
and so on.
The silent letters, such as k in know, had a short
line put under them whenever they occurred, but this mark was
rubbed out after the difficulty had been dealt with. Of course
in the final examination no marks were placed “in, on, or near”
any letter or word in phonic character.
This brings me to an important issue, namely, the characters
used by the phonic group. These were those which appear in all
printed books and newspapers, not what is now called print-
script, but real printer's print. It differs considerably from
script and, of course, very materially from phonoscript. The
a's and g's especially are very distinct from those of ordinary
script; and I have often thought that if there is anything in the
contention that children learn to read faster by using print-script
in the earlier stages of learning to write rather than cursive-script
(which is usually called writing), we might as well at first, "go
the whole hog," so to speak, and use real printed characters-un-
less we were to decide to adopt phonoscript methods of learning
to read, in which case we should use phonoscript, not script nor
ordinary print.
There is, however, nothing more at present than mere opinion
to guide us in these matters, and perhaps we had better wait until
88
TEACHING BEGINNERS TO READ
some definite experimental work is done; but there does seem
some good a priori basis for supposing that more progress might
be made in reading in its early stages if the writing or printing
taught at that age corresponded, letter for letter, with the type
used for the printed characters of the reading primer. Whether
that correspondence would involve a subsequent loss in rapidity
and fluency of writing is an undetermined question; there are
good a priori grounds for supposing that it would. In any case,
we decided to use real print for our phonic group and phono-
script for our other group and to do no writing of any kind
during the experiment. But, as I have said, the words to be
learned and the reading matter for both groups, though the type
was necessarily different, were to be the same. And as we had
no phonic readers which covered the same ground as the phono-
script, it was necessary to make out a course of lessons to be
taught phonically which should cover the same ground. To be
certain of effecting this, a complete analysis was made of the
words in the phonoscript primer, lessons 1-20, in accordance
with their vowel sounds. There were:
(1) One hundred and seventeen words with what are called the short vowel
sounds, such as: a in have; e in ten; i in pit; o in on; and u in but.
(2) One hundred and four words with what are called the long vowel
sounds, such as: a in pay; e in he; i in fine; o in so; and u in use.
(3) Twenty-five words with oo and ou, as in food and could respectively.
(4) Twenty-seven words with the au-sound, as in taught and paw.
(5) Eleven words with the broad a sound, as in father, calm, and after.
(6) Six words with a as in was and what.
(7) Fourteen words with ir and er, as in girl and fern.
(8) Ten words with a as in any.
(9) Eight words with o as in son and come.
(10) Six words with oi and oy, as in choice and boy.
(11) Seven words with ou, as in mouse.
(12) Two words with o as in one and once.
Thus a total of 337 words was taught to the children of
both groups. Only three proper nouns occur: Mary, Pat, and
Pip. Neither Pat nor Pip are well-known names in this school;
but we used them, perforce, because they occurred in the phono-
script text. No capital letters were used for these words, nor
for the words with which sentences began.
The actual teaching of the phonoscript group.—Any teacher
who wishes to understand this thoroughly should get the phono-
PHONIC VERSUS PHONOSCRIPT METHODS
89
script primers and, if possible, attend Mr. Hayes' lectures and
demonstrations. But I append a lesson or two as illustrative of
the system. In the first lesson (Step I), the following characters
in phonoscript were written on the blackboard, one at a time.
i.a. h. v
As each letter was written, the teacher sounded it and said to
the children, "You all tell me what this letter says." This was
done three times collectively. Individual children were then
called upon to sound the letter, thus: "Harry, you tell me what
this letter says," etc. The following explanations were given as
the letters were written:
i. When you see this little mark (pointing to the upper left
side of the letter) touching this letter it says i (sounding it as in
"find"). It thinks a lot of itself and says "I."
a. When this round letter has a straight line to its side,
touching it, it says a (sounding it as in fat).
h. When I say this letter, I just breathe out the sound like
this (making the sound of the aspirate).
v. When I say this letter, you watch and see what I do. I
put my top teeth to my bottom lip and then gently drop my bot-
tom lip a little way (making an emphasized articulation).
In Step II of the first lesson the letters given below were
written on the blackboard, one at a time. For the letters i, a, h,
and v the same procedure was adopted as in Step I. But there
is now a new factor, the silencing mark has been added to some
of the letters.
i.i,a,a,h,h,v, ve,i,a, h,
a.h.
v.e, ha, h,v.e.a.h.a.e
The teacher said, before beginning to sound these letters,
"When you see a letter with a little mark on its forehead, like
this, (pointing to the silencing mark on i) or like this h, or like
this v, or this a, you will know that it is a lazy letter; it does
not make any sound." The silent letter e called for special
comment, since the mark which converts it into a silent letter is
not a little horizontal line joined to the rest of the letter at the
90
TEACHING BEGINNERS TO READ
top and left, as it is in all other letters. It is placed on the left
side of the letter, but takes an upward curve from the middle
of it. Somewhat curiously, but possibly with intent, the letter
e, carrying a sound, had not yet been dealt with. The teacher
could only say, therefore, "when we see a letter like this (point-
ing to the letter e) with this little mark (pointing to the silencing
mark), we must put our fingers on our lips, and make no sound,
because it is another lazy letter."
Step III of the lesson was a revision of all the letters which
had been hitherto dealt with one at a time during Step II. The
teacher said, “I am going to point to all these letters (those shown
in Step II), one at a time; you are not to speak, but when I point
to a lazy letter, a letter that does not make any sound, you are
to put your fingers on your lips."
When this had been done collectively, individual children were
called upon to come to the blackboard and point out all the
silent letters. The teacher proceeded, "Now I am going to
print all these letters again, one at a time. I want you all to tell
me what each letter says as I point to it; but when I point to a
lazy letter, put your fingers on your lips, as you have been doing,
and do not make any sound." Individual boys and girls then
performed the exercise which the class had done collectively.
Step IV dealt with letters in the following actual words:
hie vie hive have
high have i? ihave
The procedure was identical with that for Step II. The chil-
dren were not asked to combine the letters into words. (Mr.
Hayes' system is a synthetic system throughout; it does not, like
the Dale and other phonic systems, start by analysing the known
sound of a word into its components before re-synthesis, though
I cannot see that its methods would be spoiled by such analysis).
The words were used only as a basis for repeating the sounds
of the letters and for dealing with silent letters. They were
written to accustom the class to seeing letters put into groups,
serving as a preparation for reading words later. In the word
high a new letter was introduced, the silent g, which had not been
PHONIC VERSUS PHONOSCRIPT METHODS
91
dealt with before; but as it was silent it was only necessary to
question the class thus: "What can you tell me about this letter?”
The answer was given: "It's a lazy letter; it's got a little mark
on its forehead."
In Lesson 2, Step I, the teacher wrote the following letters
on the blackboard, one at a time:
o.e.p.r
As each letter was written the teacher sounded it, then the chil-
dren pronounced it collectively three times, and then it was pro-
nounced by individuals. We have now, be it noted, a third
form of the letter a. We have already dealt with what is usually
called the short a as in have, and also with the silent a.
The teacher said, "When we see this letter with a little line
pointing upwards and touching the letter here (pointing to the
mark) we sound it like this," (giving it the sound of a in amidst).
The whole process was repeated for this letter. During Lesson 1
the letter e had been written on the blackboard, but as it then
had a silencing mark attached, no reference was made to it other
than as a silent letter. The unsilenced form now given of this
letter e is like the e of ordinary print. To this form of the letter
the phonoscript system gives the sound of e in the word pet.
The teacher said, "When this letter has no little mark in front
of it, when it is written like this (pointing to the letter), it is
not lazy, it says e (sounding it as e in pet)." The only directions
in connection with p and t were the example and precept to make
the sounds as short as possible and not drag them out, as children
are prone to do, till they almost become per and ter.
Step II dealt with actual words, which were combinations of
the letters previously dealt with. The words follow:
hap, par, per, var, rap.
pie, tie, Fight, height,
ahive, aper. arap.
The following was the method adopted for combining the
letters into words. The teacher wrote the word hap on the black-
board. (I should not personally use the word hap for reading
92
TEACHINg beginnERS TO READ
purposes at this stage of a child's mental development, but I am
following the system). As the letters of this word had been
taught previously, she did not sound them, but pointing to each
letter, one at a time, she said, "All tell me what this letter says."
She pointed first to the letter h. Having obtained the appropri-
ate sound, she next pointed to a, and obtained what is called the
short a-sound. Then, before going further, the sound of h was
repeated, followed immediately by the sound of a (short). When
these sounds had been satisfactorily made, she pointed to and
required from the class the sound of the third letter p, and finally
dealt similarly with each of the three letters in succession. The
sounding of the letter, therefore, took this form: h; a; h, a; p;
h, a, p.
I have given very fully this method of sounding letters in
order to build up words because it is unlike the synthesis adopted
in other "phonic" systems, and also because it is regarded as
essential to the phonoscript method. My opinion-it is only
opinion-is that it is a very valuable feature of the system; it
is really a spelling method, though not, of course, an alphabetic
spelling method, where names, not sounds, are used. It does
also, it appears to me, help towards the synthetic combination of
the sounds of the letters into the sound of the word. But fail-
ure of combination by a purely synthetic system which has not
been preceded by an analysis of the words into their component
sounds is almost certain at first. Though the procedure indi-
cated above was carried out three times, the children could not
combine the sounds into one word. Of course the word hap was
not known as a word and had no meaning to the children, and
should not have been used for reading purposes at this stage.
The teacher again sounded the letters of the word in phonoscript
order with exaggerated emphasis and pronounced the word her-
self, after which individual children were called upon, and were
successful.
The other words of this exercise, pat, pet, vat, and tap were
taught in the same way. The word vat was not known to the
children as a word, nor was its meaning known. The object
PHONIC VERSUS PHONOSCRIPT METHODS
93
could not be procured, so the teacher drew one on the blackboard,
but she did not think that the drawing helped matters much.
The next line of phonoscript printing was then dealt with.
Pointing to p in the word pie the teacher said, “All tell me what
this letter says." She treated i similarly, repeating what she had
said before, that this letter thinks a lot of itself. The sound of p
was readily forthcoming, and i gave little difficulty. Then the
silencing mark was referred to in the e of pie. Now came the
phonoscript combination: p; i; p, i; e (silent); p, i, e. Only
one child could pronounce the word except in imitation of the
teacher. The word tie was then dealt with in a similar way, and
this time five children could give the word after the separate
sounding and phonoscript succession of the letters.
The word tight was next taught. This does not appear to
be an easy word for children of this age in their second reading
lesson, but it is one of the excellences of the phonoscript system
that a word like this presents little difficulty. The class was
asked collectively to sound t and i, and then t, i in succession.
Then came the turn of the silent g. When this was pointed to,
seven of the children remembered about the silencing mark, and
put their fingers on their lips. The others of course did nothing,
for the sound of g had not yet occurred in the lessons. (It
might be objected theoretically that a silencing mark is not very
appropriately attached to a letter, which, so far, has no sound;
but I did not think the matter one of much practical importance.)
The teacher, for those who had forgotten, pointed out the
silencing mark on the g, and all the children put their fingers to
their lips. She then pointed to the letter h, which also had the
silencing mark attached. Many more children this time made
the appropriate gesture, but not all, and the teacher again pointed
to the mark. Then the final letter was dealt with; the remem-
brance of its sound gave little difficulty. Next came the phono-
script spelling, done by the class collectively, with the teacher
pointing. But no child could pronounce the word until the
teacher had done so. It will be noticed, of course, that there
are only three operative letters, namely t, i, t, and after the
94
TEACHING BEGINNERS TO READ
teacher had pronounced the word it was not difficult for the chil-
dren to pronounce it either collectively or individually.
No fresh point of interest arises until the combination of
words into the sentences "I have a pet." "I have a tie." "I have
a pie." But at this stage the children were not sufficiently con-
versant with the sounds of the letters and their combination into
words to be able to put the words together into sentences.
Appended to the account by the teacher, from which the
above notes are made, is a record of every lesson given to both
groups during the whole experiment, the work done each day,
and the date of each lesson. All these records must be omitted
since there were seventy lessons in all for each group. As I
have said, all the sounds and words dealt with were the same
for each group, being those given in Lessons 1-20 of the phono-
script primer.
The analysis of the sounds for the phonic group was not car-
ried out as closely and minutely as required by the phonoscript
system. For example, in phonoscript the o sounds in the word
good are distinguished from those in the word food and separate
characters are used. Again, the a in Mary is distinguished from
the a in many by means of a distinction made in the written
character of a. These distinctions are not made in other phonic
methods in current use, and were, of course, not made in the
teaching of Group A.
NOTES ON THE IMPRESSIONS OF THE TEACHER BEFORE THE
FINAL TESTING
The teacher was of the opinion that at the end of the experi-
ment, and even before, the time had come when reading from
the blackboard should be discontinued and books substituted.
There were some children in each group, she thought, who had
made scarcely any progress. (I was prepared for this by the
results of the preliminary lessons on sounds and shapes.) These
must, of necessity, start afresh and begin again from the black-
board. The reader is reminded that the whole of Grade II, of
PHONIC VERSUS PHONOSCRIPT METHODS
95
an average age at the beginning of 5 years 1 month, was set to
do this work. The great majority were so keen and success-
ful-even with a method, which, like the phonoscript, seems to
start with difficulties right away-that the teacher intended with
those who had learned something of phonoscript to continue that
method (if the final testing showed even moderately good re-
sults) throughout the rest of their infant-school career.
She noted that none of the phonoscript children, throughout
the whole series of lessons, even attempted to pronounce a word
without sounding the letters in the approved way. This was not
true of the group taught phonically. One or more children tried,
often successfully, to say some of the words without sounding
the letters, especially when a favorite word like pussy or little
was written on the blackboard. I am not sure whether this
proves ultimately advantageous or not; I mention it as an in-
teresting observation. It may be due to the greater definiteness
of the letter-sounds which phonoscript gives, and this may or
may not prove of final advantage. The point must be empha-
sized that at no time were any of the children in either group
either encouraged or discouraged in pronouncing the word with-
out sounding the letters. Of course in the early stages they could
not do so.
When the children of Group B had become accustomed to
the silencing marks, which did not take long, the practice of
putting fingers on lips was discontinued. The teacher merely
asked when dealing with a new word, "How many lazy letters
are there?" or "Where are the lazy letters?" or "What can you
tell me about these letters?" (pointing to the silent ones). And,
as the lessons proceeded, the teacher found pointing to the letters
unnecessary, and only occasionally pointed to the words. At
the end of each lesson the teacher read to the children the sub-
ject matter of it, all of which had been printed on the black-
board. Her general impression was that the phonoscript method
would prove victorious.
96
TEACHING BEGINNERS TO READ
THE FINAL EXAMINATION and the METHOD OF MARKING
This was carried out by the head teacher, assisted by the
regular teacher of the class. The tests were selected by me.
There were two tests for each group, subdivided into parts
a, b, and c, given on different occasions so that freshness might
be evenly distributed and fatigue might not supervene during
the tests. The tests a, b, and c put together made a complete
conversational narrative. They were printed on pieces of car-
tridge paper 7½ by 5 inches and multiplied by the hectograph
until we had six copies of each test (three for Test I and three
for Test II) for every child in each group. On the top of each
paper was written the child's name and his age in years and
months at the date of testing. A space was left for the time, in
minutes and seconds, required to work the exercise. At the
bottom of the paper was a space for the date of the tests and
one for the total marks gained. These papers were used by
the teacher to make an exact record of what each child did with
each word. A piece of cartridge paper with the reading matter
printed upon it was placed before the child; but, to avoid con-
fusion, no marks except the words in the tests were placed on
this paper. The head teacher who taught the children heard
them read, of course without correcting them, while the records
were being made, unknown to the children, by the usual teacher
of the class. Each child was tested individually in his own class
room, the room in which he had been taught. For the sake of
fairness to both groups, children from each group were tested
alternately. The examination began on May 30th, 1923, thir-
teen days after the teaching lessons had been concluded. This
delay was partly necessitated by the preparation of papers, and,
apart from that, was intentional.
During so long an experiment, especially with such young
children, we had lost a good many through removal or protracted
absence (a rather large proportion of these were inferior read-
ers), and only 19 children in each group, 38 in all, came suc-
cessfully through all the lessons and tests. Occasional daily ab-
sences during the lessons were disregarded. Each group at the
PHONIC VERSUS PHONOSCRIPT METHODS
97
time of examination averaged slightly over 5½ years of age.
After careful deliberation, it was decided to use the words of
Lesson 21 of the Introductory Phonoscript Primer as the test
for both groups. It was arranged that the phonoscript group
should be required to read this exactly as it was written; and
that the phonic group, whose lessons had been arranged to in-
clude all the sounds used in this lesson, should also read the same
words, though they were arranged rather differently. This
phonoscript lesson contained all the letters and sounds, and no
others, which both groups had been taught. It also contained
many words (to be exact, 53) which both groups had been
taught, and 19 words which neither group had been taught. The
untaught words contained no letters or sounds which the chil-
dren in both groups had not dealt with. The tests follow:
TEST I FOR GROUP A, THE PHONIC GROUP
(This was printed in real print, the print of the ordinary
reading book or newspaper.)
(a) may my chum buy one of your rabbits,
please mary? come here, be a kind girl. which
beauty can you spare him? you choose for him.
(28 words)
(b) he wants the fine black one. he thinks it is a
perfect gem. how much are you asking him to give
you? half a crown. (26 words)
(c) that's so cheap. thank you, mary. jack is a
nice boy. he says he has a hutch all ready. good-
bye. (18 words)
Test II for Group A was precisely the same as Test I, except
that it was written in phonoscript print. The object of Test II
was to discover how far the children who had learned by a phonic
system from ordinary print could successfully attack the same
words and sentences when presented to them in a form of print
which they had not learned, and, more especially, whether they
would be so bothered by the modifications of the shapes of the
letters which phonoscript requires that the greater part of their
knowledge would be rendered inoperative. I may say at once
that the special signs and shapes of the phonoscript characters
98
TEACHING BEGINNERS TO READ
scarcely worried them at all; they were almost as successful as
with the print they had learned. Obviously the phonoscript
system can be adopted without much loss of previous knowledge
derived from an ordinary phonic system, in which the letter
shapes are not modified.
TEST I FOR GROUP B, THE PHONOSCRIPT GROUP
(This was printed in phonoscript, exactly as given in the
Primer).
(a) mary, please come here. my chum jack
wants to buy one of your rabbits. can you spare
him one? he says you may give him which you
choose. (28 words)
(b). he thinks the black beauty is a perfect gem.
he is a nice boy, so be a kind girl. he has a fine
hutch all ready. (26 words)
(c) how much are you asking him for it? half-a-
crown! that's cheap. thank you, mary, good-bye!
(18 words)
Test II was exactly the same as Test I except that it was
printed, not in phonoscript characters, but in ordinary print.
The object of the test was to discover how far the children who
had learned to read from phonoscript characters would be able
to use the knowledge they had gained when using ordinary print.
I may say at once-the necessary statistical calculations will be
given later that they were seriously at a loss. It does not
necessarily follow that at a much later stage-these tests were
given after only six months' teaching-the loss would be equally
great; but it does seem that we must be prepared, if we adopt
phonoscript, for considerable loss when the child is required, as
of course he will be later on, to read from ordinary print. The
tests given above are by no means easy for children who have
learned to read by any method for only six months (4 lessons a
school week, 20 minutes a lesson), but they were not too hard
to enable us to get a clear answer to the questions proposed in
this research.
An analysis of the tests follows. There are in each of them
72 words, and the same words, including a, which when standing
PHONIC VERSUS PHONOSCRIPT METHODS
99
alone has a distinctive sound. Fifty-six of the words are dif-
ferent, only seven being repeated. Fifty-three of the words have
been taught to both groups, and 19 have not been taught to either
group. The untaught words are hutch, for, Jack, chum, buy,
bye, spare, perfect, gem, nice, kind, ready, may, please, asking,
thank, half, crown, and wish.
The phonoscript children had learned to distinguish nine dif-
ferent sounds of o, requiring many modifications of the ordinary
written character. (I may say, in parenthesis, that phonoscript
deals with no diphthongs or double sounds; it modifies the first
letter and silences the second or vice versa.) They had dealt
with nine different sounds of a, three different sounds of i, three
different sounds of y, and five different sounds of e. There
were three different sounds of t, two of g, two of s, two of r,
three of c, two of n, three of u, two of w, and two of f. The
remaining letters each had but one sound.
The phonic children did not carry their analyses so far; they
did not, for example, distinguish between the f-sound in for
and the f-sound in of. Phonoscript requires of to be pro-
nounced ov, if I may, inadequately, so render it. It cannot be
denied that in colloquial speech we do say ov rather than of; but
whether we should teach children in school the degraded forms
of colloquial expressions rather than the more pedantic and
fuller sounds of the letters, is a question much bigger than one
of reading method. I will, in parenthesis, put the matter by
simple illustration and statement; I shall not argue it. Shall a
child, for example, be taught to say Nelson, as he usually is,
or shall he be taught to call it Nelsen? Shall the schools take
part in the process of phonetic decay to which all civilised tongues
are prone, or shall they try to arrest it? To arrest it entirely
is a vain hope. But shall they try? The neutral vowel sound,
a sort of e in colloquial speech, is swallowing up the other fuller
and more distinctive vowel sounds. This whole matter is a
question of taste and of literary expression rather than of school
method. I leave it there, simply saying that in our examinations
any recognised pronunciation, pedantic or colloquial, was accepted
100
TEACHING BEGINNERS TO READ
as correct. But no marks were given if the child merely gave
correctly the sounds of the letters, though they were all recorded;
marks were given only if the word was correctly pronounced in an
intelligible fashion. Still, as I have said, a record was made of
precisely what was done by each child with every word and,
indeed, with every letter. If a correct sound were given for a
letter, a tick (V) was placed over the letter; if an incorrect
sound were given the symbol representing the sound was written
over the letter. Above the ticks, or letter-symbols, was written
the pronunciation (if any) which the child gave to the word as
a whole.
SOME COMMENTS ON THE FINAL TESTS MADE FROM THE
OBSERVATIONS OF THE HEAD TEACHER
The shortest time taken to read a subdivision of any test was
2½ minutes: one girl in the phonic group took this time to read
Test Ia, and one boy in the phonoscript group took exactly the
same time to read his Test Ia. In neither case was there any
stopping or hesitation. It is therefore probable that this is about
the shortest time in which this test could be read by children of
this mental level. Neither of these children attempted to read
any of the words without first sounding the letters. The boy
did well indeed to get through the test in so short a time; for,
in accordance with the phonoscript method, he had, after sound-
ing each letter, to go through the sounds again from the begin-
ning before going on to the next letter of the word. The girl in
the phonic group had to sound each letter only once before at-
tempting to pronounce the word. The phonic children generally
were very much superior in speed, but I attach very slight im-
portance, if any, to this fact. The important issue in beginning
to learn to read is not rapidity but certainty.
One interesting point was noted in connection with pointing
to the words or letters in the final examination. It can be
positively asserted that not a single child in either group ever
pointed to any word or letter of the test (they had been taught
from the blackboard, be it remembered); nor did they put their
PHONIC VERSUS PHONOSCRIPT METHODS
101
hands on the table at which they stood to read the tests. Speak-
ing generally, their hands were held loosely by their sides.
A complete analysis of the correct and incorrect responses
was, as I have said, provided by the detailed results for each
child. From these was obtained another analysis showing what
was done by both groups in Test Ia and Test IIa with each
word taught and untaught, and the kind of errors made by each
group. To quote these would require many pages of print, and
I must for the present confine myself to a few of the comments
made by the mistress who performed the analysis. There was
only one signal failure to remember the sound of a letter, namely,
that of what is called the soft c in the word nice. In the phono-
script tests it occurred in Tests Ib and IIb; in the phonic tests
it occurred in Test Ic and IIc. Of the phonoscript group only
one gave the correct sound; 17 gave it the hard sound, and one
rendered it as hard g. They had either failed to notice that the
dot at the commencement of the c had been extended upwards,
or had forgotten the sound of the c thus modified. Every child
in the phonic group sounded the c as k. The word itself was an
untaught word to both groups.
Every child in the phonoscript group sounded the letters ac-
cording to the system before attempting to pronounce the word.
And every child, except one, in the phonic group did this also in
the phonic way, though several of them during the lessons had
pronounced some words without a preliminary sounding of the
letters.
The untaught words in Test I were generally much harder for
the phonic group than for Group B. For example, the word
please was not difficult to those taught by the phonoscript method.
The difficulty of this word, in any kind of phonic method, phono-
script or not, lies in the vowels. The phonoscript marking
makes the first e into a long e and silences the a and the
final e.
Eleven children of this group pronounced the word
correctly. Although the phonic group had dealt with ea as a
composite sound in their lessons, they did not in a single case
achieve success. The word gem, another untaught word, also
102
TEACHING BEGINNERS TO READ
gave much difficulty to the phonic group. Group B had only to re-
member that the g had the characteristic phonoscript marking for
soft g and they had no further trouble: twelve of them were
correct. The phonic children had no such help in the shape of
the letter. Only one child was correct; all the rest gave the hard
sound of g.
One very young child in the phonic group, Lily W—., aged
4 years 11 months, who scored a good mark for Test I, failed to
obtain the word ready from the sound of its letters, as well she
might. She then began again, starting from the end, with the
y first. She ended up with a hearty laugh, but of course with
nothing more.
When the phonic group read their Test II, which was written
in phonoscript, they encountered some real difficulties, though
they were very successful. The phonoscript i, however, was
often sounded as j and the g was sounded as p. One boy called
it "figure 9." The phonoscript letter a in its various forms was
unknown, though often guessed, and the ordinary y took three
other shapes in phonoscript, but not very different ones.
The phonoscript group also encountered serious difficulties
in Test II, in which they were required to work from ordinary
print; and they were not, as a rule, successful. Of course, they
suffered from their ignorance of the printed letters a and
g. Doubtless they were more inconvenienced by a change of
type than the phonic group, for they had been accustomed to
give very careful attention to the exact shape of their letters.
This is necessitated by their system. They gave to the shapes,
roughly speaking, the sounds the shapes would have had in
phonoscript. In the word much, for example, the c would have
had a special marking in phonoscript, giving it a soft ch sound,
and the h would have been silenced. In the absence of these
markings, the children quite rightly gave c the hard sound and
aspirated the h, pronouncing the word "muck-h." Again, in the
word buy, the u would have had a silencing mark in phonoscript
and the tail of the y would have been turned up to show that
it had the sound, in this case, of long i. In the absence of these
PHONIC VERSUS PHONOSCRIPT METHODS
103
marks, the children sounded the b correctly, gave to u a short
sound, gave to y the sound as in you, and then pronounced the
word as "bu-y" quite correctly on the basis of their method.
Of course they scored most of the marks they did score where
the shapes of the letters, written in their customary characters,
required no special markings.
Both groups took a long time to read Test II, though they
had read it before in their own print. Their attention wandered
and they looked over the paper rather generally. They did not
like pronouncing combinations of sounds which, to their knowl-
edge, did not make words; for they were aware that they were
trying to read.
STATISTICAL RESULTS
The summarized individual results are given in Table XI.
The difference between the means of the results for the
phonic and phonoscript groups, each reading its own print, for
the words actually taught, is 0.3 in favor of the phonoscript.
With a correlation "r" coefficient of +.536 and a P.E. of 2.0,
this difference is obviously not significant. With the untaught
words, each group reading its own print, the case is very dif-
ferent. The difference in the means is 5.2 in favor of the phono-
script group: the "r" coefficient between the two series of re-
sults is .385, and the P.E. of the mean difference is 0.87.
That is to say, the difference is 6 times its P.E.
When the phonic group are required to read their test in
phonoscript print, and the phonoscript group are required to
read in ordinary print, the result is reversed. In the words that
had been taught, the phonic group have a balance in their favor
of 18.3. The "r" coefficient is + .462, and the P.E. of the mean
difference is 1.34. That is, the mean difference is 13 times its
probable error. With the untaught words the difference between
the means is 3.6 in favor of the phonic group. The "r" co-
efficient between the two series of individual results is +.567
and the P.E. of the mean difference is 0.41. Thus the differ-
ence is 9 times its own probable error. I may mention that the
"r" coefficient between the results from the preliminary series
104
TEACHING BEGINNERS TO READ
of tests in shapes and sounds is +.99 with no difference be-
tween the means. The speed of reading is throughout in favor
of the phonic group. The results for Test I are r+.294,
mean difference = 2.9, P.E. of difference = 0.938. For Test II,
r.154, mean difference 2.0, and P.E. 1.07.
=
TABLE XI. INDIVIDUAL RESULTS IN THE PRELIMINARY AND
FINAL TESTS
(Group A taught by an ordinary phonic method)

NAME
(Initials only)
AGE ON
MAY 31,
1923
MARKS
FOR
PRELIM-
INARY
SHAPES
AND
MARKS FOR TEST I
IN PRINT
TIME
TAKEN
FOR
READING
Taught Untaught TEST I
Words Words
MARKS FOR TEST II
IN PHONOSCRIPT
TIME
TAKEN
FOR
READING
TO
(Yrs. and Sounds
(Max. 53) (Max. 19)| MINUTE |(Max. 53)|(Max. 19)| MINUTE
NEAREST
Taught Untaught TEST II
Words Words
NEAREST
то
Mos.)
W. S.
5-11
38
48
C. H..
6- 0 34
40
∞∞
8
16
46
9
16
8
16
31
13
H. E.
5-11 34
42
11
14
41
9
14
S. R.
•
W. C...
D. R.
D. A.
·
5- 1 32
5-6
7 30
6- 5 29
N. A..
5- 8 26
T. G..
5-11 26
S. L...
5- 3
R. J...
1 24
·
C. A.
6- 0
H2 22227 22
34
45
38
44
44
23
40
24
38
78
873O CO
16
39
15
35
76
13
18
13
14
19
16
13
34RDR
39
45
26
39
36
47
12
35
J. A.
5-10
19
29
D. R.
6- 0 17
45
•
Mc. P...
5-10
15
33
24∞∞∞
9
47
13
20
28
3
14
29
13
39
10
18
25
W. H……….
W. R.
W. L.
5- 9 15
6
5- 7 13
34
• •
P. W...
4-11
4-11 12
12
43
38
ONCOLO
19
8
7
11
36
6
14
40
5
14
36
OOHON BITOO OO LO LO
6
14
18
18
6
15
7
18
11
1
18
4
•
16
18
6
20
17
12
18
15
Totals... 107— 2 456
712
119
284
665
115
302
Averages 5-8 24.0 37.5
6.3 14.9 35.0
6.1 15.9
The correlation "r" coefficients between the marks obtained and
the time taken is in every case negative. Thus the total marks
obtained by the phonic group, reading print, correlate with the
time taken to read it-.63. The marks for group B, reading
their own phonoscript, correlate with the time taken to the ex-
tent of ―.76. The marks of the phonic group, reading phono-
script, correlate with the time taken.33, and those for the
PHONIC VERSUS PHONOSCRIPT METHODS
105
phonoscript group, reading print, correlate with the time taken
.13. That is to say, the best readers read faster, considerably
faster, when they are attacking a task for which they are fully
competent, and to which they have become to some extent
habituated. When transferred to a task where their acquired
knowledge and method to some extent fail them, the negative cor-
relation falls; in the case of children who, after studying phono-
TABLE XI (continued)
(Group B taught by the phonoscript method)


NAME
(Initials only)
AGE ON
MAY 31,
1923
(Yrs. and
Mos.)
MARKS
FOR
PRELIM-
INARY
SHAPES
AND
MARKS FOR TEST I
IN PHONOSCRIPT
TIME
ΤΑΚΕΝ
FOR
READING
MARKS FOR TEST II TIME
IN PRINT
TAKEN
SOUNDS
TO
NEAREST
Taught Untaught TEST I
Words Words
(Max. 53) (Max. 19)] MINUTE (Max. 53)|(Max. 19)| MINUTE
FOR
READING
Taught Untaught TEST II
Words Words
ΤΟ
NEAREST
•
F. G.
W. J.
H. H..
S. D.
5- 9
36
52
19
со
8
21
•
6- 3 35
53
18
12
20
5-8
34
33
6
21
14
·
•
5-10
33
47
17
14
22
C. E.
5-
33
47
17
15
23
3503 LO
15
20
23
15
15
•
E. G.
J. L.
H. I.
G. D.
C. V...
D. E..
G. A.
N. J
B. M...
V. A..
J. D...
N. D
•
•
6—
5-
5-7
5- 4
5—
83749
31
50
14
15
8
0
29
49
15
20
19
3
11
28
36
11
20
20
2
25
44
13
16
23
24
50
16
8
11
1
5-10 24
49
17
11
35
10
11
•
5-8 23
42
9
21
16
1
29
D
5-5
18
44
14
19
22
7 1 18
47
15
17
21
5- 0 16
37
13
19
16
LO H
5
5- 6
14
3
0
35
1
0
5-10
14
20
3
27
12
0
21
•
•
N. F..
5-7
12
6
1
23
9
K. E...
4- 8
9
10
18
4
#INO2 H2=22 229S
14
17
19
12
17
13
20
12
40
17
Totals... 108- 5 456
719
218
339
317
48
341
Averages
5-8 24.0 37.8
11.5
17.8 16.7
2.5
17.9
script, were tested on real print, it falls very low indeed. This
merely means that the better children, faced with a very difficult
task, take relatively, though not absolutely, more time to achieve
what success they accomplish than the weaker ones. There is noth-
ing new in these statements of the relations between speed and ac-
106
TEACHING beginnERS TO READ
curacy; every first-rate experienced teacher knows them already;
but it may not be amiss to give them what support they may
acquire from a long experiment in actual school method, in
which the teaching and testing have been conducted with quite
exceptional care, ability, and thoroughness.
SUMMARIZED CONCLUSIONS FROM THE EXPERIMENT ON PHONIC
VERSUS PHONOSCRIPT READING
1. By either of these methods, most school children of the
average age of about 5½ years can begin to learn to read, and
make very adequate progress.
2. The results of the teaching, each group reading its own
print, as shown by the pronunciation of words already dealt with
during the lessons, appear to be almost the same for both
methods. The phonoscript children appear slightly superior, but
their advantage is not statistically valid.
3. The results of the teaching, each group reading its own
print, so far as shown by the pronunciation of words which at
the examinations, were new to both groups, are overwhelmingly
in favor of the phonoscript method.
4. The children taught by means of ordinary print suffer
very slight loss, if any, when set to read phonoscript, thus justify-
ing the claim of its author that it is intelligible to everybody.
5. The children taught by means of phonoscript suffer very
real and serious loss when they are required to read the same
matter in ordinary print. This issue is so important that I pur-
pose pursuing it farther.
6. No evidence of any certainty has been obtained in the
course of this experiment as to the relative effects of these two
methods on the learning of correct spelling. It is probable that
for spelling a method which possesses the clearest reversibility
is the best; that is to say, a method in which the sounds are most
unequivocally attached to the same definite shapes, and the shapes
most unequivocally attached to the same definite sounds, either
of letters or of words.
CHAPTER VIII
THE INFLUENCE ON SPELLING OF THE PHONO-
SCRIPT VERSUS AN ORDINARY PHONIC
METHOD OF TEACHING READING
IN ITS EARLY STAGES
Many infants' school teachers who teach by a phonic method
in the early stages of learning to read hesitate to adopt the phono-
script method, because they assert that it makes the acquisition
of correct English spelling more difficult than it already is. Let
me try to make the present educational position clear. No one
now believes that we should delay the writing of compositions
until a child can spell all, or most of the words which he wishes
to use. Children of seven, and many of six can write about their
experiences, and write fluently, even though they cannot spell
correctly the words they use. But correct spelling must be ac-
quired, in all ordinary cases, before a child leaves school, and it
is a perfectly legitimate criticism to make against any system
of teaching reading that it hinders, or insufficiently helps, the
acquisition of correct English spelling.
Head mistresses of infants' schools, as I have said, assert
that phonoscript does not help, but rather hinders. They forget,
sometimes, that they are already teaching by a phonic method of
some kind. Moreover, the head teachers of boys' and girls'
schools and of junior and senior mixed schools unanimously
condemn all phonic systems (including phonoscript) as far as
spelling is concerned, and pin their faith to the old alphabetic
system, which, though with some modified recrudescence in what
is now called "individual work," has practically disappeared in
infants' schools. We have seen reason to believe that some form
of a phonic system is more successful in the early stages so far
as learning to read is concerned; but no evidence has been pre-
sented with respect to the relative accuracy of spelling produced
by different methods. To set forth some such evidence concern-
107
108
TEACHING BEGINNERS TO READ
ing phonoscript and an ordinary phonic method is the object of
the present section.
THE CHILDREN WHO DID THE WORK
The work was done with the Grade III class of young chil-
dren, who, a year before, when in Grade II, had been divided
into "two equal, parallel, and highly correlated groups" on the
basis of their capacity to learn associated sounds and shapes.
One group was taught by an ordinary phonic method, the other
by means of phonoscript. After six months' teaching, the relative
efficiency of the phonic and phonoscript systems has been tested;
and the methods have proved about equally efficient so far as the
knowledge of words already taught was concerned. The phono-
script method, however, was much the more successful in giving
the power to attack new and untaught words. At the end of
six months' teaching, the two competing systems ceased to be
so arranged as to cover the same ground, and no further reading
tests were given for comparative purposes. But, six months
later, the phonoscript group was again tested for a comparison
of loss on transfer with that previously found.
During all this period there was no teaching of spelling, ex-
cept in so far as it was involved in the methods of learning to
read, and no tests were given in spelling either in writing or
orally. In fact, during the first six months of the year no writing
of any kind had been taught, in order that the comparative results
of the two methods of teaching reading might not be unfairly
influenced, as they might well have been. For the writing ordin-
arily taught in this infants' school in Grades II and III (not in
the Standard I class, the highest class of this infants' school, in
which cursive script or ordinary English writing was taught) is
neither real print nor phonoscript, but a cross between print and
cursive writing, in which the letters are not joined and which
rather closely, except in the slope of the letters, resembles italics.
But after July, 1923, on the completion of the comparative read-
ing tests, this print-script had been taught. The children who
were learning to read print were now being taught to write script
THE INFLUENCE OF SPELLING
109
and those who were being taught phonoscript were also being
taught to write script (not phonoscript). No lessons in spelling
were being given and no dictation tests in spelling; though some
little compositions had been set, in which the spelling was not
corrected.
Of our original groups since the second test in transference,
one more child, in Group A, Jessie R-, had left the school;
therefore the girl, Edith D—, who had been paired with her
was removed from the groups paired on the original lessons and
tests in associated shapes and sounds. It was possible, however,
to put Edith D-in the additional grouping formed on the
head teacher's estimate, since one child who had been working
in the phonic section and who had been too good to pair amongst
the five additional pairs hitherto made, could be paired with her.
We now had twenty children in each group-fourteen in each of
our original groups, and six additional cases on each side. At
the time of working the spelling tests which we were now going
to apply, late in November, 1923, the average age of our groups
paired on tests in associated shapes and sounds was for Group A
(the phonic group) 6 years 3 months and for Group B (the
phonoscript group) 6 years 3 months, and of the additional
children, paired on the head teacher's estimate (who, be it re-
membered, had also worked through the whole course, except
the preliminary lessons and tests in shapes and sounds) for
Group A was 6 years 1 month and for Group B was 6 years 1
month.
But what sort of spelling test should we give? The phono-
script group and the phonic group had now diverged as to reading
matter, and it would not be fair to take the test either from a
phonoscript reader or from a phonic reader. Nor was I particu-
larly anxious to take the tests from a reading book of any kind.
What I really wanted to see was the relative success of the two
groups with words whose meanings the children knew and which
they were accustomed to use. As a comparative test in spelling,
a set of composition exercises would scarcely have been practic-
able; for the children who tried to use the more advanced vocab-
110
TEACHING BEGINNERS TO READ
ulary, might very well have made the poorer score, whilst really
being the better spellers. How could we combine the virtues of
composition and dictation as a test? We set about that in this
way. The whole of the class (it was, and always had been, taught
as one class in writing and composition), was set to write a
composition exercise on "My Cat." Few of the homes of these
little ones are without a cat, and they set to work gleefully and
with confidence. Then, after a careful and protracted examina-
tion, the words were picked out which had been commonly used
by the children, and the following test constructed.
THE TEST IN SPELLING AND HOW IT WAS ADMINISTERED
The spelling test consisted of 33 words, excluding the words
I and a. The words of the test follow:-
My little cat is black and white. I call him Tim.
He is a pet. You can see him on top of the hay.
When we go home, I will give him some nice hot
milk.
The word him is repeated three times, and the word is is used
twice. The word hay calls for some comment. I should cer-
tainly not have used it had it not often occurred in the com-
positions. These were not country children, and I could not
understand its frequent use until I found that it had been read
and explained in the course of the reading lessons for both groups.
I can imagine every competent teacher of young children saying,
“Yes, but the test is much too long." Of course it is, and to give
it as it stands would be to court failure on our part, fatigue and
disheartenment on the part of the children. For, up to now,
they had done no dictation of any kind. First of all then, prior
to the test, two or three very short dictation exercises were
given, not, of course, involving the same words, to accustom the
children to the process. Then we divided the test into three
parts, the first exercise to contain 10 words, the second 12, and
the third 13. At the opening of the school session on each morn-
ing on which the test was to be given, the children were provided
with a good lead pencil and a piece of paper (6½″ x 8″) ruled
THE INFLUENCE OF SPELLING
111
in the style commonly known as Slate Ruling, on which they were
accustomed to write. These papers were placed out of sight
under mill-boards till needed. The tests were given on Tuesday,
Wednesday, and Thursday, the twentieth to the twenty-second
of November, 1923, at 9:40 A.M. precisely, immediately after
Scripture lesson. The first exercise was: "My little cat is black
and white. I call him Tim." Before dictating the test, the head
inistress read it through once to the children. Then she called
it out, one word at a time, the next word not being dictated until
the previous word had been written, rightly or wrongly. The
second part of the test was: "He is a pet. You can see him on
top of the hay." This and the third section of the test, which
was “When we go home, I will give him some nice hot milk,"
were given in the same way. Before beginning the second section
of the test, the mistress said, "We wrote about a cat yesterday.
Today, we are going to write some more about a cat." On the
third day the procedure was the same. The children were not
reminded to put capital letters in the appropriate places. If they
had been so reminded, it might well have happened that some
children, through not knowing how to make the capitals, would
have made no attempt to write the word, even though they knew
how to spell it. At the close of each exercise, the children were
told that they might write their names on their papers. They had
received no definite lesson in writing their names, and, of course,
some of them could not do it. In these cases the teacher of the
class, not the head mistress, wrote their names for them.
THE METHOD OF MARKING THE TESTS
Every word, correctly spelled, except 'I' and 'a' whether
repeated or not, carried one mark. It was possible, therefore,
for a child to obtain 33 marks.
These children had been taught to make capital letters, but
they were rather uncertain about them in some cases, as indeed,
is the case with children of a much larger growth. So no marks
were given when they were written correctly, and no mark was
subtracted when they were incorrect. If a letter of a word was
112
TEACHING BEGINNERS TO READ
badly made, so long as it was intelligible, the mark was allowed;
but if the letter were made, as it were, backwards, so that we
were uncertain whether the spelling of the word were correct or
not, no mark was allowed. No marks were allowed for the I's
and for a standing alone. Table XII gives the individual results.
On careful inspection, two results appear very doubtful; those
of George T. in the first exercise and of Arthur G. in the
TABLE XII. INDIVIDUAL MARKS IN SPELLING OF GROUP A AND
GROUP B COMPARED AFTER A YEAR'S LEARNING TO READ;
GROUP A BY AN ORDINARY PHONIC METHOD,
GROUP B BY THE PHONOSCRIPT METHOD
Name
GROUP A
Age, Nov.
30, 1923
Yrs. Mo.
6
6
Stanley W.
Herbert C.. 6
Ernest H.
Robert S.
Cyril W...
Rose D..
Arthur D…….
Albert N.
George T...
Louisa S...
Arthur C..
Albert J.
6
GROUP B




Age, Nov.
30, 1923
Name
Yrs. Mo.
First Test
(Max. 10)
50∞ | Second Test
9
7365OH
23
18
George F..
Joan W.
6
6
20
Herbert H. 6
21
Doris S
6
10
4 21
Edward C.
6
(Max. 12)
Third Test
(Max. 11)
Total Marks
75677
LO CO LO ON
LO
1125 O
6 11
COLO
7
6636 ~]
11
8
8
1 5
10
6
754
4
6
6 4
Rose D..
6
6
Philip McI. 6 4
3677
9
9
5 10
4576O
39243
6
∞7377
8
4625 CO
18
19
First Test
(Max. 10)
Second Test
(Max. 12)
Third
(Max. 11)
Test
355 Total Marks
25 Ethel G..
19
Lily J...
18
Irene H.
23
21
192 2282
10
20
Daisy G.
Violet C...
75 COLO
6
6 3
Arthur G.. 6
20
Jane N.. 5 11
Marjorie B 7
Annie V...
5 6
201OM 2H7O
4 6
9
5
6
11
6
6
5
LO LO
COLO GLO
8
5 10
5
5
6
19
13
15
22
12
20
22 4022
15
19
12
12
344 L
1432
Averages. 6 3 5.6 8.7 5.219.4 Averages.| 6 3 5.1 6.8 3.9 15.7
ADDITIONAL CHILDREN PAIRED ORIGINALLY ON THE HEAD
TEACHER'S ESTIMATE

Fred T..
Gladys T..
Arthur G..
6
•
William W. 6
John B..
•
5
Elizabeth A. 6
Averages. 6
11
11
6 9
6 10
1 6.810.5 6.824.1
∞∞76
082527
8
8 11
11
10
5675O ∞
24
29 Edith D... 6
Kathleen E 6
4 6 11 6
23
9
23
24
Irene Ho.. 6
6
11 6 23
24
20
8 24
Hilda K...
Ernest R.
·
Walter H. 6
LO LO CO
5
5
11
CO4L
6 8 3 17
8
5
5
CO
54
17
4
15
Averages. 6
1 | 6.0 8.8 4.8 19.6
THE INFLUENCE OF SPELLING
113
third exercise. The results show that the phonic group has
achieved a material victory. Group A originally paired on Shapes
and Sounds have scored an average mark of 19.4, whilst the
corresponding section of Group B scores 15.7: the correlation
"r" coefficient between the paired series of individual results is
+.042; the difference between the means is 3.7 and its probable
error is 1.005, about one-fourth only of that difference. For
the sections of the groups paired on the head teacher's estimate,
the "r" coefficient is +.515; the difference between the means is
4.5 and its P. E. is .804, about one-fifth of that difference. Both
the results are, therefore, statistically significant.
In the case of 7 words the spelling of the two groups was
equally good; with 5 words the phonoscript group was the better;
with 21 words the phonic group had the advantage.
A FEW COMMENTS ON THE KINDS OF ERRORS MADE BY GROUPS
A AND B
First of all, the exact results were analyzed and tabulated for
every word used in the dictation, and arranged in parallel
columns. One or two examples follow:
PHONIC GROUP
(20 IN NUMBER)
black, 8 correct
Misspellings were:
PHONOSCRIPT GROUP
(20 IN NUMBER)
black, 2 correct
Misspellings were:
blak
back
•
•
bac
blac
•
ble.
balk
•
bake
dak
daak.
•
4
blak
1
back
1
bac
•
1
•
blac
•
1
balc
•
1
dack
1
dlac
•
•
1
•
pal
•
1
peti
•
12
Nine different misspellings in each group.
white, O correct
Misspellings were:
white, 0 correct
Misspellings were:
wite
wit
•
witt
withe..
wott..
•
•
9
wit.
8
wic
1
bit
•
•
1
rtd.
.1
nothing written
Five different misspellings in each group.
•
4
1
3
5
1
1
1
1
1
18
.16
1
1
1
1
•
20
20
114
TEACHING BEGINNERS TO READ
PHONIC GROUP
top, 14 correct
Misspellings were:
pot
tob
tog
•
PHONOSCRIPT GROUP
top, 16 correct
Misspellings were:
2
tob.
2
.2
bot
1
2
tog
• •
6
4
nise
nis
nos .
nine
Three different misspellings in each group.
nice, 8 correct
Misspellings were:
4
nise
•
6
nis
•
1
nic.
nip
nest
•
12
nice, O correct
Misspellings were:
1
.16
1
1
1
2120
Four different misspellings in the phonic and five in the phonoscript group.
will, 18 correct
Misspellings were:
will, 10 correct
Misspellings were:
wllo
lliw
.1
wil .
.5
.1
wel .
.1
•
1
well
.1
2
wtll
.1
•
wli.
wis
.1
10
Two different misspellings in the phonic and six in the phonoscript group.
PHONIC GROUP
of, 13 correct
Misspellings were:
fo...
OV
nov
se
PHONOSCRIPT GROUP
of, 15 correct
Misspellings were:
..5
fo.
.1
1
OV
.3
.1
tog
.1
7
5
Three different misspellings in each group.
see, 18 correct
Misspellings were:
see, 4 correct
Misspellings were:
.2
se
sem
2
seem
since
•
•
One form of misspelling in the phonic group, and 4 forms
of misspelling in the phonoscript group. The 13 children
who wrote 'se' in the latter group wrote the e with a
phonoscript mark which gives it the long or double e-sound.
13
1
1
1
16
THE INFLUENCE OF SPELLING
115
Probably enough illustrative instances have been given to
render intelligible the following notes on the observations of
the head mistress who tested the children and analysed the whole
of the children's papers.
A tendency exists in both groups to omit the silent letters.
when writing down spellings—a tendency more marked in the
phonoscript than in the phonic group.
This relatively greater weakness in the phonoscript group
probably arises from the fact that the child does not need to
look at the shape of the silent letter in order to pronounce the
word; he has only to look at the silencing mark and then ignore
the letter. When the child is being taught to read by means of
phonoscript the teacher says, "How many lazy letters are there?",
or, printing the word on the blackboard, calls on the child to
point out the silent or lazy letters. The child's attention is really
only directed to the silencing mark, not to the letter as a whole.
The ordinary phonic method directs the attention of the child
to each letter; for example, when teaching the word will, point-
ing to the two l's, the teacher says, "Both these letters have the
same sound," and she sounds and points to each of the 1's suc-
cessively. Again, in a word like home, the child taught by
phonoscript notes the silencing mark attached to the final e, and
need not give any further consideration to that letter, as far as
pronouncing the word is concerned. In an ordinary phonic
method, the teacher says, pointing to the final e, "This letter at
the end makes no sound itself, but it makes this letter (pointing
to the o), say 'o.'" The teacher, of course, gives the long sound
of o in home. It is worth noting that no child spelled this word
correctly in the phonoscript group, whilst 8 were successful in
the phonic group.
In the case, however, of a word like know, in which the k
and w are silent and do not in any way modify the sounds of
the other letters, the phonic and phonoscript teaching are more
nearly on a level in spelling efficiency. In phonoscript the children
note the silencing marks on the k and w, and give these letters
no further attention. When teaching the phonic group, the
116
TEACHING BEGINNERS TO READ
teacher would say, pointing to the k and w, "These two letters
are lazy, they do not make any sound. We will put a little line
under them to remind us that, when we are sounding the letters
of this word, we must leave these two letters out." After the
word has been pronounced, the little lines are rubbed out. This
procedure seems to help somewhat in the subsequent spelling of
the word.
There appeared to be another source of misspelling. Every
phonic system (including phonoscript) produces difficulty with
silent letters; but there is also the non-reversibility of the shape
and sound series to be contended with, except in the case of the
International Phonetic Script. In this latter form of print one
shape has one sound attached to it and one sound only; and
each sound has one shape attached to it, and one shape only.
But this international script could hardly be taught to very young
children, and it is not intelligible to those who have not learned it.
It is, of course, one great merit of phonoscript, that, though
many of the literal symbols are modified, they are not modified
out of recognition, and any one who can read ordinary print can
read it. But though it possesses this great advantage in reading,
it loses much, in consequence of its lack of complete reversibility,
when considered in its relation to spelling. Let it not be thought,
however, that this is a defect peculiar to the phonoscript method;
it is to be found in all phonic methods.
The method of teaching reading which suffers least from this
defect is the alphabetic spelling method; for, in this system, the
unit of sound is the word, and the units of shape are the suc-
cessively visually apprehended letters (named, but never sounded
separately). And the same word-sound is always associated with
the same sequence of visual impressions; except in the few cases
in which words of different meaning, like their and there, are
sounded alike, but spelled differently.
It is also true that in a look-and-say method, the word-
sounds and word-shapes are equally reversible; but its weakness
as a spelling system is consequent upon its non-provision of a
means for the separate and successive apprehension of the literal
shapes.
THE INFLuence of spellinG
117
Perhaps the above discussion may be helpful in the elucida-
tion of some of the phonic and phonoscript errors in spelling, of
which I append one or two further illustrations.
When writing the word some from dictation, only three of
the phonic group spelled the word correctly and no child in the
phonoscript group was successful. Fourteen children in each
group spelled it "sum"; other misspellings for the phonic group
were "sun" (twice) and "san"; other misspellings for the phono-
script group were "som" (four times), "sume" and "sites." Both
failed with the silent e, and it must be remembered, further, that
the modified o of this word in phonoscript carries the short u
sound. An eccentric rendering like "sites" is quite unintelligible
to me, for the meaning and sound and use of the word some are
well known to all the children.
When writing the word you, 16 of the phonic group were
correct. They had learned that ou sounds like u in this word.
That is not quite true, but no finer distinction need be presented
to children, at this stage, at any rate. Only five of the phono-
script group were correct. Ten errors were due to the omission
of the u as if it were silent, instead of the o; and the word was
written "yo," giving a modified shape to the o which carries the
sound of u. If a modified o were written for this word, it
should have been the modified form of the first o in food, accord-
ing to the phonoscript system. One child in each group reversed
the o and the u writing the word "yuo," and one child in the
phonic group wrote the word backwards, thus, "uoy." Reversals
are evidently not confined to the look-and-say system, though
they are probably more common in that system.
ū.
With the word go, 13 of the phonic group were correct, and
17 of the phonoscript group. Five children in the phonic group
spelled the word "bo," one spelled it "og," and one "yo." The
three children who were wrong in the phonoscript group spelled
the word "po." These were, all but one (which was an inver-
sion), mistakes in shapes; and to explain their greater frequency
in the phonic group we must remember that these children had
learned to read from a printed g, whilst in phonoscript this g is
the same as in the print-script which the children learn to write.
118
TEACHING BEGINNERS TO READ
The reader may remember that, after each exercise in dicta-
tion, the children were told that they might write their names.
George Fordham, a rather clever boy in the phonoscript group,
did not do so after his first exercise; but after the second test,
he did; and he wrote it thus:
GOG fodm
After the third exercise he wrote it again, this time thus:
GAG fadm
The second letters in his rendering of George and Fordham, a
modified o in the first case, and a modified a in the second case,
are both correct; for they both, in phonoscript, carry an English
au-sound. George had not been taught to write phonoscript;
but like other children, he sometimes used it. The o's would have
been printed in the way George wrote them, had these words
occurred in a phonoscript reading book. In both cases he has
left out the silent letters. In his second signature he has again
written his name correctly as to sound, but this time he has used
a modified a, instead of keeping to the o of his first signature.
As we see, George knows that there is more than one literal
symbol for the same sound, and exhibits his knowledge.
One little girl, in a subsequent dictation, obtained a few
marks only for her words. They were all of them written in
phonoscript, and she was marked wrong by her teacher, where-
upon she said to her teacher (not the head mistress), "They
are right, teacher," and turning to another child she remarked,
"Teacher cannot read them; you see she does not understand:
we have to put in all our lazy letters." She spelled face as “faes,”
said as "saed," milk as "millke," and nice as "nies." All the e's
had the silencing mark attached. No comment is needed.
Summing up this section we may fairly conclude that all
phonic systems are weak so far as the acquisition of correct
spelling is concerned and that the phonoscript, as being in a sense
more phonic (not of course phonetic), is weaker than the other
phonic systems.
As a minor conclusion, I may say that the head mistress de-
termined that those of her children who were learning to read
THE INFLuence of SPELLING
119
from print-real print-should learn to write print-real print
—for a year or two; and that those who were learning to read
from phonoscript should learn to write phonoscript. She had
seen enough of the spelling results from the employment of two
different scripts for writing and reading.
And her opinion was that she would continue to begin cursive
script (writing) and alphabetic spelling in her Standard I class
(average age 6 yrs. 9 mos. to 7 yrs. 9 mos.). Indeed, she does
at present, without throwing away, even at that stage, a possible
appeal to phonoscript or other phonic method to help her children
out with the reading of unknown words. And she resolved
to begin the teaching of reading and writing in her lower classes
by means of phonoscript, notwithstanding the spelling results at
this stage. I quote these opinions because, though I cannot
justify all of them by scientific proof, the balanced judgment
of a well-trained experimenter and highly efficient experienced
teacher is by no means without value in itself. Each head
teacher must balance the advantages and disadvantages (here
set forth in proof) of all systems in both reading and spelling
and make her own decision.
CHAPTER IX
ON THE LOSS DUE TO TRANSFERENCE FROM
PHONOSCRIPT TO ORDINARY PRINT AFTER
ONE YEAR OF TEACHING THE
PHONOSCRIPT METHOD
This issue was reserved for further treatment. It is often
maintained that the teaching of a special print or script by means
of which children may learn to read, involves a considerable loss.
when transfer is made to ordinary print. There is, in psycho-
logical language, much "interference," and a need for "re-
adaptation." This argument is countered by the statement that
it is only at first that difficulties are experienced; that by-and-by,
we, and young children, do not notice, nor need to notice, the
special forms of the letters of the word, nor indeed the letters,
as separate entities at all; we apprehend, it is said, the appearance
of the word as a whole and associate the appropriate sound with
that visual perception. The interference rapidly diminishes and
at last disappears.
It would be rash to deny any of these assertions in any whole-
sale way. They are all true. But, just as in modern economics,
the sweeping generalizations of the classical school and the
equally sweeping generalizations of the psychological school are
being harmonized into truth and practicability by careful
statistical work, so, in modern education, we are now trying,
not to meet assertion merely with counter assertion, even though
the latter be true, but to find out exactly how far each dictum is
true on the whole and on which side the balance of truth or
advantage lies. And, of course, education has this great ad-
vantage over most of the human sciences, that we cannot only
observe and calculate, we can also manipulate our material and
experiment. It may very well happen that education, being pro-
vided by experimental psychology with its methods, and by in-
ductive logic and statistics with its means of calculation and
120
TRANSFER FROM PHONOSCRIPT TO ORDINARY PRINT 121
methodology, coupled with the outstanding advantages that it
possesses of manipulable material and refinement of individual
measurement, may become, and that at no distant date, one of
the most advanced and definite of the sciences, to which, indeed,
other biological sciences may turn for guidance and corroboration.
Perhaps I may be forgiven this exordium; it is only by some
such belief that one can be sustained through the long detail of
protracted educational research. Our question then, the edu-
cationist's question par excellence, whenever a well-accredited
dictum, psychological or educational, is advanced, is, "How much
is it true?"
Passing now to the particular issue, I am quite prepared to
admit that the loss in transference from one script or print to
another may diminish as time goes on, the question for us is,
At what rate does it diminish? Does it diminish so as to render
its employment profitable under the ordinary conditions of school
life? Mere assertions without experimental proof we shall
ignore, as we shall also ignore what is asserted to be true on the
basis of the work of selected children. This is too often, I regret
to say, the basis of work shown in educational exhibitions-
often a modern camouflage to educational inefficiency.
Just a year had now passed since the little boys and girls in
one group of Grade II children began to learn to read by means
of phonoscript. Six months before, this group had largely failed
when required to read from ordinary print, though they had
been very successful with phonoscript. This group had now an
average age of 6 years 3 months. It will be remembered that
these children began to learn to read in November, 1922, at an
average age of 5 years 1 month: it was now November, 1923,
and the children were a class higher in grade, namely in Grade
III. This was one of the equal groups into which the Grade II
class had been divided for the purpose of testing the relative
efficiency of an ordinary phonic system and the phonoscript sys-
tem as a means of beginning to learn to read. The groups were
not quite intact; several of the weaker children in each group had
not been promoted to Grade III. Their marks in the preliminary
122
TEACHING BEGINNERS TO READ
tests in shapes and sounds were all well below 20, out of a max-
imum of 48, and they had failed to learn to read by either method.
A few other children who had not taken the preliminary lessons
and examination in shapes and sounds had been placed, paired
on the estimate of the head mistress, 5 in the one group and 5
in the other. The head mistress had herself continued to teach
the phonoscript group. The phonic group, after the six months'
trial and tests, had been handed over to the ordinary teacher of
the class who understood the system well. It is not intended to
compare at present the reading of the two groups. I give these
details so that teachers may see clearly what was being done, how
Group B-the phonoscript group-had been taught and how it
was now constituted. The present issue now is: When children
have been taught phonoscript, how much loss is found in trans-
ference to ordinary print after a year's first-rate teaching? To
determine this, we shall set two tests, one in phonoscript and one
in ordinary print, both consisting of precisely the same reading
matter. Prior to this testing, the children had read three books
of phonoscript, an introductory book and a first and second
primer. The passage was selected by me from one of the phono-
script readers. Neither the head teacher nor the children knew
that any test was to be applied.
The actual words of the test follow:
I was quite wild with joy, but Minnie looked the
image of misery. "I don't want to leave this
darling cottage," she sobbed, and Tim barked
savagely when he saw her distress.
The test was first administered in phonoscript and then in
ordinary print: the method of administering the tests was pre-
cisely similar to that already described, except that we did not
keep a time register.
One mark was given for each word correctly pronounced.
No additional mark was given if the child pronounced the word
without first sounding the letters, and no mark was added or
subtracted if the child sounded the letters before pronouncing
the word. If the child sounded the letters correctly, but did not
pronounce the word rightly or at all, no mark was awarded.
TRANSFER FROM PHONOSCRIPT TO ORDINARY PRINT 123
It will be seen that the test consisted of 30 words, exclusive
of the letter I, which was not counted for the award of marks.
It was especially suitable for the following reasons:
1. No word (except 'I') was repeated.
2. Every word had phonoscript markings on one or more of its letters
when used as Test 1.
3. The passage included some capital letters. Six months earlier no
capital letters had been taught or included in the tests.
The maximum mark was, of course, 30 for each test. I give
first the individual marks for the children of the phonoscript
group, showing the marks for Tests 1 and 2, the first in phono-
script and the second the same reading matter in ordinary print,
after which I shall make a few observations-the complete
TABLE XIII. THE INDIVIDUAL RESULTs of group B, THE PHONO-
SCRIPT GROUP, TESTED AFTER TWELVE MONTHS' TEACHING
IN READING
Name
Age,
Nov., 1923
(Yrs. and Mos.)
6
Marks for
First Test,
Phonoscript
30
Marks for
Second Test,
George F...
Joan W..
Herbert H.
Doris S.
Edward C……..
Ethel G..
Lily J..
Irene H.
Daisy G..
Violet C.
•
Edith D..
Arthur G.
Jane N.
Marjorie B...
Annie V.
Averages.
*
Kathleen E...
Irene H..
Hilda K.
•
Ernest R...
Walter H...
Averages.
Ordinary Print

6
•
•
6
39243
•
•
•
•
75 COLD CO
2
9
6
5
-10
6 - 3
6
•
6
2
-11
5 - 6
≈ON** ***** SONNG
24
30
25
27
18
30
28
30
26
28
19
28
23
30
26
28
21
30
30
30
30
26
19
27
16
22
19
26
16
6 — 3
28.1
22.7
ADDITIONAL CHILDREN


6 1
•
6
5
5
8
6 — 5
28272
30
30
29
2228
30
29
20
27
22
24
18
6 - 1
28.0
23.8
124
TEACHING BEGINNERS TO READ
individual records would print into a small volume by them-
selves on the method of attack and the kind of errors made in
the transference from one form of script or print to another.
First then, I show separately the individual marks for correct
pronunciation by the two sections of the group. The first section
had taken the preliminary lessons and following examination in
shapes and sounds; the second had not.
The first observation to be made on these tables is the ex-
traordinarily high mark for the phonoscript group (both sections)
when reading their own print. I think the reader will agree,
after looking at the test and remembering that it was in no
way specially prepared, that for children of this age, admirable
progress has been made. For the first six months only 4 lessons
a week (omitting Mondays) of 20 minutes each had been given,
and during the last six months only 5 lessons a week of the same
length had been given. And it was real experimental work, under
rigid conditions. There was no bolstering up the results by
references in other lessons and no "chivvying up" or special teach-
ing of a relatively backward child. The second observation is
that there is again a serious loss when transference is made to
ordinary print. The necessary statistical calculations are: cor-
relation "r" coefficient +.755: the difference between the two
averages is 5.4 and its probable error is .57. That is to say, the
mean difference is about 10 times its probable error, and is there-
fore highly significant. For the little group of 5 children the
statistical results are, of course, not so secure; but they have
been calculated; the correlation "r" coefficient between the two
series of results is +.8: the difference in the averages is 4.2 and
its probable error is 1.065, which is about one-fourth of the
mean difference.
Summing up then our second observation, we can say that
there is still a serious loss; but there is much less loss than there
was six months before, and it may happen that, later on, it will
practically disappear.
If this be so I do not know at present-one of the great
objections to a special print or script in the early stages of learn-
TRANSFER FROM PHONOSCRIPT TO ORDINARY PRINT 125
ing to read will cease to have much force or relevance. More-
over, it may be remembered that this special print presents little
or no difficulty to those who have learned from real book print.
Could the great loss previously found after six months have been
adventitious? For the moment I reserve this issue.
Now a few words as to the method of attack employed by
the children to read the tests. When these children were tested
six months before, after having learned to read phonoscript for
six months, it was quite an exception, so rare as to be negligible,
for a child to pronounce a word without having first sounded the
letters. Now, six months later, in November, 1923, after having
learned to read phonoscript for a year, the reverse was the case.
Nine of the children in working Test 1-the phonoscript print-
ing, did not sound a single letter, and two others sounded the
letters of two words only. When reading Test 2-ordinary
print-ten children sounded no letters. This does not mean (I
make this parenthetical caution) that Test 2 was read more
accurately than Test 1: as we know, the reverse was the case.
And it may have been, when the children hesitated before pro-
nouncing a word, that they were sounding the letters to them-
selves or remembering, one by one, the sounds of the letters; but
if they were, we can at least say that no sounds were audible.
In their own print-the phonoscript—the children gave a
pronunciation which was more correct, definite, and clearly cut
when they sounded the letters than when they did not. The
opposite was the case when they were transferred to real or
ordinary print. It will be remembered that any intelligible pro-
nunciation was accepted as correct, whether of a colloquial or
academic variety. The mistress who examined the children
obtained the impression, from the above observations, that there
was going to be much less difference between the two series of
results than had existed six months before-an impression fully
justified by the statistical tables.
Let me now give a few illustrations of the kind of errors
made (a) in phonoscript and (b) in ordinary print. But let
me say, at once, that with an average mark of 28 out of 30 in
126
TEACHING BEGINNERS TO READ
phonoscript reading, these errors, though they may help us to
understand better how to teach the children, cannot, at this stage,
be regarded as serious. A few common errors follow:
In the case of the word image, seven children were incorrect. The difficulty
lies largely in the a-sound. It has a modified shape in phonoscript and carries
the short sound of i, as in tin. Since the phonoscript modification is very
much like another modified form of a which carries the sound of ar as in
the word star, some children gave the ar sound to the a, pronouncing the
word 'marge' or 'imarge.' One gave the letter the sound of the o in boy
(an oi-sound) which is a modified o in phonoscript, and pronounced the word
'moige.' Two children gave the a the short o sound, pronouncing the word
'mog.' It will be noted that the initial i has sometimes been left out in pro-
nouncing the word, not because of its own difficulty; but because attention,
concentrated on the vowel sound of the second part of the word, had lapsed
from the first part. One child sounded the modified a as if it were a modified
form of o, and pronounced the word as 'moge,' with the o long as in no. One
other child, after having some difficulty with the second vowel sound, pro-
nounced 'mige' with a short i and a soft g. Had not the sound of the initial i
been omitted, this would have been correct according to the phonoscript system.
In the word misery, the letters were always correctly sounded, even by
the eight children who pronounced the word incorrectly. It is, of course, a
hard word for young children, quite apart from the sounds of its component
letters. It is rather long; and mere length is a difficulty, however good our
system of teaching. Moreover-and this is possibly a more important factor—
it is unknown as a word to young children and they do not use it. There were
five erroneous pronunciations.
The word savagely is a hard word; it is long, and it is not used by these
young children. Moreover, as already pointed out in the case of the word
'image,' the sound of the second a is a source of great difficulty. It was
sounded variously as au in caught, as o in no, as ar in star, and as o in on,
the last a scarcely excusable error. One child who sounded all the letters
correctly, called the word 'carriage,' an obvious shot from the sounds to a
word he knew.
In the word distress, every child gave the correct sounds of the letters,
but eight children pronounced the word incorrectly. It was pronounced
variously 'estress,' 'destress,' 'tress,' 'dresses,' and 'dress' (3 times). It is
obvious that, since the sounds of the letters were correctly given by every
child, the difficulty lay either in the length or in the unknown nature of the
word, most probably the latter, since the sounds of these individual letters
are not difficult to combine.
Now let me pass to the errors made when the children, trained phonically,
were tested on the same passage in ordinary print. The reading is still good,
but the average mark falls from 28 to 23 out of a maximum of 30; and it is
worth while to say something about the errors. Eleven words out of the
thirty proved to possess the principal difficulties. A comment on the rendering
of these words follows:
The word quite, written without the modifications or phonoscript markings,
was found very hard, though the word itself is well within the vocabulary of
these young children. Four children, who did not sound the letters of the
word before pronouncing it, called it 'quit,' 'qit,' 'qt' and 'guff,' respectively.
Five children who did sound the letters, gave an incorrect sound of the letter
'i,' and did not sound the final 'e.' They pronounced the word respectively
as 'qith,' 'quit,' 'qit,' 'qith,' and 'cousin'; in each case the i was sounded
as the i in tin. The error 'cousin' is probably a shot to a well-known word.
The final e would have had a silencing mark in phonoscript but was now
without it. It would not have been surprising, therefore, had it been sounded;
TRANSFER FROM PHONOSCRIPT TO ORDINARY PRINT
127
but it was not. The error as to the i is easy to understand. There are eight
i's in phonoscript (some, of course, very rarely used), and the symbol printed
as in ordinary print, which the children were now reading, carries the sound
of the short i. If, therefore, close attention were given to the form of the i,
the children might quite rightly, on a phonoscript basis, call the word quit,
if the final e were disregarded.
With the word wild, only eight children were correct: every other child
called the word 'wild,' again giving the i the short sound. The explanation is
the same as in the word just cited. Only three children sounded (at least
audibly) the letters before pronouncing the word.
Six children pronounced the word joy incorrectly, of whom only one
sounded the letters before pronouncing the word. It was pronounced 'John,’
'joi,' 'just,' 'us,' and 'ji. Since they had not the phonoscript modification
of the 'o' to guide them, they seemed at a loss.
With the word misery, eight out of the nineteen children in the group
were wrong. It will be remembered that eight of them did not succeed even
when helped by the phonoscript markings. Of the eight children who pro-
nounced the word wrongly when trying to read it in ordinary print, not one
of them sounded the letters.
In the word want, the 'a' of real print was a difficulty. No letter a of
phonoscript, modified or unmodified, is like a printed a. They are all modifica-
tions of an italic letter a. It is therefore worth noting that the new form
of the a did not always interfere with the association of the written word with
its sound. Seven children were quite correct. No single child sounded the
letters audibly. Nine pronounced the word as
Nine pronounced the word as 'went'; two pronounced it
wǎnt,' with the a as in man; and one called it 'with.'
The word leave was difficult. Thirteen children were incorrect, of whom
only two sounded the letters. These two sounded all the letters, giving to
both e's the sound of e in men, and to the a the sound of a in man; but they
pronounced the word as 'lăv.' Eight other children who did not sound the
letters also called the word 'lăv.' The remaining three incorrect pronunciations
were 'lěvěr,' 'lavender,' and 'lovely.' It is not surprising that this word was
very inaccurately rendered; for, in phonoscript, the first e would be modified
in order to lengthen it, and the a and final e would both possess silencing
marks. In ordinary print these guides are worse than lacking, for the un-
modified form of the letters, if they are closely attended to, are misleading
to a child brought up on phonoscript. 'Lavender' and 'lovely' are errors
arising from the perception of some of the sounds of the letters, with their
assimilation into a word known to the child-a very fertile source of error
with young children-but, as this process, perhaps more accurately done, is
one basis of rapid reading later on, and is, very largely, the basis of transfer
from one script or print to another, it cannot be wholly avoided.
The word savagely had been badly rendered from the phonoscript itself,
so it was not surprising that ten children were wrong when required to read
it from an ordinary print. All the children who were wrong rendered the
vowel a as in man, and e as in men; as, indeed, on an accurately perceived
phonoscript basis, they should have done, for the second a was without its
modification, and the e was not silenced.
The word saw was a failure. Only four were correct: fifteen were wrong.
No single child sounded the letters, rightly or wrongly. Eleven called the word
'was,' an inversion well known to teachers; two called it 'so'; one said
'say'; and one said 'sat.' The phonoscript system does not prevent reversals;
nor, indeed, does any other system. Reversals, however, are much less common
with the phonoscript than with the look-and-say system; for phonoscript
aims (with some other systems, the alphabetic, for example) at the consecutive
apprehension of the letters from the beginning of the word to the end—which
the look-and-say method does not.
Even when reading from phonoscript, eight children out of nineteen had
pronounced the word distress incorrectly. On the transfer to ordinary print
128
TEACHING BEGINNERS TO READ
this number was increased to twelve. The phonoscript i in this word is the
same as that in ordinary print, and it was rendered correctly in the main. The
other errors are probably due to the length of the word and to assimilations
to well-known words.
In rendering the word image from phonoscript seven children were wrong.
This number increased to eleven when transfer was made to ordinary print.
Two of the eleven sounded all the letters, the a as in man, the i as in tin,
and the e as in men (the e was, of course, not now silenced, as in phonoscript,
and the a was not modified in shape). The remaining nine who were wrong
did not sound the letters. Six of them pronounced the word as 'imag,' giving
the a-sound as in man and g as in rag. In phonoscript print the g would
have been modified for its soft sound. And, of course, the children were now
working from the g of ordinary print, not from an italic g which the phono-
script g's closely resemble. The children certainly apprehended the printed
g as a g of some sort, though it was a shape they had not seen before. I
can only suppose that they found the printed g more like a phonoscript g than
any other phonoscript letter.

r,
With the word barked eight children were wrong; they now lacked the
modified a. Two of them sounded the letters, the a as in man, rolled the r
and sounded the e as in men; but they pronounced the word as 'backed.
The remaining six did not sound the letters. Two called the word 'backed';
two called it 'baked'; one said 'bagging'; and one said 'begin.' 'Back,' a
well-known syllable, was rather frequently in use. 'Bark' is a well-known
word also; but, as I have said, the modified a was lacking; and, moreover,
the printed a is not really like any phonoscript a, even those most commonly
recurring.
Summing up from the observations of the mistress who took
the tests and the teacher who made the records: the loss on
transfer was not very great; and the children had been fairly suc-
cessful, even with the new shapes of a and g. As I have before
remarked, sounding the letters gave an advantage when the
children were reading from the script or print they had learned,
but did not appear to do so, when the transfer was made to the
ordinary print of a printed book. The great loss on transfer
after the end of the first six months, in view of the comparatively
small loss now made, surprised us all, and we reserved this issue
for experimental work with a new class and a new teacher.
:
CHAPTER X
A FURTHER EXPERIMENT ON THE LOSS WHEN
CHILDREN TAUGHT BY THE PHONOSCRIPT
METHOD ARE REQUIRED TO READ
FROM ORDINARY PRINT
Both the head mistress and I were so impressed by the
diminution of the loss on transfer after a year's phonoscript
that we began to wonder, if, after all, the loss after six months
was always, of necessity, as great as that which we had found
after that period in our 'phonic versus phonoscript' experiment
(Chapter VII). So we started all over again with an absolutely
new class and another teacher.
This experiment was carried out with the whole of a Grade II
class, boys and girls, of an average age of 5 years 0 months at
the beginning of the educational year just before the experiment
began. No reading of any kind, nor any preparation for read-
ing, had been taught or learned; and, at the end of April, 1923,
a beginning was made of teaching phonoscript.
Five reading lessons, of exactly 20 minutes each, were given
each week, one on each morning of the five school days, Monday
to Friday inclusive. All the teaching was done by the usual
teacher of the class, not by the head mistress as in the previous
experiment with phonoscript. Up to the dates of the tests on
transference (to be described in this chapter), all the teaching
was done by means of letters, words, phrases, and sentences,
written on the blackboard; though, occasionally for five minutes,
at the beginning of the school session, while the teacher was oc-
cupied with registration, the children had been allowed to see
the reading books containing the reading matter which they were
being taught. The head mistress had instructed the teacher to
make at first some variation in the order of the words used in
the phonoscript system. She thought it desirable, and I whole-
heartedly agreed, that very short, well-known words like ran
129
130
TEACHING BEGINNERS TO READ
and not, should not be delayed while words such as visit and
engine, were being dealt with.
Accordingly, the simplest sentences on the Phonoscript Read-
ing Chart were dealt with first. After that, when the children
had become accustomed to read from phonoscript, the teacher
recommenced her lessons following the order of the chart, word
for word.
Until a week or so before the test on transference with which
this chapter deals, neither the head teacher, the teacher of the
class, nor the children knew that any special tests of the work
were to be given.
The whole of the class was tested, or at least the testing
was begun, for it was a longish business, on Wednesday, Novem-
ber 28, 1923, after some six months' teaching. The average age
of the class at the end of this November was 5 years 6 months.
Eight children then in the class were recent admissions to the
school, and they were not tested. Those who were tested num-
bered 41, of whom 20 were boys and 21 were girls.
The object of the test was to discover how much loss, if any,
there would be when these children, wholly taught to read and
write by means of phonoscript for six months, were required to
read from print. For this purpose, it was, as before, necessary
to give the same reading test twice, once in phonoscript, and once
in ordinary print-the print of the ordinary printed book or
newspaper.
THE TESTS AND THEIR ADMINISTRATION
The passage selected for the tests was one which the children
had actually read during the course of their reading lessons; not,
as on the previous occasion with another class, a passage in
phonoscript which they had not actually read. The piece se-
lected follows:
i have a little pet rat.
i call it pip.
its cage is big.
i lay hay in it.
he lets me see him eat.
FURTHER EXPERIMENT ON LOSS
131
The reader may note that there were no capital letters—capi-
tals, quite rightly in my judgment, had not yet been taught—and
that the passage contains 25 words, including i, which occurs
three times, and it, which occurs twice, no other word being
repeated.
The test, first in phonoscript, and then in ordinary print, was
hectographed until there were copies for each child to read from,
one for the teacher of the class who gave the test, and one for
each child, upon which latter copy the head mistress wrote the
child's name, his age in years and months, the date and time of
the test, and an exact and complete record of what the child did.
If the child sounded the letters before pronouncing the word,
the sounds were recorded: the record showed also when the
sounds were not given, or, if some were given, which were given
and which were omitted. Both wrong and right sounds were
recorded, and all pronunciations of the words or parts of words.
The children were tested individually in the classroom in
which they had been taught. The other children of the class
were meanwhile occupied with free drawing under another
teacher in another classroom until their turn came to be tested.
It should be mentioned here that these children were going to
be concerned with two systems of printing or writing only. They
had learned no writing in the ordinary sense, either of cursive-
script or of print-script. What little writing had been done
was done in phonoscript.
There were only three words, pip, in, and him, in which the
forms of all the letters were practically identical in phonoscript
and real print. Even the letter i, when standing alone, has, in
phonoscript, a modifying mark to give it the long i-sound. Only
correct pronunciations are scheduled in Table XIV. Any in-
telligible pronunciation was accepted, academic or colloquial.
The boys and girls are shown separately.
One word of caution: these are not selected, equal groups
of paired children; they are the boys and girls of an ordinary
Grade II class in an infants' school, in no way selected. No
conclusion can be drawn as to the relative efficiency of boys and
132
TEACHING BEGINNERS TO READ
girls in learning to read from the above table, except, perhaps,
the negative one that there is no significant difference. But since
the boys and girls are tabled separately, I will present the usual
statistical calculations separately.
TABLE XIV. INDIVIDUAL RESULTS AFTER SIX MONTHS' INSTRUC-
TION IN PHONOSCRIPT
(Maximum mark, 25)

Boys
Age in
(Yrs. and
Mos.)
Test 1, Test 2,
Phono- Ordinary
script
Age in
Girls
Print
script
Test 1, Test 2,
(Yrs. and Phono-Ordinary
Mos.)
Print
James W.
Herbert N...| 6
William N... 5
Henry C..
John H.
5
·
•
Frank T.
James H..
William C...
Henry S.
Arthur D...
Ernest B....
George P...
Joseph F....
Henry W
Eric K.
James T.
Frank P.
Cyril H.
Charles T.
·
5
5
5
6 -10
4 - 9
4.
6
5
LO CO
6
-10
4
William S.... 6 — 0
LO CO LO LO LO
LO LO LO C+
5
2233CO
3720 ●
25
10
25
10
6A132
445COLO
9
21
*** ***** 22222 09502
21
15
12
10
24
• 15536
Maud H.
•
Rose W...
24
9
23
12
LO LO CO C∞
3
LO LO LO CO
17
Alice B..
6
0
16
Marjorie S.. 5
4
15
LO 3 3 LO
5
Charlotte B. 5
Isabella McI 5
Elizabeth D. 5
Elsie L.
Elsie A..
Elizabeth H. 4
TOBIKE
7
14
19
8
17
2504
6
3
0
Averages. 5 - 6
20.4
6.6
Averages 5 — 6
19.2
5.2
Dorothy G.. 5
Emma B..
Victoria J...
•
Kathleen M.
Irene E..
Gladys J.
Doris J.
Marie H.
Florence P.
Dorothy B.. 5
Violet M..
LO CO LO NE
5
-11
6
1
4
5
2222N
25
11
25
10
25
25
25
ŁO TH LO LO LO
5
-11
4
6
2
10
5
- 3
22222 22299
25
25
24
10
24
23
HOBNO ZNO∞0
5
7
6
7
7
8
6
23
6
5
For the boys, the "r" correlation coefficient between the phono-
script and print series of marks is +.776; the difference be-
tween the means of the two results is 13.8, with a probable error
of .536: the mean difference being therefore 25 times its prob-
able error. For the girls, the "r" correlation coefficient is + .848;
the difference between the averages of the two series of results
is 14.0, with a probable error of .63; the mean difference is
therefore 22 times its probable error. Two conclusions are clear:
first, that these children, with few exceptions, have made, for
FURTHER EXPERIMENT ON LOSS
133
i
their age, very adequate progress in learning to read from phono-
script; and second, that there is again, after six months of very
good teaching, a very serious loss when transference is made to
ordinary print. Our previous conclusion in this matter is there-
fore fully justified. A few notes on the errors may be of service.
NOTES ON THE ERRORS MADE IN READING
Phonoscript.-As I have said previously, a complete analysis
of each child's work with both tests and with each word indi-
vidually had been written down. The whole of these analyses
cannot be printed, but a few results will be given as a guarantee
that our conclusions have not been formed on the evidence of
massed figures alone. Any teacher who cares to make similar
individual and systematic investigations on the results of any
system of teaching reading in its early stages, and who scrutinizes
exactly the analysed results will be amply repaid. Fortunately,
the errors fall into groups and it will not be necessary, except in
very rare cases, for the teacher to correct them by individual
tuition. A few, a very few, of the children make errors so wide
of the mark, that it is obvious that they are not yet mentally
advanced enough to begin to learn to read. It is to be under-
stood that the words printed at the left of the following com-
ments were in phonoscript.
have
a
little
This word stood alone. The modifying mark of phonoscript―
an upward stroke preceding the vertical line-gives this i a long
sound. Only 3 children made a mistake, each giving the short
sound. They did not note or act upon the specific modification
of the letter. Of course, short unmodified i's are much more
numerous in reading matter.
Eight errors. Two children sounded the letters correctly, but
pronounced the word as 'had.' One child gave the long sound
of a, and another the sound of a in amidst. One sounded
'h-a' correctly, but could not go on any further. Two
sounded all the letters correctly, but could not pronounce the
word. One sounded the letters as 'n-b-l-e,' and of course could
not pronounce the word-an obvious failure. If this child is to
learn to read, he must begin all over again and do so after an
interval, in order to forget much that he has improperly learned.
No errors.
Thirteen errors. Five sounded the letters correctly and pro-
nounced the word as 'lit.' Four sounded the letters correctly,
but gave no pronunciation. The remaining errors were scattered.
134
TEACHING BEGINNERS TO READ
pet
rat
i
call
it
pip
its
cage
is
big
Nine errors. The error is the confusion of b and p. In every
case the letters were sounded as 'b-e-t,' six children calling the
word bet, one set, and one et. One did not pronounce it.
Three errors, only one of which was serious.
Three errors made in the same way and by the same children
as before.
Fourteen errors. Six children sounded the letters correctly, three
calling the word 'caw,' one 'dell,' and one 'kill,' while another
failed to pronounce it at all. One child made no attempt at the
word, and another gave the short a-sound. Six other soundings,
with or without attempts at pronunciation, either omitted the a
or substituted for it. The apparent cause for failure with this
word, was the unusual form of the a, which is modified so as to
give an au-sound, and which presumably was not well known at
this stage.
Three errors.
Two children sounded the letters, but one called
the word 'ed,' and the other 'ter.' One child made no response.
Eleven errors. There were two main causes of failure: the 'p's'
were confused with 'b's,' and pip is not a well-known word to
these young children.
Nine errors. Six sounded the letters correctly, but pronounced
the word as 'is,' 'it,' or 'pits' or failed to pronounce it. It was
sounded without pronunciation as 't-t-s,' 'e-t-s,' and 'i-k-s.' One
child, who did not sound any letters pronounced it as 'his'-
probably not a look-and-say error, but a shot to a well-known
word from some apprehended letter sounds not audibly expressed.
Fourteen errors. Two children sounded the a as if it were the
e in me, and called the word 'kēge.' One sounded the letters
correctly, but pronounced the word as 'kag.' One could sound
the c only, another gave the a its sound in amidst, and another
sounded the final e which has the silencing mark-a very rare
case, for the silencing marks are quickly learned, and rarely
forgotten. One sounded the letters as 'k-j-o' and called the word
'Joe.' Four children, who gave no pronunciation, sounded the
letters respectively as 'k-g,' 'k-i-a,' 'd-a,' and 'k-a-g,' the last
giving the short sound of a. One other child sounded the letters
correctly, but could not give any pronunciation. (I have used
the letter k in the above to signify the hard sound of the letter c).
The g in cage is slightly modified to make it soft; this modification
was often unnoticed and the hard sound given.
Five errors-a rather large number for this word. Three children
sounded the letters correctly, pronouncing the word as 'it,' 'its,'
and 'it'-probably due to attraction by other well-known words
containing similar sounds. One called the letters 'e-s,' but did
not pronounce the word; and another, who also did not pronounce
the word, sounded the letters correctly.
Quite unexpectedly to us, 20 children, nearly half of the whole
number, failed to give a correct pronunciation. In six cases the
b was sounded and pronounced as p, in three cases the letters
were sounded as 'p-i-ē' and the word pronounced 'pie'—a very
attractive word to children of five-and-a-half years. Four times
the letters were rendered as 'd-i-g,' and the word was called 'dig'
(d's and b's are confused in phonoscript as in other systems).
FURTHER EXPERIMENT ON LOSS
135
i
lay
hay
in
it
he
lets
me
see
him
Three times the letters were sounded correctly, but the word was
mispronounced; and in four cases in which the word was not
pronounced, the letters were variously sounded.
Three errors. One child gave short i, and another short e. One
made no response.
Sixteen errors. Three children sounded the letters correctly (which
means, since the y has a silencing mark in phonoscript, that they
sounded the 1 and a correctly), but two of them called the word
'a,' and one called it 'day.' Two gave a the sound of e in me,
and pronounced the word as 'lee'; and one gave the a its sound
in amidst but could not pronounce the word at all. Two sounded
the letters as 'l-ǎ-g,' one calling the word 'lap' and the other 'lag.'
The rest sounded the letters variously, the 1-sound being usually
right. There are errors here which show that one or two children
are not fit to go on. They have begun to read too soon, at least,
by this method. The check tests, previously used, of associated
shapes and sounds, had not been administered to the children.
Thirteen errors. Two children gave the letter a the sound of e
in me, pronouncing the word 'he.' Two others sounded and
pronounced the word as 'ham' and 'hag.' Another sounded the
letters 'h-ǎ' (the y has a silencing mark) and pronounced the
word as 'hă.' The remaining incorrect pupils did not pronounce
the word, but sounded the letters variously.
Five errors. One child sounded the letters as 'i-h' and called the
word 'him'; one sounded the letters as 'i-e,' yet pronounced the
word 'it'; and the remaining three, after giving erroneous sounds,
failed to pronounce the word at all.
Two errors.
Eight errors. Two children sounded the letters correctly but
pronounced the word as 'me' and 'his.' One child sounded it as
'h-ĕ' and called it 'him'; one sounded the e as a in made and
called the word 'hay'; one sounded it as 'l-ē,' and called it 'let';
and three failed to pronounce it after giving erroneous sounds.
Nine errors. Two children sounded the letters correctly, but did
not pronounce the word; one gave no pronunciation and sounded
the letters as 'l-i-t-s'; one sounded the e as in me, and pronounced
the word as 'leets'; one, who sounded the letters as 'k-a-t-s,
did not pronounce the word; and the rest sounded the letters
correctly, but pronounced the word variously.
Eight errors. Two children got no further than sounding the m;
two sounded m as 'n'; two gave the e its short sound (as in
men); while of the remaining two, one sounded the e as 'i' and
the other as 'ā.'
Eight errors. Five children gave no pronunciation, two of them
sounding the s only, two giving each e the short e-sound, and one
sounding 's-î-ĕ.’ Two others sounded the letters correctly but
pronounced the word as 'me,' and one gave the sounds 's-a' and
pronounced correspondingly.
Six errors. One child sounded and pronounced as 'hěm,' another
as 'lit'; and three children sounded the letters correctly, one of
whom called the word 'it' while the other two attempted no
pronunciation. One sounded as 'n-i-m,' but did not pronounce
the word.
136
TEACHING BEGINNERS TO READ
eat
Twenty-one errors. The modifying mark which gives the e a
long sound was either not noticed or its significance forgotten;
and, quite contrary to the usual practice, the silencing mark on
the a received little attention. Doubtless this was because silent
a's have been rare. Four children sounded and pronounced the
word at 'ǎt' and three as 'et.' Two sounded the a as in amidst
and pronounced the word 'at.' Six did not pronounce the word
at all and sounded the letters variously; while the remaining
children gave different wrong pronunciations and sounds.
The above rather elaborate analysis will have served its pur-
pose if it induces any teacher, after a course of lessons given to
her children on any definite system, to take a careful test or
tests, write down what the children do, and then at leisure, make
a similar analysis, and attempt to discover how, where, and why
the children went wrong. If there is nothing but mere confu-
sion, the lessons and tests have failed; but mere confusion is
very rare with young children, if they have been even moderately
well taught; so, if, instead of deciding a priori what they ought
to do, we find out a posteriori what they do actually do, we shall
not lack reward. Moreover, the analysis will enable any ex-
perienced teacher to see something more than mere opinion in
a few suggestions I may presently make concerning possible
improvements.
We have now, I hope, some fairly clear ideas as to the suc-
cesses (which were many), and the failures (which were few)
which these young children exhibited after six months' teaching
of phonoscript, with only 5 twenty-minute lessons per school
week.
Ordinary print.-Records of the responses of these 41 chil-
dren were kept in similar detail when they read the same passage
in ordinary print. From the figures given in Table XIV it is
of course apparent that they made many more mistakes in
these circumstances than they did when they read from phono-
script type. I am strongly tempted to give the details for each
word of the passage as I have already given them for the phono-
cript test. The seriously minded teacher would find them abun-
dantly instructive, but the length to which such a presentation
would run perhaps makes this degree of detail inadvisable.
FURTHER EXPERIMENT ON LOSS
137
A few comments, however, will be offered concerning the
first sentence.
With the word i only one child was correct. Most of them
gave the short sound of i, and they were right in doing so on
the basis of their knowledge. On the other hand, twenty-two
children pronounced the word have correctly—a decided suc-
cess. In the word a thirty-seven children recognized the letter
as an a but only two of them gave its correct sound. These chil-
dren have not been taught any form of printed a and the phono-
script a is quite different in appearance. An effort was made
to find out whether or not the teacher had inadvertently referred
to the form as it appears in ordinary print. It became quite
evident, however, that she had not done so. We can therefore,
only suppose that the children perceived the letter as more like
the a they had learned than any other letter. Perhaps, too, they
were helped by the fact that the letter stood alone. Because of
the similarity of the t of ordinary print to the phonoscript symbol
th in the word then, 31 children sounded the t's of the word little
as th's. For similar reasons, involving similarity with the phono-
script form, a large number-29 to be exact-of the children
gave to the final e the sound of e as in men. Again, in the word
pet the printed t-form led most of the children to give the th-
sound. Six children sounded p as if it were b— an error com-
mon to all systems, and one already noted as prevalent among
these children when they were reading the test in phonoscript
form.
These illustrations will to some extent convey to the reader
the nature of the errors which the children made in transferring
from phonoscript to ordinary print. I do not apologize for giv-
ing the results in somewhat extended detail. I am myself, like
teachers generally, very uncomfortable about any scientific re-
sult unless I can see by the presentation of individual cases just
what the figures mean. I am not arguing against the statisti-
cal summaries which are given and must be given; far from it.
We have not used them enough in education; we have been too
prone to accept individual illustrations as proving a theory, in-
138
TEACHING BEGINNERS TO READ
stead of merely exemplifying it. Even jokes may be useful for
this latter purpose; but we must remember the difference be-
tween illustration and proof, and I have tried to supply both.
We see, again, that children who have learned to read by
means of phonoscript for six months only are severely handi-
capped when they are set to read the same words in ordinary
print. Words like little, pet, rat, call, it, its, cage, lay, hay, he,
lets, me, see, eat and a and i standing alone are found to be
very difficult. This is not because they are unusual words,
but because the children depend on the additional marks in
or on the literal symbols, which, in phonoscript, indicate to
them the proper sounds of the letters. Does this fact of great
loss on early transference prove phonoscript to be a bad sys-
tem of beginning to learn to read? In my judgment, it does
not. It is a defect; but if our former work with children after
one year's teaching is valid, it is a defect which gradually
remedies itself. It is quite likely that the resulting curves for
phonoscript and print, as time goes on, approach each other
asymtotically.
CHAPTER XI
ON THE LOSS AFTER EIGHTEEN MONTHS OF
TEACHING WHEN TRANSFERENCE IS MADE
FROM PHONOSCRIPT TO ORDINARY PRINT
In Chapter X it was stated that a new class of younger chil-
dren under a different teacher, after some six months teaching,
had shown very serious loss when, after learning to read by
means of phonoscript, they had been required to read from ordi-
nary print. Yet I expressed the opinion, based upon previous
experiments, that this loss would rapidly diminish, and that prob-
ably the curves of success shown in reading real print and
phonoscript respectively would in time approach each other
asymtotically. It would probably be considered that, up to the
present, I had given insufficient evidence for such an opinion.
With such a criticism I should agree; and to verify or negate
my hypothesis I proceeded to further experiments with the older
children who had been learning to read from phonoscript for a
longer time.
It may be remembered that we started at the end of Novem-
ber, 1922, after a division of a Grade II class into two equal
and parallel groups-equal, as we thought, in their potentiality
for learning to read—to teach one group by means of an ordi-
nary phonic method and the other by the highly specialized
phonic system invented by Mr. Hayes, and called phonoscript,
both groups being taught by the same teacher. For the compara-
tive results of these two methods in their early stages the reader
is referred to Chapter VII of this monograph. Suffice it to say
here, that it was clear that the group working by means of phono-
script had acquired the greater power, especially as shown by
their capacity to read successfully new and unknown words. But
that conclusion is not now the question at issue; which is,
whether the loss on "transference" to real print so offsets this
gain, that teaching by means of phonoscript is not, on the whole,
139
140
TEACHING BEGINNERS TO READ
કે
profitable? Or, to put the matter in another way, since children,
like the rest of us, will be reading all their lives, not phonoscript,
but ordinary print, we must ask ourselves at what stage they
may profitably cease to read from phonoscript and proceed hence-
forward, with a negligible loss, to read from ordinary print.
The continued work of Group B-the phonoscript group of
Chapter VII-is now in question. The reading matter of
Groups A and B had by now diverged considerably, and the
teacher of Group A had been changed, since the head mistress,
who had at first taught both groups and was still teaching
Group B, had handed over the original phonic group to one of
the class teachers. Consequently, no further quite conclusive
comparisons could now be made as to the relative success or
failure of phonoscript and the ordinary phonic method taught in
the school generally.
After six months teaching it may be remembered that a very
serious loss was suffered when the children of Group B had
been transferred from phonoscript to ordinary print; and that,
after a year's teaching, the loss was still considerable though
very much less. It remains to be seen whether now, after 18
months teaching (the children now averaging some 6 years
9 months in age), the loss on transference to real print contin-
ues to diminish, and to what extent.
Of the phonoscript group, there were still left in the school
20 children who had been tested for this issue at the end of De-
cember 1923, (see Chapter IX). All of them had been through
the whole course of learning to read since the end of November
1922. It was now June 1924, some 18 months after the begin-
ning of the reading lessons.
For the teaching of this group as to method and for the
duration of the reading and cognate lessons (if any) up to the
end of November 1923, the reader is referred to Chapters VII
and IX. The time given to each reading lesson had been in-
creased at the end of April 1924, from 20 minutes to 30 min-
utes daily on each morning of the five school days. Dictation
lessons were now given three times a week, each lesson lasting
LOSS AFTER EIGHTEEN MONTHS
141
20 minutes; but in these lessons and in the lessons in handwrit-
ing, print-script was used, not phonoscript nor ordinary print.
The passages to be read for our new test were selected by
me on Monday, June 2nd, 1924. On Thursday, June 5th, each
child read Test 1, and on Friday, June 6th, each child read
Test 2; both tests being printed in phonoscript.
The Whitsun holidays, Monday and Tuesday, June 9th and
10th, now intervened, and although school opened again on
Wednesday, June 11th, it was deemed inadvisable to proceed
with this test until the children had rested physically after the
holiday, and had become re-adapted to the mental work of the
school. Also there was thus provided a useful interval between
the tests in phonoscript and those in real print. Tests 1 and 2
were printed, not now in phonoscript, but in ordinary print, and
were given again at corresponding times for each child on Tues-
day and Wednesday, June 17th and 18th. It may be as well
to state that from the time the tests were chosen, the books con-
taining the passages selected were not used by the children; and,
that, on the days on which they were not being tested, the usual
lessons in reading were given from another phonoscript reading
book.
ters.
All further details as to the administration of the tests and
their marking were the same as those described in previous chap-
The tests were selected from a continuous reader (one
in which the same subject or story is continued throughout)
printed in phonoscript characters, called Snow-White and Rose-
Red. It was not an unseen reader, but the children had not
finished the story, and there had been no revision. The two
passages, or rather the one passage divided into two tests (we
wanted to take no risk of a fatigue factor) contained 70 words.
The actual test papers follow as given to the children.¹
¹In the first line of Test 1 (phonoscript print) the word "downstairs''
should be separated into two words.
142
TEACHING BEGINNERS TO READ
Group B.
Test 1.
Age.
they came
Name.
When
the next
morning
he had
gone
they did not
Date.
Phonoscript Print.
downstairs
they found
before
as
but
trouble to look
him this
this time. They felt
for
quite sure
that they would see
him
in the evening.
Group B.
Name.
Test 2.
Phonoscript Print.
Age.
Date.
And
SƠ
through
did
they
did, and all
the winter
not fail to visit
the same time each
and to leave them
In the
Group B.
Name.
morning.
the bear
them at
evening
very early
Test 1.
Age.
Real Print.
Date.
When they came down stairs the next morning they
found he had gone as before, but they did not trouble
to look for him this time. They felt quite sure that
they would see him in the evening. (38 words.)
LOSS AFTER EIGHTEEN MONTHS
143
Group B.
Name.
Test 2.
Age.
Real Print.
Date.
And so they did, and all through the winter the
bear did not fail to visit them at the same time each
evening and to leave them very early in the morning.
(32 words.)²
The individual results are given in Table XV. Those for
the preceding test of a similar kind, given in November 1923,
are inserted for purposes of comparison. For Tests 1 and 2
combined, when read first in phonoscript, an average mark was
obtained of 68 out of a possible mark of 70 (or 97.1 percent)
by those children who had worked through the whole of the
experiment. For the same tests in real print by the same chil-
dren, 63.7 marks were scored out of a possible 70 (or 91 per-
cent), a differential advantage in favor of phonoscript of 6.1
percent. The corresponding figures for the additional children
were 69.4 out of 70 or 99.1 percent in phonoscript, and in real
print 66.8 out of 70 or 95.4 percent, a differential advantage in
favor of phonoscript of 3.7 percent.
The "r" correlation coefficient between the totals of real print
scores and the scores for phonoscript for the first group of chil-
dren, worked out from the individual paired cases, is .655; and
the mean difference between the results is 4.3 with a probable
error of 1.072. For the second group “r” is .98; and the differ-
ence between the means is 2.6, with a probable error of 1.139.
Three children had been absent for long periods during the
period concerned. H. H-. appears to have been wrongly placed
in our preliminary tests; but was probably not, for he had been
ordered spectacles for the correction of a bad squint, an order
which was only obeyed after considerable difficulty. And when,
after many weeks, the order was obeyed, he broke his glasses,
and managed to do this three times during the six months pre-
ceding the last tests: he could not read at all without them.
2 These words in these tests were printed just as large and spaced just
as much as in the phonoscript tests.
144
TEACHING BEGINNERS TO READ
TABLE XV. INDIVIDUAL RESULTS FOR THE PHONOSCRIPT GROUP,
WHEN TESTED AFTER EIGHTEEN MONTHS' TEACHING
(First reading from phonoscript print, second reading from real print)

Possible Marks
for Tests......
30
30
38
32
70
38
32
70
NAME
AGE IN
JUNE, 1924
(Yrs. and
PHONO-
SCRIPT
PHONOSCRIPT PRINT,
REAL PRINT,
PRINT
MARKS
REAL
PRINT
JUNE, 1924
JUNE, 1924
MARKS
Mos.)
FOR TEST, FOR TEST, Test 1 Test 2 Total Test 1 Test 2 Total
Nov. 1923 Nov. 1923
G. F....
6
9
30
24
38
32
70
J. W..
7
3
30
25
38
32
70
H. H...
6
27
18
31
27
58
D. S..
6
-10
30
28
38
32
70
E. C..
6 - 9
30
26
38
32
70
E. G...
7
L. J...
6
I. H...
6
D. G.
6
V. C..
6
83749
28
19
37
31
28
23
37
32
30
26
38
32
28
21
38
32
30
30
38
32
E. D...
6-10
30
30
38
32
A. G..
J. N..
6
6
M. B...
8
891
26
19
38
32
27
16
37
30
22
19
35
28
A. V. ...
6
0
26
16
34
31
RRRR 88RRR RR586
38
32
38
32
29
24
38
32
38
32
68
34
28
69
31
23
70
38
32
70
38
32
70
38
32
70
38
32
70
35
29
67
29
63
36
27
23
24
ROBOR SHORO OJN❤+
***** **≈≈≈ 22*N*
70
70
70
70
62
54
70
70
70
70
64
52
47
Totals 101 -10
422
340
553
467 1020 521 434 955
•
Averages
6 9
28.1
22.7 36.9 31.1
31.1 68.0 34.7 28.9 63.7
CHILDREN WHO WORKED THROUGH ALL THE LESSONS AND TESTS IN READING BUT
WHO DID NOT WORK THE PRELIMINARY TESTS IN SHAPES AND SOUNDS

K. E...
6 7
30
30
38
32
70
38
32
70
I. Ho.
H. K..
7
•
6
E. R.
6
025
30
29
38
32
70
38
32
70
29
20
38
27
22
38
•
W. H... 6 -11
24
18
37
1222
32
70
38
29
67
32
70
38
32
70
30
67
32
25
57
Totals 33 1
140
119
189
158 347
·
184 | 150
334
Averages 6 7
28.0
23.8
37.8 31.6| 69.4 36.8 30.0 66.8
Moreover he was absent 72 weeks with an infectious illness.
E. G. was absent for 12½ successive weeks, an absence which
has probably affected her results. I. H-. was absent for 9
weeks, suffering from scarlet fever; her absence, apparently,
has had little or no effect.
LOSS AFTER EIGHTEEN MONTHS
145
No child pointed to the words or letters during any of these
tests; and the sounding of the letters before pronouncing the
words was the exception rather than the rule. In a few cases,
where there was hesitation before the pronunciation, it seemed
that the child was sounding the letters, though not audibly. For
Test 1 when printed in phonoscript 13 out of 20 children gained
full marks. Of these, 12 did not sound any letters during the
test, and the remaining child audibly sounded the letters of one
word only the word "trouble." The only errors in pronunci-
ation in this test were in the words stairs, next, morning, before,
trouble, sure, and evening.
In Test 2, when printed in phonoscript, 14 children out of
the 20 read all the words correctly. Eleven of these did not
sound any of the letters. One girl sounded the letters of the
words same and evening, and another girl sounded the letters
of the word visit. Only eight words out of 32 were wrongly
pronounced by any of the children. These were through, bear,
fail, visit, teach, evening, very, and early.
In Test 1, printed in ordinary type, there were 12 children
who made no errors of any kind, as against 13 when the same
passage had been read in phonoscript. None of the 12 sounded
any of the letters, apparently not even inaudibly. There were 14
words out of a total of 38 in which errors occurred as against
8 when the reading was from phonoscript. These words were
down, stairs, next, morning, found, before, trouble, look, time,
quite, sure, would, and evening. A more detailed analysis of
these errors will be given as a concrete illustration of the par-
ticular difficulties which ordinary print presented to these chil-
dren. There were 55 mistakes out of 760 pronunciations taught
by means of phonoscript:
With the word down, two children went wrong. They sounded the
letters; and, apparently missing the phonoscript modification of the o,
gave it rightly, from its shape in print, a short o-sound, and pronounced
the word as 'don.' But it must be emphasized that they were the only
children thus led into error.
With the word stairs three children were incorrect, against three
when reading this word from phonoscript. Two sounded the t as th
(for this printed 't' has the same form as the phonoscript 'th') and,
146
TEACHING BEGINNERS TO READ
quite properly, on their shapes in print, gave a and i their common
short sounds. This combination of letter-sounds was beyong them, and
one said 'thats'; and one said 'theirs.' Another child, without sound-
ing said 'street.'
The word next was mispronounced twice: One boy, who had called
it 'taxi' when reading it in phonoscript, again called it 'taxi.' The other
child gave the t the sound of th and quite rightly called the word 'nexth.'
It would be well to make another modification for the pronoscript th,
its identity with the printed form of t is a fertile source of error, es-
pecially in the earlier stages of reading.
Three children wrongly pronounced the word morning as against
one who was wrong in the phonoscript rendering; they appeared to
miss the phonoscript modification of the shape of the o. All of them,
very reasonably, gave the short o-sound; and two called the word 'mor-
row,' the other 'mŏning.'
Six children pronounced the word found as 'fond.' Five of them
did not audibly sound the letters; while the other one did so, quite cor-
rectly on the basis of his knowledge of phonoscript, sounding the o and
u as short; no child read this word wrongly in phonoscript.
With the word before, three children were incorrect; two of them
sounded all the vowels, quite reasonably giving them their common short
sounds, and called the word 'befrom.' The other gave the short sounds
to the first two vowels, silenced the final e (which he might not have
done, as it has no silencing mark in print), and called the word 'beyond.'
When read in phonoscript, there were two errors with the word
trouble; now there were five. Two children gave t its correct sound,
sounded all the letters, giving the vowels their common short sounds,
and called the word 'trunder,' which might be a look-and-say error, but
is more likely a 'shot' from the earlier sounds. Three children sounded
all the vowels as the child just cited had sounded them; but they were
misled by the t, which they sounded as th, quite reasonably. Two of
these called the word 'throuble,' and one called it 'throperly' (the word
properly had previously occurred in one of their reading lessons).
Three children were incorrect in reading the word look in ordinary
print; no one had pronounced it wrongly in phonoscript. One child
did not sound the letters; the others did as 'l-ŏ-k.' Both pronounced
it 'lock'—a very reasonable error.
The word time had been read correctly by every child in the phono-
script. In ordinary print there were two errors. The t misled the chil-
dren and they sounded it as th. Also, very reasonably, they gave the
common short vowel sounds to the letters i and e, and finally called
the word 'them' and 'thim.'
The word quite was found rather difficult in ordinary print, although
in phonoscript there had been no errors. Six children pronounced the
word incorrectly. All of them sounded the vowels short and two of
them sounded the printed t as th. These two called the word 'quith;'
and the other pronunciations were 'qut' (two) and 'qu - ĭt' (two). These
LOSS AFTER EIGHTEEN MONTHS
147
errors were reasonable for children taught to read by means of phono-
script. Curiously enough, though the final e was always sounded, it
was never pronounced as a syllable of the word—a procedure which
was correct, but unreasonable.
The word sure had given trouble, even in phonoscript. Four chil-
dren had been in error, and the number now rose to eight. Two chil-
dren did not sound any of the letters, and respectively called the word
'shiny' and 'shur' (the s of ordinary print is very, though not quite, like
the modified s of phonoscript which is sounded as sh). Six other chil-
dren sounded all the letters, giving the two vowels their common short
sounds, and one of these also sounded the s as sh. These six children
pronounced the word as 'shur,' 'surrěly,' 'sŭrě,' and 'Sunday' (this word
had recently occurred in their reading lessons).
Every child, when reading phonoscript, had pronounced the word
would correctly; but there was one error when the word was read in
ordinary print. The girl who was wrong sounded none of the letters,
and called the word 'wonderful,' a word of recent occurrence in the
reading lessons.
With the word evening, when read in phonoscript, four children
had been wrong. The number was now increased to five. One child
did not sound any of the letters, and called the word 'every.' The four
others sounded all the letters, gave the three vowels their common short
sounds, and pronounced the word as 'eving,' with the initial e as in
men, and as ‘even' and 'enving.'
Error occurred six times with the word they. One child mispro-
nounced it five times. She sounded the y as the long y in phonoscript,
which it resembles and said 'theï' each time. Another child without
sounding the letters, once and only once called the word 'then.'
In Test 2, written in ordinary print, 11 children made no
mistakes, against 14 in phonoscript. None of these 11 chil-
dren sounded any of the letters. There were 13 words in which
mispronunciations occurred, as against 8 when the same test was
read in phonoscript. The words mispronounced were through,
winter, bear, fail, they, same, time, each, evening, leave, very,
early, and morning. With four of these words, namely, winter,
fail, same, and they, errors in pronunciation occurred once only
for each word. A detailed analysis of the errors made in the
words wrongly pronounced would bring out points similar to
those which appear above in the analysis of the errors in Test 1.
I therefore omit, in the interest of brevity, the fuller treat-
ment of Test 2. I do this with some reluctance, for I am con-
vinced that teachers, who are working at the tool end of the
148
TEACHING BEGINNERS TO READ
educational machine, will never be convinced by statistical sum-
maries (whether they ought or ought not to be is another mat-
ter); and that, unless they can see for themselves how and
why children go right or go wrong, our researches will leave
them cold.
I think that supplemented to the extent that they are by de-
tailed analyses, our statistics will convince them that these young
children had made excellent progress in learning to read, and
that the loss on transference from phonoscript to real print at
this stage was very small. The latter point is the one at issue
just now. I do not assert that every teacher who teaches phono-
script about two hours a week for 18 months, to children be-
ginning to learn at five years of age, will obtain equal success.
The teacher who taught this group of children to read is one of
long experience and high capacity, with some 20 years' training
in the methods of experimental pedagogy. She knows, more-
over, how strictly to teach within the scope and up to the scope
of a systematic method; and there has been no “polishing up,"
outside the actual lessons, of backward children. What is as-
serted is, that with these poor children, under a first-rate teacher,
a phonoscript system has produced capital reading even when
the children are required to read from ordinary print.
But for there is a but-these phonoscript books had been
in the children's hands. They may have known this passage
exceptionally well; and if
so-well, if so, our conclusions fail.
So I am going to try them with a book which has never been
in the school, and this time from ordinary print first.
CHAPTER XII
AN EXPERIMENT TO CHECK THE RESULTS GIVEN
IN THE LAST CHAPTER
It was argued by some experienced teachers with whom I
discussed the results shown in Chapter XI that if the children
were allowed to read a passage in phonoscript first, and then, a
week or so afterwards, were required to read the same passage
in ordinary print, we were giving the amount of transfer an
unfair advantage; for, they said, the children might remember
the general sense of the passage and so be unduly helped when
they came to read it in print. With children of this age, I
thought that such an influence, if it existed (I was not wholly
prepared to deny its existence), would be very, very small. But
the argument should be met.
And, moreover, no completely adequate test of general power
to read can be made from a book which has already been in the
children's hands. So we decided to take further tests from
what would be called in educational phraseology an unseen
book. I selected a passage from Alice in Wonderland, a book
which was quite unknown in this school. The passage consisted
of 84 words, (exclusive of the word a), several of which had
never been read before. Since we wished to avoid the possibil-
ity of a fatigue factor-these were very young children, be it
remembered the passage was divided into two parts. After the
first 45 words, there was a slight break in the narrative; so we
used 45 words for the first test, and 39 words-the remainder-
for the second test. The first test was given on Friday, July 4th.
This was just one month after these children had worked the
first test in the series from Snow-White and Rose-Red as re-
ported in Chapter XI. Test 2 was given on Monday, July 7th.
Both of the tests were given in ordinary print.
On the following day Test 1 was read again, this time in
phonoscript; and on Friday, July 11th, Test 2 was read again,
but now also in phonoscript.
149
150
TEACHING BEGINNERS TO READ
The tests were administered and marked in the way described
in previous sections of this monograph; and it will be seen that
a two days' interval was given between Tests 1 and 2 in real
print, and between Tests 1 and 2 in phonoscript. The actual
tests follow. They are here given in ordinary type. It is to be
understood that they were also read in phonoscript type.
Group B.
Name.
Test 1.
Date.
Real Print.
Age.
There were doors all round the hall, but they were
all locked; and when Alice had been all the way down
one side and up the other trying every door, she walked
sadly down the middle wondering how she was ever
to get out again. (45 words.)
Group B.
Name.
Test 2.
Date.
Real Print.
Age.
Suddenly she came upon a little three legged table,
all made of solid glass; there was nothing on it ex-
cept a tiny golden key, and Alice's first thought was
that it might belong to one of the doors of the hall.
(39 words, exclusive of a.)
There were seven words in these passages which the children
had not read during any of their lessons; these words were
Alice, Alice's, wondering, solid, legged, locked, and key. Sev-
eral other words had only occurred once during the reading les-
sons; examples are nothing, middle, and glass.
In Test 1 there were 30 words which occurred only once.
Of the remaining 15, the words were, down, and, and she oc-
curred twice; the word all occurred three times, and the word
the occurred four times. In Test 2 there were 30 words which
occurred once: of the remaining 9, was, it, and the occurred
twice, and of three times. The word a was not counted. Table
XVI gives the individual results.
EXPERIMENT TO CHECK RESULTS
151
TABLE XVI. INDIVIDUAL RESULTS FOR THE PHONOSCRIPT group,
WHEN TESTED AFTER EIGHTEEN MONTHS' TEACHING
(First reading from real print, second reading from phonoscript print.)

POSSIBLE MARKS
FOR TESTS
45
39
84
45
39
84
AGE IN
REAL PRINT
PHONOSCRIPT PRINT
JUNE
NAME
1924
Test 1
Test 2
Test 1
Test 2
(Yrs. and July 4,
July 7,
Total
July 8,
July 11,
Total
Mos.)
1924
1924
Marks
1924
1924
Marks
G. F.
6
9
45
•
J. W..
7
3
H. H...
6
D. S
6
-10
•
E. C..
6
9
E. G...
7
L. J...
6
I. H..
6
D. G...
6
V. C...
6
8374&
9
E. D...
6
-10
A. G..
6
8
BA ABABB. AAGAA
39
84
45
39
84
36
28
45
39
45
39
40
35
40
30
45
39
42
36
45
39
45
39
40
31
J. N
6
40
31
M. B...
8
40
33
A. V...
6
36
30
FFFFF PRICE ZEN~9
45
39
84
45
39
84
64
39
34
73
84
45
39
84
84
45
70
84
45
84
45
84
45
73
66
GREGG GAGES &
39
84
43
36
79
44
37
81
39
84
45
38
83
39
84
39
84
45
38
83
44
36
80
44
34
78
45
37
82
Totals.... 101-10
629
527
1156
664
563
1227
Averages 6 - 9
41.9
35.1
77.1
44.3
37.5
81.8
CHILDREN WHO WORKED THROUGH ALL THE LESSONS AND TESTS IN READING, BUT
WHO DID NOT WORK THE PRELIMINARY TESTS IN SHAPES AND SOUNDS.

K. E...
I. Ho.
•
H. K..
E. R
W. H..
Totals....33
CON CÓ Ó
6 7
45
39
84
45
39
84
7
6
02
45
39
84
45
39
84
45
35
80
45
38
83
6
5
45
39
84
45
39
84
6
-11
36
31
67
44
36
80
1
216
183
399
224
191
415
Averages 6
7
43.2
36.6
79.8
44.8
38.2
83.0
In studying Table XVI, one must bear in mind the cases of
the children who had been absent from school for long periods.
These are noted individually in connection with Table XV. By
the 15 children who had worked through the whole of the ex-
perimental course, including the lessons and tests in associated
shapes and sounds, an average mark of 41.9 was obtained in
152
TEACHING BEGINNERS TO READ
reading Test 1 from ordinary print, and of 44.3 in subsequently
reading the same test in phonoscript. In Test 2, an average
mark of 35.1 was obtained when reading it in ordinary print,
and of 37.5 when reading it in phonoscript.
By the five children who had worked through the whole of
the experiment except the lessons and tests on associated shapes
and sounds an average mark of 43.2 out of 45 was scored when
working Test 1 from ordinary print and of 44.8 when reading
the same passage in phonoscript. The corresponding figures for
Test 2 were 36.6 and 38.2.
The average mark for both tests and both groups of children
in ordinary print is 77.7 out of 84 and for phonoscript is 82.1,
a differential advantage to the latter of 5.2 percent on passages
not previously read.
The "r" correlation coefficient between the total marks of the
two series is .75 for the first group, with a mean difference of
4.7, whose P.E. is .938.
The corresponding figures for the second group of children
are .966 for the "r" correlation coefficient, and 3.2 for the differ-
ence between the averages with a P.E. of 1.51.
As might have been expected, these children, though now
really very good readers of print for their age, still find some
advantage in the phonoscript rendering of unfamiliar words.
Some closer analysis of precisely how the children dealt with
the tests may be of service to teachers and to students of ex-
perimental pedagogy. Again, as in the tests given a month ago,
no child pointed to the words or letters while reading or sound-
ing them; and, once more, sounding the letters before pro-
nouncing the words was the exception rather than the rule.
When reading from print and sounding the letters before
saying the word, wherever double vowels or consonants oc-
curred, as, for example, oo in door and ss in glass, the children
sounded the first one only, omitting the other. This doubtless
arises from the fact that in the phonoscript system one of the
double consonants has always a silencing mark. This mark is
very easily perceived and remembered, and even in the very
EXPERIMENT TO CHECK RESULTS
153
earliest stages the children seldom failed to observe it and act
upon it; in the later stages they never failed to do so. This
marking is an admirable device in many cases; but I am doubt-
ful of its use for the breaking up of diphthongal sounds.
It was also noted that, when reading ordinary print, the chil-
dren frequently gave to the vowels in certain words their neutral
sounds as they would have done when reading from phono-
script; for example, to the a in was, and to the u in but. They
sounded both of these vowels like the a in amidst. In phono-
script these vowels would have had a modifying mark placed in
or on them; but in print, of course, they carried no such mark.
As stated before, there were seven words, and seven words
only, in the passages selected for the tests, which had not previ-
ously been read in any lesson-a testimonial to the compre-
hensiveness of the phonoscript system. This was found out
after the passages had been selected; the actual selection was
not influenced by this fact. These new words, when encountered
in print, were not found to be as difficult for children who had
learned phonoscript as might have been expected. Most of them
contained the common short vowel sounds, and the printed forms
and the phonoscript forms of these are alike. In the new word
Alice, for example, the i is short, and the soft c of phonoscript
and the c of ordinary print are very similar in shape. It is true
that the final e in Alice has a silencing mark in phonoscript and
not in print; but this appeared to give little trouble. The chil-
dren arrived at Alice from the sounds of the first four letters;
and knowing the word, they did not bother to incorporate the
e, though they had sounded it.
In the words solid, legged, and locked, which are likewise
new words, the vowels are also short. The word wondering also
occurred for the first time. The children had, however, already
read the words wonder and wonderful in their lessons, and also
the word wandering. Seven children, when reading this word
in ordinary print, called it "wandering." The a in wander is
modified in phonoscript so as to carry a short o-sound. Natu-
rally, the child, in sounding the letters of the word wondering
154
TEACHING BEGINNERS TO READ
in real print would call it "wandering." This is a case where a
phonoscript modification of the o would seem essential.
In Test 1, printed in ordinary type, there were eleven children
who gained full marks for the exercise, and none of these
sounded any of the letters before pronouncing the words. There
were 14 words out of 45 in which errors occurred. The words
pronounced incorrectly were: doors, round, hall, locked, Alice,
down, side, other, trying, every, walked, middle, wondering,
and again.
In Test 2, printed in ordinary type, there were ten children
who scored full marks, making no errors at all in pronunciation.
None of these children sounded any of the letters.
Out of a total of 39 words, there were 15 in which error
occurred, but with four of these words, mispronunciation only
occurred once; these four words were made, solid, glass, and
belong. The child who was in error with the word made audibly
sounded none of the letters and called the word "mad." He was
probably guided by his memory of the sounds of the first three
letters, but missed the phonoscript modification of the a which
would have given it a long sound as in face. The girl who failed
with the word solid sounded the letters correctly, but called it
“sodly" (sadly, but not sodly, had occurred in the reading les-
sons). The girl who was wrong with the word glass sounded
all the letters except the second s (the second letter of a double
consonant was very seldom sounded by these children), gave a
its long sound, and finally pronounced the word as "glās." The
child who mispronounced belong sounded all the letters quite
correctly, but called the word "below"-a quite unusual procedure
at this stage. In Test 2 there were 11 other words with each of
which there were two or more errors. These words were: three,
table, nothing, except, tiny, golden, key, Alice's, first, thought,
and might. The words which gave the most trouble in this read-
ing from ordinary print were nothing and except. The details
concerning these words are interesting.
nothing. Ten children were wrong,-a quite exceptionally large number.
This word had occurred once only during the course of the reading lessons,
and even then it had been divided so that 'no' stood on one line and 'thing'
on another—a reprehensible practice when printing books for beginners in
EXPERIMENT TO CHECK RESULTS
155
•
reading. The printed t gave the children the 'th' sound, as I have explained
before; for these children, with one exception, were carefully observing and
sounding the letters. They sounded all the letters correctly except the o.
To this they gave the short o-sound, probably because the printed 'o' is
precisely like the short o of phonoscript. Quite reasonably, they pronounced
the word as 'nothing' with the o as in not. The one child in error who did
not sound the letters-the other ten who did not sound the letters were all
correct-called the word 'mother,' which may have been a 'look-and-say' error.
except. Ten children were in error-again an unusually large number.
The c in ordinary print is rather more like the soft c than the hard e of
phonoscript, and this resemblance should have helped them when sounding the
letters. Two children, however, who sounded the letters called the word
'expect' and four children, who did not sound the letters, also called it
'expect. In both cases a transposition of the order of the sounds had
occurred an error, by the way, to be found whatever system of teaching
reading be adopted. Two children sounded the letters, giving the c its hard
sound, and called the word 'kept.' Mere length, as I have said before, is a
difficulty for very young children. Another child was misled by the curved-up
't,' which is very like the 'th' sound of phonoscript. Accordingly, he sounded
e-x-c-e-p-th'; but he called the word kept.' One other child, making the
fifth in error who did not sound the letters, pronounced the word 'express'-
a word which had occurred in their reading lessons in connection with an
express train.
I suggest that the most fertile source of error in this word
arises from the apprehension of the 'ex' sound, and thence a hasty passage
to a known word which contains it. Children, like the rest of us, when they
think they have enough to go upon, get on with it, erroneously or otherwise.
These children could evidently read print and read it well,
considering their age and the time they had spent in learning to
read. Some of them, a diminishing number, felt the loss of
the phonoscript modifications of the letters, for when they knew
they did not know a word (sometimes an essential piece of
knowledge) they fell back on sounding the letters-a procedure
which did not always help them. Ought they to do this?
In the first place, that is what they naturally and really do,
at least partially and unsatisfactorily, in alphabetic and word
systems; and the adoption of a phonoscript or other phonic sys-
tem only renders their procedure more explicit, systematic, and
accurate. But if any system of modified letters is adopted, the
sounding of ordinary printed letters will give rise to error as
well as to truth. The children will have something to unlearn,
it is agreed. But has not our experimental work shown that
these early aids are not as obstructive later on, as we adults
might, quite reasonably, suppose?
In the second place, the child finds out the words by a phono-
script or other phonic method; but he quickly ceases to rely on
his separate soundings and comes to know the word. True, he
156
TEACHING BEGINNERS TO READ
makes errors; and, when he does not know the word, he sounds
or may sound the letters of a system which, like print or print-
script or cursive script (handwriting), will not help him with-
out much chance of error. What is the teacher to do? Obvi-
ously, it seems that all she can possibly do is to cause the child
to revert, in the case of unknown words, to the strict system
of finding them out which she and the child already know. Such
is the mental constitution of all of us, children included, that
we do not go on sounding letters when we learn that it is safe
to say the words without doing so. That the adoption of such
a system does not delay the child but helps him rapidly to ad-
vance in dealing with real reading matter-not merely school
reading books—the whole course of our investigation goes to
show. But for unknown words, the teacher must, we think,
revert to her system, whatever it may be, phonoscript or
other phonic method, alphabetic method, look-and-say syllabic
method, or indeed, any method which gives the child, overtly
or covertly, the power to make himself at home in partially un-
familiar surroundings.
A word will now be said as to the results for the same
passages when read in phonoscript. In Test 1, there were 14
(against 11 in print) out of the 20 children who made no error.
Twelve of these did not sound the letters of any word. One
child correctly sounded the letters of Alice, walked, and wonder-
ing; while another child sounded correctly the letters of the
words locked, Alice, trying, sadly, and middle, in each case duly
observing the silencing marks.
Out of a total of 45 words, there were only 7 which were
pronounced incorrectly by any child. These were locked, Alice,
every, walked, wondering, ever, and again. Five of these words
were mispronounced only once, but Alice and wondering, both of
them new words, produced more frequent errors.
In the phonoscript form of Test 2 there were ten children
who made no error, and these were the identical ten children
who made no error in ordinary print. None of these ten chil-
dren sounded any of the letters before pronouncing the words.
EXPERIMENT TO CHECK RESULTS
157
Out of the total of 39 words, nine were wrongly pronounced
against 15 in print. These words were three, nothing, except,
golden, Alice's, first, thought, might, and belong. With five of
the words, an error in pronunciation occurred once only. These
words were: three, Alice's, thought, might, and belong. As
was the case when this passage was read in ordinary print, the
words except and nothing gave the most difficulty. With the
word except, nine children were wrong as against ten when the
reading was from print. Whether they sounded the letters or
not, their prevalent error was "expect." With nothing five chil-
dren were in error-just half the number who mispronounced
it in ordinary print. The prevalent error (3 times) was
"nothing."
We must, I think, fairly conclude that these children have
learned to read-a conclusion, however, which I do not wish
to generalize; for their teacher was an extremely good one.
But I do wish to generalize the conclusion that children who
have learned phonoscript at this age and to this degree can read
ordinary print with but little loss.
CHAPTER XIII
FINAL TESTS IN READING
There comes a time in the progress of every systematic
method of teaching reading when its special method and its
special books may and should be largely given up. It is no longer
so vitally necessary to have carefully graded books nor any special
print or diacritical marks, if any such have been employed.
Doubtless from time to time, the teacher must revert to the
method by which the children have been originally taught. When,
for example, they are faced with an unusual or very difficult
word, recourse must be had to their system, just as we older
children sometimes refer to a dictionary for correct pronuncia-
tions. But, in the main, the children now read from ordinary
reading books and ordinary print. And of course, no child who
can read only with the help of a special system can really be
said to have learned to read.
What of our competing methods, phonic versus phonoscript
in this, their final stages? In discussing our experiment, some
of my expert helpers conceded that I had shown the superiority
of the phonoscript method as compared with an ordinary phonic
one at certain stages. Yet they pointed out that I admitted, nay
asserted and proved, that there was loss when children, taught
by phonoscript, were required to read from ordinary print. Why
then, they asked, should not the loss in transference at least equal
the gain by means of phonoscript. For, of course, no transfer
is required from the phonic system to ordinary print; the chil-
dren have been reading it all the time. I tried to show them,
by elaborate calculations, that phonoscript was really the victor.
They were not convinced that I had proved it; and they retorted
with a weapon from my own armoury, by saying, "Try it and
see." So I tried it.
158
FINAL TESTS IN READING
159
THE CHILDREN WHO DID THE WORK IN THE FINAL TESTS
(PHONIC VERSUS PHONOSCRIPT)
The children who took these final tests were those who, two
years before, in November, 1922, had been divided into two
equal groups on the results of tests of their ability to learn and
remember certain arbitrary and arbitrarily associated shapes and
sounds. Each group was subsequently taught reading by dis-
tinct methods which have already been described. In addition
to these children, there were ten others in the same class who had
worked through the whole course of reading, but who had been
away from school during the whole or part of the time the pre-
liminary lessons and tests in shapes and sounds were given. They
were paired, five in each group, on the teacher's estimate of their
ability, and their work has been, and will continue to be, shown
separately.
During the progress of a long experiment-in this case, two
years the number of children originally paired, through re-
moval, sickness, and other causes, generally decreases. It was so
here; and moreover, in order to utilize adequately the accommo-
dation of the classroom, children newly admitted to the school
were included in the class from time to time. As the newcomers
were placed in the class, they were divided equally, as far as
numbers went, between the two reading groups. Though, from
their time of entry, they went through the same courses of read-
ing as the others, they had not worked through the whole course;
and the results of their work are, therefore, not tabulated. At
the time of the final testings, 54 children, boys and girls, were
on the roll of this class. Of these, 17 were children of Group A
(the original phonic group) and 20 of Group B (the original
phonoscript group). To maintain the original balance of the
groups, since three children in the phonic group had now left the
school, the work of the three children in the phonoscript group
who were originally paired with them, though it will be shown,
will not be counted for statistical purposes.
When the experiment began, these children were in a class
technically known as Grade II, a class in which the children
160
TEACHING BEGINNERS TO READ
were by no means mentally homogeneous. Owing to the
European War, and pressure on our school staffs, one room in
the Infants' Department had been closed. This left one room
only for the Grade II class, and it could not be divided into an
upper and a lower section, as had usually been the case in this
school in the second half of the educational year. I am making
these conditions clear, because an altogether false impression will
be obtained about these experiments and their results, if it sup-
posed that we were dealing only with specially selected children
under a specially selected educational environment. These
Grade II children gradually, as usual, proceeded through the
school, moving in April, 1923, into a Grade III class and in
April, 1924, into a Standard I class, in which class they were,
when these final tests were done in November, 1924. I said these
children proceeded normally through the school; but six of
them did not. In April, 1923, six children were not put for-
ward into Grade III. They had not been successful in their
usual school work; and, be it remarked also, they had failed in
their reading and in our original tests in associated shapes and
sounds. Of those originally paired on shapes and sounds the
average age in Group A (the phonic group) at the date of these
final tests was 7 years 3.3 months, and that in Group B, 7 years
3.9 months. Of the additional children (paired on the estimate
of their teacher) the average ages were, in Group A 7 years 1.0
months, and in Group B 7 years 1.2 months.
HOW THESE CHILDREN HAD BEEN TAUGHT
From November, 1922, to the end of October, 1923, the head
mistress taught both the phonic and phonoscript groups by
methods which have already been described. At the latter time—
the beginning of the second half of their Grade III year-the
phonoscript group, as before, continued to be taught by the head
teacher, but the phonic group was henceforward taught by the
usual teacher of the class. She was a very good teacher, indeed,
thoroughly experienced and very conversant with the phonic
method in use.
FINAL TESTS IN READING
161
This arrangement was not forecasted or desired by me, but
the many calls on the head mistress at a time of depleted school
staffs rendered it inevitable. I am of opinion that the work of
the phonic group did not suffer from the change; and that we
now had, in place of equal groups of children, equal conditions
and the same teacher, equal groups, equal conditions, and equal
teachers, as far as phonoscript versus phonic is concerned.
Teachers who have read this monograph so far will in-
evitably say, "You have told us about the reading of these two
groups: we understand that no other reading and spelling was
done except that involved in the reading of the two groups in
their scheduled lessons; but what about the writing and the
English composition?" Well, no writing was done at all until
October, 1923. Then writing in the form of print-script, which
is neither real print nor phonoscript nor cursive-script (hand-
writing), began to be taught to all the children in the class by
the usual teacher. When some facility in this had been acquired,
simple exercises in dictation and English composition were given.
In dictation lessons there was no spelling. Corrections were
made in print-script on the blackboard. In the early exercises in
English composition-quite wisely in any case-no orthographi-
cal corrections of any sort were made; but in the later lessons
the composition papers were corrected by the teacher who wrote
in print-script on the children's papers the correct form of the
words which were spelled wrongly. We have, I think, as far
as is humanly possible, maintained equal conditions for the two
groups. If we have not, unless we have favoured one whole
group at the expense of the other, our figures will jump and
jerk irregularly, which they do not. One other point: from
November, 1922, the reading lessons were 20 minutes long, four
a week to each group, that is, one on each school day excluding
Monday. From the end of October, 1923, five lessons were given
of 20 minutes each, Monday now being included, and from the
end of April, 1924, (the commencement of the Standard I year),
the duration of the lessons was raised to 30 minutes. Reading
lessons were entirely omitted when the head teacher was con-
ducting the terminal examinations of her school.
162
TEACHING BEGINNERS TO READ
THE FINAL TESTS
It will be remembered that tests have already been given to
these particular children for various purposes during the prog-
ress of the experiment. These final reading tests were given
on Tuesday, Wednesday, and Thursday, November 11th, 12th,
and 13th of the year 1924, a little less than two years after the
experiment began with this class.
We decided (1) that each child should now read three tests;
(2) that the tests for both groups should be printed in ordinary
print, and not in phonoscript, even for the phonoscript group;
(3) that one test should be selected from a book which only
Group A-the phonic group-had read; (4) that one test should
be selected from a book that only Group B-the phonoscript
group-had read; and (5) that one test should be taken from a
book which neither group had read. The three tests follow.
TEST I
(Fifty words chosen from a phonic book)
She was looking out of the window, when Jack
passed with the donkey on his shoulder; and the
donkey's legs were sticking up in the air. It looked
so very funny, she burst out laughing and was able
to hear and speak. Her father sent for Jack and
told him.
TEST 2
(Fifty words chosen from a phonoscript book)
Little Three Eyes had watched with her one eye
all the time, but Little Two Eyes did not know.
She felt quite sure that her sister was fast asleep.
So she ran to her and said, "It's time to go home.
Wake up, you have had such a long sleep.'
TEST 3
(Fifty words chosen from Alice in Wonderland)
And here poor Alice began to cry again, for she
felt very lonely and low-spirited. In a little while,
however, she again heard a little pattering of foot-
steps in the distance, and she looked up eagerly, half
hoping that the Mouse had changed his mind and
was coming back.
FINAL TESTS IN READING
163
The tests were administered and marked in ways that have
already been fully described. One test a day and one only was
given to avoid a fatigue factor, and paired children from each
group read successively. Only correct pronunciations carried
marks; but a complete record was made of everything each
child did, as in previous testings.
THE FINAL RESULTS
It was possible for each child to obtain 50 marks for each
test, 150 in all. The children paired originally on shapes and
sounds made the following records.
On Test 1, from the phonic book, Group A (taught phonetic-
ally) obtained an average mark of 44.9, and Group B (taught
by phonoscript) an average mark of 46.3. On Test 2, from the
phonoscript book, Group A scored an average mark of 44.4,
whilst Group B obtained an average mark of 46.8. On Test 3,
from Alice in Wonderland, Group A gained an average mark
of 40.8, and Group B an average mark of 44.8. The sum of
the marks gives the phonic group an average of 130.1 and the
phonoscript group an average of 138.0.
The additional children in Groups A and B (paired on the
teacher's estimate) scored as averages: on Test 1 A, 48.6 and
B, 48.8; on Test 2 A, 48.2 and B, 49.0; on Test 3 A, 45.8 and
B, 46.8. Averages for three tests totaled: A, 142.6; B, 144.6.
The highest mark in Group A, the phonic group, was 150
scored by one child; the lowest mark in this group was 109.
The highest mark in Group B, the phonoscript group, was like-
wise 150; but it was scored by four children. The lowest mark
in the group was 113.
It is apparent, therefore, that the phonoscript group has been
victorious all along the line. Not only have they read better
when the material was taken from their own books, but also
when it was taken from the phonic books and from Alice in
Wonderland. Moreover, the print they read when they took the
test was not their own phonoscript print, but real print, which
they had never read at all except in two or three tests previously
164
TEACHING BEGINNERS TO READ
S. W..
H. C
E. H...
R. S..
C. W.
•
given at various stages in their progress. I know we have
already shown that the loss on transference to real print from
phonoscript diminishes as time goes on; but we were hardly
prepared for so complete a victory.
True, the amounts by which the phonoscript group has won
are not great,¹ but they are consistent, and indeed, become con-
TABLE XVII.
THE INDIVIDUAL MARKS ON THE FINAL TESTS
IN READING

Name
Age 11-30-1924
(years, months)
Test 1 (Max. 50)
Test 2 (Max. 50)
Test 3 (Max. 50)
Total Marks
(Max. 150)
Marks for Shapes
and Sounds
(Max. 48)
Name
Age 11-30-1924
(years, months)
Test 1 (Max. 50)
Test 2 (Max. 50)
Test 3 (Max. 50)
Total Marks
Marks for Shapes
and Sounds
(Max. 48)
(Max. 150)
7.5
42 38
37
117
38
G. F.
7.3 50 50 49
149
36
D
•
7.6 50 50
49
149
34
J. W.
7.9 50
50 49
149
35
7.5 43 44 40
127
34
H. H
•
7.0 42 143 39 124
34
D. S..
•
16.7
50 148 43 141
30
E. C……….
7.2 36
7.4 50 50 50 150
7.350
39 138 113
34
33
50 49
149
33
R. D...
7.1 50 50 46
A. D....
A. Na.
7.11 37 38
འཚ
146
30
E. G...
8.2 40 44 39
123
31
34
109
30
L. J
6.9 45 44 143
132
29
•
7.2
•
•
•
Left the
school
26
I. Ha.
7.1 50 50 48
148
28
G. T.
7.5 42 41 35
118
26
D. G
•
•
•
•
·
6.10 48 47 43 138
25
L. S..
·
•
6.9 47 46 41 134
24
V. C.....
7.3 50 50 50
150
23
•
6.7
Left the school
24
K. D..
7.4 50
50
49
149
23
•
17.6
Left the school
24
A. G.....
7.2 46 45
39
130
23
• •
7.4 43 42 38 123
19
J. N..
6.1146 46 42
134
18
·
•
• •
•
7.6 50 50
48 148
17
M. B.
· •
8.7 50 49 46 145
18
•
·
•
15
A. V...
6.6 41 43 40 124
16
Averages 7.39 46.3 46.8 44.8138.0
J. R.*.
A. C..
A. J.
R. D
P. McI... 17.4 43 43 40 126
Averages. [7.33 44.9 44.4 40.8130.1
F. T.
G. T..
A. G..
W. W
•
J. B..
•
ADDITIONAL CHILDREN

7.0 50 150 48 148
7.8 46 47 41 134
7.2 50 50 50 150
7.5 47 44 43 134
6.2 50 50 47 147
Averages. [7.10 48.648.245.8[142.6|
K. E....
7.1 50 150 50 150
•
•
I. H..
H. K.
E. R..
• •
6.8 50 50 46 145
6.11 50 49 46 145
7.5 50 50 50 150
W. H... 7.5 44 46 42 132
Averages
•
71.2 48.8 49.0 46.8|144.6|
a The figures for these pupils were omitted in the computations.
¹I have, on previous occasions pointed out that huge and oscillating differ-
ences are always suspect; they are mostly due to intermissions of attention,
and unsatisfactory conditions of experimentation.
FINAL TESTS IN READING
165
siderable when the marks of two children who had been in-
capacitated during the course and remained partially so—with the
marks of their paired associates-are excluded.
Table XVII summarizes the individual results.
Omitting H.H. and E.G., who had been hors de combat and
remained partly so, with their paired associates E.H. and R.D.,
the "r" coefficient between the two series of preliminary shapes
and sounds is .99. That between the two series of marks for
reading from the phonic book is .53 with a difference between the
means of 3.4, whose P.E. is .78. The "r" coefficient between the
two series of marks for reading from the phonoscript book is
.48; the mean difference is 4.0 with a P.E. of .80. The "r"
coefficient between the two series for reading from Alice in
Wonderland is .39, the difference of the averages of the two
series is 5.7, with a P.E. of 1.0. For the totals, the figures are
.47; the mean difference is 13.1 with a P.E. of 2.5. That
is to say, on every test the phonoscript group gained significantly.
Not only did they read better from Alice in Wonderland, and
better from the phonoscript text, but they also read better from
the phonic text. And in all cases be it remembered, they read
from ordinary print, not from the phonoscript printing they had
learned.
r
One further interesting statistical fact must be referred to.
It will be seen from the table of individual results that, of the
additional children, i.e., those paired on the teacher's estimate,
the phonoscript group has also been victorious, but it is not a
valid victory. The mean differences are low, and their P.E.'s
high. Moreover, the correlations between these two series are
now all negative. They are -.41, —.17, and -.31 for the three
tests respectively and for the totals, -.28. That is, while the
correlation coefficients between the two series originally paired
on tests on shapes and sounds remain, after two years, quite
significantly positive, those between the children that were paired
on the teacher's estimates-a first-rate teacher too-have become
negative. Apparently we shall have to attach little importance
to estimates of any kind, when precision is required over a long
period.
166
TEACHING BEGINNERS TO READ
OBSERVATION OF THE TEACHER WHO CONDUCTED THE
FINAL TESTS
Whilst the teaching of the neutral vowel sound in the early
stages of teaching young children to read is legitimately open to
criticism, it must also be stated that when they can read, as is
now the case, their reading is most clear and natural. It was
noticeable during these final tests that the phonoscript group
retained the use of the neutral vowel when pronouncing words,
even though they were not now reading from phonoscript print.
The reading of these children sounded more distinct and fluent
than that of the phonic group. The method does, indeed, make
for exactness all round. A short illustration, not a proof, may
be given. The teacher of the class sent two children with a
message to another teacher. A return message was brought and
the child in the phonic group said, "Teacher says, "Thank yer.
The second child, who belonged to the phonoscript group, turning
to the first, said, "Teacher did not say, "Thank yer,' she said,
"Thank you."
CHAPTER XIV
FINAL TESTS IN SPELLING
A question repeatedly asked by teachers in senior depart-
ments, when I have described look-and-say, phonic, or phono-
script methods of teaching reading to very young children, is
usually phrased somewhat thus: "Yes, that may be all very
well, but what about the spelling?" It will be remembered
(see Chapter VIII) that the effects of phonic and phonoscript
methods on spelling were tested and compared after some twelve
months' teaching of two groups. Now, towards the end of the
course, a little before the tests in reading described in the last
chapter, the general results and the comparative results on spell-
ing of the two methods were again tested.
First of all, let me say that the phonic and phonoscript
methods of reading are also spelling methods. Any method
which requires successive attention to the letters of a word or
syllable is a spelling method. There are successive visual per-
ceptions of the letters of the word. These letters may be
sounded, as in a phonic or phonoscript method (excluding silent
letters), or they may be named, as in an alphabetic spelling
method. This successive visual perception, whether accompanied
by the sound values or names of the letters, is spelling. Ridicule
has been cast by people who ought to have known better on
alphabetic spelling, because it does not accomplish what it does
not set out to do. Experienced teachers maintain unanimously
that alphabetic spelling does produce good reading and does pro-
duce correct spelling, which latter, they assert, no phonic or
phonoscript method does or can do.
I conclude, on the previously given experimental evidence,
that an alphabetic reading method does not give us such good
early, reading as a phonic or phonoscript method. The question
of its comparative effects on spelling is reserved for further
treatment, and is not dealt with in this monograph.
167
168
TEACHING BEGINNERS TO READ
But there is one issue that must be faced in connection with
reading. How can one learn the sounds of the letters when only
the names are learnt, not the sounds? Does "a" always sound
the same? Is it not sometimes "ă," sometimes "a," sometimes
"ä," and sometimes other sounds? If we teach by names, not by
sounds, the child must just inaccurately and unsystematically
fudge out the sounds for himself. That the child does, as we do,
get some notion of sounds from alphabetic spelling we all know;
his mistakes show that he does (we call them phonetic errors).
But the object of alphabetic spelling is not to teach a child to read
by the sounds of the letters at all. It does not aim at an analysis
of the sound of the word or syllable into its literal components,
admittedly not a very easy operation, quite apart from irregu-
larities. It aims to establish a series of reversible associations
between the letter series and the sound of the word as a whole.
B-a-t is to be associated with the sound of the whole word or
syllable "bat." C-a-u-g-h-t is to be associated with the sound of
the whole word "caught." And these associations are—with the
exceptions of the very few words with different meanings but
sounded alike, like "their" and "there"-really reversible. No
phonic system, with the exception of the phonetic system of the
International Phonetic Script, can give us this inevitable re-
versibility; and many, if not all, of the errors in spelling aris-
ing from all phonic methods are due to this defect.
But there is spelling of some kind in both our phonic and
phonoscript methods; let us see what it is. In our phonic
method each letter of the word is sounded once in succession-
methods which pick out the middle letters first are fatal to spell-
ing. In some cases two letters are sounded together as ch in
church, sh in she, au in caught, and silent letters have been under-
lined during the reading lessons. In the phonoscript method, the
first letter of the word is sounded; then the second, then the
first and second; then the third separately, then the first, second,
and third; and so on. This rather elaborate and repetitive
method may prove of advantage in subsequent spelling as well
as in the building up of the word from its component sounds.
FINAL TESTS IN SPELLING
169
As has been shown in the previous chapter, no spelling other
than that involved in the reading lessons has been taught in this
class.
THE TESTS
The final tests in spelling, now about to be described, were
given on Tuesday, Wednesday, Thursday, and Friday, then on
the following Tuesday and Wednesday from September 9 to 17,
1924. After one week's interval, two further tests on the same
day, Wednesday, September 24, were given. Fifty-two children
worked the tests; but, as before, we shall only tabulate the work
of our groups.
Since the books which have been read by our two groups are
not identical, and since at a certain stage in our experiment the
reading matter diverged, we are faced with a difficulty if we
endeavor to try a nice issue by means of a general test. For it
may be claimed, in fact it is claimed, that the words of a general
test may, and often do, favour one method rather than another.
Against Ballard's One-Minute Reading Test it was urged that
it was specially favourable to the children who had been taught
by a look-and-say method, and that the highly specialized Dale
(phonic) system did not have a fair chance. In actual fact,
his phonic readers did better, but the argument contains a truth
relevant to the testing of any methods in any school subject by
a general test. For a rough estimate of the work of a class in
any school subject, this consideration is not of grave importance.
But for careful work on the evaluation of different methods it
is a serious consideration. How can we allow for it? We can
try to frame a general test which will not favour one method
rather than another, and we can test each group on the matter
dealt with by its own method and also on the matter dealt with
by the competing method. And, further, we can arrange our
general test so that it contains matter not already dealt with by
either group in any form. All these things we have done in the
course of the present research, and we shall adopt the same pro-
cedure now in these, our final tests in spelling.
170
TEACHING BEGINNERS TO READ
The tests and what they comprise follow:
1. A passage from the phonic reader then being used by the phonic group.
2. A passage from the phonoscript reader then being used by the phono-
script group.
3. A passage from a standard work neither read nor previously seen by
either group.
4. A specially compiled test. Since we felt that the first three tests were
somewhat difficult for these children, and rather outside the range of the
usual spoken and written vocabulary (that used in English written composition)
I compiled a fourth and easier test.
TEST 1
(From the phonic reading book)
Over their heads and down the backs of their
necks they wear a white cloth. Unless they wear
these cloths, the great heat of the sun will hurt
them. All the land is not desert. There are places
where springs of water are found. In these places
grass and trees grow.
(Fifty words, “a” not counted)
TEST 2
(From the phonoscript reading book)
Before long he met an old woman with a bundle
of sticks on her back. Piggy looked at her. Then
an idea came into his head. He would like those
sticks to build a house with. So he said to the old
woman, “Please stop one moment. I want to talk
to you."
(Fifty words excluding two "a's”)
TEST 3
(From Alice in Wonderland)
She got up and went to the table to measure her-
self by it, and found that, as nearly as she could
guess, she was now about two feet high, and was
going on shrinking rapidly. She soon found out
that the cause of this was the fan she was holding,
and she dropped it hastily, just in time to avoid
shrinking away altogether. "That was a narrow
escape!" said Alice.
(Seventy words including “a”)
FINAL TESTS IN SPELLING
171
TEST 4
(Specially compiled)
Tom saw a pretty black cat on the wall. He
called her and his mother gave her a piece of his
bread. She would not eat it. So the boy gave her
some meat and milk. Pussy drank the milk but left
the meat. Then she lay down to sleep.
(Fifty words including two "a's”)
HOW THE TESTS WERE ADMINISTERED
These tests are lengthy for young children; and since we
wished to avoid the possibility of a fatigue factor, we divided
each test into two sections. The breaks were made at natural
stops. Thus we had eight separate exercises in dictation each
worked by every member of the class at the same time. The
tests were written with pencil. The children could write with
ink, but they wrote much more readily with pencil. First the
teacher read the passage to the class and then she dictated it,
one word at a time. No child was hurried.
METHOD OF MARKING THE TESTS
For every word correctly spelled one mark was given. When
one word occurred more than once a mark was given each time
it was correctly spelled. Capital letters and stops were disre-
garded. Table XVIII gives the individual results. A brief
inspection of the table shows that both groups have failed in
spelling. This result is in striking contrast with the excellent
marks which were shortly afterwards obtained in reading and
which were given in Table XVII. Both groups have learnt to
read and to read extremely well; neither group has learnt to
spell in any way comparably.
The "r" coefficients were all worked out: they were slightly
positive in nearly all cases, but so small as to be very doubtful.
The phonoscript group has, test by test and section by section,
beaten the phonic group; but it is a victory which is not statis-
tically significant. Adding the marks for all the four tests to-
172
TEACHING BEGINNERS TO READ
gether, we have a mean difference in favour of the phonoscript
group of 12.9 (with a P.E. of 6.8) in respect of the groups
TABLE XVIII.
XVIII. FINAL RESULTS IN SPELLING

Initials
S. W.
H. C.
•
•
E. H.
R. S...
C. W...
R. D.
A. D..
A. N.
G. T.
•
L. S...
J. R..
A. C.
A. J.
•
•
•
R. D..
P. McI.
Averages..
F. T....
G. T.
A. G.
•
W. W
J. B..
Averages..
GROUP A (PHONIC)
Test 1 (Max. 50)
Test 2 (Max. 50)
Test 3 (Max. 70)
Test 4 (Max. 50)
Totals (Max. 220)
Initials
GROUP B (PHONOSCRIPT)
Test 1 (Max. 50)
Test 2 (Max. 50)
Test 3 (Max. 70)
Test 4 (Max. 50)
Totals (Max. 220)
19
16 125 26
86
G. F
•
32
25 45 140
142
J. W.
·
22
15 37
31
105
H. H.
26
•
•
26 17 40
30
113
D. S....
38
•
27
16 29
26
98
E. C....
29
·
33
26 148
41
148
E. G.
•
16
11 23 23
73
L. J.
20
·
Left the school
J. H.a
38
•
15
9 21 25
70
D. G..
•
•
21 19 133 28 101
V. C.
36
27282 22888
42 30
33333
50
42
164
37 32 44
41
154
15 29 21
91
27
44
40
149
24
24
41
37
131
15 133 26
96
16 24 31
91
32
51 41
162
24 35 27
114
28 46 37
147
•
Left the school
E. D.&.
137 31 149 41
158
•
2823
Left the school
19 34 32 107
28 54 40
A. G.a
23 16 32
24
95
•
•
J. N...
25 15 36 28
104
=
160
26 44 34
136
M. B...
A. V
25.218.9 36.131.3] 111.6
Averages..
35 25 42 35 137
22 21 42 31 116
[30.0|22.738.8 33.0 124.5
ADDITIONAL CHILDREN

42 37 51
42
172
K. E...
41
•
•
•
136 49 41
167
23 19 35 132
109
38 36 54 45
173
I. Ho.
H. K.
48
41 58 48
195
•
•
30 22 37 33
122
·
·
·
25 20 37 29
111
E. R..
37 26 47 37
147
•
•
28 20 38 29
115
W. H..
26 22 39 32 119
Averages..
36.4 29.446.038.2 150.0
&
31.2 26.4 43.0 35.4 136.0
Figures for these pupils were not used in the calculations.
originally paired on shapes and sounds; and for the additional
children, an average difference in favor of the phonoscript group
of 14.0 (whose P.E. is 13.4) with the "r" coefficient negative.
FINAL TESTS IN SPELLING
173
OBSERVATIONS OF THE TEACHERS WHO MARKED THE PAPERS
The teachers were distressed at the results; and we all went
through the papers very anxiously. The head mistress, who
made a careful and protracted study of the children's work, thus
concluded. Where the children could not perceive and associate
each sound component of the word with one definite letter they
had almost completely failed to spell at all. I subsequently went
through the papers, and concurred in her judgment. All that we
can say is that the phonoscript method appears to be a little better
than the phonic; but that the advantage cannot be counted upon.
The usual judgment of teachers, that no phonic method (in
which, of course, phonoscript is included) teaches a child to spell
decently, is thoroughly substantiated by our results. In two
years, these two groups of young children have learnt to read
extremely well-but they have yet to learn to spell.
CHAPTER XV
SUMMARY AND SOME IMPORTANT CONCLUSIONS
Over a period extending from 1905 to 1924, observations
and experiments have been made in English schools on the
methods and results of teaching beginners to read. After a
brief introduction on the present educational attitude toward
early reading, some observations are described on the compara-
tive results of phonic, alphabetic, and look-and-say systems of
beginning to learn to read. The work of other English observers
and experimenters is then referred to. The inconclusiveness of
the observations and experiments described drives us on to more
systematic and controlled experiments under exact conditions,
which are still the actual conditions of school life.
It is maintained that special tests of potentiality to learn
reading are desirable if equal groups are to be formed for trying
out various methods of learning to read. These tests are in-
vented, described, and applied, and found to work satisfactorily.
Equal groups of young children are formed, that is, equal in
their potentiality to begin to learn to read. Phonic and look-and-
say methods are put in practice and compared. The result is a
victory for the phonic method. A phonic and an alphabetic
method are compared. The result is a victory for the phonic
method. The most recent phonic system, Mr. Hayes' Phono-
script, is compared with an ordinary phonic system; the result
is a victory for the phonoscript method.
Phonoscript involves modifying marks in or on ordinary
literal symbols, and its value for correct spelling in its very early
stages is found to be rather less than that of an ordinary phonic
system. It is also found that serious loss occurs when children,
taught by phonoscript, are required to read from ordinary print.
This loss is checked by an experiment with a new class, and
again found to be considerable. Then the loss on transference
from phonoscript to real print is tested after eighteen months'
174
SUMMARY AND CONCLUSIONS
175
teaching and is still found to be statistically significant, though
much reduced. A check experiment follows to substantiate these
results; and it is suggested that the curves for reading phono-
script and real print, by children taught by means of phonoscript,
approach each other asymtotically. Then final tests in reading
are given to two equal groups of children who have been learn-
ing to read by a phonic and phonoscript system respectively, and
it is found that the phonoscript group is superior when tested
from a book which is literature, from a phonoscript reader, and
also from a phonic reader.
Comprehensive tests in spelling are again given to the same
two groups which have made such excellent progress in learning
to read. Both systems have failed, but the phonoscript group
is not now worse than the phonic; it is perhaps slightly better.
In two appendices, some criticisms of and suggestions for the
improvement of phonoscript are made.
This monograph, which began as a survey of the relative
success of phonic, look-and-say, and alphabetic methods of teach-
ing reading in its very early stages, becomes, in consequence of
the superiority of phonoscript readers in early stages with their
own form of print, an evaluation of this system as against an
ordinary phonic method.
Some important conclusions follow from the experiments
described:
1. That a division into equal groups on the basis of asso-
ciated shapes and sounds is justified pragmatically by the results,
as well as by a theoretical analysis of the psychological founda-
tions of early reading; for the statistical correlation in reading
between the groups remains adequately high for considerable
periods of time-in one case for two years.
2. That a system of teaching reading which requires modify-
ing marks in or on the letters (provided it does not, like the In-
ternational Phonetic Script, modify them out of recognition) is
amply justified by results in reading, both from its own modified
print and from ordinary print.
176
TEACHING BEGINNERS TO READ
3. That, from an analysis of errors made by children, it
appears that they are guided in their pronunciation of words
by a more or less accurate recognition of some literal shapes
and sounds; they see a letter or letters whose sounds they know
and "jump" to the nearest word they know; they do not learn
by a look-and-say method, that is by the general appearance of
the word.
4. That no phonic method (including phonoscript) produces
adequate correctness in spelling.
5. That one form of print, namely, the print from which the
child is learning to read, should be taught both for reading and
writing in their early stages, say for a year or so. This would
mean “real print" for children learning on a phonic, look-and-say,
or alphabetic system, and phonoscript print for children learning
by the phonoscript system. It would not mean print script now
taught in so many infant schools. The latter is a cross between
printing and writing.
6. That English children, under our conditions of school
organization, may quite profitably begin definitely to learn to
read at an average age of 5 years 3 months the normative age
in London for the commencement of the second half of our
Grade II educational year. In schools where the children are
well-born and of good home environment, they may begin
earlier; in schools attended by very poor children, later.
APPENDIX I
SOME SUGGESTIONS MADE BY TEACHERS AND BY
THE WRITER IN CONNECTION WITH
PHONOSCRIPT
AS TO THE ORDER IN WHICH THE SYMBOLS ARE TAUGHT
1. Letters are introduced-for example, i, e, g, w, u, k, and
y-upon which silencing marks are placed, before their unmodi-
fied shapes and sounds have been dealt with. May not this in-
duce a habit of looking only at the silencing marks? It appears
to do so. A child may note that the letter is silent but not which
letter it is that is silent, and so acquire an attitude of mind
which will not be helpful in his spelling, either in phonoscript
or in ordinary print.
2. Again, is it not desirable to teach the unmodified shapes
of the letters first? For example, the shape and sound of the
letter o which is first taught requires two modifications and car-
ries the sound of o in the word some, like the u in sum. To
exemplify the use of this symbol, the words son and wonder are
taught; and these are not easy words for young children. The
o symbol which is next taught requires one modifying mark.
The effect of this mark is to lengthen its sound, and to give it
the o-sound as in no. If, instead of beginning with these forms
of o, we began with its very common short sound, as in pot, no
modification in shape would be required when o was first pre-
sented. The letter has its most usual sound in early reading, and
several favorite words, the meanings of which are well known
to young children, could be easily learned. Likewise, if the short
sound of i, as in the word in, were taught first, no modifying
mark would be needed. As a matter of fact, the long sound of
i as in the word find, which requires a modifying mark, is given
first.
The short sound of a, with a very simple form of its literal
symbol, quite rightly in our judgment, is given first; then that
177
178
TEACHING BeginnerS TO READ
form of a which has a simple modification carrying its sound in
the word amidst. The third form of a is then more complicated.
It involves two modifying marks, one of which is a little ring
inside the body of the letter. It carries the sound of a in the
word call. This arrangement leaves the long a-sound as in made,
which is very common and which requires only one modification,
till later.
The first s to be taught is not the simplest form and sound
of the letter, but one which is modified so as to give it its flatter
sound, as in the word rise. The unmodified s might well be
taught first. It is the commoner and simpler shape and carries
the more usual sound for early reading.
The u-sound as in blue is taught first with a modified shape
of the letter, instead of the simpler, unmodified form of the let-
ter which is sounded as u in run.
The go first taught also has a modifying mark. It is printed
with a twist to its tail, carrying the soft sound of g as in gentle.
The simpler form carrying the commoner sound in early read-
ing is the g in go.
AS TO THE ORDER IN WHICH WORDS, PHRAses, and SentENCES
ARE INTRODUCED
It is, of course, fully and freely admitted that any graded
system of reading based on literal sounds and their shapes does
not very easily satisfy other criteria of early reading matter.
Words must be introduced to exemplify the sounds; but, it
seems to us, that they should also be such as fall within the
child's early vocabulary. Early reading is a problem in the
mechanics of reading. It is not literature, and it is not under-
taken for the purpose, at first, of adding to a child's spoken
vocabulary, or of increasing his knowledge of the real world.
He is to learn to read about what he already knows in other ways,
through his actual experience of life or through oral lessons.
For example, it does not seem to us desirable that in an in-
troductory primer words like existence, bureau, crozier, azure,
myrrh, lieutenant, aegis, phoebus, caesar, reservoir, hallelujah,
APPENDIX
179
quay, hiccough, norwich, and others of equal difficulty as to pro-
nunciation and meaning, should find a place at all. Even on
the phonoscript system of reverting to the first letter of the
word after each successive letter is sounded, these words give
difficulty, quite apart from the fact that the sounds of their let-
ters do not combine into words whose pronunciations and mean-
ings are known to the children. Even mere length is a difficulty
to little ones, and words like weather, biscuit, daughter, quarrel-
some, and patience should not be found in a first book.
Even if the Reading Sheets are used before the books, the
matter is hardly mended. For the first 20 lessons of the book
are identical with those of the Reading Sheets; and the words
weather, biscuit, daughter, beauty, and rejoice appear on them.
It is true that each of the above words contains one or more
silent letters; but, even then, the words are too long for the
great majority of young children who are making a beginning
in the mechanics of reading. (Syllabification might help as far
as the pronunciation of the word is concerned; but it is not
employed.) No doubt it is startling and very convincing to the
layman to hear an able young child quite successfully, with the
aid of the phonoscript markings, achieve a good pronunciation
of words like some of those above cited. He is astonished, as
well he may be; but such words are, in our judgment, too hard
for the average child at an early age, and their introduction
might well be postponed.
There is just one other point in connection with the presenta-
tion of exemplifying words and sentences. It would be better
if there were more reading matter, as compared with the num-
ber of "key words" placed at the beginning of the lessons, so as
to give additional reading practice. Curiously enough, some of
the words given at the heads of the lessons, do not appear in
the current reading matter at all. Some of these are bureau,
aegis, excel, crozier, luring, regular, bran, saviour, berkshire,
glazier, myrrh, and music. Doubtless it is not easy to bring
them in to any meaningful sentences to young children; but is
not that a condemnation of their use at all at this stage?
180
TEACHING BEGINNERS TO READ
AS TO THE ACTUAL SENTENCES USED FOR READING MATTER
To us, it seemed, in using the phonoscript books, that some
of the sentences were quite beyond young children in construc-
tion and meaning. For example, in the first two or three pages
of the first primer we find "Tap a vat at a height," "I have a
high pipe at a vat,” “I have a tight tap at a high pipe," and "Tap
a tight tap." And on page 8, more excusably perhaps, we find
'Paul pays Pat a visit.' The meaning of this sentence, well con-
structed as it is, will not be apprehended by a young child. On
pages 11 and 12, the following sentences are given "I don't like
cats." "If I find a cat in the house, I shout at it and frighten it."
Our young children in London are not, generally, unkind to ani-
mals; but I do not think that the ethics implied in these sen-
tences will help teachers to make them kinder.
AS TO THE COLLOQUIAL NATURE OF THE PHONOSCRIPT
PRONUNCIATIONS
I
The issues raised in this section are, I fear, rather funda-
mental. This is not the place to discuss the great question of
phonetic decay, to which all civilized languages are prone.
am not, indeed, competent to do it. I confine myself to those
aspects of it which are noticeable in early reading and in spell-
ing. No doubt, the fuller and older vowel sounds which teach-
ers endeavor to impress upon their young pupils seem to many
of us strained and artificial. Doubtless phonetic decay is too big
a process to be stopped in the schools, whatever we may do.
But, if we accept and teach colloquial pronunciation, in which
phonetic decay has already gone a long way, may we not be
helping on the process?
As one young girl, a London County Council Scholar, one
of a family of teachers—there are such things in London-said
at dinner one night, "If we were taught to speak clearly at
school and we talk as we do now when we are grown up, how
should we speak as grown-ups if we started at school as we
speak now?" Her argument may not be irrefutable; but the
APPENDIX
181
great consensus of opinion among school teachers would be on
her side.
Certainly many, if not most, of our London school children
hear at home vowel sounds of a blurred and non-distinctive char-
acter. In such cases, they often write and spell as they speak.
Would it not be wiser to use sounds over full and even pedantic-
ally distinct when speaking to and teaching such children? To
illustrate the position in relation to phonoscript, may I exemplify?
The vowels of the following italicized words have their sounds
different and their shapes modified in the sentences labelled (h)
from those which they possess in the sentences labelled (a). All
the instances of the vowels in the sentences labelled (b) have
approximately the neutral vowel sound, a sort of short e-sound.
(a) I want to tell you as quickly as I can.
(b) I can often set mother's machine going.
(a) We had reached our new house.
(b) He and I had never seen Minnie.
(a) So you are leaving us.
(b) Mother took us with her.
(a) What a big station it was.
(b) Father was at home.
(a) Grown-ups never seem to understand.
(b) I'll try to deserve your good opinion.
(a) I liked that photograph.
(b) One night father told us that he had met a naval lieutenant.
(a) He threw himself at us.
(b) Tim simply flew about with joy at being free again.
(a) I asked Phil where the water came from.
(b) "Some from springs,
(a) Some from big reservoirs,'' he said.
(b) Some boys-choir-boys from Westminster Abbey, Phil said-were
teasing the monkeys.
No one can deny that an educated adult, speaking good
English, does give a blurred and neutral sound to the vowels of
the words underlined in the above sentences labelled (b)—a
much less full and distinctive vowel sound than in the same
words in the sentences labelled (a).
Nor can anyone deny that, in whatever way we teach our
vowel and other sounds, these sounds will become modified by the
literal combinations into which they enter, and by the emphasis,
or lack of it, required by the word in order to convey our full
meaning. The problems for us as teachers are: Shall we help
on the process of decay? Shall we make so many distinctions
182
TEACHING BEGINNERS TO READ
and modifications in early reading? Shall we, at the risk of
being called academic and pedantic, teach the fuller sounds and
leave the process of decay to the influence of time?
It is for teachers to answer these questions. I am raising
them because I think an excellent system of teaching reading-
which phonoscript most certainly is-may be rejected, for early
reading at least, because it contains, both for shapes and sounds,
distinctions and modifications which might well be eliminated
or postponed.
APPENDIX II
A SUGGESTED INTRODUCTION TO THE PHONO-
SCRIPT SYSTEM OF READING IN ITS
VERY EARLY STAGES
I have previously mentioned, I think, that the head mistress
of the infants' school (not a kindergarten, be it understood)
in which the later experiments had been conducted, had decided
to use phonoscript for the future as her school method of
teaching early reading and writing.
She, if anyone, knew its weaknesses. She was aware that
there was loss, although a rapidly diminishing one-on transfer-
ence to ordinary print. She was aware that, in spelling results,
it had been beaten, in its early stages, by an ordinary phonic
system. She believed that it contained distinctions, both in shapes
and sounds, which were not readily discernible by young chil-
dren and were, perhaps, unnecessary. But she was so much im-
pressed by the interest, vigor, and success with which her children
attacked reading matter that they had not been taught, and by
the desirability of using one system of print for reading and
writing in their early stages that she intended henceforward to
teach phonoscript to Grade II (official age, 4 years 9 months to
5 years 9 months) at least in the second half of the educational
year, and to Grade III (official age, 5 years 9 months to 6 years
9 months). In her Standard I class—a year older-however, she
intended to make a transference, if possible, to ordinary print and
cursive script or handwriting.
Accordingly, she started with a new Grade II class, a little
before the second half of its educational year. In the first half,
no names or sounds or drawings of shapes relating to the teach-
ing of reading had been dealt with. All that had been done
that could help (I am not sure whether in the best way) was
the printing of phonoscript letters on little cards which the chil-
dren had sorted out and matched in their individual occupations.
183
184
TEACHING BEGINNERS TO READ
Then she began, or rather the teacher of the class began, un-
der her direction, to teach phonoscript; and for the purpose
of making the matter clear to her staff, she divided her scheme
into steps, as follow:
Step 1. Before using the phonoscript charts or books, a
series of preliminary lessons was given. The teacher of the
class composed short narratives about matters with which the
children were well acquainted, and told them, speaking slowly
and distinctly, to her little ones. She then printed certain parts
of them on the blackboard in phonoscript characters. The chil-
dren, with the teacher's help, spelled through the words (all of
which were well-known to them in pronunciation and meaning)
in phonoscript fashion, reverting to the first letter after each
successive letter was sounded. The reading matter contained
only the simplest form of the letters (those without any modify-
ing marks); there were no silent letters; and the words were
short, consisting of only two or three letters. When this step
was completed, the children knew practically all the unmodified
consonantal sounds, and all the short sounds of the vowels.
Step 2. At this stage, silent letters were introduced and
some long vowels with the necessary modifying marks. The
teacher used the phonoscript primer for the reading matter,
which she placed upon the blackboard. She did not, however,
adhere to the exact text of the book, but selected the simpler
sentences, such as the following: "i have a pet." "it is a rat.”
"i pat it." The sentence "i visit it" was not used; since the word
visit, it was thought, was unsuitable in meaning and perhaps in
length at this stage. A selection of sentences as far as the
twentieth lesson was thus printed and dealt with.
Step 3.—After the above preliminary work, the children were
sufficiently practised and advanced to begin with the Phono-
script Charts (Set 1) and the Introductory Primer, the reading
matter of both being identically the same, and to go straight
forward, the text being followed precisely.
The lessons proceeded thus: One day the teacher used the
Phonoscript Chart, and taught from it enough reading matter
APPENDIX
185
for one lesson. The next day books were given to the children;
and they read over again, being called upon individually, the exact
text which they had read from the chart on the previous day.
Step 4.-When the children had finished with these charts.
(Set 1), books were used in every lesson and the class went
steadily forward through the primers and readers. All the les-
sons in each step were 20 minutes in length and five of them
were given each week, one at a good pedagogical time in the
morning of each school day.
By the method of approach above indicated the head teacher
believes that average and dull children will learn the system bet-
ter than they will if so many modifications of sounds and shapes
and difficult words are introduced to them at the outset.
35570-

UNIVERSITY OF MICHIGAN
3 9015 00399 0465
#
DO NOT REMOVE
OR
MUTILATE CARD