The 13th Easy Way to Make Learning to Read Difficult: A Reaction to Gleitman and Rozin

Author(s): Kenneth S. Goodman

Source: Reading Research Quarterly, Vol. 8, No. 4 (Summer, 1973), pp. 484-493
Published by: Wiley on behalf of the International Reading Association
Stable URL: http://www.jstor.org/stable/747170 .
Accessed: 22/06/2014 20:46

Your use of the JSTOR archive indicates your acceptance of the Terms & Conditions of Use, available at .
http://www.jstor.org/page/info/about/policies/terms.jsp

.
JSTOR is a not-for-profit service that helps scholars, researchers, and students discover, use, and build upon a wide range of
content in a trusted digital archive. We use information technology and tools to increase productivity and facilitate new forms
of scholarship. For more information about JSTOR, please contact support@jstor.org.

Wiley and International Reading Association are collaborating with JSTOR to digitize, preserve and extend
access to Reading Research Quarterly.

http://www.jstor.org

This content downloaded from 185.44.78.144 on Sun, 22 Jun 2014 20:46:38 PM

All use subject to JSTOR Terms and Conditions
The 13th easy way to make learning to read difficult:
a reaction to Gleitman and Rozin

KENNETH S. GOODMAN
Wayne State University

CHALLENGES THE BASES for the Gleitman-Rozin recommendation

that children be taught to read English as a syllabary. The sig-
nificance of the experiment using Chinese characters with Amer-
ican children and the central Gleitman-Rozin thesis that lan-
guage is psychoacoustic are challenged. It is noted that the syl-
lable is subject to the complex relationships between morphology
and dialect. The contention that children can understand what
they can pronounce is also refuted. It is stressed that children
are masters of the morpheme and display linguistic competence
with words and larger elements more abstract than syllables.
The syllabary method, it is argued, does not consider reading as
a psycholinguistic process.

La 13&memaniere aisle
de rendre difficile l'acquisitionde la lecture
CETTE ETUDE MET en question les bases de la recommendation
faite par Gleitman et par Rozin que l'on enseigne aux enfants la
lecture de l'anglais comme un syllabaire. On critique la signifi-
'
cation de l'exp6rience ofi l'on a pr6sent6 des enfants americains
des caracteres chinois, ainsi que la these principale de Gleitman
et de Rozin que le language est psycho-acoustique. On fait remar-
quer que la syllabe subit l'influence des rapports complexes qui
existent entre la morphologie et l'idiome. La pr6tention que las
enfants comprennent ce qu'il peuvent prononcer est r6fut6e. On
souligne le fait que les enfants possedent a fond le morpheme et
d6montrent une comp6tence linguistique avec des mots et des
6l6ments bien plus larges et plus abstraits que des syllabes. On
pr6tend que l'emploi d'un syllabaire est une m6thode qui ne con-
sidere pas la lecture comme un proc6d6 psycho-linguistique.

484

This content downloaded from 185.44.78.144 on Sun, 22 Jun 2014 20:46:38 PM

All use subject to JSTOR Terms and Conditions
A reaction to Gleitman and Rozin GOODMAN 485

La 13* forma fdcil de hacer que el aprendizajede la lectura

resulte dificil
DESAFIA EL FUNDAMENTO de la recomendaci6n Gleitman-Rozin de
que a los nifios se les ensefie a leer ingl6s como silabario. Son
desafiadas la importancia del experimento de utilizar caracteres
chinos con nifios americanos y la tesis central de Gleitman-Rozin
de que el lenguaje es psico-acfistico. Se hace notar que la silaba
depende de las complejas relaciones entre la morfologia y el dia-
lecto. El argumento de que los nifios pueden entender lo que
pronuncian, tambi6n es refutado. Se pone de relieve el hecho de
que los nifios son duefios del morfema y despliegan una capaci-
dad lingiiistica con palabras y con elementos mayores ye mais
abstractos que las silabas. Se sostiene que el m6todo silibico no
considera a la lectura como un proceso psico-lingiiistico.

This content downloaded from 185.44.78.144 on Sun, 22 Jun 2014 20:46:38 PM

All use subject to JSTOR Terms and Conditions
486 READING RESEARCH QUARTERLY * Summer 1973 VIII/4

How can one explain the phenomenon, here repeated one

more time, of knowledgeable researchers and students of language-
or learning, or human development, or medicine, or society-aban-
doning the rigor and logic of their own disciplines once they turn their
attention to reading and reading instruction? Here we have authors
who, we must assume, are enlightened in the latest theories of lan-
guage, language acquisition, and psycholinguistics saying "[given
what] little is known of the reading process" and later, "since no one
knows how much of what is learned about reading is a consequence
of what is explicitly taught about reading," and again "no one has
any idea of how to do this." Can it be imagined that they would pref-
ace proposals in theoretical linguistics or psycholinguistics by a plea
of total ignorance? Would they stand before an audience of their own
professional associations and justify their theses by proclaiming that
no knowledge exists on which to base them?
The Gleitman-Rozin paper is an almost classic example of
the phenomenon I have described. The authors start with an interest-
ing and well designed bit of research. Some kids who couldn't
read were asked to read a series of Chinese symbols, each of which
represented a word. Chinese symbols, like the numerals we use, do
not represent speech sounds. Rather they represent meaning directly.
The children in the study could get the meaning from the Chinese
symbols and report it in oral English. This is interesting but no great
surprise. Why would the researchers think that English speaking chil-
dren would have any more or less trouble than Chinese speaking chil-
dren with this task? Chinese children have been learning to do this
for thousands of years. Remember, the signs represent meaning, not
the oral Chinese language.
One conclusion which could be drawn from this research is
that kids could learn to read without any connections being built
between oral English and written English, treating each English writ-
ten word as if it were an ideograph for the meaning. In fact, the
authors have overlooked a rather simple logical relationship which
can be easily demonstrated. In an alphabetic writing system, at least
in origin, a graphic system represents an oral system which in turn
represents meaning:
and O-mthen
G-Since
Since g--- o and o---- m then g---•m

This content downloaded from 185.44.78.144 on Sun, 22 Jun 2014 20:46:38 PM

All use subject to JSTOR Terms and Conditions
A reaction to Gleitman and Rozin GOODMAN 487

Even though the alphabetic system starts out to represent the

speech system, it comes, as written language, to represent meaning
directly as well. In the latter part of the article, when the authors are
building a tenuous case for their syllabic system of teaching reading,
they indicate some awareness of this direct relationship to meaning.
They assert (since no reference is cited we must assume an assertion)
"Some children end up at what is called 'the fourth grade level' in
reading skill, with a vocabulary of at best a few thousand words. But
this is exactly what we would expect as the outcome of learning a
logographic system: there is a slow accretion of items and, in the
absence of overriding motivation, a diminishing return as the number
of items increases."
Curiously this statement, while appearing to recognize that
English written words may stand directly for meaning, implies two
very strange conclusions:
1] That no Chinese has ever achieved readinglevel above"thefourth
grade level" without "overridingmotivation"(perhaps as a nation
they have overridingmotivation?).
2] That reading is either naming words or figuring them out. This
implicit assumption permeates their article. It is a curious posi-
tion for linguist and psycholinguistto take since it ignores every-
thing that is known about how language works, how it is ac-
quired, and how it relates to meaning. Can it be that the authors
think reading is not a receptivelanguage process? Do they believe
that one must go from a surface representationin listening to a
deep structure and meaning but that in reading one must fool-
ishly say the names of words to oneself and then derive meaning
from this string of words somehow transformedto oral language?
Why would anyone learn so clumsy a procedure when deep
structure and meaning are directly accessible from the written
language?
Here is the essence of their argument: "Ourfindings are that
children have little or no difficulty in comprehending the basic prin-
ciples of logographic systems; they can readily grasp the idea that
language can be represented by a sequence of written signs, if the
signs correspond directly to meaning. They do have some difficulty
with the acquisition of a syllabary; that is, it is relatively hard to learn
the concept of phoneticization (the idea that the sound system medi-
ates the relation between sign and meaning). They have particular
difficulty with an alphabetic system; it is especially hard to learn

This content downloaded from 185.44.78.144 on Sun, 22 Jun 2014 20:46:38 PM

All use subject to JSTOR Terms and Conditions
488 READING RESEARCH QUARTERLY * Summer 1973 VIII/4

phoneticization at the level that corresponds to alphabetic signs: the

phoneme".
This would appear to be an argument for a non-phonics ap-
proach reading instruction. Somehow the authors find it a justifica-
to
tion for their syllable approach. The reason is that they have accepted
as a proven fact that children must somehow turn print to speech in
order to comprehend written language. Over and over this assertion
is made:
In short, we are claiming that the fundamental conceptual prob-
lem in reading acquisition is psychoacoustic: it has to do with
the awareness of phonological segmentation, and it has very
little to do with the writing system itself (i.e., the visual input).
Apparentlysignificant difficultyin learning to read is apprehend-
ing that the sound system is to mediate between sign and mean-
ing.
Some children apparently fail to realize that sound is involved
at all in the reading task."
And finally they state "as the phoneme is the gross unit of
alphabetic writing, it is not surprising that some people don't learn
to read; it is amazing that so many do." Only their lack of knowledge
can explain the authors' amazement.
Too much is known about the reading process to summarize
it in a short simple refutation of this central but unwarranted as-
sumption. Frank Smith in Understanding Reading (1971) has syn-
thesized research in psycholinguistics and in reading to demonstrate
that reading, like listening, involves the use of minimal graphic cues
to get through language to meaning. In their article in the Reading
Research Quarterly, Smith and Holms (1971) show that a reader
must use far more graphic information to get to letter identifications
than to words and far more to name words than to get to meaning
(Pp. 394-415). Modern models and theories of reading have come to
recognize reading not as a secondary representation of language but
as language itself. At many points in the article the authors refer to
the process of relating print to speech as decoding-as if speech were
any less a code than writing. Frequently they use the terms sign and
symbol to refer to writing but never to refer to oral language. They
equate speech with language. I must conclude that they have not
done their homework, and they have not sufficiently thought through
their positions to warrant the claims that they make for their "fifth

This content downloaded from 185.44.78.144 on Sun, 22 Jun 2014 20:46:38 PM

All use subject to JSTOR Terms and Conditions
A reaction to Gleitman and Rozin GOODMAN 489

method" of teaching reading. At no point do they seem to see the learn-

er's language competence and the conceptual background he brings to
his task as important in his success in reading. They found it "interest-
ing" when a child put together child +s and read children. The child
was telling them that morphemes, not syllables, are the units of lan-
guage. And the child was telling them also that he was a master of
the English morphemic system.
One might wonder, on the basis of the authors' view of the
central problem in reading instruction, how they think children learn
to talk. If one can learn speech "without a fully developed phonemic
organization" why can't one learn reading without it? Youngsters have
built their perceptual categories from whole language and not by
learning each phoneme in isolation or in syllables.
The research cited to support their position on phonics is the
same sort used by Jeanne Chall to argue that "code-emphasis" pro-
grams are needed. It demonstrates that some children have difficulty
abstracting sounds from sound sequences and that such children, as
a group, tend to have more difficulty learning to read than children
who can abstract sounds. The work of Venezky and Calfee (1968)
suggests that this is not a matter of "auditory discrimination" at all
but of the relative difficulty of dealing with abstraction. Rozin and
Gleitman would probably agree that this is so, since they seem to have
trouble reconciling the research they cite with linguistic reality: "Note
that the issue for our purposes is not whether a phoneme identifica-
tion and generation machinery exists in the brain (clearly it must)

For them the question becomes a matter of accessibility. But

if the problem is one of dealing with abstraction, syllables abstracted
from language will be only slightly less difficult to deal with than
sounds, as indeed their own study has shown. Words, with some con-
crete referents, are less abstract than syllables. But whole natural
language is treated by the child almost as if it were part of the con-
crete world. So why the syllable?
Perhaps the least scholarly section of this article is the one
in which the authors attempt, through a number of third hand and
impressionistic reports, to demonstrate that a syllabic system really
works. "Japan is reported to have a very low rate of illiteracy" and
"the Cherokee were 90% literate in their native language in the
1830's." That the Cherokees were literate is known and remarkable.
What the rate of literacy was must certainly be conjecture-the world

This content downloaded from 185.44.78.144 on Sun, 22 Jun 2014 20:46:38 PM

All use subject to JSTOR Terms and Conditions
490 READING RESEARCH QUARTERLY * Summer 1973 VIII/4

still has no universally acceptable standard for literacy-and that the

literacy has anything to do with the form of the writing system is not
demonstrated at all. The Japanese are literate people. The extent may
be exaggerated: reading difficulties are known to exist in Japan;
Gray's UNESCO studies indicated concern for them (Gray, 1956, pp.
37-38). But the extension of the authors' argument would be that if
Japan is a model of success in reading instruction, then all nations
need triple systems: alphabet, logograph, syllable; since that is what
Japan has. Why don't the authors cite Spanish as an example of
methods based on syllables? Spanish reading instruction uses a sylla-
bic focus to a great extent. Spanish syllables are far more discernible
than English ones; yet Spain and Latin America have not achieved
high literacy rates.
The syllable as a unit for learning to read English bears ex-
amination. The authors admit that "English has a complex syllable
structure." Articles by Roger Shuy (1969) and Ronald Wardhaugh
(1969) have discussed just how complex this structure is. Both have
argued that the English syllable is only important to the printer in
determining where to break words. My own research has shown in
fact that hyphenating words at ends of lines causes reading miscues
no matter how the words are divided. The authors never consider-or
at least never discuss-the complex relationships between English
morphology and inflection and the syllable. Walked is a one-syllable
word with 2 morphemes. Toward is a single syllable in some dialects
but has 2 in others (as well as an alternate towards). Shuy suggests
that in dividing words into syllables, meaning units and inflectional
units be given priority over sound: -ing would be treated as a unit
without the second consonant in words like hopping, running, etc.
The English syllable is very much influenced by the stress it
receives. All vowels in unaccented syllables become schwa. The authors
have not seemed to note this since they use in their program: Can a
bee get a can? The first can is a function word and unaccented so it
sounds not like the can meaning container but like kin. A is also an
unaccented syllable; it becomes schwa too. Since stress shifts within
and between words, the English syllable becomes quite unstable, par-
ticularly in its vowels. But there are also consonant shifts which result
from combining certain morphemes to produce inflected words.
For example site becomes situate with the/t/shifting to /c/
and then situation with the second /t/ becoming /s/. The spelling
system preserves the derivational relationships while the sounds shift.

This content downloaded from 185.44.78.144 on Sun, 22 Jun 2014 20:46:38 PM

All use subject to JSTOR Terms and Conditions
A reaction to Gleitman and Rozin GOODMAN 491

A child may well be able to accommodate these shifts even within the
semi-rebus, semi-letter system the authors use, but if he does so, he
is using his linguistic competence to overcome a barrier that their pro-
gram has placed in his way; for they are teaching him that the syl-
lables are constant when represented by the same sign. Frank Smith
talks about 12 easy ways to make learning to read difficult (Smith
1973). Give the authors credit. They invented a thirteenth.
The syllable is also not a constant across dialects. Bin, Ben,
been, bean, being are words some of which sound alike in all dialects
of English. But the ones that sound alike vary from dialect to dialect.
The authors report that "rhyme relation was significantly more useful
to suburban children." Thus their inner city subjects didn't seem to
be helped by rhyming. That's true, of course, particularly if the
rhymes aren't rhymes. Studies at the University of Wisconsin over the
past several decades have attempted to use Shakespeare's rhymed
couplets to reconstruct Elizabethan phonology. They rhymed for him
but they didn't necessarily rhyme for us.
So much for the fickle English syllable. Let's examine the
"fifth method" of teaching reading the authors propose.
The essence of their method has little to do with syllables but
goes back to their unexamined and unproven assumption: "Teaching
the child 'meaning' as part of teaching him to read is thus a red her-
ring. On the contrary, teaching him to decode is teaching him a major
essential of reading." To rationalize this assertion they have intro-
duced a prior assertion: "It is only during a brief transition period
that children with basic decoding skills cannot understand a passage
as they pronounce it." The authors cannot mean this in any full
sense. Surely they do not believe that a reader can understand every-
thing he pronounces. If that were so, then you and I would indeed be
wise and no teacher would need do more than teach children to say
the names of new words in their reading. So let us assume that the
authors mean this in a more limited sense. We are then back to a
classic argument for a narrow focus on phonics: If print is recoded
as sound, the reader will be able to get meaning. Yet research over
half a century has shown that there is little relationship between
phonics and comprehension beyond the fourth or fifth grade--even on
standardized tests. Remedial reading classes are filled with youngsters
in late elementary and secondary schools who can sound out words
but get little meaning from their reading. Research shows low profi-
ciency readers in second, fourth, sixth, eighth, and tenth grades be-

This content downloaded from 185.44.78.144 on Sun, 22 Jun 2014 20:46:38 PM

All use subject to JSTOR Terms and Conditions
492 READING RESEARCH QUARTERLY * Summer 1973 VIII/4

coming increasingly more able to sound out and produce phonic near
misses for unknown words but with little increase in their ability to
comprehend what they read. (Goodman & Burke, 1973)
In countries like Finland and Israel, where reading instruc-
tion relies heavily on phonics, many youngsters in the third year of
school can bark at print with great effectiveness but have not acquired
the ability to get sense from what they read.
The authors agree that nobody has been able to find a way
of teaching phonics or to develop a rationale for a phonics method
that is productive. Thus they take another cop-out, this the ulti-
mate one resorted to by all who offer a reading panacea. "There is no
coherent theoretical description of the decoding process and its me-
chanics. Thus, the only question askable is a practical one: Is one
known method or another more likely to help some children become
fluent readers?" Let trial and error be the judge, or "try it, you'll like
it," as the TV man says.
The authors claim that there are only 4 methods of teach-
ing reading which have ever been proposed and that they are introduc-
ing "only the fifth." Perhaps they will recognize the folly of lecturing
to an audience of reading researchers on the history of methodology
in reading. I hope so. The 4 methods they recognized are 1) learn-
ing whole words, 2) learning "meaningful units" larger than words,
3) learning phoneme-alphabet correspondences, and 4) "the lingui-
stic method, in which phonemes are blended onto 'spelling units' (syl-
lables) rather than onto other phonetic units."
Space precludes dealing in depth with the variations in read-
ing methodology that the authors don't seem to be aware of. They have
offered not another method but another place to begin, the syllable,
which is not a new place, as they themselves have pointed out. They
have incorporated into their method or approach the rebus, which
again is not a novel introduction in reading instruction.
The problem with most reading methods is that they are not
soundly based in theories of reading and language learning and that
they are not thought through and fully developed. Most are built in-
completely around minimal propositions and most are prematurely
put into practice. The authors have offered a phonics program that
starts with the syllable. They have not thought beyond the word,
failing to consider reading a psycholinguistic process. It is not sur-
prising that their catalogue of methods includes only some that are
based on getting children to say names of words. To quote the au-

This content downloaded from 185.44.78.144 on Sun, 22 Jun 2014 20:46:38 PM

All use subject to JSTOR Terms and Conditions
A reaction to Gleitman and Rozin GOODMAN 493

thors, "This theoretical and practical ignorance has led to madcap

fashions in education." They have certainly thrown their mad caps
into the ring.
Much progress has been made in the last decade in the field
of reading. The Targeted Research program's literature search (see
Reading Research Quarterly, Fall, 1971, Summer 1972, and Winter
1973 issues) has demonstrated the interdisciplinary basis of this
progress. Scholars from linguistics, psychology, and many other fields
have joined those of us in education in clarifying and redefining the
issues and beginning the task of making sense out of the reading proc-
ess and its acquisition. This progress has been based on a mutual
respect. The day is over for the Lone Linguist to gallop out of the East
on his great horse, Syllable, to save us from ourselves. There's work to
be done and a lot to learn. Come and bring knowledge, but leave the
silly syllables at home.

REFERENCES

GOODMAN, KENNETH S. & BURKE, CAROLYN ing identification in reading. Reading

L. Theoretically based studies of patterns Research Quarterly, Spring 1971, 6, 394-
of miscues in oral reading performance. 415.
USOE Final Report, Project No. 9-0375, SMITH, FRANK. Twelve easy ways to make
U.S. Department of Health, Education learning to read difficult. In Frank
and Welfare, April, 1973. Smith (Ed.) Psycholinguistics and
GRAY, WILLIAM. The teaching of reading Reading. Holt, Rinehart & Winston. Pp.
and writing. UNESCO, Scott Foresman, 183-196.
1956. SMITH, FRANK. Understanding Reading.
SHUY, ROGER W. Some language and cul- Holt, Rinehart & Winston, 1971.
tural differences in a theory of reading. VENEZKY, RICHARD; CALFEE, ROBERT; &
In Kenneth S. Goodman and J. T. Flem- CHAPMAN, ROBIN. Skills required for
ing (Eds.) Psycholinguistics and the learning to read: a preliminary analysis.
Teaching of Reading. Newark, Dela- Working paper No. 10, R&D Center,
ware: The International Reading Asso- University of Wisconsin, 1968.
ciation, 1969. Pp. 34-47. WARDHAUGH, RONALD W. Reading, a lin-
SMITH, FRANK & HOLMES, DEBORAH. The guistic perspective. Harcourt, Brace and
independence of letter, word, and mean- World, 1969.

This content downloaded from 185.44.78.144 on Sun, 22 Jun 2014 20:46:38 PM

All use subject to JSTOR Terms and Conditions