LINGUIST List 26.1067

Tue Feb 24 2015

Review: Applied Ling; Cog Sci; Psycholing; Writing Systems: Treiman, Kessler (2014)

Editor for this issue: Sara Couture <>

Date: 24-Oct-2014
From: Geoffrey Sampson <>
Subject: How Children Learn to Write Words
E-mail this message to a friend

Discuss this message

Book announced at

AUTHOR: Rebecca Treiman
AUTHOR: Brett Kessler
TITLE: How Children Learn to Write Words
PUBLISHER: Oxford University Press
YEAR: 2014

REVIEWER: Geoffrey Sampson, University of South Africa

Review's Editor: Helen Aristar-Dry


Studies of the process of becoming literate often focus mainly on learning to
read. The excellent survey volume by Rayner et al. (2012), for instance (as
its title indicates) is exclusively about that aspect of literacy. And of
course, for any child, reading comes before writing. Nevertheless, to qualify
as literate a person must be able to write as well as read his language. The
book under review aims to draw together all aspects of what modern research
can tell us about the process of learning to write, in English or in other
languages, whether they use an alphabet or some other type of script.

In the English-speaking world, learning to write words is commonly identified
with learning to spell. But Treiman and Kessler only begin to address the
latter topic on p. 216 (in a book of 318 pages, excluding end-matter). They
acknowledge (p. 21) that this ''may strike readers as quite late'', but they
point out that children ''have to learn such things as how writing differs
from drawing and that it stands for language'' -- a great deal of learning has
to happen before getting spellings right can become an issue.

After an introductory chapter, Treiman and Kessler's Chapter 2 surveys the
different kinds of script used for various languages, including phonographic
scripts like our own and logographic scripts such as that of Chinese. Chapter
3 covers various considerations relating to the general concepts of learning
and of teaching, and Chapter 4 expounds rival theories about the learning
process. For instance, ''constructivism'', based on the ideas of Jean Piaget,
envisages children as passing through a fixed sequence of stages, at each of
which they hold a distinctive concept of how writing works. ''Connectionism''
holds that the literacy-acquisition process can be modelled by neural-network
computer software. Various other theoretical approaches have been advocated
in the literature; already in Chapter 1 Treiman and Kessler drew attention to
a movement which has been influential among the school teaching profession
called ''whole language'', associated with Kenneth Goodman and Stephen
Krashen, according to which explicit teaching of any general principles or
rules about orthography is unnecessary, perhaps even harmful, because children
will automatically acquire the ability to write correctly through mere
exposure to real-life written material. Treiman and Kessler oppose the
whole-language movement, holding that literacy acquisition depends on
abstracting patterning of many different kinds, absolute and statistical, from
the written form of one's mother tongue, and that formal teaching plays an
important positive role in helping children to absorb those patterns. They
themselves advocate what they call an Integration of Multiple Patterns (IMP)
theory, according to which learning to write correctly depends not on one
well-defined type of linguistic knowledge but on the ability to deploy many
kinds simultaneously.

Chapters 5 to 9 cover the learning of issues such as the surface properties of
writing, e.g. its one-dimensional linearity, the concept of symbolization, and
the fact that letters of the alphabet have names and a fixed order. Then
Chapters 10 and 11 move on to the learning of spelling -- which, as the
introduction points out, is a harder task than learning to read even with a
regularly-spelled language such as Italian, because it is not self-correcting.
If a child's attempt to translate a letter-sequence into a spoken form yields
the wrong phoneme string, the child will probably know that that phoneme
sequence is not a familiar word of his language, but going in the other
direction the child has no way of knowing whether an attempt to spell a word
familiar in speech is right or wrong, unless a teacher gives feedback.

A particularly interesting conclusion is that individuals do not reach a
''steady state'' of spelling knowledge, akin to the steady state of linguistic
competence which generative linguists believe speakers attain by puberty. We
never stop learning about spelling (p. 277).

Chapter 12 discusses acquisition of the skills of punctuation and
capitalization. And finally Chapter 13 draws together some of the general
conclusions which emerge from the research surveyed, with particular attention
to implications for desirable teaching styles. (The authors aim throughout
the book to avoid technical linguistic terminology where possible, or at least
to explain it clearly where it is unavoidable, so as to make the book
accessible to a wide readership, including primary school teachers, rather
than exclusively to academics.)

The authors themselves have done considerable original research on the topics
covered, but in this book they set out to survey the state of research as a
whole rather than to focus on the particular issues they happen to have
investigated personally. They quote work relating to literacy acquisition in
many languages; if the coverage of English is fuller than for other languages,
that is merely because, as they rightly say, more research internationally has
been done on English than on literacy in any other language. Their list of
references comprises some 700 publications, many quite recent.


To my knowledge no book similar to this has appeared before, so it fills an
important gap in the market. And in my judgement it does a very good job. It
is surely destined to become the obvious first place to look for specific
information on many of the topics it covers. Some might feel that a
disproportionate amount of space is devoted to areas which Treiman and Kessler
regard as preconditions for learning to spell, such as the contrast between
drawing and writing, or the idea that letters come in a fixed alphabetical
order. But, for laymen, it is too easy to suppose that these things are
''obvious'' and that the only real issue is learning to spell; it is perhaps
healthy to be exposed to careful analysis of things which seem obvious to
adults but which have to be learned from scratch by children.

A particular strength of the book is that it draws on classroom experience,
not just on academic literature and experiments in psychology labs. For
instance, one aspect of the book which interested me as an Englishman was
discussion of the results of contrasting approaches to literacy-teaching in
Britain and the USA. One difference relates to letter-names. When I learned
to read and write in the late 1940s, teachers taught us to name the letters as
adults name them, ''ay, bee, see,'' etc. Apparently this is still usual in
the USA; but in Britain, at least since my children were in primary school in
the early 1980s, pupils have initially been taught to name letters by the
sounds they commonly make, e.g. /æ/, /b/, /k/ (with the stops followed by a
minimal shwa to make them pronounceable). The adult names are introduced only
much later. This change was made in the belief that it would help children
understand how to use letters. Treiman and Kessler cite detailed data from
classrooms in the two countries which go to show how far and in what respects
that belief is true.

Linguists will value findings about literacy acquisition that bear on
theoretical controversies in our discipline. For instance, generative
phonologists since Chomsky and Halle's _The Sound Pattern of English_ (1968)
have held that where a language contains regular morphophonemic alternations,
speakers store words mentally in an underlying form, and apply phonological
rules to generate surface forms as these are needed in speaking. One
implication is that English spelling is much more regular than it appears.
Many linguists have found this theory implausible. But Treiman and Kessler,
while they would not claim to give us reason to believe in the entire complex
edifice of _Sound Pattern of English_ phonology, do cite more than one kind of
evidence that ''even young children appear to ... consider morphology when
deciding on spellings'' (p. 266).

There are weak points in the book. While Treiman and Kessler's attempt to
give due consideration to scripts different from ours is laudable, they make
mistakes about languages which, probably, they have not themselves studied.
They say (p. 54) that the Hebrew form for ''the'' is written as a prefix to
the following word because of a graphotactic rule that no word may consist of
just one letter; but it is clear from the phonology of Biblical Hebrew that
the definite article must be a prefix rather than a separate word,
independently of script considerations. They quote an example of ''simplified
Chinese'' (pp. 28-9), i.e. the new script variant used in the People's
Republic since the 1950s, but nothing in their example differs from
traditional Chinese script. They quote /kat/ as an example of a Japanese
syllable for which Japanese syllabic _kana_ script provides no symbol (p. 37);
but no Japanese syllables end in /t/. (The closest Japanese approximation to
English ''cat'' is /katsu/, which would be written with the _kana_ symbols for
/ka/ and /tsu/.) The statement on p. 142 that ''Korean han'geul ... abandoned
its original Brahmi-based ordering in favor of one based more on visual
similarity'' is doubly misleading. The Korean han'geul alphabet was an
independent invention, not based on an Indian model, and its original order
was not the same as that of the Brahmi alphabet. (Both alphabets did show
phonetic logic in their sequencing, unlike the Roman alphabet, but that was
merely coincidental.)

However, in the context of the book as a whole these points and others like
them are minor blemishes.

A deeper objection relates to Treiman and Kessler's Integration of Multiple
Patterns idea. They call this a ''theory'', which is said to make empirical
''claims'' (p. 241) about the literacy acquisition process. But if it is a
theory, it is left quite vague. The term ''theory'' strikes me as an
unnecessarily highflown way of referring to what is simply a commonsensical
assumption that, if any kind of patterning whatever is visible in the data
available to a child, the child may well latch on to it and use it alongside
all the rest of his linguistic knowledge in deciding how to write words. But
to say that is not to diminish Treiman and Kessler's achievement. Pedagogy
could probably benefit from less theorizing and more application of common
sense. Certainly Treiman and Kessler are persuasive when they point out the
limited value of other ''theories'' described in their Chapter 4.

Much of Treiman and Kessler's advice about suitable methods for teaching
writing skills seems sensible and well-argued. I felt sceptical, though, when
they urged (pp. 311-12) that:

children want to know why things are the way they are ... there are reasons
why words are spelled as they are. In orthography, as in other domains,
children can learn about the reasons ... generating reasons about why _photo_
is spelled with "ph" helps children to understand orthography.

The reason why _photo_ has "ph" is that in Ancient Greek, the word for
''light'' began not with a labiodental fricative but with a bilabial aspirated
plosive. I do not want to be patronizing, but I believe this explanation
might go over the heads of some teachers of five-year-olds, let alone of the
children themselves. Is there really any way it could be got across to a
primary-school class without creating more confusion than enlightenment?

However, a book that never took a position with which one felt inclined to
disagree might be tedious and empty. Treiman and Kessler's book is a very
worthwhile addition to the literature of linguistics and literacy-teaching. I
have learned much from it, and I expect to consult it frequently in the


Chomsky, Noam and Morris Halle. 1968. The Sound Pattern of English. New
York: Harper and Row.

Rayner, Keith, Alexander Pollatsek, Jane Ashby, and Charles Clifton Jr. 2012.
The Psychology of Reading (2nd edn). New York: Psychology Press.


After graduating in Chinese from Cambridge University, Geoffrey Sampson
studied linguistics as a graduate student at Yale and as a Fellow of Queen's
College, Oxford. He spent the bulk of his career at the universities of
Lancaster, Leeds, and Sussex successively. Since becoming Emeritus at Sussex,
he has been a research fellow in the Linguistics department of the University
of South Africa. A new edition of Sampson's book _Writing Systems_ will appear
in January 2015.

Page Updated: 24-Feb-2015