The EFL Dictionary Pioneers
and their Legacies
[This paper is an abridged version of
the presentation made at the inauguration of the Israel Association
for Lexicography (ISRALEX) at Seminar Levinsky, Tel Aviv on 3 February
It should come as no surprise that, in recent years, ELT dictionaries
of all kinds have become a vitally important focus of applied
linguistic research, innovative design and substantial investment.
Learners' dictionaries are a crucial part of the response to
a worldwide demand for English that is constantly expanding.
Within the broader field of EFL dictionary development, the monolingual
learner's dictionary (MLD) for advanced students occupies a special
place. The use of the foreign language as the language
of definition in the MLD - with the actual choice of defining
words being carefully controlled - is an acknowledgement that
the typical user has a relatively firm footing in the semantic
structure of the L2. The broad scope of the listed vocabulary
reflects the global importance of English as the language of
science, commerce and mass communication. And, not least, the
attention given to grammatical words and patterns, and to example
sentences, is a clear indication that MLDs are designed to meet
the needs of writers as well as readers. The way in which this
information is presented (making it, as it were, more 'learner-friendly')
has been enhanced, in recent years, by a growing body of research
into the use which students actually make of their dictionaries
(Atkins 1998, Cowie 1999). And the authenticity of the grammatical
claims made about English, and of the examples selected, has
been improved beyond recognition by the use, since the early
1980s, of large-scale computer-stored corpora of English, the
best known of which are the British National Corpus and the Bank
Because of the marked improvement in the resources available
to EFL lexicographers, it is tempting to assume that their products
have undergone fundamental changes, and that they now have little
in common with the very earliest MLDs - those of the 1930s and
1940s. But as recent research has revealed, the 'founding fathers'
of the MLD - Harold Palmer, Michael West and A.S. Hornby - had
even at that early stage added to the established features of
mother-tongue English dictionaries a new set of elements
that were inspired by the needs of non-native learners (Cowie
1998b, 1999). With the passage of time, those new elements in
turn 'acquired the status of convention, as the monolingual learner's
dictionary developed into a distinct genre' (Rundell 1998). It
is important to bear in mind, then, that despite the considerable
advances of the 1980s and 1990s, especially in the quality of
information available to lexicographers and in the way this is
presented, those pivotal elements and basic needs have changed
very little. So it is worthwhile to go back to those early years,
to ask what those original features were, and to enquire into
the sources from which they sprang.
The key figures who come into focus are three expatriate Englishmen,
teaching and conducting innovative programmes of research in
Japan and what was then British India. Harold E. Palmer and his
assistant (and later successor) A.S. Hornby, were based in Tokyo,
at the Institute for Research in English Teaching (IRET); Michael
West served in Bengal. It is a curious fact, by the way, that
though these programmes of lexical research led almost inevitably
to the MLD - we could in fact, without exaggeration, claim that
they gave birth to the learner's dictionary - the intended goal
of all this effort was the development of classroom syllabuses
and simplified readers. Palmer, for instance, was not aware until
the mid-1930s that 'a learner's dictionary' (his term) would
be the most complete and natural end-point of the lexical research
he had started up (Palmer 1934).
Here, I shall look at various aspects of this research programme
and trace its connections with the earliest achievements in EFL
dictionary design. The period covered is from about 1927 to 1942.
I begin with the so-called 'vocabulary control' movement. Then
I move on to consider Michael West and the idea - which was not
taken up by Palmer and Hornby - of using a controlled vocabulary
for defining. Afterwards, I shall consider the development
of phraseological research, and assess its subsequent impact
on pedagogical lexicography. In this project, A.S. Hornby played
a key role. Finally I shall try to show how the research into
collocations and idioms is firmly tied up with the design of
2. The vocabulary control movement
Harold Palmer had been interested in vocabulary control (or vocabulary
limitation) since 1903, when he had run a language school at
Verviers, in Eastern Belgium. His interest in vocabulary control
is explained by a desire to ease the learning burden of the foreign
learner by pinpointing those relatively few words which carried
the main weight of everyday communication. This interest was
closely linked to the preparation of simplified readers. In fact,
Michael West, researching in India, was largely inspired by the
desire to produce schemes of simplified readers for schoolchildren
(Howatt 1984). The purpose of preparing the limited vocabularies
was chiefly educational, then, but the scientific soundness of
any word-list could later be tested by using it to simplify other
unabridged texts. Words which occurred seldom or never in those
texts would probably not be kept, while words which occurred
frequently but were not already part of the list might be considered
for inclusion. West, as we shall see, applied similar principles
when testing successive versions of his limited defining vocabulary
for the New Method English Dictionary (NMED) of
It is not difficult to see now why the word-lists produced in
the early 1930s by Palmer, or by Palmer and Hornby working together,
should evolve into designs of dictionary entries suitable for
foreign learners (Cowie 1998b, 1999). When Palmer was asked by
IRET, in 1927, to compile a controlled vocabulary for middle-grade
Japanese schools, he was already aware that drawing up a word-list
was a more complex affair than producing an alphabetical inventory
of spelling-forms, based on frequency of occurrence. He knew,
for instance, that the form act could embrace different
meanings ('act', as in a play, 'act', in the sense "pretend",
etc) as well as different parts of speech ('act', noun or 'act',
verb). To run all these differences together, and count simply
one form, act, would be to gloss over distinctions that
were crucial to the learner (and also to the dictionary-maker).
Taking careful account of the meanings and grammatical functions
of words, Palmer, West and later Hornby succeeded in producing,
from 1930 onwards, what I have called 'structured lexicons' (Cowie
We can see this approach at work in Thousand-Word English
(TWE), a word-list begun by Hornby, then refined with
Palmer's help, and finally published in 1937. First, the entries
in the list were words, or 'lexemes', as they are in most dictionaries.
They were not inflected forms, like acts, acting,
acted, for instance. Those forms would be arranged inside
the entry (in this case for ACT), where the reader would also find a cluster
of derivatives (e.g. the nouns 'actor', 'actress'). To see how
this works out in more detail in TWE, consider the entry
at (1). In this entry, irregular inflected forms are picked out
in italic, and basic meanings, at 1 and 2, are conveyed by 'a
picture', 'a line', which are actually collocates of 'draw'.
The derivative 'drawing' is placed at the end of the entry.
(1) DRAW [drc:], v.
drew [dru:], pret.
drawn [drc:n], past ppl.
(1. e.g., a picture)
(2. e.g., a line)
drawing ['drc:in], n.
(Palmer and Hornby 1937)
Harold Palmer drew on the word-list
of TWE when he published his own MLD, A Grammar of
English Words (GEW), in 1938. The entry structures
of the two works also share certain features, including the positioning
of meanings and derivatives. But the dictionary has examples
and idioms - these do not appear in the word-list - and it also
has some interesting features of arrangement designed to help
the user grasp connections of meaning and form across the entry.
Notice in the SOFT entry at (2), below, how the definition at
soften contains the object nouns 'leather', 'one's voice',
'a person's heart'. These enable the user to link that word to
meanings 1, 3 and 4 of soft, but not meaning 2. These
arrangements are of great help when writing, because placing
the verb soften in the same entry as the adjective soft
helps the writer to avoid repeating the adjective in the same
sentence (like this: 'If it isn't soft, soften it!'). However,
if you are reading, 'nesting' words such as soften,
softly or softness in the entry for SOFT
probably makes it more difficult to locate the individual derivatives.
For the reader, they should arguably be placed in separate entries.
[scft], softer ['scfte], softest ['scftist], adj.
1. = not hard
a soft bed.
soft leather [wood, etc.].
The ground is soft after the rain.
Which of the two chairs is softer?
soft to the touch. ...
2. = smooth
a soft hand.
a soft skin.
as soft a silk.
soft to the feel [touch]. ...
3. Said of the voice and other sounds =
low, not harsh
a soft voice [sound, etc.].
4. Said of the character
a soft heart.
the softer side of his nature.
soften leather [one's voice, a person's heart, etc.].
This excerpt from GEW illustrates
only a few of the many connections between IRET research and
the earliest learners' dictionaries. But perhaps it is enough
to show that the direction taken by that research encouraged
the development of dictionaries that strongly favoured production
(encoding). First, extreme vocabulary limitation would give special
prominence to structural words (e.g. the, may,
in) and heavy-duty verbs (e.g. make, send,
bring) and these are, of course, the basic building-blocks
of sentence construction. Then, as we have just seen,
the Palmer-Hornby approach to lexical analysis gave us an entry
structure in which derivatives were clustered around their roots,
also with potential benefits for encoding. Of course, within
the vocabulary-control project there was also scope for prioritizing
the needs of the reader. This could be achieved by adapting the
word-lists produced in the course of research to compose controlled
defining vocabularies for dictionaries. In fact, there
was only one attempt made, in those early days, to develop and
utilize a limited vocabulary for defining. It appeared in NMED
(1935), which was first and foremost a dictionary for the reader.
3. Michael West's limited defining
The New Method English Dictionary, jointly compiled by
Michael West and J.G. Endicott, was the first MLD to be published.
The dictionary contained definitions based on a 'minimum adequate
definition vocabulary'. Also in 1935, West published Definition
Vocabulary, an account of how the defining vocabulary had
been systematically chosen, checked and revised. The research
involved compiling a preliminary version of the dictionary, in
which a defining vocabulary of 1799 words - eventually to be
reduced to 1490 - was used to define 23,898 vocabulary items
(West 1935: 34-41).
The definition vocabulary devised by West was to prove enormously
influential - it was the basis of the controlled vocabulary used
in the first Longman MLD, of 1978 - and in his 1935 essay West
put his finger on several of the problems that would later face
lexicographers wishing to devise defining vocabularies of their
own. The discussion was in fact remarkable for the range of issues
that he raised. West identified several of the characteristic
weaknesses of definitions in mother-tongue dictionaries - the
fondness for defining the known (say, pencil) in terms
of the unknown ('instrument'?, 'tapering'?), and the tendency
to resort to 'scatter-gun' techniques, whereby 'one fires off
a number of near or approximate synonyms in the hope that one
or other will hit the mark and be understood', as in the example
at (3) (1935: 8):
(3) sinuate tortuous, wavy, winding
One important question that concerned
West was how far one could depend on the learner to use prefixes
and suffixes as building-bricks. He included some of the commonest
prefixes and suffixes (e.g. dis-, in-, -able,
-en) in the defining vocabulary, and in the definitions
he allowed these to be attached to various words - provided their
meanings were regular. So the deverbal suffix -able can
be added to drink, eat, read, etc, on the
assumption that the user will infer the meanings of drinkable,
eatable and readable (cf. West 1935: 16). In this
way great economies can be made.
One of the lessons that West was quick to learn - and that others
have not been slow to profit from - was that to arrive at natural
and precise definitions of very many words, he had to include
in his definition vocabulary a number of very general ('genus')
words, including behaviour, belief, engine,
insect, instrument, metal, noun,
quality, relation, science, skill,
solid, surface, vegetable. These do not
necessarily occur very frequently, but their importance is easily
seen if we try to define onion or parsnip without
using vegetable, or bee or fly without using
4. Research into collocations and
Harold Palmer had set up a programme of research into phraseology
at IRET in 1927 - at the same time as it was decided to compile
a limited word-list for middle-school children. The result was
the first, large-scale analysis of English phraseology to be
undertaken with the needs of the foreign learner in mind. The
project was directed by Palmer, but much of the actual collecting
and classification was carried out by A.S. Hornby. The first
detailed findings were published as the Second Interim Report
on English Collocations, in 1933. The importance of the Interim
Report cannot be overstated. It consisted of a meticulous
classification of word-combinations in English, but it also showed
how much of everyday speech and writing is in fact made up of
'fixed phrases', and it helped pave the way for the strong growth
of interest in phraseology in the 1980s and 1990s (Cowie 1998a,
The expressions classified in the Interim Report were
familiar word-combinations (called 'collocations' by Palmer and
Hornby) that could function as elements in simple sentences.
These could be broken down initially into verb-collocations (to
toe the line), noun-collocations (a tidy amount),
adjective-collocations (as pleased as Punch), and so on,
but much finer sub-categories could be recognized within those
broad divisions. One sub-category, shown at (4), was the verb-collocation
'VERB x SPECIFIC NOUN (x PREP x N3)' (no. 31211). This
(4) To catch a cold
To entertain a belief
To give notice (x of x N3)
To hold one's tongue
To keep good [bad, etc.] company
The Palmer-Hornby approach was not perfect.
For example, the term 'collocation' was applied not only to the
very large groups of 'word-like' combinations - which the Report
actually treated - but to proverbs, slogans and catchphrases
as well. Nowadays, most phraseologists would restrict the term
collocation to word-like combinations such as to catch a cold
or a tidy amount. But in doing this they would also be
saying that each two-word combination consists of one word used
in a normal, familiar, sense (cold, amount), and
another word (catch, tidy) whose special meaning
is confined to that context and a few similar ones (cf. to
catch a chill, a tidy sum). And, of course, they would
be implying that these are not idioms - that is, not fixed phrases
that are difficult to explain in terms of all of the individual
words (Cowie 1998a). Now, neither Palmer nor Hornby recognized
this important distinction.
The most obvious practical effect was that, in Palmer's GEW
and the earliest editions of the Advanced Learner's Dictionary,
idioms were not always given the special prominence that they
deserve. In the list at (4), the phrase to hold one's tongue
is both fixed and 'unmotivated', and is therefore an idiom, while
to catch a cold can be internally modified, as in to
catch a fever/chill, and is therefore a collocation. Ideally,
the former should appear in bold print, but not the latter.
All the same, the Palmer-Hornby approach had a number of enduring
strengths. Remember that the Interim Report was a very
detailed grammatical classification. This impressed many
dictionary-makers and explains the emphasis given in several
British phraseological dictionaries to the grammatical treatment
of idioms and collocations. Once the notion had caught on of
classifying collocations and idioms according to form and structure
(see again the heading of the examples at (4)), the natural next
step was to provide a more detailed description, particularly
by indicating whether idioms could be used in the passive and
other 'transformations'. We find this finer detail in both volumes
of the Oxford Dictionary of Current Idiomatic English
5. Phraseology and the design of
Another benefit of this early interest in phraseology was that
it undoubtedly influenced the design of illustrative examples
in learners' dictionaries. Palmer, Hornby and West were all aware
that examples would help to show the learner what words (in their
various uses) meant. Palmer, though, and later Hornby,
were particularly interested in examples which showed the lexical
and grammatical contexts in which words typically occurred.
In a paper on vocabulary lay-out published in 1936, Palmer illustrated
this point by referring to the adjective used to:
(5) to be used to something or somebody
to get used to something or somebody
to be used to doing something
to get used to doing something
Now, we can argue that these are strictly
not examples at all, if by examples we mean instances of performance,
whether real (that is, taken from a corpus) or imitated. They
are simplifications (there is no grammatical subject)
and abstractions ('something' stands for a whole range
of possible noun phrases). I call these 'minimal lexicalized
patterns' (Cowie 1995, 1996), and their value has long been recognized
in French and Italian monolingual dictionaries as a basis for
imitation and expansion when writing. Palmer referred to them
as 'skeleton-type examples' and he and Hornby were responsible
for introducing them into the MLD (Cowie 1998a).
Hornby, very interestingly, developed for the Idiomatic and
Syntactic English Dictionary (ISED), of 1942, later
to become the Advanced Learner's Dictionary, a kind of
skeleton clause example which is less abstract than those I have
just shown. All the same, it was sufficiently simplified to provide
a good model for imitation and expansion. Here, at (6), are some
examples of this kind from ISED:
(6) to cut steps in a rock
to cut a figure in stone
to man a ship
to manage a horse
You can see that these are subjectless
clauses, that the verb is in the infinitive, and that modification
of object nouns (a rock, a ship, etc) is cut to
the minimum. And notice how close these examples are to the verb-collocations
we looked at earlier (cf. to catch a cold, to man a
ship). Some of the ISED examples are indeed part of
We have had to wait until the last few years of the twentieth
century and the first of this to gain a reliable, rounded picture
of the early history of the learner's dictionary and a true sense
of how fully it drew on the linguistic research carried out by
a small group of Englishmen working at a remote distance from
Europe in the 1920s and 1930s. It is only fitting that much of
the opening-up of a neglected episode in lexicographical history
has taken place in Japan, or has drawn extensively on Japanese
sources (Imura 1997, Cowie 1999, Smith 1999a, b). It is an astonishing
fact, too, that Palmer, West and Hornby were not aware until
about the mid-1930s that the true and natural goal of their research
would be lexicographical. In the end, all the key innovative
features of the new dictionaries sprang from this programme -
West's defining vocabulary, the verb-patterns of Palmer and Hornby,
the skillfully designed examples, and not least the collocations
and idioms, an area in which Hornby made such an immensely important
contribution. And once these corner-stones had been put in place,
they acquired the status of convention, giving the MLD a unique
lexicographic character. This is the enduring foundation, upon
which all the subsequent developments have been built.
The degree of interest is reflected in the intensity of the competition
(cf. Herbst 1996). Dictionaries from four major publishers at
present share the field: the Cambridge International Dictionary
of English (1995), the Collins Cobuild English Language
Dictionary (2/e, 1995), the Longman Dictionary of Contemporary
English (3/e, 1995), and the Oxford Advanced Learner's
Dictionary (6/e, 2000). The latter is the newest edition
of the work initially compiled by Hornby and his associates.
A. The early monolingual learners' dictionaries
Hornby, A.S., Gatenby, E.V., and Wakefield, H. 1942. Idiomatic and Syntactic English Dictionary.
(Photographically reprinted and published as A Learner's Dictionary
of Current English by Oxford University Press, 1948; subsequently,
in 1952, retitled The Advanced Learner's Dictionary of Current
English.) Tokyo: Kaitakusha.
Palmer, H.E. 1938. A Grammar of English Words.
London: Longmans, Green.
West, M.P. and Endicott, J.G. 1935. The New Method
English Dictionary. London: Longmans, Green.
B. Other references
Atkins, B.T.S. (ed.) 1998. Using
Dictionaries: Studies of Dictionary Use by Language Learners
and Translators. (Lexicographica, Series Maior 88.) Tübingen:
Max Niemeyer Verlag.
Cowie, A.P. 1995. 'The learner's dictionary in a changing
cultural perspective.' In B.B. Kachru and H. Kahane (eds.), Cultures,
Ideologies and the Dictionary. (Lexicographica, Series Maior
64.) Tübingen: Max Niemeyer Verlag, 283-95.
Cowie, A.P. 1996. 'The "dizionario scolastico":
a learner's dictionary for native speakers.' International
Journal of Lexicography 9.2: 118-31.
Cowie, A.P. (ed.) 1998a. Phraseology: Theory, Analysis
and Applications. Oxford: Clarendon Press. Cowie, A.P.
1998b. 'A.S. Hornby, 1898-1998; A Centenary Tribute.' International
Journal of Lexicography 11.4: 251-68.
Cowie, A.P. 1999. English Dictionaries for Foreign
Learners - a History. Oxford: Clarendon Press.
Herbst, T. 1996. 'On the way to the perfect learners'
dictionary: a first comparison of OALD5, LDOCE3, COBUILD2 and
CIDE.' International Journal of Lexicography 9.4: 321-57.
Howatt, A.P.R. 1984. A History of English Language
Teaching. Oxford: Oxford University Press.
Imura, M. 1997. Palmer to nihon no eigokyoiku.
(Harold E. Palmer and Teaching English in Japan.) Tokyo:
Palmer, H.E. 1933. Second Interim Report on English Collocations.
Palmer, H.E. 1934. An Essay in Lexicology. Tokyo:
Palmer, H.E. 1936. 'The art of vocabulary lay-out.' IRET
Bulletin 121: 1-8; 14-19.
Palmer, H.E. and Hornby, A.S. 1937. Thousand-Word English.
London: George Harrap.
Rundell, M. 1998. 'Recent trends in English pedagogical
lexicography.' International Journal of Lexicography 11.4:
Smith, R.C. 1999a. 'The Palmer-Hornby contribution to
English teaching in Japan.' International Journal of Lexicography
Smith, R.C. 1999b. The Writings of Harold E. Palmer:
An Overview. Tokyo: Hon-no-Tomosha.
West, M.P. 1935. Definition Vocabulary. (Bulletin
No. 4 of the Department of Educational Research.) Toronto: University
About the author
K Dictionaries Ltd
After reading Modern Languages at Oxford and training as an EFL
teacher in London, Anthony
Paul (Tony) Cowie taught English in
Western Nigeria, later becoming an English Language specialist
with the British Council. In 1964, he was appointed Lecturer
(later Senior Lecturer) in the School of English, University
of Leeds. He has since been involved in six major ELT dictionary
projects. From 1980-83, and as part of that continuing involvement,
he directed at Leeds an OUP-funded Lexical Research Unit. In
1992, he was appointed Reader in Lexicography. He was a founding
Executive Board member of the European Association for Lexicography
(EURALEX), and continues to serve there and on the Editorial
Board of the New OED. He is Editor of the International Journal
of Lexicography and author of English Dictionaries for
Foreign Learners - a History (Oxford 1999). He organized
the first of three International Symposia on Phraseology, and
spoke at all (Leeds 1994, Moscow 1996, Stuttgart 1998). His research
in the field has led to the establishment of a major bibliography
on the EURALEX website http://www.ims.uni-sttutgart/euralex/bibweb and the publication of Phraseology: Theory,
Analysis and Applications (Oxford 1998).
10 Nahum Street, Tel Aviv 63503 Israel
tel: 972-3-5468102 fax: 972-3-5468103