[Valerie Yule: see Journals, Newsletters, Media, Personal View, Anthology, Bulletins, Web links.]

Pronunciation Guides in Children's Dictionaries.

Valerie Yule.*

*Paper for presentation to the Australian Style conference.
16 October 2005.

My fields of study as a psychologist are literacy and imagination. Dictionaries are fascinating on both accounts.

How can learners find out from dictionaries how to pronounce new vocabulary they find in reading? Migrants to our shores include many who can speak English but cannot read it, but also others who can read and write English well - yet cannot speak it. Voice technologies can help out, but a book dictionary is always handy. Pronunciation keys for adults can use the International Phonetic Alphabet and other sophisticated guides, but these are too hard for children and for most migrants still learning English.

In the 1980s I looked at forty children's dictionaries and word-books and found out that only four gave any help about how to pronounce the most tricky words, and only two gave a pronunciation key for all words. No wonder children often dared not use their full vocabulary because they could be laughed at. (Yule, 1989)

Now I have updated this study, 25 years on. I have looked at 36 dictionaries and word books currently on the market for children and schools. [TABLE 1.] Half of them still give no guide at all, especially those for younger children. I suppose it is assumed that adults will read the words to them. Of the remaining half, 8 dictionaries sometimes give a guide to pronunciation, 3 often do and only 7 always do. That is, the advantage is given to the lucky children who are already most adept in English.

The eleven 'Occasional' guides differ widely in the words that they select as needing an aid. Usually they give no clue to the most common irregular words, such as who, was, are, so again the advantage is for English-speaking children who are already fairly literate and leave out beginners and learners of English language.

This table [TABLE 2] shows four major types of pronunciation guide provided for children. It also shows, on the left, the guide used in the Concise Oxford Dictionary, 1934, and on the right, a suggestion to consider for children in the future. I used the old Concise Dictionary for adults, because its guide is simple and looks most like ordinary spelling. It has a few extra clues, that I can't show very well here because they are not all on my keyboard - its use of the clues is not altogether satisfactory, but you might say fair enough.

Four school dictionaries illustrate different approaches to representing pronunciation for children - the Macquarie Junior, Heinemann Lower Primary, Macmillan Primary and the Jolly Dictionary. It is striking how they diverge in how they represent the spoken word. Which are the most helpful for learners?

I have looked at five aspects - How to show stress in a word, Unexpected initial letters, Speech sounds that English spelling does not discriminate, Dialect issues, and long vowels.

The first issue is, How to show stress in a word. Showing where the stress goes in a word is important, because children can feel laughed at when they get it wrong. Melanncoly and elly-jibble. - ha ha.

There are five ways that dictionaries commonly show stress placement in words - bold print, underlining, macrons, apostrophes or stops after the stressed syllable. Heavy macron diacritics and even more, apostrophes or stops, disrupt the visual line of the word and are more difficult for children to use than bold or underlined syllables. Macquarie Dictionary underlines the stressed syllable, the other three children's dictionaries use bold type. Underlining is easy to copy in handwriting. Bold print is probably least disruptive of the visual line of the word; it is intuitively understood, and it lessens the difference in appearance between the pronunciation guide and real spelling.

But some of the ways used to make clear the stress in a word can make a pronunciation guide look very unlike the spelling that the children must learn to write themselves. Two problem issues are how to represent syllables, and the obscure unstressed vowel schwa, in IPA the upside-down e.

Syllables. Where should syllables be cut in showing where the stress goes? Curiously, dictionaries often differ in what is a syllable in a word. For example, should it be tem-pru-cher or temp-ruh-chur, rest-a-ron or res-tuh-ront? For the sake of children's understanding of the linguistic structure of words, I would hope that pronunciation guides would segment words by their morphemes - into the segments of meaning.

We know that longer words are harder for children to read and spell. Cutting words into syllables may try to make them easier to tackle. In my youth, children's comics like Tiger Tim never worried about using long words. They just split them up with hyphens and then the little bits were easy. So children who might still be stuck at school in their Year 2 reading book, out of class could be enjoying de-tect-ive stories with brill-iant hero-ic char-act-ers hav-ing a-maz-ing ad-vent-ures. Educators heaped scorn on this practice, but I have never seen any research that justified the scorn.

Three of the four dictionaries also split up the longer words with hyphens. Why do I feel less happy about this than I did about Tiger Tim? One reason is that Tiger Tim's hyphens aimed to make it easier to read words silently for meaning. But do hyphens make it easier to speak the words aloud smoothly? I understand that there has not yet been comparative research to find out whether children find easier pronunciation guides that are chopped up with hyphens, or the Jolly Dictionary which uses no hyphens at all, and relies on the bold print that marks the stressed syllable to make visible a sort of segmentation, while the word itself remains whole and neat-looking. It may be that the more letters used in the pronunciation guide, making it longer, the more that children will need hyphens to read the syllables, and Jolly's guide uses fewer letters, - although when it does have a really long word, how do children cope with a pronunciation like exclemaishen?

All the dictionaries represent the slurred syllables in one way or another, but some of their representations of the slurred vowel, the schwa, are more labored than in the casually spoken word. It is hard not to pronounce the guide fuh-roh-shus as fuh-roh-shus, despite the underlining of the roh, or temp-ruh-chuh as temp-ruh-chuh despite the bold print for temp. Heinemann Lower Primary may be more help to a child to pronounce ak-tew-el so that it sounds like actual than Macmillan Primary with ak-choo-uhl.

Three of the children's dictionaries represent the unstressed syllable in a word with up to five letters, which can make quite a business of them. d-u-h-n-t for the final syllable in accident - is that really necessary? The word bicentenary has 11 letters but the pronunciation guide in Macquarie Junior takes 20 characters - nearly twice as long buy-suhn-teen-uh-ree! The Jolly Dictionary takes the simple approach of bold letters for the first letters of a stressed syllable, then following with spellings resembling conventional spelling, as in temperecher. When spoken with the stress on the bold letters, the slurring follows naturally - temperecher becomes temprecher, and acsident becomes acsident.

The Jolly Dictionary simplifies further by representing most slurred vowels with a different shape of e, a single character, which may make for less labored pronunciation, and also has the effect of making the guide look more like a normally spelled word. I think this approach could be tested as to whether it might be the most effective as an aid for children - to be as visually concise and as close to the conventional spelling as possible - that is, rely on bold print to show the stressed syllable, and for the rest, give the closest possible approximation to the lexical form of the word as in the example of acsident rather than ak-suh-dunt.

The second question is - How can learners find the words in a dictionary when the initial letters are not the same as the initial sound? - for example, knock? Macmillan Primary provides a handy table that sets out misleading initial letters and letter pairs. (For your information, by the way, the list is incomplete - most alphabet letters can be misleading in initial place, and nearly two thirds could be solved if spelling simply dropped the misleading initial letters when they are really surplus as in ghost, guess, who, gnaw, knight, rhinoceros, write, scheme and who.

Moseley's Aurally Coded Dictionary solves the problem by categorising words according to vowel sounds, and then setting them out alphabetically on each list, which is easy to scan down - but this is a specialised spelling dictionary, which does not include definitions.

Third question. How to show English speech sounds that English spelling does not discriminate. English spelling uses th for both this and thing, and has no distinctive spelling for the sound - - as in put, book and wolf.

Some dictionaries simply ignore these differences. Others make distinctions such as dh as in this and th as in thin, oo as in book and ooh as in boot. The Jolly Dictionary uses thinner and thicker letter-shapes for th - th is thick as in thair, and thin as in thin. Strong OO as in boot is larger, weak oo as in book is smaller. These are intuitively easy to understand, but are not easy to copy or to type on a standard keyboard. However, this may be the most helpful way to distinguish these pronunciations in a dictionary, and again, it keeps the dictionary guide looking similar to the real spelling.

Fourth - differences in dialect. What about English pronunciation guides internationally and in multicultural societies like ours, which has such a wide range of accents on the street? For children and English learners pronunciation guides have not so much a descriptive function as prescriptive - a guide to how they can speak the words and be understood by the community around them, and how, hearing the words or seeing them in print, they can find them in the dictionary with how to speak them.

Well, in fact, pronunciation keys in dictionaries show how little dialects and accents really matter when it comes to spelling - and hence, it should follow, to spelling reform, where this is always raised as a reason why improvement is impossible. Spelling is a standardised convention, like a line-sketch, and people then pronounce it in their own dialect like their own individual photograph - you could say an audio-sketch, contrasted with an audiograph. There are no problems in the spelling dog regardless of whether it is pronounced like dog/dawg/ or daag according to individual accents. The pronunciation differences that do occur between English dictionaries and Australian are in fact trivial. [1] A general sort of broadband works. And this will apply in designing spelling improvement too.

The fifth and last important issue concerns the so-called long vowels A E I O U. They are the biggest problem in English spelling - as well as for any attempts to make it more user-frendly. Long vowels have the greatest number of different ways to spell them. The children's dictionary guides may spell them in some of the more unusual ways - for example, uy oh ooh.

The Jolly Dictionary uses ligatured letters, ai, ee, ie, oa, ue, which make the spellings look shorter, though I gather there has been no research on whether this tactic is worth the trouble. The patterns, ai, ee, ie, oe, ue, are common in real spelling, and so can make the pronunciation guides look more familiar - less strange than as in, say, meg-uh-luh-may-nee-uh. It can still look rather odd with - for example mechooer, and oaaisis, even though that one resembles the real spelling better than oh-ay-sus.

I would like experiments that simply placed a grav accent as diacritic over long vowels - that is, the same letters as for short vowels, a e i o u, but with accents over them. This would achieve two things - align A E I O U with the short vowels a e i o u that they so often toggle with in words like nàtion, national, fìnìte and infinite, and so help to visually clarify their meaning and relationships. It would also be concise, unlike tthe present common clumsy 2-3 letter representations in dictionary guides. This tactic has seemed to work well in informal trial of pieces of text, but more experiment is needed to see how intuitive and easy such a marking would be for children.


- How useful to different categories of learners, including English-language learners, are the different ways to represent speech sounds in children's dictionaries? Research is needed if it does not exist - I have not found it.

But some things can be said.

First. The pronunciation guides of most children's dictionaries cater more for the advantaged children in what they assume learners already know, and so what they leave out. How can they be made most helpful to those who need them most - the educationally disadvantaged children with small vocabularies and poor reading skills, and the non-native learners of English language (EFL) - who now number more than its native speakers.

The Jolly Dictionary may be the way to go - trying to make the pronunciation guide look as close to present English spelling as possible. But how useful are its special characters which are simply like different fonts of the same letters, and are ligatured letters worth the bother? Its phonemic pronunciation guide looks so like present spelling patterns - does that promote literacy or make it likely for users to confuse the two, and start spelling phonemically themselves?

Secondly, looking to the future. There are many signs that informal spelling practices are changing rapidly, and also that Anglo-American literacy crises are by no means solved, and spelling difficulties are proven to be implicated.

Many spelling reformers have thought that the ideal way to begin spelling reform would be as a dictionary pronunciation key, because it is completely phonemic. Children's dictionaries in many other modern countries need no pronunciation guides for vocabulary because their spelling is consistent. Nevertheless this is not so simple for designing more user-friendly English spelling for the English language. You can see from the examples I have given that existing pronunciation keys for children's dictionaries would be clearly impossible candidates as spelling reform - the Jolly Dictionary perhaps comes closest.

A dictionary pronunciation key might represent the first level of a spelling reform - the alphabetic principle of letters representing sounds, - but there are good psycholinguistic reasons why this would be inadequate as a full reform. It is a practical necessity that any English spelling improvement must maintain the general appearance of present print. It must take into account the special features of the English language, and it must better match the needs and abilities of readers, spellers and learners. That is, primarily, it must facilitate reading and writing by eye as well as by ear, and reading for meaning as well as representing the spoken language. This means a visual appearance for print that can show word relationships, word structure and grammar.

There are many reasons, including commercial, why it is in the interests of dictionaries to seek to promote spelling improvement that would promote wider literacy by reducing its difficulties. What an increasing demand for dictionaries there could then be, from the wide swathe of populations world-wide that currently cannot read well enough to use one.

[1] Only Moseley's aurally coded spelling dictionary does have to take account of dialect because it categories spelling according to their pronunciation. It gets round this by placing words twice, with markers to show the Scottish pronunciations.

Pronunciation Guides
in 25 Dictionaries and 11 Word Books for children and schools.

September 2005.

None N 7Sometimes N 8 Usually N 3Always or mostly N 7
Collins Australian D.
Collins first school D. 2005
Kingfisher First D.
Ladybird pocket D.
Ladybird Read it yourself school D.
Oxford Very First D.
Times Chambers First Learners D. 1999

Eleven Word Books for children
Collins Australian Ringbook D. 2003
Collins Shorter School D.
Macmillan Australian Primary D. 2005
Macquarie Junior D. 1992
Oxford Junior D. 7+
Oxford Australian Primary D. 2002
Oxford Children's D. 2003
Penguin Pocket English D. 2004
McGraw-Hill Junior School D. (Singapore) 1974
Oxford Basic School D.
Oxford My First Australian D. & Thesaurus
Heinemann Lower Primary D. 2003
Heinemann Primary D 2004
Jolly D. 2003
Learning Development Aids Moseley: Aurally Coded English Spelling D. No date.
Macmillan Australian School D.
Macquarie Budget D. 2004 (IPA)
Oxford My first Australian D. 2004

Examples of Pronunciation Guides
in dictionaries for children & schools

(omitting IPA International Phonetic Alphabet).

Concise Oxford 1934 (adapted to keyboard) Macquarie junior Heinemann Lower Primary Macmillan Primary Jolly* Suggested
a.cident (-ks) ak-suh-duhntak-sa-dent ak-suh-duhntacsident acsident
a.ctual- ak-tew-elak-chooh-uhl acchooelactùal
age-ayj -aijàj
a.ny (e -) en-ee en-eeen-ee eneeeni
.remony - se-ra-mo-neese-ruh-muh-nee seremeneeseremoni
chauffeur shôfer. sho-fuh or sho-fersho-fa not innot insho-fer
Fe.bruary feb-yoo-uh-ree -feb-rooh-oh-ree or feb-yooh-uh-ree febrooereefebruari
ferö.cious fuh-roh-shuhs not in-feroashes fer-o-shus
occa.sion (äzhn)- o-kay-zhonuh-kay-zhuhn ecaizheno-cà-zhon
occur. (-rr-) -o-ker -not inocur
phy.sical (f) - fiz-i-kulfiz-i-kuhl fizicelfizical
psycho.logy (s) suy-kol-uh-jeenot in not innot incoloji
re.staurant (-or) res-tuh-ront rest-a-ront res-tuh-ront resteront restaront
scare -skair -sceerscair
te.mperature (tsher,-tür-) temp-ruh-chuhtem-pru-cher tempruh-chuhtemperecher tempratùr
their (dhär)-thair -theerthair
va.cuum vak-yoohm vak-yoohmvak-yoohm vacuemvacùm
varî.ety - --not in
vâr.ïîêgated vair-ree-uh-gayt-uhd va-rye-a-ted vuh-ruy-uh-ted - vairigàted

* Not shown: 1. Ligatures for ai ah ar aw ee er ieng oa oo oi or ou ue qu ch sh th zh.
2. Special characters distinguishing a a and e e.
3. two sizes for oo OO and th th

Back to the top.