logo ELT Concourse teacher training
Concourse 2

English spelling

shcool nohing excellense

"a conspiracy to undermine a country that won't tow the western line" – BBC News website 17 December 2014

"Angry Hamilton accuses Vettel of breaking rules with break testing" – Daily Mail, 01 May 2018

duel Earlier this month David Cameron, the prime minister, visited Stonehenge ... and announced plans to duel the A303
The Daily Telegraph website 21 December 2014

We all make mistakes.  There are three sorts above.  Which is which?

  1. Slips and typographical errors
  2. Errors caused by the lack of sound-spelling consistency in English
  3. Errors caused by simply not knowing how a word is spelled


  1. The first two pictures (shcool and nohing) are just slips and they happen frequently.  For the purposes of this guide, we can ignore them because they have nothing to do with spelling rules in English.
  2. The third picture (excellense) is the result of someone simply not knowing how the word excellence is spelled in English.
  3. The quotations from the BBC, Daily Mail and Daily Telegraph websites are errors caused by the lack of sound-spelling consistency in English.  They are called homophone errors because toe and tow, break and brake and duel and dual sound the same but are differently written and mean different things.  News sites are a rich source of such errors because many news stories are submitted by journalists over the telephone.  Another example, from the Daily Telegraph, sadly since corrected, is the description of shelves which "grown with academic tomes".


English spelling is wholly irregular and random.  Right?


It is often claimed that English spelling is impossible, or at least very difficult, to learn.  For example:

English spelling is probably the most irregular spelling system of all those based on the alphabetic system.  Not only can't you tell how to spell a word from hearing it spoken, you can't predict how a word is spoken from the written word either.
The English Spelling Society

In fact, that is only very partially right because there are clear connections between how a word is pronounced and its spelling despite the fact, which is not arguable, that there are numerous exceptions.
It is also not a very well-informed opinion because many other languages have a deep orthography (see below) in which the connections between sound and spelling are difficult always to discern.
Among other languages with a deep orthography like English are, incidentally, Korean (in which only one consonant is pronounced as it is written), Hungarian, Mongolian, Thai and French.
Even Italian, which has a shallow, i.e., predictable, spelling system has areas of full unpredictability.

To see how intuition and the application of simple rules allows you to predict the sound of a word, try pronouncing these nonsense words.
Click here when you have tried the test.

flubber flooksendy flimmax greancumbler
drimble bentle heasering chumblegrint

To understand what follows, you need to understand that we are considering sounds not letters.
So, for example, the single letter 'p' is represented as a single sound (transcribed, i.e., written as /p/) and even a double 'p' as in pepper still only has a single sound /p/ in the middle and at the beginning.
Pairs of letters like 'sh', 'ch', 'th', 'ck', 'sc' etc. are also often only single sounds as they are in, for example:
    shave (transcribed as /ʃeɪv/ with a single symbol [/ʃ/] representing the first two letters)
    school (transcribed as /skuːl/ with a single symbol [/k/] representing 'ch')
    this (transcribed as /ðɪs/ with a single symbol [/ð/] representing the first two letters)
    sock (transcribed as /sɒk/ with a single symbol [/k/] representing the last two letters)
    science (transcribed as /ˈsaɪəns/ with a single symbol [/s/] representing the first two letters)

You do not need to be able to read the phonemic transcriptions to understand what follows (although it will help).


The rules

Here's a short list of the rules:

Long and short vowels

  1. short vowels only usually require a single letter: bit, bat, sit, hat, kit, cut and so on.  This is the Consonant-Vowel-Consonant principle or CVC.  It is adhered to in thousands of words and knowing it allows us to predict from the spelling how, for example, all of the following will be pronounced.  The CVC principle also allows us to infer the spelling from the sound, even if we have never heard the word.

    and innumerable more words.
  2. long vowels are represented in two ways:
    1. by inserting a vowel, often 'e' or 'i', after the consonant: bite, bate, site, stating, hate, kite, cute, driving and so on.  This is the Vowel-Consonant-Vowel (VCV) principle.
    2. by inserting a second vowel letter before the consonant: beat, boat, heat etc.  This is the Consonant-Vowel-Vowel-Consonant (CVVC) principle.

Applying to these two principles, along with principle a., allows us to predict how another huge range of words will be pronounced, including, for example:

and thousands more.  Again, this will work in reverse: we can make a very good guess at the spellings of these words when we hear them by applying the principles.

There are exceptions (of course) and they are common words, usually.  They include:
    come, done, give, gone, have, live, love, none, one, some
All these have short vowels but end VCV.  They are quite simple to learn and learners usually have little trouble pronouncing or spelling them.

Consonant doubling

  1. consonant doubling often occurs in an effort to 'protect' the short vowel sound.
    If the consonant were not doubled in hatter, for example, the word would be pronounced as hater (because the VCV pattern would compel it) and the same goes for very many other words or word pairs such as diner-dinner, taper-tapper, cuter-cutter, fate, fatter, later-latter etc.  This is the VCCV principle.
  2. the consonant is doubled when adding an -ed or -ing ending only if the stress is on the second syllable: befit-befitting, remit-remitting but offer-offering etc.  This rule is often broken so we get, e.g., focussed and focused, targetting and targeting.
  3. regardless of stress, British English doubles the consonant if it is 'p' or 'l'.  American English usually does not (see below).

Spelling the /k/ sound

The /k/ sound (as in, e.g., click, black, corn, bacon etc.) also obeys principles:

  1. single 'c' is the most common (and the way to bet) and can occur anywhere: actor, cactus, cart, because etc.
  2. doubling of the 'c' follows the same principle as any other consonant doubling: it protects the short vowel.  So we have tobacco, Mecca etc.  This is the VCCV principle, again.
  3. if the /k/ sound is followed by 'e', 'i' or 'y', 'k' is used instead of 'c': make, sketch, token, sake, slinky, skin etc.
  4. 'ck' similarly replaces 'cc' if it is followed by 'e', 'i' or 'y': panicky, tarmacked etc.
  5. 'ck' always follows a short vowel: luck, duck, back, fickle etc.  This obeys the CVC principle because, as we saw, the 'ck' combination is actually a single consonant sound (/k/).
  6. 'k' follows any other vowel or consonant: soak, coke, musk, lark etc.
  7. /kw/ is always spelled 'qu': queen, quince, equine etc.

Spelling the /dʒ/ sound (as in the 'g' in gentle)

  1. if the sound is followed by 'a', 'o' or 'u', the letter 'j' is preferred: injure, just, jangle, jumble etc.
  2. if the sound is followed by 'e', 'i' or 'y', the letter 'g' is preferred: aged, ginger, gauge, original, mangy etc.
  3. 'j' cannot be doubled (to protect a short vowel, again) so 'dg' is used instead: judged, budgerigar


No English words end in 'v' so we always have a following 'e': thieve, save, shave etc.

ie or ei?

Schoolchildren learn to chant I before E except after C and for hundreds of words like believe, perceive, ceiling, thieving etc., the rule works just fine.  Unfortunately, there are exceptions, as usual.

  1. Before gh, gn and ght: the spelling is ei not ie:
    weigh, neigh, foreign, feign, reign, height, freight
  2. There are numerous exceptions which include:
    beige, counterfeit, either, forfeit, heir, leisure, neither, rein, seize, their, veil, vein, weir, weird


  1. words ending in 'y'
    1. if you add an 's', the 'y' changes to 'ie': worry-worries, cry-cries etc.
    2. words ending in 'ey' usually do not change when an 's' or 'd' is added: valley-valleys, player-players-played etc.  (But lay-laid, say-said etc.)
    3. 'y' does not change when it's followed by 'i': dry-drying, cry-crying, spy-spying etc.
  2. words ending in 'e'
    1. words ending in a consonant + 'e' drop the 'e' when adding anything beginning with a vowel: love-loving, response-responsible etc.  (But, in this case, the long vowel is often protected by retaining the 'e': likeable, mileage etc.)
    2. the soft 'c' (/s/) and 'g' (/dʒ/) sounds also mean that the 'e' is retained: manage-manageable, outrage-outrageous, trace-traceable etc.
    3. Some adjectives ending in -able can have alternative spellings, with and without the retention of the 'e'.  For example:
          likeable / likable
          loveable / lovable
          saleable / salable
          sizeable / sizable
          useable / usable
  3. words ending in 'ie' change to 'y': die-dying, lie-lying
  4. words ending in 'ue' drop the 'e': argue-argument, true-truly
  5. words ending in 's', 'z', 'x', 'ch' or 'sh' take 'es': watch-watches, match-matches, fizz-fizzes, fix-fixes, bush-bushes, bus-buses etc.

(There is a little more on the spelling and pronunciation of verb endings in the guide to basic verb forms.)

So there are rules.  Rather too many of them, in fact, and the list above is nowhere near complete.


Teaching the rules

You can't of course, teach all the rules listed here at one go.  Many would say that learning English spelling rules is actually impossible, and unnecessary because people will pick up the patterns by exposure.  That's certainly arguable and it is clear that the more exposure people have to examples of the patterns, the more likely they are to absorb them.
It is, however, worth taking each of the regular patterns, one at a time, and focusing 20 minutes or so on them now and again (weekly, perhaps).
Spelling rules lend themselves to light-hearted games, races and competitions in the classroom because they have Right/Wrong answers.



As we saw, not even the BBC is immune to the homophone trap.  Many native speakers trip up in this area.
If you'd like a list of a couple of hundred common ones, click here.
It is helpful, then, to look through the language you are presenting just to see if there are any homophone traps the authors or you have fallen into.
It is, however, not usually worth focusing on them because that can confuse as much as it enlightens.

There are, however, a set of homophones in English on which it is worth focusing.  In the following, can you detect the problem to describe in the right-hand column?  Click on the table when you have a few notes.

spelling task


Why is English spelling irregular?

OK, so spelling in English is tough because a) there are lots of confusing rules and b) even when you have mastered the rules, there's always an exception.  If you look back at the list above, you can probably find an exception to every single rule.
English spelling is something of a mess because of its history:

good news

English spelling: the good news

One advantage of having a rather odd spelling system is that meaning can often be inferred from spelling.  For example, in English, these three words are pronounced the same (they are homophones):

rite, write, right

But, because of the different spellings, we can easily see which are connected to the following words and if we know the base word, we can have a good stab at inferring the meaning of the others:

written, ritual, rightness, rights, writing, rites, righteous

Spelling and meaning, in English, are connected.
Reforming English spelling would mean that this handy reference would disappear and all three words would probably be spelled rite.
That, many argue, would be a loss.  English, they point out, is not meant to have a phonetic spelling system.

So, next time your students despair of being able to spell correctly, it may be worth pointing this out.

Another thing to point out to despairing learners is that native speakers also have trouble.  We often see, for example, dessicate, acommodation (or accomodation), embarass, wierd, definately and so on.
You, of course, can correct all those.



The term diacritic refers to the accent marks which exist in many languages and show a difference in pronunciation (or used to).
Modern English retains the use of some accent marks on imported words, almost always those from French, but it is often a matter of personal taste.  Using role instead of rôle is common and hôtel is almost never seen, for example.  However, in other cases, words more recently or more obviously borrowed from French retain the accents: cliché, café, vis-à-vis, tête-à-tête etc.
When there is a possibility of ambiguity, the convention is to retain the accent markers, so we have résumé and resume.
Loan words from other languages usually get stripped of their diacritics, although über from German is seen alongside uber.  The words doppelganger and doppelgänger are in free variation and English usually removes the capital letter obligatory in German on all nouns.  The Turkish döner kebap is almost always spelled as doner kebab in English.
The dieresis signalled by two dots over a vowel to denote a separate pronunciation rather than a diphthong is becoming rarer and it does not appear on most computer keyboards so, for example, naive and naïve are both seen.  The mark appears over the second of two vowels unless it is intended to signify that the final vowel is not silent (as, for example in Brontë).  It is retained on many proper names so we see, for example, Chloë, Eloïse, Zoë, Noël, Raphaël etc.
Some publications, notably the New Yorker, retain the mark for separately pronounced vowels in, e.g., reëntry and coöperate but this is increasingly rare and inconsistent.

union jack usa

English (BrE) and American (AmE)

English (BrE) and American spelling (AmE) differ as everyone knows but the differences are not huge.  The number of words affected is probably fewer than 400.  (Given that the 20-volume Oxford English Dictionary contains entries for 171,476 words in current use, and the Oxford website estimates that there are at least 250,000 distinct words in English, that represents a tiny proportion, less than 0.25%.)
In most cases, the AmE variant is simpler and rooted in Noah Webster's concern to eliminate inconsistencies.  In this he was only partially successful.
Here are the most commonly cited differences with the BrE version given first followed by the AmE version:

As a rule of thumb, Commonwealth countries tend to follow the BrE rules although Canadian use is more often influenced by AmE.

There's no test on all this but there are plenty of spelling tests to be had online (with a good one at http://www.oxforddictionaries.com/spelling-challenge/) and two on this site.  One for B1/B2-level learners, here, and another for A1/A2-level learners here.
There is also a short lesson for learners with tests and exercises covering the essentials of this guide.


Writing systems

This survey of English spelling would not be complete (as far as teaching the system is concerned) without some consideration of other people's writing systems.
All written systems probably developed from an early pictographic system in which the item referred to was represented by a picture (or icon) recognisably depicting the word.  The assumption is also made that all European writing systems derive ultimately from the system in use in Mesopotamia from around 4000BCE.  So, for example, the sign for mouth could be interpreted as mouth, speak, say, word and so on.
Later, when writing with wedge-shaped tools was adopted, the signs were simplified to what is known as cuneiform but remained essentially ideographic.  Gradually, Sumerian took on some alphabetic characteristics because the simplified cuneiform script was developed, through what is known as the rebus principle, to make the sound of a word stand for all instances of that sound in other words.  For example the symbol eye can be made to stand for eye, I and aye, and for the same sound in words such as ice, idle, belie, decry and so on.
By the rebus principle, the famous
    To be or not to be?
can be represented as:
to be
Ideographic and pictographic symbols are still very widely used, of course, or you would know which was the appropriate toilet to use in a culture whose language you don't speak but no modern languages rely on a pictographic script.
Modern scripts, by contrast, rely on graphemes and they are defined in two ways:

  1. as the smallest units which represent a sound in a language so, for example, in English:
        th is the grapheme representing the /ð/ in this and the /θ/ in heath.
  2. as the smallest recognisable units of the written system so, in English:
        t and h are separate graphemes.

You may encounter the word glyph which has a connected meaning.  A glyph is a readable unit of a writing system and may not be a member of an alphabet in itself so, for example, the letter g in English can also be written as G and g.  The alternative ways to write the letter are called alloglyphs.

There are three fundamental sorts of writing systems which rely on graphemes and glyphs:

  1. Alphabetic systems
    English has ostensibly an alphabetic system in which individual letters represent individual sounds (phonemes) but, as we saw above, spelling in English is not a good guide often to pronunciation and vice versa.  Some sounds in English are represented by digraphs and even trigraphs so, for example:
        th is this represents the single sound /ð/
        sh in hush represents /ʃ/
        ugh in cough represents /f/
    and so on.
    English may be described as having deep orthography.  This means that the system is poor in relation to shallow languages such as German or Serbo-Croat in which there is a good correspondence between spelling and pronunciation.
    Languages having a deep orthography include French, Mongolian, Modern Greek and, of course, English.  Languages with shallower orthography include Spanish, Italian and Czech (and even Italian has some notable irregularities).
    Most European languages and many languages which have only recently acquired a written form use an alphabetic system of writing although scripts vary from Latin (in which this is written) through Greek and Cyrillic to many other systems.  For obvious reasons, alphabetic systems are sometimes referred to as segmental systems because each word can be broken down into sound segments.
    Although many languages add a range of accents and diacritics to letters to represent certain sounds there are only three alphabets in use in Europe: Cyrillic, Greek and Latin.
  2. Syllabic systems
    In these writing systems, each grapheme (i.e., independent symbol) represents a syllable rather than a single sound.  Examples include Arabic, the Japanese Hiragana and Katakana scripts, Hebrew and Amharic among others.  Languages which deploy scripts like these usually have a syllabary (i.e., the total number of available syllables) which is quite limited.  English would be ill-suited to such as system as the number of possible syllables is well over 10,000 and each would require a separate grapheme to represent it.  Japanese, by contrast, makes do with around 100 syllables so the language is suited to this form of written representation.
    The Arabic script is used incidentally, for many other languages unrelated to Arabic (or only distantly so).
    Both Hebrew and Arabic are described as abjad scripts because vowels are not separately represented.
  3. Logographic systems
    In these systems, each word is represented by a single symbol (or glyph).  The most obvious and famous examples are the Chinese languages and these are the only common surviving languages to use a logographic system.  For example, the symbol sun represents the concept of sun or day.  Egyptian hieroglyphics are another familiar form of logographic writing (developed from a pictographic system).  Chinese characters are extensively used in written Japanese, Korean and Vietnamese.
    Despite the fact that to be literate in Chinese it is necessary to learn to recognise around 5000 symbols (which places something of a strain on the educational system), the great advantage of the system is that symbols represent meanings, not sounds, so, even though various Chinese languages are not mutually comprehensible in the spoken form, they are in the written form.

Each type of writing system is processed by readers in a different way and there is good reason to suppose that these processing strategies will, especially in the early stages of learning, be carried over to attempts to read in English.  For example, Arabic speakers may be indifferent to vowels (as they are not represented in Arabic) and may confuse, e.g., glass with gallons.  Even speakers of languages which use a Latin script such as Spanish may have problems processing words in English because Spanish has a relatively shallow orthography.


Spelling sounds

One area of clear irregularity in English concerns the spelling of individual sounds rather than whole words.  English is not unique in this respect but it is, perhaps, an extreme example.
Many other languages, including Portuguese and Modern Greek have very weak relationships between sound and spelling.  Portuguese, for example, has only 5 vowel symbols but these can be pronounced in 17 different ways and the language's 18 consonants have 30 pronunciations.  Modern Greek has no fewer than 6 ways to spell the long 'ee' (/iː/) sound.

This makes writing unknown words down from dictation in English a very challenging and often disheartening business.  It is much easier to write down accurately what you hear in some other languages, especially those that have undergone occasional reform of the writing system to get closer to a one-to-one relationship between the sound and the spelling, such as German or Spanish.

In the in-service section of this site, some detail is provided concerning how consonant and vowel sounds are conventionally spelled in English.  For consonants, for example:

Vowel sounds are more variable so, for example:

Both the guide to consonants and the guide to vowels, linked below, consider how the sounds are conventionally written.

Related guides
punctuation for the guide to a related area
pronunciation for the initial plus index to pronunciation guides
pronunciation for the in-service index to pronunciation guides
types of languages for the guide that considers many other differences between languages
consonants for a guide which considers how these are pronounced in English and how they are spelled
vowels for a guide which considers how these are pronounced in English and how they are spelled
the roots of English for a guide to how English has developed over 1500 years and why its spelling is so odd

BBC News: http://www.bbc.com/news/world-europe-30492348 [accessed 17/12/2014]
Campbell, G, 1998, Concise Compendium of the World's Languages, London: Routledge
Chalker, S, 1987, Current English Grammar, London: Macmillan
Daily Telegraph: http://www.telegraph.co.uk/news/earth/environment/archaeology/11303127/Stonehenge-discovery-could-rewrite-British-pre-history.html [accessed 21/12/2014]

Oxford dictionaries on line: http://www.oxforddictionaries.com [accessed 17/12/2014]
Proctor, P (Ed.), 1995, Cambridge International Dictionary of English, Cambridge: Press Syndicate of the University of Cambridge
The English Spelling Society: http://www.spellingsociety.org/ [accessed 17/12/2014]